diff options
Diffstat (limited to 'arch/parisc')
164 files changed, 6260 insertions, 2309 deletions
diff --git a/arch/parisc/Kbuild b/arch/parisc/Kbuild index a6d3b280ba0c..749b195f2894 100644 --- a/arch/parisc/Kbuild +++ b/arch/parisc/Kbuild @@ -1,5 +1,5 @@ # SPDX-License-Identifier: GPL-2.0-only -obj-y += mm/ kernel/ math-emu/ +obj-y += mm/ kernel/ math-emu/ net/ # for cleaning subdir- += boot diff --git a/arch/parisc/Kconfig b/arch/parisc/Kconfig index a98940e64243..47fd9662d800 100644 --- a/arch/parisc/Kconfig +++ b/arch/parisc/Kconfig @@ -8,25 +8,32 @@ config PARISC select HAVE_FUNCTION_GRAPH_TRACER select HAVE_SYSCALL_TRACEPOINTS select ARCH_WANT_FRAME_POINTERS + select ARCH_HAS_CPU_CACHE_ALIASING + select ARCH_HAS_DMA_ALLOC if PA11 + select ARCH_HAS_DMA_OPS select ARCH_HAS_ELF_RANDOMIZE select ARCH_HAS_STRICT_KERNEL_RWX select ARCH_HAS_STRICT_MODULE_RWX - select ARCH_HAS_UBSAN_SANITIZE_ALL + select ARCH_HAS_UBSAN select ARCH_HAS_PTE_SPECIAL select ARCH_NO_SG_CHAIN + select ARCH_SPLIT_ARG64 if !64BIT select ARCH_SUPPORTS_HUGETLBFS if PA20 select ARCH_SUPPORTS_MEMORY_FAILURE select ARCH_STACKWALK + select ARCH_HAS_CACHE_LINE_SIZE select ARCH_HAS_DEBUG_VM_PGTABLE select HAVE_RELIABLE_STACKTRACE - select DMA_OPS select RTC_CLASS select RTC_DRV_GENERIC select INIT_ALL_POSSIBLE select BUG - select BUILDTIME_TABLE_SORT + select HAVE_KERNEL_UNCOMPRESSED select HAVE_PCI select HAVE_PERF_EVENTS + select HAVE_PERF_REGS + select HAVE_PERF_USER_STACK_DUMP + select PERF_USE_VMALLOC select HAVE_KERNEL_BZIP2 select HAVE_KERNEL_GZIP select HAVE_KERNEL_LZ4 @@ -36,30 +43,40 @@ config PARISC select GENERIC_ATOMIC64 if !64BIT select GENERIC_IRQ_PROBE select GENERIC_PCI_IOMAP + select GENERIC_IOREMAP select ARCH_HAVE_NMI_SAFE_CMPXCHG select GENERIC_SMP_IDLE_THREAD select GENERIC_ARCH_TOPOLOGY if SMP + select ARCH_SUPPORTS_SCHED_MC if SMP && PA8X00 select GENERIC_CPU_DEVICES if !SMP select GENERIC_LIB_DEVMEM_IS_ALLOWED select SYSCTL_ARCH_UNALIGN_ALLOW + select SYSCTL_ARCH_UNALIGN_NO_WARN select SYSCTL_EXCEPTION_TRACE select HAVE_MOD_ARCH_SPECIFIC select MODULES_USE_ELF_RELA select CLONE_BACKWARDS select TTY # Needed for pdc_cons.c + select HAS_IOPORT if PCI || EISA select HAVE_DEBUG_STACKOVERFLOW + select ARCH_WANT_DEFAULT_TOPDOWN_MMAP_LAYOUT + select HAVE_ARCH_MMAP_RND_COMPAT_BITS if COMPAT + select HAVE_ARCH_MMAP_RND_BITS select HAVE_ARCH_AUDITSYSCALL select HAVE_ARCH_HASH - select HAVE_ARCH_JUMP_LABEL - select HAVE_ARCH_JUMP_LABEL_RELATIVE + # select HAVE_ARCH_JUMP_LABEL + # select HAVE_ARCH_JUMP_LABEL_RELATIVE select HAVE_ARCH_KFENCE select HAVE_ARCH_SECCOMP_FILTER select HAVE_ARCH_TRACEHOOK + select HAVE_EBPF_JIT + select ARCH_WANT_DEFAULT_BPF_JIT select HAVE_REGS_AND_STACK_ACCESS_API + select HOTPLUG_CORE_SYNC_DEAD if HOTPLUG_CPU select GENERIC_SCHED_CLOCK select GENERIC_IRQ_MIGRATION if SMP select HAVE_UNSTABLE_SCHED_CLOCK if SMP - select LEGACY_TIMER_TICK + select GENERIC_CLOCKEVENTS select CPU_NO_EFFICIENT_FFS select THREAD_INFO_IN_TASK select NEED_DMA_MAP_STATE @@ -68,13 +85,13 @@ config PARISC select HAVE_KPROBES select HAVE_KRETPROBES select HAVE_DYNAMIC_FTRACE if $(cc-option,-fpatchable-function-entry=1,1) - select HAVE_FTRACE_MCOUNT_RECORD if HAVE_DYNAMIC_FTRACE select FTRACE_MCOUNT_USE_PATCHABLE_FUNCTION_ENTRY if DYNAMIC_FTRACE select HAVE_KPROBES_ON_FTRACE select HAVE_DYNAMIC_FTRACE_WITH_REGS select HAVE_SOFTIRQ_ON_OWN_STACK if IRQSTACKS select TRACE_IRQFLAGS_SUPPORT select HAVE_FUNCTION_DESCRIPTORS if 64BIT + select PCI_MSI_ARCH_FALLBACKS if PCI_MSI help The PA-RISC microprocessor is designed by Hewlett-Packard and used @@ -105,9 +122,12 @@ config ARCH_HAS_ILOG2_U64 default n config GENERIC_BUG - bool - default y + def_bool y depends on BUG + select GENERIC_BUG_RELATIVE_POINTERS if 64BIT + +config GENERIC_BUG_RELATIVE_POINTERS + bool config GENERIC_HWEIGHT bool @@ -122,6 +142,20 @@ config TIME_LOW_RES depends on SMP default y +config ARCH_MMAP_RND_BITS_MIN + default 18 if 64BIT + default 8 + +config ARCH_MMAP_RND_COMPAT_BITS_MIN + default 8 + +config ARCH_MMAP_RND_BITS_MAX + default 18 if 64BIT + default 13 + +config ARCH_MMAP_RND_COMPAT_BITS_MAX + default 13 + # unless you want to implement ACPI on PA-RISC ... ;-) config PM bool @@ -129,6 +163,10 @@ config PM config STACKTRACE_SUPPORT def_bool y +config LOCKDEP_SUPPORT + bool + default y + config ISA_DMA_API bool @@ -207,9 +245,9 @@ config PARISC_HUGE_KERNEL def_bool y if !MODULES || UBSAN || FTRACE || COMPILE_TEST config MLONGCALLS - def_bool y if PARISC_HUGE_KERNEL bool "Enable the -mlong-calls compiler option for big kernels" if !PARISC_HUGE_KERNEL depends on PA8X00 + default PARISC_HUGE_KERNEL help If you configure the kernel to include many drivers built-in instead as modules, the kernel executable may become too big, so that the @@ -224,9 +262,9 @@ config MLONGCALLS Enabling this option will probably slow down your kernel. config 64BIT - def_bool y if "$(ARCH)" = "parisc64" bool "64-bit kernel" if "$(ARCH)" = "parisc" depends on PA8X00 + default "$(ARCH)" = "parisc64" help Enable this if you want to support 64bit kernel on PA-RISC platform. @@ -243,6 +281,7 @@ choice config PARISC_PAGE_SIZE_4KB bool "4KB" + select HAVE_PAGE_SIZE_4KB help This lets you select the page size of the kernel. For best performance, a page size of 16KB is recommended. For best @@ -258,10 +297,12 @@ config PARISC_PAGE_SIZE_4KB config PARISC_PAGE_SIZE_16KB bool "16KB" + select HAVE_PAGE_SIZE_16KB depends on PA8X00 && BROKEN && !KFENCE config PARISC_PAGE_SIZE_64KB bool "64KB" + select HAVE_PAGE_SIZE_64KB depends on PA8X00 && BROKEN && !KFENCE endchoice @@ -282,14 +323,6 @@ config SMP If you don't know what to do here, say N. -config SCHED_MC - bool "Multi-core scheduler support" - depends on GENERIC_ARCH_TOPOLOGY && PA8X00 - help - Multi-core scheduler support improves the CPU scheduler's decision - making when dealing with multi-core CPU chips at a cost of slightly - increased overhead in some places. If unsure say N here. - config IRQSTACKS bool "Use separate kernel stacks when processing interrupts" default y @@ -298,16 +331,6 @@ config IRQSTACKS for handling hard and soft interrupts. This can help avoid overflowing the process kernel stacks. -config TLB_PTLOCK - bool "Use page table locks in TLB fault handler" - depends on SMP - default n - help - Select this option to enable page table locking in the TLB - fault handler. This ensures that page table entries are - updated consistently on SMP machines at the expense of some - loss in performance. - config HOTPLUG_CPU bool default y if SMP @@ -340,32 +363,20 @@ config NR_CPUS int "Maximum number of CPUs (2-32)" range 2 32 depends on SMP - default "4" if 64BIT + default "8" if 64BIT default "16" -config KEXEC - bool "Kexec system call" - select KEXEC_CORE - help - kexec is a system call that implements the ability to shutdown your - current kernel, and to start another kernel. It is like a reboot - but it is independent of the system firmware. And like a reboot - you can start any kernel with it, not just Linux. - - It is an ongoing process to be certain the hardware in a machine - shutdown, so do not be surprised if this code does not - initially work for you. - -config KEXEC_FILE - bool "kexec file based system call" - select KEXEC_CORE - select KEXEC_ELF - help - This enables the kexec_file_load() System call. This is - file based and takes file descriptors as system call argument - for kernel and initramfs as opposed to list of segments as - accepted by previous system call. - endmenu +config ARCH_SUPPORTS_KEXEC + def_bool y + +config ARCH_SUPPORTS_KEXEC_FILE + def_bool y + +config ARCH_SELECTS_KEXEC_FILE + def_bool y + depends on KEXEC_FILE + select KEXEC_ELF + source "drivers/parisc/Kconfig" diff --git a/arch/parisc/Kconfig.debug b/arch/parisc/Kconfig.debug index f66554cd5c45..f4f164eb12df 100644 --- a/arch/parisc/Kconfig.debug +++ b/arch/parisc/Kconfig.debug @@ -1 +1,23 @@ # SPDX-License-Identifier: GPL-2.0 +# +config LIGHTWEIGHT_SPINLOCK_CHECK + bool "Enable lightweight spinlock checks" + depends on DEBUG_KERNEL && SMP && !DEBUG_SPINLOCK + default y + help + Add checks with low performance impact to the spinlock functions + to catch memory overwrites at runtime. For more advanced + spinlock debugging you should choose the DEBUG_SPINLOCK option + which will detect unitialized spinlocks too. + If unsure say Y here. + +config TLB_PTLOCK + bool "Use page table locks in TLB fault handler" + depends on DEBUG_KERNEL && SMP + default n + help + Select this option to enable page table locking in the TLB + fault handler. This ensures that page table entries are + updated consistently on SMP machines at the expense of some + loss in performance. + diff --git a/arch/parisc/Makefile b/arch/parisc/Makefile index a2d8600521f9..48ae3c79557a 100644 --- a/arch/parisc/Makefile +++ b/arch/parisc/Makefile @@ -11,7 +11,7 @@ # Copyright (C) 1994 by Linus Torvalds # Portions Copyright (C) 1999 The Puffin Group # -# Modified for PA-RISC Linux by Paul Lahaie, Alex deVries, +# Modified for PA-RISC Linux by Paul Lahaie, Alex deVries, # Mike Shaver, Helge Deller and Martin K. Petersen # @@ -39,7 +39,9 @@ endif export LD_BFD -# Set default 32 bits cross compilers for vdso +# Set default 32 bits cross compilers for vdso. +# This means that for 64BIT, both the 64-bit tools and the 32-bit tools +# need to be in the path. CC_ARCHES_32 = hppa hppa2.0 hppa1.1 CC_SUFFIXES = linux linux-gnu unknown-linux-gnu suse-linux CROSS32_COMPILE := $(call cc-cross-prefix, \ @@ -50,12 +52,12 @@ export CROSS32CC # Set default cross compiler for kernel build ifdef cross_compiling - ifeq ($(CROSS_COMPILE),) + ifeq ($(CROSS_COMPILE),) CC_SUFFIXES = linux linux-gnu unknown-linux-gnu suse-linux CROSS_COMPILE := $(call cc-cross-prefix, \ $(foreach a,$(CC_ARCHES), \ $(foreach s,$(CC_SUFFIXES),$(a)-$(s)-))) - endif + endif endif ifdef CONFIG_DYNAMIC_FTRACE @@ -119,6 +121,8 @@ export LIBGCC libs-y += arch/parisc/lib/ $(LIBGCC) +drivers-$(CONFIG_VIDEO) += arch/parisc/video/ + boot := arch/parisc/boot PALO := $(shell if (which palo 2>&1); then : ; \ @@ -137,7 +141,7 @@ palo lifimage: vmlinuz fi @if test ! -f "$(PALOCONF)"; then \ cp $(srctree)/arch/parisc/defpalo.conf $(objtree)/palo.conf; \ - echo 'A generic palo config file ($(objree)/palo.conf) has been created for you.'; \ + echo 'A generic palo config file ($(objtree)/palo.conf) has been created for you.'; \ echo 'You should check it and re-run "make palo".'; \ echo 'WARNING: the "lifimage" file is now placed in this directory by default!'; \ false; \ @@ -175,12 +179,8 @@ vdso_prepare: prepare0 $(Q)$(MAKE) $(build)=arch/parisc/kernel/vdso32 include/generated/vdso32-offsets.h endif -PHONY += vdso_install - -vdso_install: - $(Q)$(MAKE) $(build)=arch/parisc/kernel/vdso $@ - $(if $(CONFIG_COMPAT_VDSO), \ - $(Q)$(MAKE) $(build)=arch/parisc/kernel/vdso32 $@) +vdso-install-y += arch/parisc/kernel/vdso32/vdso32.so +vdso-install-$(CONFIG_64BIT) += arch/parisc/kernel/vdso64/vdso64.so install: KBUILD_IMAGE := vmlinux zinstall: KBUILD_IMAGE := vmlinuz diff --git a/arch/parisc/boot/Makefile b/arch/parisc/boot/Makefile index b873ee4720ca..657f967240ee 100644 --- a/arch/parisc/boot/Makefile +++ b/arch/parisc/boot/Makefile @@ -10,7 +10,7 @@ subdir- := compressed $(obj)/image: vmlinux FORCE $(call if_changed,objcopy) -$(obj)/bzImage: $(obj)/compressed/vmlinux FORCE +$(obj)/bzImage: $(if $(CONFIG_KERNEL_UNCOMPRESSED),$(objtree)/vmlinux,$(obj)/compressed/vmlinux) FORCE $(call if_changed,objcopy) $(obj)/compressed/vmlinux: FORCE diff --git a/arch/parisc/boot/compressed/Makefile b/arch/parisc/boot/compressed/Makefile index a294a1b58ee7..f8481e4e9d21 100644 --- a/arch/parisc/boot/compressed/Makefile +++ b/arch/parisc/boot/compressed/Makefile @@ -5,10 +5,6 @@ # create a compressed self-extracting vmlinux image from the original vmlinux # -KCOV_INSTRUMENT := n -GCOV_PROFILE := n -UBSAN_SANITIZE := n - OBJECTS := head.o real2.o firmware.o misc.o piggy.o targets := vmlinux.lds vmlinux vmlinux.bin vmlinux.bin.gz vmlinux.bin.bz2 targets += vmlinux.bin.xz vmlinux.bin.lzma vmlinux.bin.lzo vmlinux.bin.lz4 @@ -22,6 +18,7 @@ KBUILD_CFLAGS += -fno-PIE -mno-space-regs -mdisable-fpregs -Os ifndef CONFIG_64BIT KBUILD_CFLAGS += -mfast-indirect-calls endif +KBUILD_CFLAGS += -std=gnu11 -fms-extensions LDFLAGS_vmlinux := -X -e startup --as-needed -T $(obj)/vmlinux: $(obj)/vmlinux.lds $(addprefix $(obj)/, $(OBJECTS)) $(LIBGCC) FORCE diff --git a/arch/parisc/boot/compressed/misc.c b/arch/parisc/boot/compressed/misc.c index 7ee49f5881d1..9c83bd06ef15 100644 --- a/arch/parisc/boot/compressed/misc.c +++ b/arch/parisc/boot/compressed/misc.c @@ -6,7 +6,7 @@ #include <linux/uaccess.h> #include <linux/elf.h> -#include <asm/unaligned.h> +#include <linux/unaligned.h> #include <asm/page.h> #include "sizes.h" @@ -117,7 +117,7 @@ char *strchr(const char *s, int c) return NULL; } -int puts(const char *s) +static int puts(const char *s) { const char *nuline = s; @@ -172,7 +172,7 @@ static int print_num(unsigned long num, int base) return 0; } -int printf(const char *fmt, ...) +static int printf(const char *fmt, ...) { va_list args; int i = 0; @@ -204,13 +204,13 @@ void abort(void) } #undef malloc -void *malloc(size_t size) +static void *malloc(size_t size) { return malloc_gzip(size); } #undef free -void free(void *ptr) +static void free(void *ptr) { return free_gzip(ptr); } @@ -278,7 +278,7 @@ static void parse_elf(void *output) free(phdrs); } -unsigned long decompress_kernel(unsigned int started_wide, +asmlinkage unsigned long __visible decompress_kernel(unsigned int started_wide, unsigned int command_line, const unsigned int rd_start, const unsigned int rd_end) diff --git a/arch/parisc/configs/generic-32bit_defconfig b/arch/parisc/configs/generic-32bit_defconfig index 8ce0ae370680..5444ce6405f3 100644 --- a/arch/parisc/configs/generic-32bit_defconfig +++ b/arch/parisc/configs/generic-32bit_defconfig @@ -12,7 +12,6 @@ CONFIG_USER_NS=y CONFIG_BLK_DEV_INITRD=y CONFIG_EXPERT=y CONFIG_PERF_EVENTS=y -CONFIG_SLAB=y CONFIG_PA7100LC=y CONFIG_SMP=y CONFIG_HZ_100=y @@ -83,7 +82,6 @@ CONFIG_TUN=m # CONFIG_NET_VENDOR_AMD is not set # CONFIG_NET_VENDOR_ATHEROS is not set # CONFIG_NET_VENDOR_BROADCOM is not set -# CONFIG_NET_VENDOR_BROCADE is not set # CONFIG_NET_VENDOR_CHELSIO is not set # CONFIG_NET_VENDOR_CISCO is not set CONFIG_NET_TULIP=y @@ -98,6 +96,7 @@ CONFIG_LASI_82596=y # CONFIG_NET_VENDOR_NVIDIA is not set # CONFIG_NET_VENDOR_OKI is not set # CONFIG_NET_VENDOR_QLOGIC is not set +# CONFIG_NET_VENDOR_BROCADE is not set # CONFIG_NET_VENDOR_RDC is not set # CONFIG_NET_VENDOR_REALTEK is not set # CONFIG_NET_VENDOR_SEEQ is not set @@ -120,7 +119,6 @@ CONFIG_INPUT_MISC=y CONFIG_INPUT_UINPUT=m CONFIG_LEGACY_PTY_COUNT=64 CONFIG_SERIAL_8250=y -# CONFIG_SERIAL_8250_DEPRECATED_OPTIONS is not set CONFIG_SERIAL_8250_CONSOLE=y CONFIG_SERIAL_8250_NR_UARTS=8 CONFIG_SERIAL_8250_EXTENDED=y @@ -130,17 +128,53 @@ CONFIG_PRINTER=m CONFIG_PPDEV=m # CONFIG_HW_RANDOM is not set CONFIG_I2C=y -# CONFIG_HWMON is not set +CONFIG_HWMON=m +CONFIG_DRM=m +CONFIG_DRM_DISPLAY_DP_AUX_CEC=y +CONFIG_DRM_RADEON=m +CONFIG_DRM_NOUVEAU=m +# CONFIG_DRM_NOUVEAU_BACKLIGHT is not set +# CONFIG_DRM_NOUVEAU_CH7006 is not set +# CONFIG_DRM_NOUVEAU_SIL164 is not set +CONFIG_DRM_VGEM=m +CONFIG_DRM_UDL=m +CONFIG_DRM_MGAG200=m CONFIG_FB=y CONFIG_FB_FOREIGN_ENDIAN=y -CONFIG_FB_MODE_HELPERS=y +CONFIG_FB_PM2=m +CONFIG_FB_PM2_FIFO_DISCONNECT=y +CONFIG_FB_NVIDIA=m +CONFIG_FB_NVIDIA_I2C=y +# CONFIG_FB_NVIDIA_BACKLIGHT is not set +CONFIG_FB_RIVA=m +CONFIG_FB_RIVA_I2C=y +# CONFIG_FB_RIVA_BACKLIGHT is not set CONFIG_FB_MATROX=m +CONFIG_FB_MATROX_MILLENIUM=y +CONFIG_FB_MATROX_MYSTIQUE=y CONFIG_FB_MATROX_G=y +CONFIG_FB_MATROX_I2C=m +CONFIG_FB_MATROX_MAVEN=m +CONFIG_FB_ATY128=m +# CONFIG_FB_ATY128_BACKLIGHT is not set +CONFIG_FB_ATY=m +CONFIG_FB_ATY_CT=y +CONFIG_FB_ATY_GX=y +# CONFIG_FB_ATY_BACKLIGHT is not set +CONFIG_FB_S3=m +CONFIG_FB_SAVAGE=m +CONFIG_FB_SAVAGE_I2C=y +CONFIG_FB_SAVAGE_ACCEL=y +CONFIG_FB_SIS=m +CONFIG_FB_SIS_300=y +CONFIG_FB_SIS_315=y CONFIG_FB_VOODOO1=m +CONFIG_FB_TRIDENT=m +CONFIG_FB_SMSCUFX=m +CONFIG_FB_UDL=m CONFIG_DUMMY_CONSOLE_COLUMNS=128 CONFIG_DUMMY_CONSOLE_ROWS=48 CONFIG_FRAMEBUFFER_CONSOLE=y -CONFIG_FRAMEBUFFER_CONSOLE_DETECT_PRIMARY=y CONFIG_LOGO=y # CONFIG_LOGO_LINUX_MONO is not set # CONFIG_LOGO_LINUX_VGA16 is not set @@ -197,12 +231,12 @@ CONFIG_AUXDISPLAY=y CONFIG_EXT2_FS=y CONFIG_EXT2_FS_XATTR=y CONFIG_EXT2_FS_SECURITY=y -CONFIG_EXT3_FS=y -CONFIG_EXT3_FS_SECURITY=y +CONFIG_EXT4_FS=y +CONFIG_EXT4_FS_SECURITY=y CONFIG_QUOTA=y CONFIG_QUOTA_NETLINK_INTERFACE=y CONFIG_QFMT_V2=y -CONFIG_AUTOFS4_FS=y +CONFIG_AUTOFS_FS=y CONFIG_ISO9660_FS=y CONFIG_JOLIET=y CONFIG_VFAT_FS=y @@ -216,21 +250,19 @@ CONFIG_CIFS=m CONFIG_CIFS_XATTR=y CONFIG_CIFS_POSIX=y # CONFIG_CIFS_DEBUG is not set -CONFIG_CRYPTO_TEST=m -CONFIG_CRYPTO_HMAC=y -CONFIG_CRYPTO_MD5=y -CONFIG_CRYPTO_MICHAEL_MIC=m -CONFIG_CRYPTO_SHA1=y -CONFIG_CRYPTO_WP512=m +CONFIG_CRYPTO_BENCHMARK=m CONFIG_CRYPTO_BLOWFISH=m CONFIG_CRYPTO_CAST5=m CONFIG_CRYPTO_CAST6=m CONFIG_CRYPTO_DES=y CONFIG_CRYPTO_SERPENT=m CONFIG_CRYPTO_TWOFISH=m +CONFIG_CRYPTO_HMAC=y +CONFIG_CRYPTO_MD5=y +CONFIG_CRYPTO_MICHAEL_MIC=m +CONFIG_CRYPTO_SHA1=y +CONFIG_CRYPTO_WP512=m CONFIG_CRYPTO_DEFLATE=y -CONFIG_CRC_CCITT=m -CONFIG_CRC_T10DIF=y CONFIG_FONTS=y CONFIG_PRINTK_TIME=y CONFIG_MAGIC_SYSRQ=y diff --git a/arch/parisc/configs/generic-64bit_defconfig b/arch/parisc/configs/generic-64bit_defconfig index 57501b0aed92..ce91f9d1fdbf 100644 --- a/arch/parisc/configs/generic-64bit_defconfig +++ b/arch/parisc/configs/generic-64bit_defconfig @@ -20,9 +20,6 @@ CONFIG_USER_NS=y CONFIG_RELAY=y CONFIG_BLK_DEV_INITRD=y CONFIG_CC_OPTIMIZE_FOR_SIZE=y -# CONFIG_COMPAT_BRK is not set -CONFIG_PA8X00=y -CONFIG_64BIT=y CONFIG_SMP=y CONFIG_HPPB=y CONFIG_IOMMU_CCIO=y @@ -37,6 +34,7 @@ CONFIG_MODULE_FORCE_UNLOAD=y CONFIG_MODVERSIONS=y CONFIG_BLK_DEV_INTEGRITY=y CONFIG_BINFMT_MISC=m +# CONFIG_COMPAT_BRK is not set # CONFIG_COMPACTION is not set CONFIG_MEMORY_FAILURE=y CONFIG_NET=y @@ -103,7 +101,6 @@ CONFIG_TUN=y # CONFIG_NET_VENDOR_AMD is not set # CONFIG_NET_VENDOR_ATHEROS is not set # CONFIG_NET_VENDOR_BROADCOM is not set -# CONFIG_NET_VENDOR_BROCADE is not set # CONFIG_NET_VENDOR_CHELSIO is not set # CONFIG_NET_VENDOR_CISCO is not set CONFIG_NET_TULIP=y @@ -121,6 +118,7 @@ CONFIG_E1000=y # CONFIG_NET_VENDOR_OKI is not set CONFIG_QLA3XXX=m CONFIG_QLCNIC=m +# CONFIG_NET_VENDOR_BROCADE is not set # CONFIG_NET_VENDOR_RDC is not set # CONFIG_NET_VENDOR_REALTEK is not set # CONFIG_NET_VENDOR_SEEQ is not set @@ -160,7 +158,6 @@ CONFIG_SERIO_SERPORT=m CONFIG_SERIO_RAW=m # CONFIG_LEGACY_PTYS is not set CONFIG_SERIAL_8250=y -# CONFIG_SERIAL_8250_DEPRECATED_OPTIONS is not set CONFIG_SERIAL_8250_CONSOLE=y CONFIG_SERIAL_8250_NR_UARTS=8 CONFIG_SERIAL_8250_RUNTIME_UARTS=8 @@ -186,7 +183,6 @@ CONFIG_WATCHDOG=y CONFIG_SOFT_WATCHDOG=m CONFIG_SSB=m CONFIG_SSB_DRIVER_PCICORE=y -CONFIG_HTC_PASIC3=m CONFIG_LPC_SCH=m CONFIG_MFD_SM501=m CONFIG_REGULATOR=y @@ -197,13 +193,45 @@ CONFIG_AGP=y CONFIG_AGP_PARISC=y CONFIG_DRM=y CONFIG_DRM_RADEON=y +CONFIG_DRM_NOUVEAU=m +# CONFIG_DRM_NOUVEAU_BACKLIGHT is not set +# CONFIG_DRM_NOUVEAU_CH7006 is not set +# CONFIG_DRM_NOUVEAU_SIL164 is not set +CONFIG_DRM_MGAG200=m CONFIG_FB=y -CONFIG_FB_MATROX=y +CONFIG_FB_PM2=m +CONFIG_FB_PM2_FIFO_DISCONNECT=y +CONFIG_FB_NVIDIA=m +CONFIG_FB_NVIDIA_I2C=y +# CONFIG_FB_NVIDIA_BACKLIGHT is not set +CONFIG_FB_RIVA=m +CONFIG_FB_RIVA_I2C=y +# CONFIG_FB_RIVA_BACKLIGHT is not set +CONFIG_FB_MATROX=m +CONFIG_FB_MATROX_MILLENIUM=y CONFIG_FB_MATROX_MYSTIQUE=y CONFIG_FB_MATROX_G=y -CONFIG_FB_MATROX_I2C=y -CONFIG_FB_MATROX_MAVEN=y +CONFIG_FB_MATROX_I2C=m +CONFIG_FB_MATROX_MAVEN=m CONFIG_FB_RADEON=y +# CONFIG_FB_RADEON_BACKLIGHT is not set +CONFIG_FB_ATY128=m +# CONFIG_FB_ATY128_BACKLIGHT is not set +CONFIG_FB_ATY=m +CONFIG_FB_ATY_CT=y +CONFIG_FB_ATY_GX=y +# CONFIG_FB_ATY_BACKLIGHT is not set +CONFIG_FB_S3=m +CONFIG_FB_SAVAGE=m +CONFIG_FB_SAVAGE_I2C=y +CONFIG_FB_SAVAGE_ACCEL=y +CONFIG_FB_SIS=m +CONFIG_FB_SIS_300=y +CONFIG_FB_SIS_315=y +CONFIG_FB_VOODOO1=m +CONFIG_FB_TRIDENT=m +CONFIG_FB_SMSCUFX=m +CONFIG_FB_UDL=m CONFIG_LOGO=y # CONFIG_LOGO_LINUX_CLUT224 is not set CONFIG_HIDRAW=y @@ -219,18 +247,17 @@ CONFIG_UIO_AEC=m CONFIG_UIO_SERCOS3=m CONFIG_UIO_PCI_GENERIC=m CONFIG_STAGING=y -CONFIG_QLGE=m CONFIG_EXT2_FS=y CONFIG_EXT2_FS_XATTR=y CONFIG_EXT2_FS_SECURITY=y -CONFIG_EXT3_FS=y -CONFIG_EXT3_FS_SECURITY=y +CONFIG_EXT4_FS=y +CONFIG_EXT4_FS_SECURITY=y CONFIG_XFS_FS=m CONFIG_BTRFS_FS=m CONFIG_QUOTA=y CONFIG_QUOTA_NETLINK_INTERFACE=y CONFIG_QFMT_V2=y -CONFIG_AUTOFS4_FS=y +CONFIG_AUTOFS_FS=y CONFIG_FUSE_FS=y CONFIG_CUSE=y CONFIG_ISO9660_FS=y @@ -240,7 +267,6 @@ CONFIG_PROC_KCORE=y CONFIG_TMPFS=y CONFIG_TMPFS_XATTR=y CONFIG_CONFIGFS_FS=y -CONFIG_SYSV_FS=y CONFIG_NFS_FS=m CONFIG_NFS_V4=m CONFIG_NFS_V4_1=y @@ -256,17 +282,14 @@ CONFIG_NLS_ASCII=m CONFIG_NLS_ISO8859_1=m CONFIG_NLS_ISO8859_2=m CONFIG_NLS_UTF8=m -CONFIG_CRYPTO_MANAGER=y +CONFIG_CRYPTO_FCRYPT=m CONFIG_CRYPTO_ECB=m CONFIG_CRYPTO_PCBC=m CONFIG_CRYPTO_MD4=m CONFIG_CRYPTO_MD5=y CONFIG_CRYPTO_MICHAEL_MIC=m -CONFIG_CRYPTO_FCRYPT=m CONFIG_CRYPTO_DEFLATE=m # CONFIG_CRYPTO_HW is not set -CONFIG_CRC_CCITT=m -CONFIG_LIBCRC32C=y CONFIG_PRINTK_TIME=y CONFIG_DEBUG_KERNEL=y CONFIG_STRIP_ASM_SYMS=y diff --git a/arch/parisc/include/asm/Kbuild b/arch/parisc/include/asm/Kbuild index e6e7f74c8ac9..4fb596d94c89 100644 --- a/arch/parisc/include/asm/Kbuild +++ b/arch/parisc/include/asm/Kbuild @@ -1,6 +1,7 @@ # SPDX-License-Identifier: GPL-2.0 generated-y += syscall_table_32.h generated-y += syscall_table_64.h +generic-y += agp.h generic-y += kvm_para.h generic-y += mcs_spinlock.h generic-y += user.h diff --git a/arch/parisc/include/asm/agp.h b/arch/parisc/include/asm/agp.h deleted file mode 100644 index 14ae54cfd368..000000000000 --- a/arch/parisc/include/asm/agp.h +++ /dev/null @@ -1,21 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef _ASM_PARISC_AGP_H -#define _ASM_PARISC_AGP_H - -/* - * PARISC specific AGP definitions. - * Copyright (c) 2006 Kyle McMartin <kyle@parisc-linux.org> - * - */ - -#define map_page_into_agp(page) do { } while (0) -#define unmap_page_from_agp(page) do { } while (0) -#define flush_agp_cache() mb() - -/* GATT allocation. Returns/accepts GATT kernel virtual address. */ -#define alloc_gatt_pages(order) \ - ((char *)__get_free_pages(GFP_KERNEL, (order))) -#define free_gatt_pages(table, order) \ - free_pages((unsigned long)(table), (order)) - -#endif /* _ASM_PARISC_AGP_H */ diff --git a/arch/parisc/include/asm/alternative.h b/arch/parisc/include/asm/alternative.h index 1ed45fd085d3..1601ae4b888d 100644 --- a/arch/parisc/include/asm/alternative.h +++ b/arch/parisc/include/asm/alternative.h @@ -13,7 +13,7 @@ #define INSN_PxTLB 0x02 /* modify pdtlb, pitlb */ #define INSN_NOP 0x08000240 /* nop */ -#ifndef __ASSEMBLY__ +#ifndef __ASSEMBLER__ #include <linux/init.h> #include <linux/types.h> @@ -34,7 +34,8 @@ void apply_alternatives(struct alt_instr *start, struct alt_instr *end, /* Alternative SMP implementation. */ #define ALTERNATIVE(cond, replacement) "!0:" \ - ".section .altinstructions, \"aw\" !" \ + ".section .altinstructions, \"a\" !" \ + ".align 4 !" \ ".word (0b-4-.) !" \ ".hword 1, " __stringify(cond) " !" \ ".word " __stringify(replacement) " !" \ @@ -44,7 +45,8 @@ void apply_alternatives(struct alt_instr *start, struct alt_instr *end, /* to replace one single instructions by a new instruction */ #define ALTERNATIVE(from, to, cond, replacement)\ - .section .altinstructions, "aw" ! \ + .section .altinstructions, "a" ! \ + .align 4 ! \ .word (from - .) ! \ .hword (to - from)/4, cond ! \ .word replacement ! \ @@ -52,12 +54,13 @@ void apply_alternatives(struct alt_instr *start, struct alt_instr *end, /* to replace multiple instructions by new code */ #define ALTERNATIVE_CODE(from, num_instructions, cond, new_instr_ptr)\ - .section .altinstructions, "aw" ! \ + .section .altinstructions, "a" ! \ + .align 4 ! \ .word (from - .) ! \ .hword -num_instructions, cond ! \ .word (new_instr_ptr - .) ! \ .previous -#endif /* __ASSEMBLY__ */ +#endif /* __ASSEMBLER__ */ #endif /* __ASM_PARISC_ALTERNATIVE_H */ diff --git a/arch/parisc/include/asm/assembly.h b/arch/parisc/include/asm/assembly.h index 0f0d4a496fef..c20261604f09 100644 --- a/arch/parisc/include/asm/assembly.h +++ b/arch/parisc/include/asm/assembly.h @@ -53,7 +53,7 @@ #define SR_TEMP2 2 #define SR_USER 3 -#ifdef __ASSEMBLY__ +#ifdef __ASSEMBLER__ #ifdef CONFIG_64BIT #define LDREG ldd @@ -90,10 +90,6 @@ #include <asm/asmregs.h> #include <asm/psw.h> - sp = 30 - gp = 27 - ipsw = 22 - /* * We provide two versions of each macro to convert from physical * to virtual and vice versa. The "_r1" versions take one argument @@ -101,26 +97,28 @@ * version takes two arguments: a src and destination register. * However, the source and destination registers can not be * the same register. + * + * We use add,l to avoid clobbering the C/B bits in the PSW. */ .macro tophys grvirt, grphys - ldil L%(__PAGE_OFFSET), \grphys - sub \grvirt, \grphys, \grphys + ldil L%(-__PAGE_OFFSET), \grphys + addl \grvirt, \grphys, \grphys .endm - + .macro tovirt grphys, grvirt ldil L%(__PAGE_OFFSET), \grvirt - add \grphys, \grvirt, \grvirt + addl \grphys, \grvirt, \grvirt .endm .macro tophys_r1 gr - ldil L%(__PAGE_OFFSET), %r1 - sub \gr, %r1, \gr + ldil L%(-__PAGE_OFFSET), %r1 + addl \gr, %r1, \gr .endm - + .macro tovirt_r1 gr ldil L%(__PAGE_OFFSET), %r1 - add \gr, %r1, \gr + addl \gr, %r1, \gr .endm .macro delay value @@ -578,9 +576,11 @@ */ #define ASM_EXCEPTIONTABLE_ENTRY(fault_addr, except_addr) \ .section __ex_table,"aw" ! \ + .align 4 ! \ .word (fault_addr - .), (except_addr - .) ! \ + or %r0,%r0,%r0 ! \ .previous -#endif /* __ASSEMBLY__ */ +#endif /* __ASSEMBLER__ */ #endif diff --git a/arch/parisc/include/asm/atomic.h b/arch/parisc/include/asm/atomic.h index dd5a299ada69..d4f023887ff8 100644 --- a/arch/parisc/include/asm/atomic.h +++ b/arch/parisc/include/asm/atomic.h @@ -73,10 +73,6 @@ static __inline__ int arch_atomic_read(const atomic_t *v) return READ_ONCE((v)->counter); } -/* exported interface */ -#define arch_atomic_cmpxchg(v, o, n) (arch_cmpxchg(&((v)->counter), (o), (n))) -#define arch_atomic_xchg(v, new) (arch_xchg(&((v)->counter), new)) - #define ATOMIC_OP(op, c_op) \ static __inline__ void arch_atomic_##op(int i, atomic_t *v) \ { \ @@ -122,6 +118,11 @@ static __inline__ int arch_atomic_fetch_##op(int i, atomic_t *v) \ ATOMIC_OPS(add, +=) ATOMIC_OPS(sub, -=) +#define arch_atomic_add_return arch_atomic_add_return +#define arch_atomic_sub_return arch_atomic_sub_return +#define arch_atomic_fetch_add arch_atomic_fetch_add +#define arch_atomic_fetch_sub arch_atomic_fetch_sub + #undef ATOMIC_OPS #define ATOMIC_OPS(op, c_op) \ ATOMIC_OP(op, c_op) \ @@ -131,6 +132,10 @@ ATOMIC_OPS(and, &=) ATOMIC_OPS(or, |=) ATOMIC_OPS(xor, ^=) +#define arch_atomic_fetch_and arch_atomic_fetch_and +#define arch_atomic_fetch_or arch_atomic_fetch_or +#define arch_atomic_fetch_xor arch_atomic_fetch_xor + #undef ATOMIC_OPS #undef ATOMIC_FETCH_OP #undef ATOMIC_OP_RETURN @@ -185,6 +190,11 @@ static __inline__ s64 arch_atomic64_fetch_##op(s64 i, atomic64_t *v) \ ATOMIC64_OPS(add, +=) ATOMIC64_OPS(sub, -=) +#define arch_atomic64_add_return arch_atomic64_add_return +#define arch_atomic64_sub_return arch_atomic64_sub_return +#define arch_atomic64_fetch_add arch_atomic64_fetch_add +#define arch_atomic64_fetch_sub arch_atomic64_fetch_sub + #undef ATOMIC64_OPS #define ATOMIC64_OPS(op, c_op) \ ATOMIC64_OP(op, c_op) \ @@ -194,6 +204,10 @@ ATOMIC64_OPS(and, &=) ATOMIC64_OPS(or, |=) ATOMIC64_OPS(xor, ^=) +#define arch_atomic64_fetch_and arch_atomic64_fetch_and +#define arch_atomic64_fetch_or arch_atomic64_fetch_or +#define arch_atomic64_fetch_xor arch_atomic64_fetch_xor + #undef ATOMIC64_OPS #undef ATOMIC64_FETCH_OP #undef ATOMIC64_OP_RETURN @@ -218,11 +232,6 @@ arch_atomic64_read(const atomic64_t *v) return READ_ONCE((v)->counter); } -/* exported interface */ -#define arch_atomic64_cmpxchg(v, o, n) \ - ((__typeof__((v)->counter))arch_cmpxchg(&((v)->counter), (o), (n))) -#define arch_atomic64_xchg(v, new) (arch_xchg(&((v)->counter), new)) - #endif /* !CONFIG_64BIT */ diff --git a/arch/parisc/include/asm/barrier.h b/arch/parisc/include/asm/barrier.h index c705decf2bed..519b1903c5ed 100644 --- a/arch/parisc/include/asm/barrier.h +++ b/arch/parisc/include/asm/barrier.h @@ -4,7 +4,7 @@ #include <asm/alternative.h> -#ifndef __ASSEMBLY__ +#ifndef __ASSEMBLER__ /* The synchronize caches instruction executes as a nop on systems in which all memory references are performed in order. */ @@ -93,5 +93,5 @@ do { \ }) #include <asm-generic/barrier.h> -#endif /* !__ASSEMBLY__ */ +#endif /* !__ASSEMBLER__ */ #endif /* __ASM_BARRIER_H */ diff --git a/arch/parisc/include/asm/bitops.h b/arch/parisc/include/asm/bitops.h index 0ec9cfc5131f..bd1280a8a5ec 100644 --- a/arch/parisc/include/asm/bitops.h +++ b/arch/parisc/include/asm/bitops.h @@ -123,7 +123,7 @@ static __inline__ int test_and_change_bit(int nr, volatile unsigned long * addr) * cycles for each mispredicted branch. */ -static __inline__ unsigned long __ffs(unsigned long x) +static __inline__ __attribute_const__ unsigned long __ffs(unsigned long x) { unsigned long ret; @@ -161,7 +161,7 @@ static __inline__ unsigned long __ffs(unsigned long x) * This is defined the same way as the libc and compiler builtin * ffs routines, therefore differs in spirit from the above ffz (man ffs). */ -static __inline__ int ffs(int x) +static __inline__ __attribute_const__ int ffs(int x) { return x ? (__ffs((unsigned long)x) + 1) : 0; } @@ -171,7 +171,7 @@ static __inline__ int ffs(int x) * fls(0) = 0, fls(1) = 1, fls(0x80000000) = 32. */ -static __inline__ int fls(unsigned int x) +static __inline__ __attribute_const__ int fls(unsigned int x) { int ret; if (!x) diff --git a/arch/parisc/include/asm/bug.h b/arch/parisc/include/asm/bug.h index 4b6d60b94124..5cf35489ad80 100644 --- a/arch/parisc/include/asm/bug.h +++ b/arch/parisc/include/asm/bug.h @@ -2,8 +2,6 @@ #ifndef _PARISC_BUG_H #define _PARISC_BUG_H -#include <linux/kernel.h> /* for BUGFLAG_TAINT */ - /* * Tell the user there is some problem. * The offending file and line are encoded in the __bug_table section. @@ -17,24 +15,27 @@ #define PARISC_BUG_BREAK_ASM "break 0x1f, 0x1fff" #define PARISC_BUG_BREAK_INSN 0x03ffe01f /* PARISC_BUG_BREAK_ASM */ -#if defined(CONFIG_64BIT) -#define ASM_WORD_INSN ".dword\t" +#ifdef CONFIG_GENERIC_BUG_RELATIVE_POINTERS +# define __BUG_REL(val) ".word " __stringify(val) " - ." #else -#define ASM_WORD_INSN ".word\t" +# define __BUG_REL(val) ".word " __stringify(val) #endif + #ifdef CONFIG_DEBUG_BUGVERBOSE #define BUG() \ do { \ asm volatile("\n" \ "1:\t" PARISC_BUG_BREAK_ASM "\n" \ - "\t.pushsection __bug_table,\"aw\"\n" \ - "2:\t" ASM_WORD_INSN "1b, %c0\n" \ - "\t.short %c1, %c2\n" \ - "\t.org 2b+%c3\n" \ + "\t.pushsection __bug_table,\"a\"\n" \ + "\t.align 4\n" \ + "2:\t" __BUG_REL(1b) "\n" \ + "\t" __BUG_REL(%c0) "\n" \ + "\t.short %1, %2\n" \ + "\t.blockz %3-2*4-2*2\n" \ "\t.popsection" \ : : "i" (__FILE__), "i" (__LINE__), \ - "i" (0), "i" (sizeof(struct bug_entry)) ); \ + "i" (0), "i" (sizeof(struct bug_entry)) ); \ unreachable(); \ } while(0) @@ -47,28 +48,31 @@ #endif #ifdef CONFIG_DEBUG_BUGVERBOSE -#define __WARN_FLAGS(flags) \ +#define __WARN_FLAGS(cond_str, flags) \ do { \ asm volatile("\n" \ "1:\t" PARISC_BUG_BREAK_ASM "\n" \ - "\t.pushsection __bug_table,\"aw\"\n" \ - "2:\t" ASM_WORD_INSN "1b, %c0\n" \ - "\t.short %c1, %c2\n" \ - "\t.org 2b+%c3\n" \ + "\t.pushsection __bug_table,\"a\"\n" \ + "\t.align 4\n" \ + "2:\t" __BUG_REL(1b) "\n" \ + "\t" __BUG_REL(%c0) "\n" \ + "\t.short %1, %2\n" \ + "\t.blockz %3-2*4-2*2\n" \ "\t.popsection" \ - : : "i" (__FILE__), "i" (__LINE__), \ + : : "i" (WARN_CONDITION_STR(cond_str) __FILE__), "i" (__LINE__), \ "i" (BUGFLAG_WARNING|(flags)), \ "i" (sizeof(struct bug_entry)) ); \ } while(0) #else -#define __WARN_FLAGS(flags) \ +#define __WARN_FLAGS(cond_str, flags) \ do { \ asm volatile("\n" \ "1:\t" PARISC_BUG_BREAK_ASM "\n" \ - "\t.pushsection __bug_table,\"aw\"\n" \ - "2:\t" ASM_WORD_INSN "1b\n" \ - "\t.short %c0\n" \ - "\t.org 2b+%c1\n" \ + "\t.pushsection __bug_table,\"a\"\n" \ + "\t.align 4\n" \ + "2:\t" __BUG_REL(1b) "\n" \ + "\t.short %0\n" \ + "\t.blockz %1-4-2\n" \ "\t.popsection" \ : : "i" (BUGFLAG_WARNING|(flags)), \ "i" (sizeof(struct bug_entry)) ); \ diff --git a/arch/parisc/include/asm/bugs.h b/arch/parisc/include/asm/bugs.h deleted file mode 100644 index 0a7f9db6bd1c..000000000000 --- a/arch/parisc/include/asm/bugs.h +++ /dev/null @@ -1,20 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -/* - * include/asm-parisc/bugs.h - * - * Copyright (C) 1999 Mike Shaver - */ - -/* - * This is included by init/main.c to check for architecture-dependent bugs. - * - * Needs: - * void check_bugs(void); - */ - -#include <asm/processor.h> - -static inline void check_bugs(void) -{ -// identify_cpu(&boot_cpu_data); -} diff --git a/arch/parisc/include/asm/cache.h b/arch/parisc/include/asm/cache.h index e23d06b51a20..3f8d3be6ef24 100644 --- a/arch/parisc/include/asm/cache.h +++ b/arch/parisc/include/asm/cache.h @@ -16,11 +16,20 @@ #define L1_CACHE_BYTES 16 #define L1_CACHE_SHIFT 4 -#ifndef __ASSEMBLY__ +#ifndef __ASSEMBLER__ #define SMP_CACHE_BYTES L1_CACHE_BYTES -#define ARCH_DMA_MINALIGN L1_CACHE_BYTES +#ifdef CONFIG_PA20 +#define ARCH_DMA_MINALIGN 128 +#else +#define ARCH_DMA_MINALIGN 32 +#endif +#define ARCH_KMALLOC_MINALIGN 16 /* ldcw requires 16-byte alignment */ + +#define arch_slab_minalign() ((unsigned)dcache_stride) +#define cache_line_size() dcache_stride +#define dma_get_cache_alignment cache_line_size #define __read_mostly __section(".data..read_mostly") @@ -37,6 +46,7 @@ extern int split_tlb; extern int dcache_stride; extern int icache_stride; extern struct pdc_cache_info cache_info; +extern struct pdc_btlb_info btlb_info; void parisc_setup_cache_timing(void); #define pdtlb(sr, addr) asm volatile("pdtlb 0(%%sr%0,%1)" \ @@ -56,7 +66,7 @@ void parisc_setup_cache_timing(void); ALTERNATIVE(ALT_COND_NO_IOC_FDC, INSN_NOP) :::"memory") #define asm_syncdma() asm volatile("syncdma" :::"memory") -#endif /* ! __ASSEMBLY__ */ +#endif /* ! __ASSEMBLER__ */ /* Classes of processor wrt: disabling space register hashing */ diff --git a/arch/parisc/include/asm/cacheflush.h b/arch/parisc/include/asm/cacheflush.h index 0bdee6724132..8394718870e1 100644 --- a/arch/parisc/include/asm/cacheflush.h +++ b/arch/parisc/include/asm/cacheflush.h @@ -31,28 +31,36 @@ void flush_cache_all_local(void); void flush_cache_all(void); void flush_cache_mm(struct mm_struct *mm); -void flush_kernel_dcache_page_addr(const void *addr); - #define flush_kernel_dcache_range(start,size) \ flush_kernel_dcache_range_asm((start), (start)+(size)); +/* The only way to flush a vmap range is to flush whole cache */ #define ARCH_IMPLEMENTS_FLUSH_KERNEL_VMAP_RANGE 1 void flush_kernel_vmap_range(void *vaddr, int size); void invalidate_kernel_vmap_range(void *vaddr, int size); -#define flush_cache_vmap(start, end) flush_cache_all() -#define flush_cache_vunmap(start, end) flush_cache_all() +void flush_cache_vmap(unsigned long start, unsigned long end); +#define flush_cache_vmap_early(start, end) do { } while (0) +void flush_cache_vunmap(unsigned long start, unsigned long end); +void flush_dcache_folio(struct folio *folio); +#define flush_dcache_folio flush_dcache_folio #define ARCH_IMPLEMENTS_FLUSH_DCACHE_PAGE 1 -void flush_dcache_page(struct page *page); +static inline void flush_dcache_page(struct page *page) +{ + flush_dcache_folio(page_folio(page)); +} #define flush_dcache_mmap_lock(mapping) xa_lock_irq(&mapping->i_pages) #define flush_dcache_mmap_unlock(mapping) xa_unlock_irq(&mapping->i_pages) +#define flush_dcache_mmap_lock_irqsave(mapping, flags) \ + xa_lock_irqsave(&mapping->i_pages, flags) +#define flush_dcache_mmap_unlock_irqrestore(mapping, flags) \ + xa_unlock_irqrestore(&mapping->i_pages, flags) -#define flush_icache_page(vma,page) do { \ - flush_kernel_dcache_page_addr(page_address(page)); \ - flush_kernel_icache_page(page_address(page)); \ -} while (0) +void flush_icache_pages(struct vm_area_struct *vma, struct page *page, + unsigned int nr); +#define flush_icache_pages flush_icache_pages #define flush_icache_range(s,e) do { \ flush_kernel_dcache_range_asm(s,e); \ @@ -68,17 +76,11 @@ void flush_cache_page(struct vm_area_struct *vma, unsigned long vmaddr, void flush_cache_range(struct vm_area_struct *vma, unsigned long start, unsigned long end); -/* defined in pacache.S exported in cache.c used by flush_anon_page */ -void flush_dcache_page_asm(unsigned long phys_addr, unsigned long vaddr); - #define ARCH_HAS_FLUSH_ANON_PAGE void flush_anon_page(struct vm_area_struct *vma, struct page *page, unsigned long vmaddr); #define ARCH_HAS_FLUSH_ON_KUNMAP -static inline void kunmap_flush_on_unmap(const void *addr) -{ - flush_kernel_dcache_page_addr(addr); -} +void kunmap_flush_on_unmap(const void *addr); #endif /* _PARISC_CACHEFLUSH_H */ diff --git a/arch/parisc/include/asm/cachetype.h b/arch/parisc/include/asm/cachetype.h new file mode 100644 index 000000000000..e0868a1d3c47 --- /dev/null +++ b/arch/parisc/include/asm/cachetype.h @@ -0,0 +1,9 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __ASM_PARISC_CACHETYPE_H +#define __ASM_PARISC_CACHETYPE_H + +#include <linux/types.h> + +#define cpu_dcache_is_aliasing() true + +#endif diff --git a/arch/parisc/include/asm/checksum.h b/arch/parisc/include/asm/checksum.h index 3c43baca7b39..2aceebcd695c 100644 --- a/arch/parisc/include/asm/checksum.h +++ b/arch/parisc/include/asm/checksum.h @@ -40,7 +40,7 @@ static inline __sum16 ip_fast_csum(const void *iph, unsigned int ihl) " addc %0, %5, %0\n" " addc %0, %3, %0\n" "1: ldws,ma 4(%1), %3\n" -" addib,< 0, %2, 1b\n" +" addib,> -1, %2, 1b\n" " addc %0, %3, %0\n" "\n" " extru %0, 31, 16, %4\n" @@ -126,6 +126,7 @@ static __inline__ __sum16 csum_ipv6_magic(const struct in6_addr *saddr, ** Try to keep 4 registers with "live" values ahead of the ALU. */ +" depdi 0, 31, 32, %0\n"/* clear upper half of incoming checksum */ " ldd,ma 8(%1), %4\n" /* get 1st saddr word */ " ldd,ma 8(%2), %5\n" /* get 1st daddr word */ " add %4, %0, %0\n" @@ -137,8 +138,8 @@ static __inline__ __sum16 csum_ipv6_magic(const struct in6_addr *saddr, " add,dc %3, %0, %0\n" /* fold in proto+len | carry bit */ " extrd,u %0, 31, 32, %4\n"/* copy upper half down */ " depdi 0, 31, 32, %0\n"/* clear upper half */ -" add %4, %0, %0\n" /* fold into 32-bits */ -" addc 0, %0, %0\n" /* add carry */ +" add,dc %4, %0, %0\n" /* fold into 32-bits, plus carry */ +" addc 0, %0, %0\n" /* add final carry */ #else @@ -163,7 +164,8 @@ static __inline__ __sum16 csum_ipv6_magic(const struct in6_addr *saddr, " ldw,ma 4(%2), %7\n" /* 4th daddr */ " addc %6, %0, %0\n" " addc %7, %0, %0\n" -" addc %3, %0, %0\n" /* fold in proto+len, catch carry */ +" addc %3, %0, %0\n" /* fold in proto+len */ +" addc 0, %0, %0\n" /* add carry */ #endif : "=r" (sum), "=r" (saddr), "=r" (daddr), "=r" (len), diff --git a/arch/parisc/include/asm/cmpxchg.h b/arch/parisc/include/asm/cmpxchg.h index 5f274be10567..bf0a0f1189eb 100644 --- a/arch/parisc/include/asm/cmpxchg.h +++ b/arch/parisc/include/asm/cmpxchg.h @@ -22,7 +22,7 @@ extern unsigned long __xchg64(unsigned long, volatile unsigned long *); /* optimizer better get rid of switch since size is a constant */ static inline unsigned long -__xchg(unsigned long x, volatile void *ptr, int size) +__arch_xchg(unsigned long x, volatile void *ptr, int size) { switch (size) { #ifdef CONFIG_64BIT @@ -49,33 +49,31 @@ __xchg(unsigned long x, volatile void *ptr, int size) __typeof__(*(ptr)) __ret; \ __typeof__(*(ptr)) _x_ = (x); \ __ret = (__typeof__(*(ptr))) \ - __xchg((unsigned long)_x_, (ptr), sizeof(*(ptr))); \ + __arch_xchg((unsigned long)_x_, (ptr), sizeof(*(ptr))); \ __ret; \ }) /* bug catcher for when unsupported size is used - won't link */ extern void __cmpxchg_called_with_bad_pointer(void); -/* __cmpxchg_u32/u64 defined in arch/parisc/lib/bitops.c */ -extern unsigned long __cmpxchg_u32(volatile unsigned int *m, unsigned int old, - unsigned int new_); -extern u64 __cmpxchg_u64(volatile u64 *ptr, u64 old, u64 new_); +/* __cmpxchg_u... defined in arch/parisc/lib/bitops.c */ extern u8 __cmpxchg_u8(volatile u8 *ptr, u8 old, u8 new_); +extern u16 __cmpxchg_u16(volatile u16 *ptr, u16 old, u16 new_); +extern u32 __cmpxchg_u32(volatile u32 *m, u32 old, u32 new_); +extern u64 __cmpxchg_u64(volatile u64 *ptr, u64 old, u64 new_); /* don't worry...optimizer will get rid of most of this */ static inline unsigned long __cmpxchg(volatile void *ptr, unsigned long old, unsigned long new_, int size) { - switch (size) { + return #ifdef CONFIG_64BIT - case 8: return __cmpxchg_u64((u64 *)ptr, old, new_); + size == 8 ? __cmpxchg_u64(ptr, old, new_) : #endif - case 4: return __cmpxchg_u32((unsigned int *)ptr, - (unsigned int)old, (unsigned int)new_); - case 1: return __cmpxchg_u8((u8 *)ptr, old & 0xff, new_ & 0xff); - } - __cmpxchg_called_with_bad_pointer(); - return old; + size == 4 ? __cmpxchg_u32(ptr, old, new_) : + size == 2 ? __cmpxchg_u16(ptr, old, new_) : + size == 1 ? __cmpxchg_u8(ptr, old, new_) : + (__cmpxchg_called_with_bad_pointer(), old); } #define arch_cmpxchg(ptr, o, n) \ diff --git a/arch/parisc/include/asm/current.h b/arch/parisc/include/asm/current.h index dc7aea07c3f3..2814529a4c28 100644 --- a/arch/parisc/include/asm/current.h +++ b/arch/parisc/include/asm/current.h @@ -2,7 +2,7 @@ #ifndef _ASM_PARISC_CURRENT_H #define _ASM_PARISC_CURRENT_H -#ifndef __ASSEMBLY__ +#ifndef __ASSEMBLER__ struct task_struct; static __always_inline struct task_struct *get_current(void) @@ -16,6 +16,6 @@ static __always_inline struct task_struct *get_current(void) #define current get_current() -#endif /* __ASSEMBLY__ */ +#endif /* __ASSEMBLER__ */ #endif /* _ASM_PARISC_CURRENT_H */ diff --git a/arch/parisc/include/asm/dma-mapping.h b/arch/parisc/include/asm/dma-mapping.h index d5bd94247371..635665004fe6 100644 --- a/arch/parisc/include/asm/dma-mapping.h +++ b/arch/parisc/include/asm/dma-mapping.h @@ -21,7 +21,7 @@ extern const struct dma_map_ops *hppa_dma_ops; -static inline const struct dma_map_ops *get_arch_dma_ops(struct bus_type *bus) +static inline const struct dma_map_ops *get_arch_dma_ops(void) { return hppa_dma_ops; } diff --git a/arch/parisc/include/asm/dma.h b/arch/parisc/include/asm/dma.h index 9e8c101de902..582fb5d1a5d5 100644 --- a/arch/parisc/include/asm/dma.h +++ b/arch/parisc/include/asm/dma.h @@ -14,6 +14,8 @@ #define dma_outb outb #define dma_inb inb +extern unsigned long pcxl_dma_start; + /* ** DMA_CHUNK_SIZE is used by the SCSI mid-layer to break up ** (or rather not merge) DMAs into manageable chunks. diff --git a/arch/parisc/include/asm/dwarf.h b/arch/parisc/include/asm/dwarf.h index f4512db86a19..526f4a79262c 100644 --- a/arch/parisc/include/asm/dwarf.h +++ b/arch/parisc/include/asm/dwarf.h @@ -6,7 +6,7 @@ #ifndef _ASM_PARISC_DWARF_H #define _ASM_PARISC_DWARF_H -#ifdef __ASSEMBLY__ +#ifdef __ASSEMBLER__ #define CFI_STARTPROC .cfi_startproc #define CFI_ENDPROC .cfi_endproc @@ -15,6 +15,6 @@ #define CFI_REL_OFFSET .cfi_rel_offset #define CFI_UNDEFINED .cfi_undefined -#endif /* __ASSEMBLY__ */ +#endif /* __ASSEMBLER__ */ #endif /* _ASM_PARISC_DWARF_H */ diff --git a/arch/parisc/include/asm/elf.h b/arch/parisc/include/asm/elf.h index cc426d365892..2d73d3c3cd37 100644 --- a/arch/parisc/include/asm/elf.h +++ b/arch/parisc/include/asm/elf.h @@ -163,8 +163,7 @@ typedef struct elf32_fdesc { /* Format for the Elf64 Function descriptor */ typedef struct elf64_fdesc { - __u64 dummy[2]; /* FIXME: nothing uses these, why waste - * the space */ + __u64 dummy[2]; /* used by 64-bit eBPF and tracing functions */ __u64 addr; __u64 gp; } Elf64_Fdesc; @@ -350,15 +349,7 @@ struct pt_regs; /* forward declaration... */ #define ELF_HWCAP 0 -/* Masks for stack and mmap randomization */ -#define BRK_RND_MASK (is_32bit_task() ? 0x07ffUL : 0x3ffffUL) -#define MMAP_RND_MASK (is_32bit_task() ? 0x1fffUL : 0x3ffffUL) -#define STACK_RND_MASK MMAP_RND_MASK - -struct mm_struct; -extern unsigned long arch_randomize_brk(struct mm_struct *); -#define arch_randomize_brk arch_randomize_brk - +#define STACK_RND_MASK 0x7ff /* 8MB of VA */ #define ARCH_HAS_SETUP_ADDITIONAL_PAGES 1 struct linux_binprm; diff --git a/arch/parisc/include/asm/extable.h b/arch/parisc/include/asm/extable.h new file mode 100644 index 000000000000..4ea23e3d79dc --- /dev/null +++ b/arch/parisc/include/asm/extable.h @@ -0,0 +1,64 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __PARISC_EXTABLE_H +#define __PARISC_EXTABLE_H + +#include <asm/ptrace.h> +#include <linux/compiler.h> + +/* + * The exception table consists of three addresses: + * + * - A relative address to the instruction that is allowed to fault. + * - A relative address at which the program should continue (fixup routine) + * - An asm statement which specifies which CPU register will + * receive -EFAULT when an exception happens if the lowest bit in + * the fixup address is set. + * + * Note: The register specified in the err_opcode instruction will be + * modified at runtime if a fault happens. Register %r0 will be ignored. + * + * Since relative addresses are used, 32bit values are sufficient even on + * 64bit kernel. + */ + +struct pt_regs; +int fixup_exception(struct pt_regs *regs); + +#define ARCH_HAS_RELATIVE_EXTABLE +struct exception_table_entry { + int insn; /* relative address of insn that is allowed to fault. */ + int fixup; /* relative address of fixup routine */ + int err_opcode; /* sample opcode with register which holds error code */ +}; + +#define ASM_EXCEPTIONTABLE_ENTRY( fault_addr, except_addr, opcode )\ + ".section __ex_table,\"aw\"\n" \ + ".align 4\n" \ + ".word (" #fault_addr " - .), (" #except_addr " - .)\n" \ + opcode "\n" \ + ".previous\n" + +/* + * ASM_EXCEPTIONTABLE_ENTRY_EFAULT() creates a special exception table entry + * (with lowest bit set) for which the fault handler in fixup_exception() will + * load -EFAULT on fault into the register specified by the err_opcode instruction, + * and zeroes the target register in case of a read fault in get_user(). + */ +#define ASM_EXCEPTIONTABLE_VAR(__err_var) \ + int __err_var = 0 +#define ASM_EXCEPTIONTABLE_ENTRY_EFAULT( fault_addr, except_addr, register )\ + ASM_EXCEPTIONTABLE_ENTRY( fault_addr, except_addr + 1, "or %%r0,%%r0," register) + +static inline void swap_ex_entry_fixup(struct exception_table_entry *a, + struct exception_table_entry *b, + struct exception_table_entry tmp, + int delta) +{ + a->fixup = b->fixup + delta; + b->fixup = tmp.fixup - delta; + a->err_opcode = b->err_opcode; + b->err_opcode = tmp.err_opcode; +} +#define swap_ex_entry_fixup swap_ex_entry_fixup + +#endif diff --git a/arch/parisc/include/asm/fb.h b/arch/parisc/include/asm/fb.h deleted file mode 100644 index 55d29c4f716e..000000000000 --- a/arch/parisc/include/asm/fb.h +++ /dev/null @@ -1,24 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef _ASM_FB_H_ -#define _ASM_FB_H_ - -#include <linux/fb.h> -#include <linux/fs.h> -#include <asm/page.h> - -static inline void fb_pgprotect(struct file *file, struct vm_area_struct *vma, - unsigned long off) -{ - pgprot_val(vma->vm_page_prot) |= _PAGE_NO_CACHE; -} - -#if defined(CONFIG_FB_STI) -int fb_is_primary_device(struct fb_info *info); -#else -static inline int fb_is_primary_device(struct fb_info *info) -{ - return 0; -} -#endif - -#endif /* _ASM_FB_H_ */ diff --git a/arch/parisc/include/asm/fixmap.h b/arch/parisc/include/asm/fixmap.h index 5cd80ce1163a..9cafa449c4a7 100644 --- a/arch/parisc/include/asm/fixmap.h +++ b/arch/parisc/include/asm/fixmap.h @@ -39,7 +39,7 @@ #define KERNEL_MAP_START (GATEWAY_PAGE_SIZE) #define KERNEL_MAP_END (FIXMAP_START) -#ifndef __ASSEMBLY__ +#ifndef __ASSEMBLER__ enum fixed_addresses { @@ -59,6 +59,6 @@ extern void *parisc_vmalloc_start; void set_fixmap(enum fixed_addresses idx, phys_addr_t phys); void clear_fixmap(enum fixed_addresses idx); -#endif /*__ASSEMBLY__*/ +#endif /*__ASSEMBLER__*/ #endif /*_ASM_FIXMAP_H*/ diff --git a/arch/parisc/include/asm/floppy.h b/arch/parisc/include/asm/floppy.h index b318a7df52f6..f15b69fea901 100644 --- a/arch/parisc/include/asm/floppy.h +++ b/arch/parisc/include/asm/floppy.h @@ -8,9 +8,9 @@ #ifndef __ASM_PARISC_FLOPPY_H #define __ASM_PARISC_FLOPPY_H +#include <linux/sizes.h> #include <linux/vmalloc.h> - /* * The DMA channel used by the floppy controller cannot access data at * addresses >= 16MB @@ -20,15 +20,12 @@ * floppy accesses go through the track buffer. */ #define _CROSS_64KB(a,s,vdma) \ -(!(vdma) && ((unsigned long)(a)/K_64 != ((unsigned long)(a) + (s) - 1) / K_64)) - -#define CROSS_64KB(a,s) _CROSS_64KB(a,s,use_virtual_dma & 1) - + (!(vdma) && \ + ((unsigned long)(a) / SZ_64K != ((unsigned long)(a) + (s) - 1) / SZ_64K)) #define SW fd_routine[use_virtual_dma&1] #define CSW fd_routine[can_use_virtual_dma & 1] - #define fd_inb(base, reg) readb((base) + (reg)) #define fd_outb(value, base, reg) writeb(value, (base) + (reg)) @@ -206,7 +203,7 @@ static int vdma_dma_setup(char *addr, unsigned long size, int mode, int io) static int hard_dma_setup(char *addr, unsigned long size, int mode, int io) { #ifdef FLOPPY_SANITY_CHECK - if (CROSS_64KB(addr, size)) { + if (_CROSS_64KB(addr, size, use_virtual_dma & 1)) { printk("DMA crossing 64-K boundary %p-%p\n", addr, addr+size); return -1; } diff --git a/arch/parisc/include/asm/ftrace.h b/arch/parisc/include/asm/ftrace.h index a7cf0d05ccf4..8b89d2b642eb 100644 --- a/arch/parisc/include/asm/ftrace.h +++ b/arch/parisc/include/asm/ftrace.h @@ -2,7 +2,7 @@ #ifndef _ASM_PARISC_FTRACE_H #define _ASM_PARISC_FTRACE_H -#ifndef __ASSEMBLY__ +#ifndef __ASSEMBLER__ extern void mcount(void); #define MCOUNT_ADDR ((unsigned long)mcount) @@ -12,6 +12,10 @@ extern void mcount(void); extern unsigned long sys_call_table[]; extern unsigned long return_address(unsigned int); +struct ftrace_regs; +extern void ftrace_function_trampoline(unsigned long parent, + unsigned long self_addr, unsigned long org_sp_gr3, + struct ftrace_regs *fregs); #ifdef CONFIG_DYNAMIC_FTRACE extern void ftrace_caller(void); @@ -25,6 +29,6 @@ unsigned long ftrace_call_adjust(unsigned long addr); #define ftrace_return_address(n) return_address(n) -#endif /* __ASSEMBLY__ */ +#endif /* __ASSEMBLER__ */ #endif /* _ASM_PARISC_FTRACE_H */ diff --git a/arch/parisc/include/asm/grfioctl.h b/arch/parisc/include/asm/grfioctl.h index a740844a1581..597201530d20 100644 --- a/arch/parisc/include/asm/grfioctl.h +++ b/arch/parisc/include/asm/grfioctl.h @@ -59,42 +59,4 @@ #define CRT_ID_LEGO 0x35ACDA30 /* Lego FX5, FX10 ... */ #define CRT_ID_PINNACLE 0x35ACDA16 /* Pinnacle FXe */ -/* structure for ioctl(GCDESCRIBE) */ - -#define gaddr_t unsigned long /* FIXME: PA2.0 (64bit) portable ? */ - -struct grf_fbinfo { - unsigned int id; /* upper 32 bits of graphics id */ - unsigned int mapsize; /* mapped size of framebuffer */ - unsigned int dwidth, dlength;/* x and y sizes */ - unsigned int width, length; /* total x and total y size */ - unsigned int xlen; /* x pitch size */ - unsigned int bpp, bppu; /* bits per pixel and used bpp */ - unsigned int npl, nplbytes; /* # of planes and bytes per plane */ - char name[32]; /* name of the device (from ROM) */ - unsigned int attr; /* attributes */ - gaddr_t fbbase, regbase;/* framebuffer and register base addr */ - gaddr_t regions[6]; /* region bases */ -}; - -#define GCID _IOR('G', 0, int) -#define GCON _IO('G', 1) -#define GCOFF _IO('G', 2) -#define GCAON _IO('G', 3) -#define GCAOFF _IO('G', 4) -#define GCMAP _IOWR('G', 5, int) -#define GCUNMAP _IOWR('G', 6, int) -#define GCMAP_HPUX _IO('G', 5) -#define GCUNMAP_HPUX _IO('G', 6) -#define GCLOCK _IO('G', 7) -#define GCUNLOCK _IO('G', 8) -#define GCLOCK_MINIMUM _IO('G', 9) -#define GCUNLOCK_MINIMUM _IO('G', 10) -#define GCSTATIC_CMAP _IO('G', 11) -#define GCVARIABLE_CMAP _IO('G', 12) -#define GCTERM _IOWR('G',20,int) /* multi-headed Tomcat */ -#define GCDESCRIBE _IOR('G', 21, struct grf_fbinfo) -#define GCFASTLOCK _IO('G', 26) - #endif /* __ASM_PARISC_GRFIOCTL_H */ - diff --git a/arch/parisc/include/asm/hugetlb.h b/arch/parisc/include/asm/hugetlb.h index f7f078c2872c..21e9ace17739 100644 --- a/arch/parisc/include/asm/hugetlb.h +++ b/arch/parisc/include/asm/hugetlb.h @@ -6,26 +6,11 @@ #define __HAVE_ARCH_HUGE_SET_HUGE_PTE_AT void set_huge_pte_at(struct mm_struct *mm, unsigned long addr, - pte_t *ptep, pte_t pte); + pte_t *ptep, pte_t pte, unsigned long sz); #define __HAVE_ARCH_HUGE_PTEP_GET_AND_CLEAR pte_t huge_ptep_get_and_clear(struct mm_struct *mm, unsigned long addr, - pte_t *ptep); - -/* - * If the arch doesn't supply something else, assume that hugepage - * size aligned regions are ok without further preparation. - */ -#define __HAVE_ARCH_PREPARE_HUGEPAGE_RANGE -static inline int prepare_hugepage_range(struct file *file, - unsigned long addr, unsigned long len) -{ - if (len & ~HPAGE_MASK) - return -EINVAL; - if (addr & ~HPAGE_MASK) - return -EINVAL; - return 0; -} + pte_t *ptep, unsigned long sz); #define __HAVE_ARCH_HUGE_PTEP_CLEAR_FLUSH static inline pte_t huge_ptep_clear_flush(struct vm_area_struct *vma, diff --git a/arch/parisc/include/asm/ide.h b/arch/parisc/include/asm/ide.h deleted file mode 100644 index 7aa75b93a1b6..000000000000 --- a/arch/parisc/include/asm/ide.h +++ /dev/null @@ -1,54 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -/* - * linux/include/asm-parisc/ide.h - * - * Copyright (C) 1994-1996 Linus Torvalds & authors - */ - -/* - * This file contains the PARISC architecture specific IDE code. - */ - -#ifndef __ASM_PARISC_IDE_H -#define __ASM_PARISC_IDE_H - -/* Generic I/O and MEMIO string operations. */ - -#define __ide_insw insw -#define __ide_insl insl -#define __ide_outsw outsw -#define __ide_outsl outsl - -static __inline__ void __ide_mm_insw(void __iomem *port, void *addr, u32 count) -{ - while (count--) { - *(u16 *)addr = __raw_readw(port); - addr += 2; - } -} - -static __inline__ void __ide_mm_insl(void __iomem *port, void *addr, u32 count) -{ - while (count--) { - *(u32 *)addr = __raw_readl(port); - addr += 4; - } -} - -static __inline__ void __ide_mm_outsw(void __iomem *port, void *addr, u32 count) -{ - while (count--) { - __raw_writew(*(u16 *)addr, port); - addr += 2; - } -} - -static __inline__ void __ide_mm_outsl(void __iomem *port, void *addr, u32 count) -{ - while (count--) { - __raw_writel(*(u32 *)addr, port); - addr += 4; - } -} - -#endif /* __ASM_PARISC_IDE_H */ diff --git a/arch/parisc/include/asm/io.h b/arch/parisc/include/asm/io.h index c05e781be2f5..f01ad3ad60b5 100644 --- a/arch/parisc/include/asm/io.h +++ b/arch/parisc/include/asm/io.h @@ -125,19 +125,15 @@ static inline void gsc_writeq(unsigned long long val, unsigned long addr) /* * The standard PCI ioremap interfaces */ -void __iomem *ioremap(unsigned long offset, unsigned long size); -#define ioremap_wc ioremap -#define ioremap_uc ioremap -#define pci_iounmap pci_iounmap +#define ioremap_prot ioremap_prot + +#define _PAGE_IOREMAP (_PAGE_PRESENT | _PAGE_RW | _PAGE_DIRTY | \ + _PAGE_ACCESSED | _PAGE_NO_CACHE) -extern void iounmap(const volatile void __iomem *addr); +#define ioremap_wc(addr, size) \ + ioremap_prot((addr), (size), __pgprot(_PAGE_IOREMAP)) -void memset_io(volatile void __iomem *addr, unsigned char val, int count); -void memcpy_fromio(void *dst, const volatile void __iomem *src, int count); -void memcpy_toio(volatile void __iomem *dst, const void *src, int count); -#define memset_io memset_io -#define memcpy_fromio memcpy_fromio -#define memcpy_toio memcpy_toio +#define pci_iounmap pci_iounmap /* Port-space IO */ @@ -228,46 +224,58 @@ extern void outsl (unsigned long port, const void *src, unsigned long count); #define F_EXTEND(x) ((unsigned long)((x) | (0xffffffff00000000ULL))) #ifdef CONFIG_64BIT -#define ioread64 ioread64 -#define ioread64be ioread64be -#define iowrite64 iowrite64 -#define iowrite64be iowrite64be extern u64 ioread64(const void __iomem *addr); extern u64 ioread64be(const void __iomem *addr); +#define ioread64 ioread64 +#define ioread64be ioread64be + extern void iowrite64(u64 val, void __iomem *addr); extern void iowrite64be(u64 val, void __iomem *addr); +#define iowrite64 iowrite64 +#define iowrite64be iowrite64be #endif -#include <asm-generic/iomap.h> -/* - * These get provided from <asm-generic/iomap.h> since parisc does not - * select GENERIC_IOMAP. - */ +extern void __iomem *ioport_map(unsigned long port, unsigned int nr); +extern void ioport_unmap(void __iomem *); #define ioport_map ioport_map #define ioport_unmap ioport_unmap + +extern unsigned int ioread8(const void __iomem *); +extern unsigned int ioread16(const void __iomem *); +extern unsigned int ioread16be(const void __iomem *); +extern unsigned int ioread32(const void __iomem *); +extern unsigned int ioread32be(const void __iomem *); #define ioread8 ioread8 #define ioread16 ioread16 #define ioread32 ioread32 #define ioread16be ioread16be #define ioread32be ioread32be + +extern void iowrite8(u8, void __iomem *); +extern void iowrite16(u16, void __iomem *); +extern void iowrite16be(u16, void __iomem *); +extern void iowrite32(u32, void __iomem *); +extern void iowrite32be(u32, void __iomem *); #define iowrite8 iowrite8 #define iowrite16 iowrite16 #define iowrite32 iowrite32 #define iowrite16be iowrite16be #define iowrite32be iowrite32be + +extern void ioread8_rep(const void __iomem *port, void *buf, unsigned long count); +extern void ioread16_rep(const void __iomem *port, void *buf, unsigned long count); +extern void ioread32_rep(const void __iomem *port, void *buf, unsigned long count); #define ioread8_rep ioread8_rep #define ioread16_rep ioread16_rep #define ioread32_rep ioread32_rep + +extern void iowrite8_rep(void __iomem *port, const void *buf, unsigned long count); +extern void iowrite16_rep(void __iomem *port, const void *buf, unsigned long count); +extern void iowrite32_rep(void __iomem *port, const void *buf, unsigned long count); #define iowrite8_rep iowrite8_rep #define iowrite16_rep iowrite16_rep #define iowrite32_rep iowrite32_rep -/* - * Convert a physical pointer to a virtual kernel pointer for /dev/mem - * access - */ -#define xlate_dev_mem_ptr(p) __va(p) - extern int devmem_is_allowed(unsigned long pfn); #include <asm-generic/io.h> diff --git a/arch/parisc/include/asm/irqflags.h b/arch/parisc/include/asm/irqflags.h index 38a19c0bac3a..00fd87724588 100644 --- a/arch/parisc/include/asm/irqflags.h +++ b/arch/parisc/include/asm/irqflags.h @@ -31,6 +31,11 @@ static inline unsigned long arch_local_irq_save(void) static inline void arch_local_irq_restore(unsigned long flags) { + /* warn if IRQs are on although they should be off */ + if (IS_ENABLED(CONFIG_LIGHTWEIGHT_SPINLOCK_CHECK)) + if (arch_local_save_flags() & PSW_I) + asm volatile("break 6,6\n"); /* SPINLOCK_BREAK_INSN */ + asm volatile("mtsm %0" : : "r" (flags) : "memory"); } diff --git a/arch/parisc/include/asm/jump_label.h b/arch/parisc/include/asm/jump_label.h index af2a598bc0f8..f325ae3c622f 100644 --- a/arch/parisc/include/asm/jump_label.h +++ b/arch/parisc/include/asm/jump_label.h @@ -2,7 +2,7 @@ #ifndef _ASM_PARISC_JUMP_LABEL_H #define _ASM_PARISC_JUMP_LABEL_H -#ifndef __ASSEMBLY__ +#ifndef __ASSEMBLER__ #include <linux/types.h> #include <linux/stringify.h> @@ -12,13 +12,15 @@ static __always_inline bool arch_static_branch(struct static_key *key, bool branch) { - asm_volatile_goto("1:\n\t" + asm goto("1:\n\t" "nop\n\t" ".pushsection __jump_table, \"aw\"\n\t" + ".align %1\n\t" ".word 1b - ., %l[l_yes] - .\n\t" __stringify(ASM_ULONG_INSN) " %c0 - .\n\t" ".popsection\n\t" - : : "i" (&((char *)key)[branch]) : : l_yes); + : : "i" (&((char *)key)[branch]), "i" (sizeof(long)) + : : l_yes); return false; l_yes: @@ -27,18 +29,20 @@ l_yes: static __always_inline bool arch_static_branch_jump(struct static_key *key, bool branch) { - asm_volatile_goto("1:\n\t" + asm goto("1:\n\t" "b,n %l[l_yes]\n\t" ".pushsection __jump_table, \"aw\"\n\t" + ".align %1\n\t" ".word 1b - ., %l[l_yes] - .\n\t" __stringify(ASM_ULONG_INSN) " %c0 - .\n\t" ".popsection\n\t" - : : "i" (&((char *)key)[branch]) : : l_yes); + : : "i" (&((char *)key)[branch]), "i" (sizeof(long)) + : : l_yes); return false; l_yes: return true; } -#endif /* __ASSEMBLY__ */ +#endif /* __ASSEMBLER__ */ #endif diff --git a/arch/parisc/include/asm/kexec.h b/arch/parisc/include/asm/kexec.h index 87e174006995..bf31e2d50df9 100644 --- a/arch/parisc/include/asm/kexec.h +++ b/arch/parisc/include/asm/kexec.h @@ -14,7 +14,7 @@ #define KEXEC_ARCH KEXEC_ARCH_PARISC #define ARCH_HAS_KIMAGE_ARCH -#ifndef __ASSEMBLY__ +#ifndef __ASSEMBLER__ struct kimage_arch { unsigned long initrd_start; @@ -28,6 +28,6 @@ static inline void crash_setup_regs(struct pt_regs *newregs, /* Dummy implementation for now */ } -#endif /* __ASSEMBLY__ */ +#endif /* __ASSEMBLER__ */ #endif /* _ASM_PARISC_KEXEC_H */ diff --git a/arch/parisc/include/asm/kgdb.h b/arch/parisc/include/asm/kgdb.h index f23e7f8f13a5..9ece98bc6d9d 100644 --- a/arch/parisc/include/asm/kgdb.h +++ b/arch/parisc/include/asm/kgdb.h @@ -17,9 +17,11 @@ #define NUMREGBYTES sizeof(struct parisc_gdb_regs) #define BUFMAX 4096 +#define KGDB_MAX_BREAKPOINTS 40 + #define CACHE_FLUSH_IS_SAFE 1 -#ifndef __ASSEMBLY__ +#ifndef __ASSEMBLER__ static inline void arch_kgdb_breakpoint(void) { diff --git a/arch/parisc/include/asm/kprobes.h b/arch/parisc/include/asm/kprobes.h index 0a175ac87698..0f42f5c8e3b6 100644 --- a/arch/parisc/include/asm/kprobes.h +++ b/arch/parisc/include/asm/kprobes.h @@ -10,9 +10,10 @@ #ifndef _PARISC_KPROBES_H #define _PARISC_KPROBES_H +#include <asm-generic/kprobes.h> + #ifdef CONFIG_KPROBES -#include <asm-generic/kprobes.h> #include <linux/types.h> #include <linux/ptrace.h> #include <linux/notifier.h> diff --git a/arch/parisc/include/asm/ldcw.h b/arch/parisc/include/asm/ldcw.h index 6d28b5514699..47ebc4c91eaf 100644 --- a/arch/parisc/include/asm/ldcw.h +++ b/arch/parisc/include/asm/ldcw.h @@ -2,39 +2,42 @@ #ifndef __PARISC_LDCW_H #define __PARISC_LDCW_H -#ifndef CONFIG_PA20 /* Because kmalloc only guarantees 8-byte alignment for kmalloc'd data, and GCC only guarantees 8-byte alignment for stack locals, we can't be assured of 16-byte alignment for atomic lock data even if we specify "__attribute ((aligned(16)))" in the type declaration. So, we use a struct containing an array of four ints for the atomic lock type and dynamically select the 16-byte aligned int from the array - for the semaphore. */ + for the semaphore. */ + +/* From: "Jim Hull" <jim.hull of hp.com> + I've attached a summary of the change, but basically, for PA 2.0, as + long as the ",CO" (coherent operation) completer is implemented, then the + 16-byte alignment requirement for ldcw and ldcd is relaxed, and instead + they only require "natural" alignment (4-byte for ldcw, 8-byte for + ldcd). + + Although the cache control hint is accepted by all PA 2.0 processors, + it is only implemented on PA8800/PA8900 CPUs. Prior PA8X00 CPUs still + require 16-byte alignment. If the address is unaligned, the operation + of the instruction is undefined. The ldcw instruction does not generate + unaligned data reference traps so misaligned accesses are not detected. + This hid the problem for years. So, restore the 16-byte alignment dropped + by Kyle McMartin in "Remove __ldcw_align for PA-RISC 2.0 processors". */ #define __PA_LDCW_ALIGNMENT 16 -#define __PA_LDCW_ALIGN_ORDER 4 #define __ldcw_align(a) ({ \ unsigned long __ret = (unsigned long) &(a)->lock[0]; \ __ret = (__ret + __PA_LDCW_ALIGNMENT - 1) \ & ~(__PA_LDCW_ALIGNMENT - 1); \ (volatile unsigned int *) __ret; \ }) -#define __LDCW "ldcw" -#else /*CONFIG_PA20*/ -/* From: "Jim Hull" <jim.hull of hp.com> - I've attached a summary of the change, but basically, for PA 2.0, as - long as the ",CO" (coherent operation) completer is specified, then the - 16-byte alignment requirement for ldcw and ldcd is relaxed, and instead - they only require "natural" alignment (4-byte for ldcw, 8-byte for - ldcd). */ - -#define __PA_LDCW_ALIGNMENT 4 -#define __PA_LDCW_ALIGN_ORDER 2 -#define __ldcw_align(a) (&(a)->slock) +#ifdef CONFIG_PA20 #define __LDCW "ldcw,co" - -#endif /*!CONFIG_PA20*/ +#else +#define __LDCW "ldcw" +#endif /* LDCW, the only atomic read-write operation PA-RISC has. *sigh*. We don't explicitly expose that "*a" may be written as reload @@ -52,7 +55,7 @@ }) #ifdef CONFIG_SMP -# define __lock_aligned __section(".data..lock_aligned") +# define __lock_aligned __section(".data..lock_aligned") __aligned(16) #endif #endif /* __PARISC_LDCW_H */ diff --git a/arch/parisc/include/asm/led.h b/arch/parisc/include/asm/led.h index 6de13d08a388..0aea47eff48d 100644 --- a/arch/parisc/include/asm/led.h +++ b/arch/parisc/include/asm/led.h @@ -11,8 +11,8 @@ #define LED1 0x02 #define LED0 0x01 /* bottom (or furthest left) LED */ -#define LED_LAN_TX LED0 /* for LAN transmit activity */ -#define LED_LAN_RCV LED1 /* for LAN receive activity */ +#define LED_LAN_RCV LED0 /* for LAN receive activity */ +#define LED_LAN_TX LED1 /* for LAN transmit activity */ #define LED_DISK_IO LED2 /* for disk activity */ #define LED_HEARTBEAT LED3 /* heartbeat */ @@ -25,19 +25,13 @@ #define LED_CMD_REG_NONE 0 /* NULL == no addr for the cmd register */ /* register_led_driver() */ -int __init register_led_driver(int model, unsigned long cmd_reg, unsigned long data_reg); - -/* registers the LED regions for procfs */ -void __init register_led_regions(void); +int register_led_driver(int model, unsigned long cmd_reg, unsigned long data_reg); #ifdef CONFIG_CHASSIS_LCD_LED /* writes a string to the LCD display (if possible on this h/w) */ -int lcd_print(const char *str); +void lcd_print(const char *str); #else -#define lcd_print(str) +#define lcd_print(str) do { } while (0) #endif -/* main LED initialization function (uses PDC) */ -int __init led_init(void); - #endif /* LED_H */ diff --git a/arch/parisc/include/asm/linkage.h b/arch/parisc/include/asm/linkage.h index cd6fe4febead..d4cad492b971 100644 --- a/arch/parisc/include/asm/linkage.h +++ b/arch/parisc/include/asm/linkage.h @@ -15,7 +15,7 @@ */ #define ASM_NL ! -#ifdef __ASSEMBLY__ +#ifdef __ASSEMBLER__ #define ENTRY(name) \ ALIGN !\ @@ -35,6 +35,6 @@ name: ASM_NL\ .procend ASM_NL\ ENDPROC(name) -#endif /* __ASSEMBLY__ */ +#endif /* __ASSEMBLER__ */ #endif /* __ASM_PARISC_LINKAGE_H */ diff --git a/arch/parisc/include/asm/machdep.h b/arch/parisc/include/asm/machdep.h deleted file mode 100644 index 215d2c43989d..000000000000 --- a/arch/parisc/include/asm/machdep.h +++ /dev/null @@ -1,17 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef _PARISC_MACHDEP_H -#define _PARISC_MACHDEP_H - -#include <linux/notifier.h> - -#define MACH_RESTART 1 -#define MACH_HALT 2 -#define MACH_POWER_ON 3 -#define MACH_POWER_OFF 4 - -extern struct notifier_block *mach_notifier; -extern void pa7300lc_init(void); - -extern void (*cpu_lpmc)(int, struct pt_regs *); - -#endif diff --git a/arch/parisc/include/asm/mckinley.h b/arch/parisc/include/asm/mckinley.h deleted file mode 100644 index 1314390b9034..000000000000 --- a/arch/parisc/include/asm/mckinley.h +++ /dev/null @@ -1,8 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef ASM_PARISC_MCKINLEY_H -#define ASM_PARISC_MCKINLEY_H - -/* declared in arch/parisc/kernel/setup.c */ -extern struct proc_dir_entry * proc_mckinley_root; - -#endif /*ASM_PARISC_MCKINLEY_H*/ diff --git a/arch/parisc/include/asm/mman.h b/arch/parisc/include/asm/mman.h new file mode 100644 index 000000000000..663f587dc789 --- /dev/null +++ b/arch/parisc/include/asm/mman.h @@ -0,0 +1,29 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __ASM_MMAN_H__ +#define __ASM_MMAN_H__ + +#include <linux/fs.h> +#include <uapi/asm/mman.h> + +/* PARISC cannot allow mdwe as it needs writable stacks */ +static inline bool arch_memory_deny_write_exec_supported(void) +{ + return false; +} +#define arch_memory_deny_write_exec_supported arch_memory_deny_write_exec_supported + +static inline unsigned long arch_calc_vm_flag_bits(struct file *file, unsigned long flags) +{ + /* + * The stack on parisc grows upwards, so if userspace requests memory + * for a stack, mark it with VM_GROWSUP so that the stack expansion in + * the fault handler will work. + */ + if (flags & MAP_STACK) + return VM_GROWSUP; + + return 0; +} +#define arch_calc_vm_flag_bits(file, flags) arch_calc_vm_flag_bits(file, flags) + +#endif /* __ASM_MMAN_H__ */ diff --git a/arch/parisc/include/asm/page.h b/arch/parisc/include/asm/page.h index 6faaaa3ebe9b..8f4e51071ea1 100644 --- a/arch/parisc/include/asm/page.h +++ b/arch/parisc/include/asm/page.h @@ -4,20 +4,11 @@ #include <linux/const.h> -#if defined(CONFIG_PARISC_PAGE_SIZE_4KB) -# define PAGE_SHIFT 12 -#elif defined(CONFIG_PARISC_PAGE_SIZE_16KB) -# define PAGE_SHIFT 14 -#elif defined(CONFIG_PARISC_PAGE_SIZE_64KB) -# define PAGE_SHIFT 16 -#else -# error "unknown default kernel page size" -#endif -#define PAGE_SIZE (_AC(1,UL) << PAGE_SHIFT) -#define PAGE_MASK (~(PAGE_SIZE-1)) +#include <vdso/page.h> +#define HAVE_ARCH_HUGETLB_UNMAPPED_AREA -#ifndef __ASSEMBLY__ +#ifndef __ASSEMBLER__ #include <asm/types.h> #include <asm/cache.h> @@ -102,7 +93,7 @@ typedef struct __physmem_range { extern physmem_range_t pmem_ranges[]; extern int npmem_ranges; -#endif /* !__ASSEMBLY__ */ +#endif /* !__ASSEMBLER__ */ /* WARNING: The definitions below must match exactly to sizeof(pte_t) * etc @@ -148,17 +139,13 @@ extern int npmem_ranges; #define KERNEL_BINARY_TEXT_START (__PAGE_OFFSET + 0x100000) /* These macros don't work for 64-bit C code -- don't allow in C at all */ -#ifdef __ASSEMBLY__ +#ifdef __ASSEMBLER__ # define PA(x) ((x)-__PAGE_OFFSET) # define VA(x) ((x)+__PAGE_OFFSET) #endif #define __pa(x) ((unsigned long)(x)-PAGE_OFFSET) #define __va(x) ((void *)((unsigned long)(x)+PAGE_OFFSET)) -#ifndef CONFIG_SPARSEMEM -#define pfn_valid(pfn) ((pfn) < max_mapnr) -#endif - #ifdef CONFIG_HUGETLB_PAGE #define HPAGE_SHIFT PMD_SHIFT /* fixed for transparent huge pages */ #define HPAGE_SIZE ((1UL) << HPAGE_SHIFT) @@ -179,7 +166,6 @@ extern int npmem_ranges; #define virt_addr_valid(kaddr) pfn_valid(__pa(kaddr) >> PAGE_SHIFT) -#define page_to_phys(page) (page_to_pfn(page) << PAGE_SHIFT) #define virt_to_page(kaddr) pfn_to_page(__pa(kaddr) >> PAGE_SHIFT) #include <asm-generic/memory_model.h> diff --git a/arch/parisc/include/asm/parisc-device.h b/arch/parisc/include/asm/parisc-device.h index 4de3b391d812..9e74cef4d774 100644 --- a/arch/parisc/include/asm/parisc-device.h +++ b/arch/parisc/include/asm/parisc-device.h @@ -41,7 +41,7 @@ struct parisc_driver { #define to_parisc_device(d) container_of(d, struct parisc_device, dev) -#define to_parisc_driver(d) container_of(d, struct parisc_driver, drv) +#define to_parisc_driver(d) container_of_const(d, struct parisc_driver, drv) #define parisc_parent(d) to_parisc_device(d->dev.parent) static inline const char *parisc_pathname(struct parisc_device *d) @@ -61,7 +61,7 @@ parisc_get_drvdata(struct parisc_device *d) return dev_get_drvdata(&d->dev); } -extern struct bus_type parisc_bus_type; +extern const struct bus_type parisc_bus_type; int iosapic_serial_irq(struct parisc_device *dev); diff --git a/arch/parisc/include/asm/pdc.h b/arch/parisc/include/asm/pdc.h index 40793bef8429..6080a1516b34 100644 --- a/arch/parisc/include/asm/pdc.h +++ b/arch/parisc/include/asm/pdc.h @@ -4,7 +4,7 @@ #include <uapi/asm/pdc.h> -#if !defined(__ASSEMBLY__) +#if !defined(__ASSEMBLER__) extern int parisc_narrow_firmware; @@ -44,10 +44,11 @@ int pdc_model_capabilities(unsigned long *capabilities); int pdc_model_platform_info(char *orig_prod_num, char *current_prod_num, char *serial_no); int pdc_cache_info(struct pdc_cache_info *cache); int pdc_spaceid_bits(unsigned long *space_bits); -#ifndef CONFIG_PA20 int pdc_btlb_info(struct pdc_btlb_info *btlb); +int pdc_btlb_insert(unsigned long long vpage, unsigned long physpage, unsigned long len, + unsigned long entry_info, unsigned long slot); +int pdc_btlb_purge_all(void); int pdc_mem_map_hpa(struct pdc_memory_map *r_addr, struct pdc_module_path *mod_path); -#endif /* !CONFIG_PA20 */ int pdc_pim_toc11(struct pdc_toc_pim_11 *ret); int pdc_pim_toc20(struct pdc_toc_pim_20 *ret); int pdc_lan_station_id(char *lan_addr, unsigned long net_hpa); @@ -80,6 +81,7 @@ int pdc_do_firm_test_reset(unsigned long ftc_bitmap); int pdc_do_reset(void); int pdc_soft_power_info(unsigned long *power_reg); int pdc_soft_power_button(int sw_control); +int pdc_soft_power_button_panic(int sw_control); void pdc_io_reset(void); void pdc_io_reset_devices(void); int pdc_iodc_getc(void); @@ -87,8 +89,8 @@ int pdc_iodc_print(const unsigned char *str, unsigned count); void pdc_emergency_unlock(void); int pdc_sti_call(unsigned long func, unsigned long flags, - unsigned long inptr, unsigned long outputr, - unsigned long glob_cfg); + unsigned long inptr, unsigned long outputr, + unsigned long glob_cfg, int do_call64); int __pdc_cpu_rendezvous(void); void pdc_cpu_rendezvous_lock(void); @@ -107,5 +109,5 @@ static inline char * os_id_to_string(u16 os_id) { } } -#endif /* !defined(__ASSEMBLY__) */ +#endif /* !defined(__ASSEMBLER__) */ #endif /* _PARISC_PDC_H */ diff --git a/arch/parisc/include/asm/pdcpat.h b/arch/parisc/include/asm/pdcpat.h index 8f160375b865..84ac81b1adde 100644 --- a/arch/parisc/include/asm/pdcpat.h +++ b/arch/parisc/include/asm/pdcpat.h @@ -210,7 +210,7 @@ #define PDC_PAT_SYSTEM_INFO 76L /* PDC_PAT_SYSTEM_INFO uses the same options as PDC_SYSTEM_INFO function. */ -#ifndef __ASSEMBLY__ +#ifndef __ASSEMBLER__ #include <linux/types.h> #ifdef CONFIG_64BIT @@ -389,6 +389,6 @@ extern int pdc_pat_mem_get_dimm_phys_location( struct pdc_pat_mem_phys_mem_location *pret, unsigned long phys_addr); -#endif /* __ASSEMBLY__ */ +#endif /* __ASSEMBLER__ */ #endif /* ! __PARISC_PATPDC_H */ diff --git a/arch/parisc/include/asm/perf_event.h b/arch/parisc/include/asm/perf_event.h index 1e0fd8ba6c03..8a2925029d15 100644 --- a/arch/parisc/include/asm/perf_event.h +++ b/arch/parisc/include/asm/perf_event.h @@ -1,6 +1,12 @@ #ifndef __ASM_PARISC_PERF_EVENT_H #define __ASM_PARISC_PERF_EVENT_H -/* Empty, just to avoid compiling error */ +#include <asm/psw.h> + +#define perf_arch_fetch_caller_regs(regs, __ip) { \ + (regs)->gr[0] = KERNEL_PSW; \ + (regs)->iaoq[0] = (__ip); \ + asm volatile("copy %%sp, %0\n":"=r"((regs)->gr[30])); \ +} #endif /* __ASM_PARISC_PERF_EVENT_H */ diff --git a/arch/parisc/include/asm/pgalloc.h b/arch/parisc/include/asm/pgalloc.h index e3e142b1c5c5..3b84ee93edaa 100644 --- a/arch/parisc/include/asm/pgalloc.h +++ b/arch/parisc/include/asm/pgalloc.h @@ -11,27 +11,12 @@ #include <asm/cache.h> #define __HAVE_ARCH_PMD_ALLOC_ONE -#define __HAVE_ARCH_PMD_FREE -#define __HAVE_ARCH_PGD_FREE #include <asm-generic/pgalloc.h> /* Allocate the top level pgd (page directory) */ static inline pgd_t *pgd_alloc(struct mm_struct *mm) { - pgd_t *pgd; - - pgd = (pgd_t *) __get_free_pages(GFP_KERNEL, PGD_TABLE_ORDER); - if (unlikely(pgd == NULL)) - return NULL; - - memset(pgd, 0, PAGE_SIZE << PGD_TABLE_ORDER); - - return pgd; -} - -static inline void pgd_free(struct mm_struct *mm, pgd_t *pgd) -{ - free_pages((unsigned long)pgd, PGD_TABLE_ORDER); + return __pgd_alloc(mm, PGD_TABLE_ORDER); } #if CONFIG_PGTABLE_LEVELS == 3 @@ -46,17 +31,19 @@ static inline void pud_populate(struct mm_struct *mm, pud_t *pud, pmd_t *pmd) static inline pmd_t *pmd_alloc_one(struct mm_struct *mm, unsigned long address) { - pmd_t *pmd; - - pmd = (pmd_t *)__get_free_pages(GFP_PGTABLE_KERNEL, PMD_TABLE_ORDER); - if (likely(pmd)) - memset ((void *)pmd, 0, PAGE_SIZE << PMD_TABLE_ORDER); - return pmd; -} + struct ptdesc *ptdesc; + gfp_t gfp = GFP_PGTABLE_USER; -static inline void pmd_free(struct mm_struct *mm, pmd_t *pmd) -{ - free_pages((unsigned long)pmd, PMD_TABLE_ORDER); + if (mm == &init_mm) + gfp = GFP_PGTABLE_KERNEL; + ptdesc = pagetable_alloc(gfp, PMD_TABLE_ORDER); + if (!ptdesc) + return NULL; + if (!pagetable_pmd_ctor(mm, ptdesc)) { + pagetable_free(ptdesc); + return NULL; + } + return ptdesc_address(ptdesc); } #endif diff --git a/arch/parisc/include/asm/pgtable.h b/arch/parisc/include/asm/pgtable.h index ea357430aafe..2c139a4dbf4b 100644 --- a/arch/parisc/include/asm/pgtable.h +++ b/arch/parisc/include/asm/pgtable.h @@ -12,7 +12,7 @@ #include <asm/fixmap.h> -#ifndef __ASSEMBLY__ +#ifndef __ASSEMBLER__ /* * we simulate an x86-style page table for the linux mm code */ @@ -73,16 +73,7 @@ extern void __update_cache(pte_t pte); mb(); \ } while(0) -#define set_pte_at(mm, addr, pteptr, pteval) \ - do { \ - if (pte_present(pteval) && \ - pte_user(pteval)) \ - __update_cache(pteval); \ - *(pteptr) = (pteval); \ - purge_tlb_entries(mm, addr); \ - } while (0) - -#endif /* !__ASSEMBLY__ */ +#endif /* !__ASSEMBLER__ */ #define pte_ERROR(e) \ printk("%s:%d: bad pte %08lx.\n", __FILE__, __LINE__, pte_val(e)) @@ -218,6 +209,9 @@ extern void __update_cache(pte_t pte); #define _PAGE_KERNEL_RWX (_PAGE_KERNEL_EXEC | _PAGE_WRITE) #define _PAGE_KERNEL (_PAGE_KERNEL_RO | _PAGE_WRITE) +/* We borrow bit 23 to store the exclusive marker in swap PTEs. */ +#define _PAGE_SWP_EXCLUSIVE _PAGE_ACCESSED + /* The pgd/pmd contains a ptr (in phys addr space); since all pgds/pmds * are page-aligned, we don't care about the PAGE_OFFSET bits, except * for a few meta-information bits, so we shift the address to be @@ -232,7 +226,7 @@ extern void __update_cache(pte_t pte); #define PxD_FLAG_SHIFT (4) #define PxD_VALUE_SHIFT (PFN_PTE_SHIFT-PxD_FLAG_SHIFT) -#ifndef __ASSEMBLY__ +#ifndef __ASSEMBLER__ #define PAGE_NONE __pgprot(_PAGE_PRESENT | _PAGE_USER) #define PAGE_SHARED __pgprot(_PAGE_PRESENT | _PAGE_USER | _PAGE_READ | _PAGE_WRITE) @@ -282,7 +276,7 @@ extern unsigned long *empty_zero_page; #define pte_none(x) (pte_val(x) == 0) #define pte_present(x) (pte_val(x) & _PAGE_PRESENT) #define pte_user(x) (pte_val(x) & _PAGE_USER) -#define pte_clear(mm, addr, xp) set_pte_at(mm, addr, xp, __pte(0)) +#define pte_clear(mm, addr, xp) set_pte_at((mm), (addr), (xp), __pte(0)) #define pmd_flag(x) (pmd_val(x) & PxD_FLAG_MASK) #define pmd_address(x) ((unsigned long)(pmd_val(x) &~ PxD_FLAG_MASK) << PxD_VALUE_SHIFT) @@ -328,7 +322,7 @@ static inline pte_t pte_mkold(pte_t pte) { pte_val(pte) &= ~_PAGE_ACCESSED; retu static inline pte_t pte_wrprotect(pte_t pte) { pte_val(pte) &= ~_PAGE_WRITE; return pte; } static inline pte_t pte_mkdirty(pte_t pte) { pte_val(pte) |= _PAGE_DIRTY; return pte; } static inline pte_t pte_mkyoung(pte_t pte) { pte_val(pte) |= _PAGE_ACCESSED; return pte; } -static inline pte_t pte_mkwrite(pte_t pte) { pte_val(pte) |= _PAGE_WRITE; return pte; } +static inline pte_t pte_mkwrite_novma(pte_t pte) { pte_val(pte) |= _PAGE_WRITE; return pte; } static inline pte_t pte_mkspecial(pte_t pte) { pte_val(pte) |= _PAGE_SPECIAL; return pte; } /* @@ -344,10 +338,6 @@ static inline pte_t pte_mkspecial(pte_t pte) { pte_val(pte) |= _PAGE_SPECIAL; re #endif -/* - * Conversion functions: convert a page and protection to a page entry, - * and a page entry and page directory to the page they refer to. - */ #define __mk_pte(addr,pgprot) \ ({ \ pte_t __pte; \ @@ -357,8 +347,6 @@ static inline pte_t pte_mkspecial(pte_t pte) { pte_val(pte) |= _PAGE_SPECIAL; re __pte; \ }) -#define mk_pte(page, pgprot) pfn_pte(page_to_pfn(page), (pgprot)) - static inline pte_t pfn_pte(unsigned long pfn, pgprot_t pgprot) { pte_t pte; @@ -388,31 +376,84 @@ static inline unsigned long pmd_page_vaddr(pmd_t pmd) extern void paging_init (void); +static inline void set_ptes(struct mm_struct *mm, unsigned long addr, + pte_t *ptep, pte_t pte, unsigned int nr) +{ + if (pte_present(pte) && pte_user(pte)) + __update_cache(pte); + for (;;) { + *ptep = pte; + purge_tlb_entries(mm, addr); + if (--nr == 0) + break; + ptep++; + pte_val(pte) += 1 << PFN_PTE_SHIFT; + addr += PAGE_SIZE; + } +} +#define set_ptes set_ptes +#define set_pte_at(mm, addr, ptep, pte) set_ptes(mm, addr, ptep, pte, 1) + /* Used for deferring calls to flush_dcache_page() */ #define PG_dcache_dirty PG_arch_1 -#define update_mmu_cache(vms,addr,ptep) __update_cache(*ptep) - -/* Encode and de-code a swap entry */ +#define update_mmu_cache_range(vmf, vma, addr, ptep, nr) __update_cache(*ptep) +#define update_mmu_cache(vma, addr, ptep) __update_cache(*ptep) +/* + * Encode/decode swap entries and swap PTEs. Swap PTEs are all PTEs that + * are !pte_none() && !pte_present(). + * + * Format of swap PTEs (32bit): + * + * 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 3 3 + * 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 + * <---------------- offset -----------------> P E <ofs> < type -> + * + * E is the exclusive marker that is not stored in swap entries. + * _PAGE_PRESENT (P) must be 0. + * + * For the 64bit version, the offset is extended by 32bit. + */ #define __swp_type(x) ((x).val & 0x1f) -#define __swp_offset(x) ( (((x).val >> 6) & 0x7) | \ - (((x).val >> 8) & ~0x7) ) -#define __swp_entry(type, offset) ((swp_entry_t) { (type) | \ - ((offset & 0x7) << 6) | \ - ((offset & ~0x7) << 8) }) +#define __swp_offset(x) ( (((x).val >> 5) & 0x7) | \ + (((x).val >> 10) << 3) ) +#define __swp_entry(type, offset) ((swp_entry_t) { \ + ((type) & 0x1f) | \ + ((offset & 0x7) << 5) | \ + ((offset >> 3) << 10) }) #define __pte_to_swp_entry(pte) ((swp_entry_t) { pte_val(pte) }) #define __swp_entry_to_pte(x) ((pte_t) { (x).val }) +static inline bool pte_swp_exclusive(pte_t pte) +{ + return pte_val(pte) & _PAGE_SWP_EXCLUSIVE; +} + +static inline pte_t pte_swp_mkexclusive(pte_t pte) +{ + pte_val(pte) |= _PAGE_SWP_EXCLUSIVE; + return pte; +} + +static inline pte_t pte_swp_clear_exclusive(pte_t pte) +{ + pte_val(pte) &= ~_PAGE_SWP_EXCLUSIVE; + return pte; +} + +static inline pte_t ptep_get(pte_t *ptep) +{ + return READ_ONCE(*ptep); +} +#define ptep_get ptep_get + static inline int ptep_test_and_clear_young(struct vm_area_struct *vma, unsigned long addr, pte_t *ptep) { pte_t pte; - if (!pte_young(*ptep)) - return 0; - - pte = *ptep; + pte = ptep_get(ptep); if (!pte_young(pte)) { return 0; } @@ -420,17 +461,10 @@ static inline int ptep_test_and_clear_young(struct vm_area_struct *vma, unsigned return 1; } -struct mm_struct; -static inline pte_t ptep_get_and_clear(struct mm_struct *mm, unsigned long addr, pte_t *ptep) -{ - pte_t old_pte; - - old_pte = *ptep; - set_pte_at(mm, addr, ptep, __pte(0)); - - return old_pte; -} +int ptep_clear_flush_young(struct vm_area_struct *vma, unsigned long addr, pte_t *ptep); +pte_t ptep_clear_flush(struct vm_area_struct *vma, unsigned long addr, pte_t *ptep); +struct mm_struct; static inline void ptep_set_wrprotect(struct mm_struct *mm, unsigned long addr, pte_t *ptep) { set_pte_at(mm, addr, ptep, pte_wrprotect(*ptep)); @@ -438,10 +472,7 @@ static inline void ptep_set_wrprotect(struct mm_struct *mm, unsigned long addr, #define pte_same(A,B) (pte_val(A) == pte_val(B)) -struct seq_file; -extern void arch_report_meminfo(struct seq_file *m); - -#endif /* !__ASSEMBLY__ */ +#endif /* !__ASSEMBLER__ */ /* TLB page size encoding - see table 3-1 in parisc20.pdf */ @@ -471,7 +502,8 @@ extern void arch_report_meminfo(struct seq_file *m); #define HAVE_ARCH_UNMAPPED_AREA_TOPDOWN #define __HAVE_ARCH_PTEP_TEST_AND_CLEAR_YOUNG -#define __HAVE_ARCH_PTEP_GET_AND_CLEAR +#define __HAVE_ARCH_PTEP_CLEAR_YOUNG_FLUSH +#define __HAVE_ARCH_PTEP_CLEAR_FLUSH #define __HAVE_ARCH_PTEP_SET_WRPROTECT #define __HAVE_ARCH_PTE_SAME diff --git a/arch/parisc/include/asm/prefetch.h b/arch/parisc/include/asm/prefetch.h index 6e63f720024d..748eefb27c68 100644 --- a/arch/parisc/include/asm/prefetch.h +++ b/arch/parisc/include/asm/prefetch.h @@ -16,7 +16,7 @@ #ifndef __ASM_PARISC_PREFETCH_H #define __ASM_PARISC_PREFETCH_H -#ifndef __ASSEMBLY__ +#ifndef __ASSEMBLER__ #ifdef CONFIG_PREFETCH #define ARCH_HAS_PREFETCH @@ -40,6 +40,6 @@ static inline void prefetchw(const void *addr) #endif /* CONFIG_PA20 */ #endif /* CONFIG_PREFETCH */ -#endif /* __ASSEMBLY__ */ +#endif /* __ASSEMBLER__ */ #endif /* __ASM_PARISC_PROCESSOR_H */ diff --git a/arch/parisc/include/asm/processor.h b/arch/parisc/include/asm/processor.h index a608970b249a..dd0b5e199559 100644 --- a/arch/parisc/include/asm/processor.h +++ b/arch/parisc/include/asm/processor.h @@ -9,8 +9,9 @@ #ifndef __ASM_PARISC_PROCESSOR_H #define __ASM_PARISC_PROCESSOR_H -#ifndef __ASSEMBLY__ +#ifndef __ASSEMBLER__ #include <linux/threads.h> +#include <linux/irqreturn.h> #include <asm/assembly.h> #include <asm/prefetch.h> @@ -19,7 +20,7 @@ #include <asm/ptrace.h> #include <asm/types.h> #include <asm/percpu.h> -#endif /* __ASSEMBLY__ */ +#endif /* __ASSEMBLER__ */ #define HAVE_ARCH_PICK_MMAP_LAYOUT @@ -44,8 +45,10 @@ #define STACK_TOP TASK_SIZE #define STACK_TOP_MAX DEFAULT_TASK_SIZE -#ifndef __ASSEMBLY__ +#ifndef __ASSEMBLER__ +struct rlimit; +unsigned long mmap_upper_limit(const struct rlimit *rlim_stack); unsigned long calc_max_stack_size(unsigned long stack_max); /* @@ -286,12 +289,42 @@ extern int _parisc_requires_coherency; #endif extern int running_on_qemu; +extern int parisc_narrow_firmware; extern void __noreturn toc_intr(struct pt_regs *regs); extern void toc_handler(void); extern unsigned int toc_handler_size; extern unsigned int toc_handler_csum; - -#endif /* __ASSEMBLY__ */ +extern void do_cpu_irq_mask(struct pt_regs *); +extern irqreturn_t timer_interrupt(int, void *); +extern irqreturn_t ipi_interrupt(int, void *); +extern void parisc_clockevent_init(void); +extern void handle_interruption(int, struct pt_regs *); + +/* called from assembly code: */ +extern void start_parisc(void); +extern void smp_callin(unsigned long); +extern void sys_rt_sigreturn(struct pt_regs *, int); +extern void do_notify_resume(struct pt_regs *, long); +extern long do_syscall_trace_enter(struct pt_regs *); +extern void do_syscall_trace_exit(struct pt_regs *); + +/* CPU startup and info */ +struct seq_file; +extern void early_trap_init(void); +extern void collect_boot_cpu_data(void); +extern void btlb_init_per_cpu(void); +extern int show_cpuinfo (struct seq_file *m, void *v); + +/* driver code in driver/parisc */ +extern void processor_init(void); +struct parisc_device; +struct resource; +extern void sba_distributed_lmmio(struct parisc_device *, struct resource *); +extern void sba_directed_lmmio(struct parisc_device *, struct resource *); +extern void lba_set_iregs(struct parisc_device *lba, u32 ibase, u32 imask); +extern void ccio_cujo20_fixup(struct parisc_device *dev, u32 iovp); + +#endif /* __ASSEMBLER__ */ #endif /* __ASM_PARISC_PROCESSOR_H */ diff --git a/arch/parisc/include/asm/psw.h b/arch/parisc/include/asm/psw.h index 46921ffcc407..9140e1ab7e63 100644 --- a/arch/parisc/include/asm/psw.h +++ b/arch/parisc/include/asm/psw.h @@ -60,7 +60,7 @@ #define USER_PSW_MASK (WIDE_PSW | PSW_T | PSW_N | PSW_X | PSW_B | PSW_V | PSW_CB) #define USER_PSW (PSW_C | PSW_Q | PSW_P | PSW_D | PSW_I) -#ifndef __ASSEMBLY__ +#ifndef __ASSEMBLER__ /* The program status word as bitfields. */ struct pa_psw { @@ -99,6 +99,6 @@ struct pa_psw { #define pa_psw(task) ((struct pa_psw *) ((char *) (task) + TASK_PT_PSW)) #endif -#endif /* !__ASSEMBLY__ */ +#endif /* !__ASSEMBLER__ */ #endif diff --git a/arch/parisc/include/asm/ropes.h b/arch/parisc/include/asm/ropes.h index 8e51c775c80a..e2d2d7e9bfde 100644 --- a/arch/parisc/include/asm/ropes.h +++ b/arch/parisc/include/asm/ropes.h @@ -29,7 +29,7 @@ struct ioc { void __iomem *ioc_hpa; /* I/O MMU base address */ char *res_map; /* resource map, bit == pdir entry */ - u64 *pdir_base; /* physical base address */ + __le64 *pdir_base; /* physical base address */ unsigned long ibase; /* pdir IOV Space base - shared w/lba_pci */ unsigned long imask; /* pdir IOV Space mask - shared w/lba_pci */ #ifdef ZX1_SUPPORT @@ -86,6 +86,9 @@ struct sba_device { struct ioc ioc[MAX_IOC]; }; +/* list of SBA's in system, see drivers/parisc/sba_iommu.c */ +extern struct sba_device *sba_list; + #define ASTRO_RUNWAY_PORT 0x582 #define IKE_MERCED_PORT 0x803 #define REO_MERCED_PORT 0x804 @@ -110,7 +113,7 @@ static inline int IS_PLUTO(struct parisc_device *d) { #define SBA_PDIR_VALID_BIT 0x8000000000000000ULL -#define SBA_AGPGART_COOKIE 0x0000badbadc0ffeeULL +#define SBA_AGPGART_COOKIE (__force __le64) 0x0000badbadc0ffeeULL #define SBA_FUNC_ID 0x0000 /* function id */ #define SBA_FCLASS 0x0008 /* function class, bist, header, rev... */ @@ -252,7 +255,7 @@ static inline int agp_mode_mercury(void __iomem *hpa) { ** fixup_irq is to initialize PCI IRQ line support and ** virtualize pcidev->irq value. To be called by pci_fixup_bus(). */ -extern void *iosapic_register(unsigned long hpa); +extern void *iosapic_register(unsigned long hpa, void __iomem *vaddr); extern int iosapic_fixup_irq(void *obj, struct pci_dev *pcidev); #define LBA_FUNC_ID 0x0000 /* function id */ diff --git a/arch/parisc/include/asm/runway.h b/arch/parisc/include/asm/runway.h index 5cf061376ddb..2837f0223d6d 100644 --- a/arch/parisc/include/asm/runway.h +++ b/arch/parisc/include/asm/runway.h @@ -2,9 +2,6 @@ #ifndef ASM_PARISC_RUNWAY_H #define ASM_PARISC_RUNWAY_H -/* declared in arch/parisc/kernel/setup.c */ -extern struct proc_dir_entry * proc_runway_root; - #define RUNWAY_STATUS 0x10 #define RUNWAY_DEBUG 0x40 diff --git a/arch/parisc/include/asm/shmparam.h b/arch/parisc/include/asm/shmparam.h index 74f74e4d35b7..5a95b0f62b87 100644 --- a/arch/parisc/include/asm/shmparam.h +++ b/arch/parisc/include/asm/shmparam.h @@ -2,6 +2,21 @@ #ifndef _ASMPARISC_SHMPARAM_H #define _ASMPARISC_SHMPARAM_H +/* + * PA-RISC uses virtually indexed & physically tagged (VIPT) caches + * which has strict requirements when two pages to the same physical + * address are accessed through different mappings. Read the section + * "Address Aliasing" in the arch docs for more detail: + * PA-RISC 1.1 (page 3-6): + * https://parisc.wiki.kernel.org/images-parisc/6/68/Pa11_acd.pdf + * PA-RISC 2.0 (page F-5): + * https://parisc.wiki.kernel.org/images-parisc/7/73/Parisc2.0.pdf + * + * For Linux we allow kernel and userspace to map pages on page size + * granularity (SHMLBA) but have to ensure that, if two pages are + * mapped to the same physical address, the virtual and physical + * addresses modulo SHM_COLOUR are identical. + */ #define SHMLBA PAGE_SIZE /* attach addr a multiple of this */ #define SHM_COLOUR 0x00400000 /* shared mappings colouring */ diff --git a/arch/parisc/include/asm/signal.h b/arch/parisc/include/asm/signal.h index 715c96ba2ec8..85c3d7409bbc 100644 --- a/arch/parisc/include/asm/signal.h +++ b/arch/parisc/include/asm/signal.h @@ -4,24 +4,12 @@ #include <uapi/asm/signal.h> -#define _NSIG 64 -/* bits-per-word, where word apparently means 'long' not 'int' */ -#define _NSIG_BPW BITS_PER_LONG -#define _NSIG_WORDS (_NSIG / _NSIG_BPW) - -# ifndef __ASSEMBLY__ +# ifndef __ASSEMBLER__ /* Most things should be clean enough to redefine this at will, if care is taken to make libc match. */ -typedef unsigned long old_sigset_t; /* at least 32 bits */ - -typedef struct { - /* next_signal() assumes this is a long - no choice */ - unsigned long sig[_NSIG_WORDS]; -} sigset_t; - #include <asm/sigcontext.h> -#endif /* !__ASSEMBLY */ +#endif /* !__ASSEMBLER__ */ #endif /* _ASM_PARISC_SIGNAL_H */ diff --git a/arch/parisc/include/asm/smp.h b/arch/parisc/include/asm/smp.h index 94d1f21ce99a..0cf1c3a2696a 100644 --- a/arch/parisc/include/asm/smp.h +++ b/arch/parisc/include/asm/smp.h @@ -12,7 +12,7 @@ extern int init_per_cpu(int cpuid); #define PDC_OS_BOOT_RENDEZVOUS 0x10 #define PDC_OS_BOOT_RENDEZVOUS_HI 0x28 -#ifndef ASSEMBLY +#ifndef __ASSEMBLER__ #include <linux/bitops.h> #include <linux/threads.h> /* for NR_CPUS */ #include <linux/cpumask.h> @@ -34,7 +34,7 @@ extern void arch_send_call_function_ipi_mask(const struct cpumask *mask); #define raw_smp_processor_id() (current_thread_info()->cpu) -#endif /* !ASSEMBLY */ +#endif /* !__ASSEMBLER__ */ #else /* CONFIG_SMP */ diff --git a/arch/parisc/include/asm/special_insns.h b/arch/parisc/include/asm/special_insns.h index c822bd0c0e3c..1013eeba31e5 100644 --- a/arch/parisc/include/asm/special_insns.h +++ b/arch/parisc/include/asm/special_insns.h @@ -8,7 +8,8 @@ "copy %%r0,%0\n" \ "8:\tlpa %%r0(%1),%0\n" \ "9:\n" \ - ASM_EXCEPTIONTABLE_ENTRY(8b, 9b) \ + ASM_EXCEPTIONTABLE_ENTRY(8b, 9b, \ + "or %%r0,%%r0,%%r0") \ : "=&r" (pa) \ : "r" (va) \ : "memory" \ @@ -22,7 +23,8 @@ "copy %%r0,%0\n" \ "8:\tlpa %%r0(%%sr3,%1),%0\n" \ "9:\n" \ - ASM_EXCEPTIONTABLE_ENTRY(8b, 9b) \ + ASM_EXCEPTIONTABLE_ENTRY(8b, 9b, \ + "or %%r0,%%r0,%%r0") \ : "=&r" (pa) \ : "r" (va) \ : "memory" \ @@ -30,6 +32,34 @@ pa; \ }) +/** + * prober_user() - Probe user read access + * @sr: Space regster. + * @va: Virtual address. + * + * Return: Non-zero if address is accessible. + * + * Due to the way _PAGE_READ is handled in TLB entries, we need + * a special check to determine whether a user address is accessible. + * The ldb instruction does the initial access check. If it is + * successful, the probe instruction checks user access rights. + */ +#define prober_user(sr, va) ({ \ + unsigned long read_allowed; \ + __asm__ __volatile__( \ + "copy %%r0,%0\n" \ + "8:\tldb 0(%%sr%1,%2),%%r0\n" \ + "\tproberi (%%sr%1,%2),%3,%0\n" \ + "9:\n" \ + ASM_EXCEPTIONTABLE_ENTRY(8b, 9b, \ + "or %%r0,%%r0,%%r0") \ + : "=&r" (read_allowed) \ + : "i" (sr), "r" (va), "i" (PRIV_USER) \ + : "memory" \ + ); \ + read_allowed; \ +}) + #define CR_EIEM 15 /* External Interrupt Enable Mask */ #define CR_CR16 16 /* CR16 Interval Timer */ #define CR_EIRR 23 /* External Interrupt Request Register */ diff --git a/arch/parisc/include/asm/spinlock.h b/arch/parisc/include/asm/spinlock.h index a6e5d66a7656..0b326e52255e 100644 --- a/arch/parisc/include/asm/spinlock.h +++ b/arch/parisc/include/asm/spinlock.h @@ -7,10 +7,24 @@ #include <asm/processor.h> #include <asm/spinlock_types.h> +static inline void arch_spin_val_check(int lock_val) +{ + if (IS_ENABLED(CONFIG_LIGHTWEIGHT_SPINLOCK_CHECK)) + asm volatile( "andcm,= %0,%1,%%r0\n" + ".word %2\n" + : : "r" (lock_val), "r" (__ARCH_SPIN_LOCK_UNLOCKED_VAL), + "i" (SPINLOCK_BREAK_INSN)); +} + static inline int arch_spin_is_locked(arch_spinlock_t *x) { - volatile unsigned int *a = __ldcw_align(x); - return READ_ONCE(*a) == 0; + volatile unsigned int *a; + int lock_val; + + a = __ldcw_align(x); + lock_val = READ_ONCE(*a); + arch_spin_val_check(lock_val); + return (lock_val == 0); } static inline void arch_spin_lock(arch_spinlock_t *x) @@ -18,9 +32,18 @@ static inline void arch_spin_lock(arch_spinlock_t *x) volatile unsigned int *a; a = __ldcw_align(x); - while (__ldcw(a) == 0) + do { + int lock_val_old; + + lock_val_old = __ldcw(a); + arch_spin_val_check(lock_val_old); + if (lock_val_old) + return; /* got lock */ + + /* wait until we should try to get lock again */ while (*a == 0) continue; + } while (1); } static inline void arch_spin_unlock(arch_spinlock_t *x) @@ -29,15 +52,19 @@ static inline void arch_spin_unlock(arch_spinlock_t *x) a = __ldcw_align(x); /* Release with ordered store. */ - __asm__ __volatile__("stw,ma %0,0(%1)" : : "r"(1), "r"(a) : "memory"); + __asm__ __volatile__("stw,ma %0,0(%1)" + : : "r"(__ARCH_SPIN_LOCK_UNLOCKED_VAL), "r"(a) : "memory"); } static inline int arch_spin_trylock(arch_spinlock_t *x) { volatile unsigned int *a; + int lock_val; a = __ldcw_align(x); - return __ldcw(a) != 0; + lock_val = __ldcw(a); + arch_spin_val_check(lock_val); + return lock_val != 0; } /* diff --git a/arch/parisc/include/asm/spinlock_types.h b/arch/parisc/include/asm/spinlock_types.h index ca39ee350c3f..8e6889bc23cc 100644 --- a/arch/parisc/include/asm/spinlock_types.h +++ b/arch/parisc/include/asm/spinlock_types.h @@ -2,14 +2,17 @@ #ifndef __ASM_SPINLOCK_TYPES_H #define __ASM_SPINLOCK_TYPES_H +#define __ARCH_SPIN_LOCK_UNLOCKED_VAL 0x1a46 + +#define SPINLOCK_BREAK_INSN 0x0000c006 /* break 6,6 */ + +#ifndef __ASSEMBLER__ + typedef struct { -#ifdef CONFIG_PA20 - volatile unsigned int slock; -# define __ARCH_SPIN_LOCK_UNLOCKED { 1 } -#else volatile unsigned int lock[4]; -# define __ARCH_SPIN_LOCK_UNLOCKED { { 1, 1, 1, 1 } } -#endif +# define __ARCH_SPIN_LOCK_UNLOCKED \ + { { __ARCH_SPIN_LOCK_UNLOCKED_VAL, __ARCH_SPIN_LOCK_UNLOCKED_VAL, \ + __ARCH_SPIN_LOCK_UNLOCKED_VAL, __ARCH_SPIN_LOCK_UNLOCKED_VAL } } } arch_spinlock_t; @@ -23,6 +26,8 @@ typedef struct { volatile unsigned int counter; } arch_rwlock_t; +#endif /* __ASSEMBLER__ */ + #define __ARCH_RW_LOCK_UNLOCKED__ 0x01000000 #define __ARCH_RW_LOCK_UNLOCKED { .lock_mutex = __ARCH_SPIN_LOCK_UNLOCKED, \ .counter = __ARCH_RW_LOCK_UNLOCKED__ } diff --git a/arch/parisc/include/asm/syscall.h b/arch/parisc/include/asm/syscall.h index 00b127a5e09b..c11222798ab2 100644 --- a/arch/parisc/include/asm/syscall.h +++ b/arch/parisc/include/asm/syscall.h @@ -17,6 +17,13 @@ static inline long syscall_get_nr(struct task_struct *tsk, return regs->gr[20]; } +static inline void syscall_set_nr(struct task_struct *tsk, + struct pt_regs *regs, + int nr) +{ + regs->gr[20] = nr; +} + static inline void syscall_get_arguments(struct task_struct *tsk, struct pt_regs *regs, unsigned long *args) @@ -29,6 +36,18 @@ static inline void syscall_get_arguments(struct task_struct *tsk, args[0] = regs->gr[26]; } +static inline void syscall_set_arguments(struct task_struct *tsk, + struct pt_regs *regs, + unsigned long *args) +{ + regs->gr[21] = args[5]; + regs->gr[22] = args[4]; + regs->gr[23] = args[3]; + regs->gr[24] = args[2]; + regs->gr[25] = args[1]; + regs->gr[26] = args[0]; +} + static inline long syscall_get_error(struct task_struct *task, struct pt_regs *regs) { diff --git a/arch/parisc/include/asm/patch.h b/arch/parisc/include/asm/text-patching.h index 400d84c6e504..400d84c6e504 100644 --- a/arch/parisc/include/asm/patch.h +++ b/arch/parisc/include/asm/text-patching.h diff --git a/arch/parisc/include/asm/thread_info.h b/arch/parisc/include/asm/thread_info.h index 1a58795f785c..b283738bb6da 100644 --- a/arch/parisc/include/asm/thread_info.h +++ b/arch/parisc/include/asm/thread_info.h @@ -2,7 +2,7 @@ #ifndef _ASM_PARISC_THREAD_INFO_H #define _ASM_PARISC_THREAD_INFO_H -#ifndef __ASSEMBLY__ +#ifndef __ASSEMBLER__ #include <asm/processor.h> #include <asm/special_insns.h> @@ -20,7 +20,7 @@ struct thread_info { .preempt_count = INIT_PREEMPT_COUNT, \ } -#endif /* !__ASSEMBLY */ +#endif /* !__ASSEMBLER__ */ /* thread information allocation */ diff --git a/arch/parisc/include/asm/traps.h b/arch/parisc/include/asm/traps.h index 0ccdb738a9a3..10c8fb68e404 100644 --- a/arch/parisc/include/asm/traps.h +++ b/arch/parisc/include/asm/traps.h @@ -4,7 +4,7 @@ #define PARISC_ITLB_TRAP 6 /* defined by architecture. Do not change. */ -#if !defined(__ASSEMBLY__) +#if !defined(__ASSEMBLER__) struct pt_regs; /* traps.c */ diff --git a/arch/parisc/include/asm/uaccess.h b/arch/parisc/include/asm/uaccess.h index 2bf660eabe42..6c531d2c847e 100644 --- a/arch/parisc/include/asm/uaccess.h +++ b/arch/parisc/include/asm/uaccess.h @@ -7,6 +7,7 @@ */ #include <asm/page.h> #include <asm/cache.h> +#include <asm/extable.h> #include <linux/bug.h> #include <linux/string.h> @@ -26,36 +27,6 @@ #define STD_USER(sr, x, ptr) __put_user_asm(sr, "std", x, ptr) #endif -/* - * The exception table contains two values: the first is the relative offset to - * the address of the instruction that is allowed to fault, and the second is - * the relative offset to the address of the fixup routine. Since relative - * addresses are used, 32bit values are sufficient even on 64bit kernel. - */ - -#define ARCH_HAS_RELATIVE_EXTABLE -struct exception_table_entry { - int insn; /* relative address of insn that is allowed to fault. */ - int fixup; /* relative address of fixup routine */ -}; - -#define ASM_EXCEPTIONTABLE_ENTRY( fault_addr, except_addr )\ - ".section __ex_table,\"aw\"\n" \ - ".word (" #fault_addr " - .), (" #except_addr " - .)\n\t" \ - ".previous\n" - -/* - * ASM_EXCEPTIONTABLE_ENTRY_EFAULT() creates a special exception table entry - * (with lowest bit set) for which the fault handler in fixup_exception() will - * load -EFAULT into %r29 for a read or write fault, and zeroes the target - * register in case of a read fault in get_user(). - */ -#define ASM_EXCEPTIONTABLE_REG 29 -#define ASM_EXCEPTIONTABLE_VAR(__variable) \ - register long __variable __asm__ ("r29") = 0 -#define ASM_EXCEPTIONTABLE_ENTRY_EFAULT( fault_addr, except_addr )\ - ASM_EXCEPTIONTABLE_ENTRY( fault_addr, except_addr + 1) - #define __get_user_internal(sr, val, ptr) \ ({ \ ASM_EXCEPTIONTABLE_VAR(__gu_err); \ @@ -71,9 +42,24 @@ struct exception_table_entry { __gu_err; \ }) -#define __get_user(val, ptr) \ -({ \ - __get_user_internal(SR_USER, val, ptr); \ +#define __probe_user_internal(sr, error, ptr) \ +({ \ + __asm__("\tproberi (%%sr%1,%2),%3,%0\n" \ + "\tcmpiclr,= 1,%0,%0\n" \ + "\tldi %4,%0\n" \ + : "=r"(error) \ + : "i"(sr), "r"(ptr), "i"(PRIV_USER), \ + "i"(-EFAULT)); \ +}) + +#define __get_user(val, ptr) \ +({ \ + register long __gu_err; \ + \ + __gu_err = __get_user_internal(SR_USER, val, ptr); \ + if (likely(!__gu_err)) \ + __probe_user_internal(SR_USER, __gu_err, ptr); \ + __gu_err; \ }) #define __get_user_asm(sr, val, ldx, ptr) \ @@ -82,7 +68,7 @@ struct exception_table_entry { \ __asm__("1: " ldx " 0(%%sr%2,%3),%0\n" \ "9:\n" \ - ASM_EXCEPTIONTABLE_ENTRY_EFAULT(1b, 9b) \ + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(1b, 9b, "%1") \ : "=r"(__gu_val), "+r"(__gu_err) \ : "i"(sr), "r"(ptr)); \ \ @@ -114,8 +100,8 @@ struct exception_table_entry { "1: ldw 0(%%sr%2,%3),%0\n" \ "2: ldw 4(%%sr%2,%3),%R0\n" \ "9:\n" \ - ASM_EXCEPTIONTABLE_ENTRY_EFAULT(1b, 9b) \ - ASM_EXCEPTIONTABLE_ENTRY_EFAULT(2b, 9b) \ + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(1b, 9b, "%1") \ + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(2b, 9b, "%1") \ : "=&r"(__gu_tmp.l), "+r"(__gu_err) \ : "i"(sr), "r"(ptr)); \ \ @@ -173,7 +159,7 @@ struct exception_table_entry { __asm__ __volatile__ ( \ "1: " stx " %1,0(%%sr%2,%3)\n" \ "9:\n" \ - ASM_EXCEPTIONTABLE_ENTRY_EFAULT(1b, 9b) \ + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(1b, 9b, "%0") \ : "+r"(__pu_err) \ : "r"(x), "i"(sr), "r"(ptr)) @@ -185,15 +171,14 @@ struct exception_table_entry { "1: stw %1,0(%%sr%2,%3)\n" \ "2: stw %R1,4(%%sr%2,%3)\n" \ "9:\n" \ - ASM_EXCEPTIONTABLE_ENTRY_EFAULT(1b, 9b) \ - ASM_EXCEPTIONTABLE_ENTRY_EFAULT(2b, 9b) \ + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(1b, 9b, "%0") \ + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(2b, 9b, "%0") \ : "+r"(__pu_err) \ : "r"(__val), "i"(sr), "r"(ptr)); \ } while (0) #endif /* !defined(CONFIG_64BIT) */ - /* * Complex access routines -- external declarations */ @@ -215,7 +200,4 @@ unsigned long __must_check raw_copy_from_user(void *dst, const void __user *src, #define INLINE_COPY_TO_USER #define INLINE_COPY_FROM_USER -struct pt_regs; -int fixup_exception(struct pt_regs *regs); - #endif /* __PARISC_UACCESS_H */ diff --git a/arch/parisc/include/asm/unaligned.h b/arch/parisc/include/asm/unaligned.h deleted file mode 100644 index c0621295100d..000000000000 --- a/arch/parisc/include/asm/unaligned.h +++ /dev/null @@ -1,11 +0,0 @@ -/* SPDX-License-Identifier: GPL-2.0 */ -#ifndef _ASM_PARISC_UNALIGNED_H -#define _ASM_PARISC_UNALIGNED_H - -#include <asm-generic/unaligned.h> - -struct pt_regs; -void handle_unaligned(struct pt_regs *regs); -int check_unaligned(struct pt_regs *regs); - -#endif /* _ASM_PARISC_UNALIGNED_H */ diff --git a/arch/parisc/include/asm/unistd.h b/arch/parisc/include/asm/unistd.h index e38f9a90ac15..3e46c6ea9df6 100644 --- a/arch/parisc/include/asm/unistd.h +++ b/arch/parisc/include/asm/unistd.h @@ -6,7 +6,7 @@ #define __NR_Linux_syscalls __NR_syscalls -#ifndef __ASSEMBLY__ +#ifndef __ASSEMBLER__ #define SYS_ify(syscall_name) __NR_##syscall_name @@ -20,7 +20,7 @@ * sysdeps/unix/sysv/linux/hppa/sysdep.h */ -#ifdef PIC +#ifndef DONT_USE_PIC /* WARNING: CANNOT BE USED IN A NOP! */ # define K_STW_ASM_PIC " copy %%r19, %%r4\n" # define K_LDW_ASM_PIC " copy %%r4, %%r19\n" @@ -43,7 +43,7 @@ across the syscall. */ #define K_CALL_CLOB_REGS "%r1", "%r2", K_USING_GR4 \ - "%r20", "%r29", "%r31" + "%r20", "%r29", "%r31" #undef K_INLINE_SYSCALL #define K_INLINE_SYSCALL(name, nr, args...) ({ \ @@ -58,7 +58,7 @@ " ldi %1, %%r20\n" \ K_LDW_ASM_PIC \ : "=r" (__res) \ - : "i" (SYS_ify(name)) K_ASM_ARGS_##nr \ + : "i" (name) K_ASM_ARGS_##nr \ : "memory", K_CALL_CLOB_REGS K_CLOB_ARGS_##nr \ ); \ __sys_res = (long)__res; \ @@ -104,42 +104,18 @@ #define K_CLOB_ARGS_1 K_CLOB_ARGS_2, "%r25" #define K_CLOB_ARGS_0 K_CLOB_ARGS_1, "%r26" -#define _syscall0(type,name) \ -type name(void) \ -{ \ - return K_INLINE_SYSCALL(name, 0); \ -} - -#define _syscall1(type,name,type1,arg1) \ -type name(type1 arg1) \ -{ \ - return K_INLINE_SYSCALL(name, 1, arg1); \ -} - -#define _syscall2(type,name,type1,arg1,type2,arg2) \ -type name(type1 arg1, type2 arg2) \ -{ \ - return K_INLINE_SYSCALL(name, 2, arg1, arg2); \ -} - -#define _syscall3(type,name,type1,arg1,type2,arg2,type3,arg3) \ -type name(type1 arg1, type2 arg2, type3 arg3) \ -{ \ - return K_INLINE_SYSCALL(name, 3, arg1, arg2, arg3); \ -} - -#define _syscall4(type,name,type1,arg1,type2,arg2,type3,arg3,type4,arg4) \ -type name(type1 arg1, type2 arg2, type3 arg3, type4 arg4) \ -{ \ - return K_INLINE_SYSCALL(name, 4, arg1, arg2, arg3, arg4); \ -} - -/* select takes 5 arguments */ -#define _syscall5(type,name,type1,arg1,type2,arg2,type3,arg3,type4,arg4,type5,arg5) \ -type name(type1 arg1, type2 arg2, type3 arg3, type4 arg4, type5 arg5) \ -{ \ - return K_INLINE_SYSCALL(name, 5, arg1, arg2, arg3, arg4, arg5); \ -} +#define syscall0(name) \ + K_INLINE_SYSCALL(name, 0) +#define syscall1(name, arg1) \ + K_INLINE_SYSCALL(name, 1, arg1) +#define syscall2(name, arg1, arg2) \ + K_INLINE_SYSCALL(name, 2, arg1, arg2) +#define syscall3(name, arg1, arg2, arg3) \ + K_INLINE_SYSCALL(name, 3, arg1, arg2, arg3) +#define syscall4(name, arg1, arg2, arg3, arg4) \ + K_INLINE_SYSCALL(name, 4, arg1, arg2, arg3, arg4) +#define syscall5(name, arg1, arg2, arg3, arg4, arg5) \ + K_INLINE_SYSCALL(name, 5, arg1, arg2, arg3, arg4, arg5) #define __ARCH_WANT_NEW_STAT #define __ARCH_WANT_STAT64 @@ -160,7 +136,6 @@ type name(type1 arg1, type2 arg2, type3 arg3, type4 arg4, type5 arg5) \ #define __ARCH_WANT_SYS_FORK #define __ARCH_WANT_SYS_VFORK #define __ARCH_WANT_SYS_CLONE -#define __ARCH_WANT_SYS_CLONE3 #define __ARCH_WANT_COMPAT_SYS_SENDFILE #define __ARCH_WANT_COMPAT_STAT @@ -169,7 +144,7 @@ type name(type1 arg1, type2 arg2, type3 arg3, type4 arg4, type5 arg5) \ #define __ARCH_WANT_SYS_UTIME #endif -#endif /* __ASSEMBLY__ */ +#endif /* __ASSEMBLER__ */ #undef STR diff --git a/arch/parisc/include/asm/vdso.h b/arch/parisc/include/asm/vdso.h index ef8206193f82..81bc1d42802a 100644 --- a/arch/parisc/include/asm/vdso.h +++ b/arch/parisc/include/asm/vdso.h @@ -2,7 +2,7 @@ #ifndef __PARISC_VDSO_H__ #define __PARISC_VDSO_H__ -#ifndef __ASSEMBLY__ +#ifndef __ASSEMBLER__ #ifdef CONFIG_64BIT #include <generated/vdso64-offsets.h> @@ -12,13 +12,11 @@ #define VDSO64_SYMBOL(tsk, name) ((tsk)->mm->context.vdso_base + (vdso64_offset_##name)) #define VDSO32_SYMBOL(tsk, name) ((tsk)->mm->context.vdso_base + (vdso32_offset_##name)) -extern struct vdso_data *vdso_data; - -#endif /* __ASSEMBLY __ */ +#endif /* __ASSEMBLER__ */ /* Default link addresses for the vDSOs */ #define VDSO_LBASE 0 -#define VDSO_VERSION_STRING LINUX_5.18 +#define VDSO_VERSION_STRING LINUX_6.11 #endif /* __PARISC_VDSO_H__ */ diff --git a/arch/parisc/include/asm/video.h b/arch/parisc/include/asm/video.h new file mode 100644 index 000000000000..a9d50ebd6e76 --- /dev/null +++ b/arch/parisc/include/asm/video.h @@ -0,0 +1,16 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef _ASM_VIDEO_H_ +#define _ASM_VIDEO_H_ + +#include <linux/types.h> + +struct device; + +#if defined(CONFIG_STI_CORE) && defined(CONFIG_VIDEO) +bool video_is_primary_device(struct device *dev); +#define video_is_primary_device video_is_primary_device +#endif + +#include <asm-generic/video.h> + +#endif /* _ASM_VIDEO_H_ */ diff --git a/arch/parisc/include/uapi/asm/cachectl.h b/arch/parisc/include/uapi/asm/cachectl.h new file mode 100644 index 000000000000..68d6b455498b --- /dev/null +++ b/arch/parisc/include/uapi/asm/cachectl.h @@ -0,0 +1,12 @@ +/* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note */ +#ifndef _ASM_CACHECTL +#define _ASM_CACHECTL + +/* + * Options for cacheflush system call + */ +#define ICACHE (1<<0) /* flush instruction cache */ +#define DCACHE (1<<1) /* writeback and flush data cache */ +#define BCACHE (ICACHE|DCACHE) /* flush both caches */ + +#endif /* _ASM_CACHECTL */ diff --git a/arch/parisc/include/uapi/asm/errno.h b/arch/parisc/include/uapi/asm/errno.h index 87245c584784..8d94739d75c6 100644 --- a/arch/parisc/include/uapi/asm/errno.h +++ b/arch/parisc/include/uapi/asm/errno.h @@ -75,7 +75,6 @@ /* We now return you to your regularly scheduled HPUX. */ -#define ENOSYM 215 /* symbol does not exist in executable */ #define ENOTSOCK 216 /* Socket operation on non-socket */ #define EDESTADDRREQ 217 /* Destination address required */ #define EMSGSIZE 218 /* Message too long */ @@ -101,7 +100,6 @@ #define ETIMEDOUT 238 /* Connection timed out */ #define ECONNREFUSED 239 /* Connection refused */ #define EREFUSED ECONNREFUSED /* for HP's NFS apparently */ -#define EREMOTERELEASE 240 /* Remote peer released connection */ #define EHOSTDOWN 241 /* Host is down */ #define EHOSTUNREACH 242 /* No route to host */ diff --git a/arch/parisc/include/uapi/asm/ioctls.h b/arch/parisc/include/uapi/asm/ioctls.h index 82d1148c6379..74b4027a4e80 100644 --- a/arch/parisc/include/uapi/asm/ioctls.h +++ b/arch/parisc/include/uapi/asm/ioctls.h @@ -10,10 +10,10 @@ #define TCSETS _IOW('T', 17, struct termios) /* TCSETATTR */ #define TCSETSW _IOW('T', 18, struct termios) /* TCSETATTRD */ #define TCSETSF _IOW('T', 19, struct termios) /* TCSETATTRF */ -#define TCGETA _IOR('T', 1, struct termio) -#define TCSETA _IOW('T', 2, struct termio) -#define TCSETAW _IOW('T', 3, struct termio) -#define TCSETAF _IOW('T', 4, struct termio) +#define TCGETA 0x40125401 +#define TCSETA 0x80125402 +#define TCSETAW 0x80125403 +#define TCSETAF 0x80125404 #define TCSBRK _IO('T', 5) #define TCXONC _IO('T', 6) #define TCFLSH _IO('T', 7) diff --git a/arch/parisc/include/uapi/asm/mman.h b/arch/parisc/include/uapi/asm/mman.h index 68c44f99bc93..b6a709506987 100644 --- a/arch/parisc/include/uapi/asm/mman.h +++ b/arch/parisc/include/uapi/asm/mman.h @@ -75,6 +75,9 @@ #define MADV_HWPOISON 100 /* poison a page for testing */ #define MADV_SOFT_OFFLINE 101 /* soft offline page for testing */ +#define MADV_GUARD_INSTALL 102 /* fatal signal on access to range */ +#define MADV_GUARD_REMOVE 103 /* unguard range */ + /* compatibility flags */ #define MAP_FILE 0 diff --git a/arch/parisc/include/uapi/asm/pdc.h b/arch/parisc/include/uapi/asm/pdc.h index 7a90070136e8..65031ddf8372 100644 --- a/arch/parisc/include/uapi/asm/pdc.h +++ b/arch/parisc/include/uapi/asm/pdc.h @@ -58,8 +58,8 @@ #define PDC_MODEL_NVA_SUPPORTED (0 << 4) #define PDC_MODEL_NVA_SLOW (1 << 4) #define PDC_MODEL_NVA_UNSUPPORTED (3 << 4) -#define PDC_MODEL_GET_BOOT__OP 8 /* returns boot test options */ -#define PDC_MODEL_SET_BOOT__OP 9 /* set boot test options */ +#define PDC_MODEL_FIRM_TEST_GET 8 /* returns boot test options */ +#define PDC_MODEL_FIRM_TEST_SET 9 /* set boot test options */ #define PDC_MODEL_GET_PLATFORM_INFO 10 /* returns platform info */ #define PDC_MODEL_GET_INSTALL_KERNEL 11 /* returns kernel for installation */ @@ -361,7 +361,7 @@ /* size of the pdc_result buffer for firmware.c */ #define NUM_PDC_RESULT 32 -#if !defined(__ASSEMBLY__) +#if !defined(__ASSEMBLER__) /* flags for hardware_path */ #define PF_AUTOBOOT 0x80 @@ -472,6 +472,7 @@ struct pdc_model { /* for PDC_MODEL */ unsigned long arch_rev; unsigned long pot_key; unsigned long curr_key; + unsigned long width; /* default of PSW_W bit (1=enabled) */ }; struct pdc_cache_cf { /* for PDC_CACHE (I/D-caches) */ @@ -609,6 +610,12 @@ struct pdc_system_map_addr_info { /* PDC_SYSTEM_MAP/FIND_ADDRESS */ unsigned long mod_pgs; }; +struct pdc_relocate_info_block { /* PDC_RELOCATE_INFO */ + unsigned long pdc_size; + unsigned long pdc_alignment; + unsigned long pdc_address; +}; + struct pdc_initiator { /* PDC_INITIATOR */ int host_id; int factor; @@ -717,6 +724,23 @@ struct pdc_toc_pim_20 { struct pim_cpu_state_cf cpu_state; }; -#endif /* !defined(__ASSEMBLY__) */ +/* for SpeedyBoot/firm_ctl funtionality */ +struct pdc_firm_test_get_rtn_block { /* PDC_MODEL/PDC_FIRM_TEST_GET */ + unsigned long current_tests; /* u_R_addr Raddr_ints[0] */ + unsigned long tests_supported; /* u_R_addr Raddr_ints[1] */ + unsigned long default_tests; /* u_R_addr Raddr_ints[2] */ +}; + +#define TORNADO_CPU_ID 0xB +#define PCXL_CPU_ID 0xD +#define PCXU_CPU_ID 0xE /* U and U+ for all but C-class with bug */ +#define VR_CPU_ID 0xF +#define PCXU_PLUS_CPU_ID 0x10 /* U+ only on C-class with bug */ +#define PCXW_CPU_ID 0x11 +#define PCXW_PLUS_CPU_ID 0x12 +#define PIRANHA_CPU_ID 0x13 +#define MAKO_CPU_ID 0x14 + +#endif /* !defined(__ASSEMBLER__) */ #endif /* _UAPI_PARISC_PDC_H */ diff --git a/arch/parisc/include/uapi/asm/perf_regs.h b/arch/parisc/include/uapi/asm/perf_regs.h new file mode 100644 index 000000000000..1ae687bb3d3c --- /dev/null +++ b/arch/parisc/include/uapi/asm/perf_regs.h @@ -0,0 +1,63 @@ +/* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note */ +#ifndef _UAPI_ASM_PARISC_PERF_REGS_H +#define _UAPI_ASM_PARISC_PERF_REGS_H + +/* see struct user_regs_struct */ +enum perf_event_parisc_regs { + PERF_REG_PARISC_R0, /* PSW is in gr[0] */ + PERF_REG_PARISC_R1, + PERF_REG_PARISC_R2, + PERF_REG_PARISC_R3, + PERF_REG_PARISC_R4, + PERF_REG_PARISC_R5, + PERF_REG_PARISC_R6, + PERF_REG_PARISC_R7, + PERF_REG_PARISC_R8, + PERF_REG_PARISC_R9, + PERF_REG_PARISC_R10, + PERF_REG_PARISC_R11, + PERF_REG_PARISC_R12, + PERF_REG_PARISC_R13, + PERF_REG_PARISC_R14, + PERF_REG_PARISC_R15, + PERF_REG_PARISC_R16, + PERF_REG_PARISC_R17, + PERF_REG_PARISC_R18, + PERF_REG_PARISC_R19, + PERF_REG_PARISC_R20, + PERF_REG_PARISC_R21, + PERF_REG_PARISC_R22, + PERF_REG_PARISC_R23, + PERF_REG_PARISC_R24, + PERF_REG_PARISC_R25, + PERF_REG_PARISC_R26, + PERF_REG_PARISC_R27, + PERF_REG_PARISC_R28, + PERF_REG_PARISC_R29, + PERF_REG_PARISC_R30, + PERF_REG_PARISC_R31, + + PERF_REG_PARISC_SR0, + PERF_REG_PARISC_SR1, + PERF_REG_PARISC_SR2, + PERF_REG_PARISC_SR3, + PERF_REG_PARISC_SR4, + PERF_REG_PARISC_SR5, + PERF_REG_PARISC_SR6, + PERF_REG_PARISC_SR7, + + PERF_REG_PARISC_IAOQ0, + PERF_REG_PARISC_IAOQ1, + PERF_REG_PARISC_IASQ0, + PERF_REG_PARISC_IASQ1, + + PERF_REG_PARISC_SAR, /* CR11 */ + PERF_REG_PARISC_IIR, /* CR19 */ + PERF_REG_PARISC_ISR, /* CR20 */ + PERF_REG_PARISC_IOR, /* CR21 */ + PERF_REG_PARISC_IPSW, /* CR22 */ + + PERF_REG_PARISC_MAX +}; + +#endif /* _UAPI_ASM_PARISC_PERF_REGS_H */ diff --git a/arch/parisc/include/uapi/asm/signal.h b/arch/parisc/include/uapi/asm/signal.h index 8e4895c5ea5d..d99accf37341 100644 --- a/arch/parisc/include/uapi/asm/signal.h +++ b/arch/parisc/include/uapi/asm/signal.h @@ -57,10 +57,20 @@ #include <asm-generic/signal-defs.h> -# ifndef __ASSEMBLY__ +#define _NSIG 64 +#define _NSIG_BPW (sizeof(unsigned long) * 8) +#define _NSIG_WORDS (_NSIG / _NSIG_BPW) + +# ifndef __ASSEMBLER__ # include <linux/types.h> +typedef unsigned long old_sigset_t; /* at least 32 bits */ + +typedef struct { + unsigned long sig[_NSIG_WORDS]; +} sigset_t; + /* Avoid too many header ordering problems. */ struct siginfo; @@ -70,5 +80,5 @@ typedef struct sigaltstack { __kernel_size_t ss_size; } stack_t; -#endif /* !__ASSEMBLY */ +#endif /* !__ASSEMBLER__ */ #endif /* _UAPI_ASM_PARISC_SIGNAL_H */ diff --git a/arch/parisc/include/uapi/asm/socket.h b/arch/parisc/include/uapi/asm/socket.h index f486d3dfb6bb..c16ec36dfee6 100644 --- a/arch/parisc/include/uapi/asm/socket.h +++ b/arch/parisc/include/uapi/asm/socket.h @@ -129,6 +129,24 @@ #define SO_RCVMARK 0x4049 +#define SO_PASSPIDFD 0x404A +#define SO_PEERPIDFD 0x404B + +#define SCM_TS_OPT_ID 0x404C + +#define SO_RCVPRIORITY 0x404D + +#define SO_DEVMEM_LINEAR 0x404E +#define SCM_DEVMEM_LINEAR SO_DEVMEM_LINEAR +#define SO_DEVMEM_DMABUF 0x404F +#define SCM_DEVMEM_DMABUF SO_DEVMEM_DMABUF +#define SO_DEVMEM_DONTNEED 0x4050 + +#define SO_PASSRIGHTS 0x4051 + +#define SO_INQ 0x4052 +#define SCM_INQ SO_INQ + #if !defined(__KERNEL__) #if __BITS_PER_LONG == 64 diff --git a/arch/parisc/install.sh b/arch/parisc/install.sh index 933d031c249a..664c2d77f776 100755 --- a/arch/parisc/install.sh +++ b/arch/parisc/install.sh @@ -16,6 +16,8 @@ # $3 - kernel map file # $4 - default install path (blank if root directory) +set -e + if [ "$(basename $2)" = "vmlinuz" ]; then # Compressed install echo "Installing compressed kernel" diff --git a/arch/parisc/kernel/Makefile b/arch/parisc/kernel/Makefile index 3d138c9cf9ce..9157bc8bdf41 100644 --- a/arch/parisc/kernel/Makefile +++ b/arch/parisc/kernel/Makefile @@ -3,10 +3,10 @@ # Makefile for arch/parisc/kernel # -extra-y := vmlinux.lds +always-$(KBUILD_BUILTIN) := vmlinux.lds obj-y := head.o cache.o pacache.o setup.o pdt.o traps.o time.o irq.o \ - pa7300lc.o syscall.o entry.o sys_parisc.o firmware.o \ + syscall.o entry.o sys_parisc.o firmware.o \ ptrace.o hardware.o inventory.o drivers.o alternative.o \ signal.o hpmc.o real2.o parisc_ksyms.o unaligned.o \ process.o processor.o pdc_cons.o pdc_chassis.o unwind.o \ @@ -21,6 +21,9 @@ CFLAGS_REMOVE_unwind.o = $(CC_FLAGS_FTRACE) CFLAGS_REMOVE_patch.o = $(CC_FLAGS_FTRACE) endif +CFLAGS_REMOVE_sys_parisc.o = -Wmissing-prototypes -Wmissing-declarations +CFLAGS_REMOVE_sys_parisc32.o = -Wmissing-prototypes -Wmissing-declarations + obj-$(CONFIG_SMP) += smp.o obj-$(CONFIG_PA11) += pci-dma.o obj-$(CONFIG_PCI) += pci.o @@ -35,6 +38,7 @@ obj-$(CONFIG_GENERIC_ARCH_TOPOLOGY) += topology.o obj-$(CONFIG_FUNCTION_TRACER) += ftrace.o obj-$(CONFIG_FUNCTION_GRAPH_TRACER) += ftrace.o obj-$(CONFIG_JUMP_LABEL) += jump_label.o +obj-$(CONFIG_PERF_EVENTS) += perf_event.o perf_regs.o obj-$(CONFIG_KGDB) += kgdb.o obj-$(CONFIG_KPROBES) += kprobes.o obj-$(CONFIG_KEXEC_CORE) += kexec.o relocate_kernel.o diff --git a/arch/parisc/kernel/alternative.c b/arch/parisc/kernel/alternative.c index 66f5672c70bd..25c4d6c3375d 100644 --- a/arch/parisc/kernel/alternative.c +++ b/arch/parisc/kernel/alternative.c @@ -25,7 +25,7 @@ void __init_or_module apply_alternatives(struct alt_instr *start, { struct alt_instr *entry; int index = 0, applied = 0; - int num_cpus = num_online_cpus(); + int num_cpus = num_present_cpus(); u16 cond_check; cond_check = ALT_COND_ALWAYS | diff --git a/arch/parisc/kernel/asm-offsets.c b/arch/parisc/kernel/asm-offsets.c index 94652e13c260..3de4b5933b10 100644 --- a/arch/parisc/kernel/asm-offsets.c +++ b/arch/parisc/kernel/asm-offsets.c @@ -13,6 +13,7 @@ * Copyright (C) 2002 Randolph Chung <tausq with parisc-linux.org> * Copyright (C) 2003 James Bottomley <jejb at parisc-linux.org> */ +#define COMPILE_OFFSETS #include <linux/types.h> #include <linux/sched.h> @@ -257,6 +258,8 @@ int main(void) BLANK(); DEFINE(TIF_BLOCKSTEP_PA_BIT, 31-TIF_BLOCKSTEP); DEFINE(TIF_SINGLESTEP_PA_BIT, 31-TIF_SINGLESTEP); + DEFINE(TIF_32BIT_PA_BIT, 31-TIF_32BIT); + BLANK(); DEFINE(ASM_PMD_SHIFT, PMD_SHIFT); DEFINE(ASM_PGDIR_SHIFT, PGDIR_SHIFT); @@ -275,6 +278,8 @@ int main(void) * and kernel data on physical huge pages */ #ifdef CONFIG_HUGETLB_PAGE DEFINE(HUGEPAGE_SIZE, 1UL << REAL_HPAGE_SHIFT); +#elif !defined(CONFIG_64BIT) + DEFINE(HUGEPAGE_SIZE, 4*1024*1024); #else DEFINE(HUGEPAGE_SIZE, PAGE_SIZE); #endif diff --git a/arch/parisc/kernel/audit.c b/arch/parisc/kernel/audit.c index f420b5552140..375cd73b5281 100644 --- a/arch/parisc/kernel/audit.c +++ b/arch/parisc/kernel/audit.c @@ -40,11 +40,6 @@ int audit_classify_arch(int arch) int audit_classify_syscall(int abi, unsigned syscall) { -#ifdef CONFIG_COMPAT - extern int parisc32_classify_syscall(unsigned); - if (abi == AUDIT_ARCH_PARISC) - return parisc32_classify_syscall(syscall); -#endif switch (syscall) { case __NR_open: return AUDITSC_OPEN; @@ -55,6 +50,10 @@ int audit_classify_syscall(int abi, unsigned syscall) case __NR_openat2: return AUDITSC_OPENAT2; default: +#ifdef CONFIG_COMPAT + if (abi == AUDIT_ARCH_PARISC) + return AUDITSC_COMPAT; +#endif return AUDITSC_NATIVE; } } diff --git a/arch/parisc/kernel/cache.c b/arch/parisc/kernel/cache.c index 1d3b8bc8a623..4c5240d3a3c7 100644 --- a/arch/parisc/kernel/cache.c +++ b/arch/parisc/kernel/cache.c @@ -19,6 +19,8 @@ #include <linux/pagemap.h> #include <linux/sched.h> #include <linux/sched/mm.h> +#include <linux/syscalls.h> +#include <linux/vmalloc.h> #include <asm/pdc.h> #include <asm/cache.h> #include <asm/cacheflush.h> @@ -28,21 +30,33 @@ #include <asm/sections.h> #include <asm/shmparam.h> #include <asm/mmu_context.h> +#include <asm/cachectl.h> + +#define PTR_PAGE_ALIGN_DOWN(addr) PTR_ALIGN_DOWN(addr, PAGE_SIZE) + +/* + * When nonzero, use _PAGE_ACCESSED bit to try to reduce the number + * of page flushes done flush_cache_page_if_present. There are some + * pros and cons in using this option. It may increase the risk of + * random segmentation faults. + */ +#define CONFIG_FLUSH_PAGE_ACCESSED 0 int split_tlb __ro_after_init; int dcache_stride __ro_after_init; int icache_stride __ro_after_init; EXPORT_SYMBOL(dcache_stride); +/* Internal implementation in arch/parisc/kernel/pacache.S */ void flush_dcache_page_asm(unsigned long phys_addr, unsigned long vaddr); EXPORT_SYMBOL(flush_dcache_page_asm); void purge_dcache_page_asm(unsigned long phys_addr, unsigned long vaddr); void flush_icache_page_asm(unsigned long phys_addr, unsigned long vaddr); - -/* Internal implementation in arch/parisc/kernel/pacache.S */ void flush_data_cache_local(void *); /* flushes local data-cache only */ void flush_instruction_cache_local(void); /* flushes local code-cache only */ +static void flush_kernel_dcache_page_addr(const void *addr); + /* On some machines (i.e., ones with the Merced bus), there can be * only a single PxTLB broadcast at a time; this must be guaranteed * by software. We need a spinlock around all TLB flushes to ensure @@ -56,7 +70,7 @@ int pa_serialize_tlb_flushes __ro_after_init; struct pdc_cache_info cache_info __ro_after_init; #ifndef CONFIG_PA20 -static struct pdc_btlb_info btlb_info __ro_after_init; +struct pdc_btlb_info btlb_info; #endif DEFINE_STATIC_KEY_TRUE(parisc_has_cache); @@ -92,11 +106,11 @@ static inline void flush_data_cache(void) /* Kernel virtual address of pfn. */ #define pfn_va(pfn) __va(PFN_PHYS(pfn)) -void -__update_cache(pte_t pte) +void __update_cache(pte_t pte) { unsigned long pfn = pte_pfn(pte); - struct page *page; + struct folio *folio; + unsigned int nr; /* We don't have pte special. As a result, we can be called with an invalid pfn and we don't need to flush the kernel dcache page. @@ -104,13 +118,17 @@ __update_cache(pte_t pte) if (!pfn_valid(pfn)) return; - page = pfn_to_page(pfn); - if (page_mapping_file(page) && - test_bit(PG_dcache_dirty, &page->flags)) { - flush_kernel_dcache_page_addr(pfn_va(pfn)); - clear_bit(PG_dcache_dirty, &page->flags); + folio = page_folio(pfn_to_page(pfn)); + pfn = folio_pfn(folio); + nr = folio_nr_pages(folio); + if (folio_flush_mapping(folio) && + test_bit(PG_dcache_dirty, &folio->flags.f)) { + while (nr--) + flush_kernel_dcache_page_addr(pfn_va(pfn + nr)); + clear_bit(PG_dcache_dirty, &folio->flags.f); } else if (parisc_requires_coherency()) - flush_kernel_dcache_page_addr(pfn_va(pfn)); + while (nr--) + flush_kernel_dcache_page_addr(pfn_va(pfn + nr)); } void @@ -258,11 +276,9 @@ parisc_cache_init(void) icache_stride = CAFL_STRIDE(cache_info.ic_conf); #undef CAFL_STRIDE -#ifndef CONFIG_PA20 - if (pdc_btlb_info(&btlb_info) < 0) { - memset(&btlb_info, 0, sizeof btlb_info); - } -#endif + /* stride needs to be non-zero, otherwise cache flushes will not work */ + WARN_ON(cache_info.dc_size && dcache_stride == 0); + WARN_ON(cache_info.ic_size && icache_stride == 0); if ((boot_cpu_data.pdc.capabilities & PDC_MODEL_NVA_MASK) == PDC_MODEL_NVA_UNSUPPORTED) { @@ -317,6 +333,18 @@ __flush_cache_page(struct vm_area_struct *vma, unsigned long vmaddr, { if (!static_branch_likely(&parisc_has_cache)) return; + + /* + * The TLB is the engine of coherence on parisc. The CPU is + * entitled to speculate any page with a TLB mapping, so here + * we kill the mapping then flush the page along a special flush + * only alias mapping. This guarantees that the page is no-longer + * in the cache for any process and nor may it be speculatively + * read in (until the user or kernel specifically accesses it, + * of course). + */ + flush_tlb_page(vma, vmaddr); + preempt_disable(); flush_dcache_page_asm(physaddr, vmaddr); if (vma->vm_flags & VM_EXEC) @@ -324,46 +352,61 @@ __flush_cache_page(struct vm_area_struct *vma, unsigned long vmaddr, preempt_enable(); } -static void flush_user_cache_page(struct vm_area_struct *vma, unsigned long vmaddr) +static void flush_kernel_dcache_page_addr(const void *addr) { - unsigned long flags, space, pgd, prot; -#ifdef CONFIG_TLB_PTLOCK - unsigned long pgd_lock; -#endif + unsigned long vaddr = (unsigned long)addr; + unsigned long flags; - vmaddr &= PAGE_MASK; + /* Purge TLB entry to remove translation on all CPUs */ + purge_tlb_start(flags); + pdtlb(SR_KERNEL, addr); + purge_tlb_end(flags); + /* Use tmpalias flush to prevent data cache move-in */ preempt_disable(); + flush_dcache_page_asm(__pa(vaddr), vaddr); + preempt_enable(); +} - /* Set context for flush */ - local_irq_save(flags); - prot = mfctl(8); - space = mfsp(SR_USER); - pgd = mfctl(25); -#ifdef CONFIG_TLB_PTLOCK - pgd_lock = mfctl(28); -#endif - switch_mm_irqs_off(NULL, vma->vm_mm, NULL); - local_irq_restore(flags); - - flush_user_dcache_range_asm(vmaddr, vmaddr + PAGE_SIZE); - if (vma->vm_flags & VM_EXEC) - flush_user_icache_range_asm(vmaddr, vmaddr + PAGE_SIZE); - flush_tlb_page(vma, vmaddr); +static void flush_kernel_icache_page_addr(const void *addr) +{ + unsigned long vaddr = (unsigned long)addr; + unsigned long flags; - /* Restore previous context */ - local_irq_save(flags); -#ifdef CONFIG_TLB_PTLOCK - mtctl(pgd_lock, 28); -#endif - mtctl(pgd, 25); - mtsp(space, SR_USER); - mtctl(prot, 8); - local_irq_restore(flags); + /* Purge TLB entry to remove translation on all CPUs */ + purge_tlb_start(flags); + pdtlb(SR_KERNEL, addr); + purge_tlb_end(flags); + /* Use tmpalias flush to prevent instruction cache move-in */ + preempt_disable(); + flush_icache_page_asm(__pa(vaddr), vaddr); preempt_enable(); } +void kunmap_flush_on_unmap(const void *addr) +{ + flush_kernel_dcache_page_addr(addr); +} +EXPORT_SYMBOL(kunmap_flush_on_unmap); + +void flush_icache_pages(struct vm_area_struct *vma, struct page *page, + unsigned int nr) +{ + void *kaddr = page_address(page); + + for (;;) { + flush_kernel_dcache_page_addr(kaddr); + flush_kernel_icache_page_addr(kaddr); + if (--nr == 0) + break; + kaddr += PAGE_SIZE; + } +} + +/* + * Walk page directory for MM to find PTEP pointer for address ADDR. + */ static inline pte_t *get_ptep(struct mm_struct *mm, unsigned long addr) { pte_t *ptep = NULL; @@ -386,32 +429,71 @@ static inline pte_t *get_ptep(struct mm_struct *mm, unsigned long addr) return ptep; } -static inline bool pte_needs_flush(pte_t pte) +static inline bool pte_needs_cache_flush(pte_t pte) { return (pte_val(pte) & (_PAGE_PRESENT | _PAGE_ACCESSED | _PAGE_NO_CACHE)) == (_PAGE_PRESENT | _PAGE_ACCESSED); } -void flush_dcache_page(struct page *page) +/* + * Return user physical address. Returns 0 if page is not present. + */ +static inline unsigned long get_upa(struct mm_struct *mm, unsigned long addr) +{ + unsigned long flags, space, pgd, prot, pa; +#ifdef CONFIG_TLB_PTLOCK + unsigned long pgd_lock; +#endif + + /* Save context */ + local_irq_save(flags); + prot = mfctl(8); + space = mfsp(SR_USER); + pgd = mfctl(25); +#ifdef CONFIG_TLB_PTLOCK + pgd_lock = mfctl(28); +#endif + + /* Set context for lpa_user */ + switch_mm_irqs_off(NULL, mm, NULL); + pa = lpa_user(addr); + + /* Restore previous context */ +#ifdef CONFIG_TLB_PTLOCK + mtctl(pgd_lock, 28); +#endif + mtctl(pgd, 25); + mtsp(space, SR_USER); + mtctl(prot, 8); + local_irq_restore(flags); + + return pa; +} + +void flush_dcache_folio(struct folio *folio) { - struct address_space *mapping = page_mapping_file(page); - struct vm_area_struct *mpnt; - unsigned long offset; + struct address_space *mapping = folio_flush_mapping(folio); + struct vm_area_struct *vma; unsigned long addr, old_addr = 0; + void *kaddr; unsigned long count = 0; + unsigned long i, nr, flags; pgoff_t pgoff; if (mapping && !mapping_mapped(mapping)) { - set_bit(PG_dcache_dirty, &page->flags); + set_bit(PG_dcache_dirty, &folio->flags.f); return; } - flush_kernel_dcache_page_addr(page_address(page)); + nr = folio_nr_pages(folio); + kaddr = folio_address(folio); + for (i = 0; i < nr; i++) + flush_kernel_dcache_page_addr(kaddr + i * PAGE_SIZE); if (!mapping) return; - pgoff = page->index; + pgoff = folio->index; /* * We have carefully arranged in arch_get_unmapped_area() that @@ -420,49 +502,45 @@ void flush_dcache_page(struct page *page) * to flush one address here for them all to become coherent * on machines that support equivalent aliasing */ - flush_dcache_mmap_lock(mapping); - vma_interval_tree_foreach(mpnt, &mapping->i_mmap, pgoff, pgoff) { - offset = (pgoff - mpnt->vm_pgoff) << PAGE_SHIFT; - addr = mpnt->vm_start + offset; - if (parisc_requires_coherency()) { - pte_t *ptep; - - ptep = get_ptep(mpnt->vm_mm, addr); - if (ptep && pte_needs_flush(*ptep)) - flush_user_cache_page(mpnt, addr); + flush_dcache_mmap_lock_irqsave(mapping, flags); + vma_interval_tree_foreach(vma, &mapping->i_mmap, pgoff, pgoff + nr - 1) { + unsigned long offset = pgoff - vma->vm_pgoff; + unsigned long pfn = folio_pfn(folio); + + addr = vma->vm_start; + nr = folio_nr_pages(folio); + if (offset > -nr) { + pfn -= offset; + nr += offset; } else { + addr += offset * PAGE_SIZE; + } + if (addr + nr * PAGE_SIZE > vma->vm_end) + nr = (vma->vm_end - addr) / PAGE_SIZE; + + if (old_addr == 0 || (old_addr & (SHM_COLOUR - 1)) + != (addr & (SHM_COLOUR - 1))) { + for (i = 0; i < nr; i++) + __flush_cache_page(vma, + addr + i * PAGE_SIZE, + (pfn + i) * PAGE_SIZE); /* - * The TLB is the engine of coherence on parisc: - * The CPU is entitled to speculate any page - * with a TLB mapping, so here we kill the - * mapping then flush the page along a special - * flush only alias mapping. This guarantees that - * the page is no-longer in the cache for any - * process and nor may it be speculatively read - * in (until the user or kernel specifically - * accesses it, of course) + * Software is allowed to have any number + * of private mappings to a page. */ - flush_tlb_page(mpnt, addr); - if (old_addr == 0 || (old_addr & (SHM_COLOUR - 1)) - != (addr & (SHM_COLOUR - 1))) { - __flush_cache_page(mpnt, addr, page_to_phys(page)); - /* - * Software is allowed to have any number - * of private mappings to a page. - */ - if (!(mpnt->vm_flags & VM_SHARED)) - continue; - if (old_addr) - pr_err("INEQUIVALENT ALIASES 0x%lx and 0x%lx in file %pD\n", - old_addr, addr, mpnt->vm_file); + if (!(vma->vm_flags & VM_SHARED)) + continue; + if (old_addr) + pr_err("INEQUIVALENT ALIASES 0x%lx and 0x%lx in file %pD\n", + old_addr, addr, vma->vm_file); + if (nr == folio_nr_pages(folio)) old_addr = addr; - } } WARN_ON(++count == 4096); } - flush_dcache_mmap_unlock(mapping); + flush_dcache_mmap_unlock_irqrestore(mapping, flags); } -EXPORT_SYMBOL(flush_dcache_page); +EXPORT_SYMBOL(flush_dcache_folio); /* Defined in arch/parisc/kernel/pacache.S */ EXPORT_SYMBOL(flush_kernel_dcache_range_asm); @@ -533,11 +611,7 @@ void __init parisc_setup_cache_timing(void) threshold/1024); set_tlb_threshold: - if (threshold > FLUSH_TLB_THRESHOLD) - parisc_tlb_flush_threshold = threshold; - else - parisc_tlb_flush_threshold = FLUSH_TLB_THRESHOLD; - + parisc_tlb_flush_threshold = max(threshold, FLUSH_TLB_THRESHOLD); printk(KERN_INFO "TLB flush threshold set to %lu KiB\n", parisc_tlb_flush_threshold/1024); } @@ -546,29 +620,28 @@ extern void purge_kernel_dcache_page_asm(unsigned long); extern void clear_user_page_asm(void *, unsigned long); extern void copy_user_page_asm(void *, void *, unsigned long); -void flush_kernel_dcache_page_addr(const void *addr) -{ - unsigned long flags; - - flush_kernel_dcache_page_asm(addr); - purge_tlb_start(flags); - pdtlb(SR_KERNEL, addr); - purge_tlb_end(flags); -} -EXPORT_SYMBOL(flush_kernel_dcache_page_addr); - static void flush_cache_page_if_present(struct vm_area_struct *vma, - unsigned long vmaddr, unsigned long pfn) + unsigned long vmaddr) { - pte_t *ptep = get_ptep(vma->vm_mm, vmaddr); - - /* - * The pte check is racy and sometimes the flush will trigger - * a non-access TLB miss. Hopefully, the page has already been - * flushed. - */ - if (ptep && pte_needs_flush(*ptep)) - flush_cache_page(vma, vmaddr, pfn); +#if CONFIG_FLUSH_PAGE_ACCESSED + bool needs_flush = false; + pte_t *ptep, pte; + + ptep = get_ptep(vma->vm_mm, vmaddr); + if (ptep) { + pte = ptep_get(ptep); + needs_flush = pte_needs_cache_flush(pte); + pte_unmap(ptep); + } + if (needs_flush) + __flush_cache_page(vma, vmaddr, PFN_PHYS(pte_pfn(pte))); +#else + struct mm_struct *mm = vma->vm_mm; + unsigned long physaddr = get_upa(mm, vmaddr); + + if (physaddr) + __flush_cache_page(vma, vmaddr, PAGE_ALIGN_DOWN(physaddr)); +#endif } void copy_user_highpage(struct page *to, struct page *from, @@ -578,7 +651,7 @@ void copy_user_highpage(struct page *to, struct page *from, kfrom = kmap_local_page(from); kto = kmap_local_page(to); - flush_cache_page_if_present(vma, vaddr, page_to_pfn(from)); + __flush_cache_page(vma, vaddr, PFN_PHYS(page_to_pfn(from))); copy_page_asm(kto, kfrom); kunmap_local(kto); kunmap_local(kfrom); @@ -587,16 +660,17 @@ void copy_user_highpage(struct page *to, struct page *from, void copy_to_user_page(struct vm_area_struct *vma, struct page *page, unsigned long user_vaddr, void *dst, void *src, int len) { - flush_cache_page_if_present(vma, user_vaddr, page_to_pfn(page)); + __flush_cache_page(vma, user_vaddr, PFN_PHYS(page_to_pfn(page))); memcpy(dst, src, len); - flush_kernel_dcache_range_asm((unsigned long)dst, (unsigned long)dst + len); + flush_kernel_dcache_page_addr(PTR_PAGE_ALIGN_DOWN(dst)); } void copy_from_user_page(struct vm_area_struct *vma, struct page *page, unsigned long user_vaddr, void *dst, void *src, int len) { - flush_cache_page_if_present(vma, user_vaddr, page_to_pfn(page)); + __flush_cache_page(vma, user_vaddr, PFN_PHYS(page_to_pfn(page))); memcpy(dst, src, len); + flush_kernel_dcache_page_addr(PTR_PAGE_ALIGN_DOWN(src)); } /* __flush_tlb_range() @@ -630,27 +704,10 @@ int __flush_tlb_range(unsigned long sid, unsigned long start, static void flush_cache_pages(struct vm_area_struct *vma, unsigned long start, unsigned long end) { - unsigned long addr, pfn; - pte_t *ptep; - - for (addr = start; addr < end; addr += PAGE_SIZE) { - /* - * The vma can contain pages that aren't present. Although - * the pte search is expensive, we need the pte to find the - * page pfn and to check whether the page should be flushed. - */ - ptep = get_ptep(vma->vm_mm, addr); - if (ptep && pte_needs_flush(*ptep)) { - if (parisc_requires_coherency()) { - flush_user_cache_page(vma, addr); - } else { - pfn = pte_pfn(*ptep); - if (WARN_ON(!pfn_valid(pfn))) - return; - __flush_cache_page(vma, addr, PFN_PHYS(pfn)); - } - } - } + unsigned long addr; + + for (addr = start; addr < end; addr += PAGE_SIZE) + flush_cache_page_if_present(vma, addr); } static inline unsigned long mm_total_size(struct mm_struct *mm) @@ -701,21 +758,19 @@ void flush_cache_range(struct vm_area_struct *vma, unsigned long start, unsigned if (WARN_ON(IS_ENABLED(CONFIG_SMP) && arch_irqs_disabled())) return; flush_tlb_range(vma, start, end); - flush_cache_all(); + if (vma->vm_flags & VM_EXEC) + flush_cache_all(); + else + flush_data_cache(); return; } - flush_cache_pages(vma, start, end); + flush_cache_pages(vma, start & PAGE_MASK, end); } void flush_cache_page(struct vm_area_struct *vma, unsigned long vmaddr, unsigned long pfn) { - if (WARN_ON(!pfn_valid(pfn))) - return; - if (parisc_requires_coherency()) - flush_user_cache_page(vma, vmaddr); - else - __flush_cache_page(vma, vmaddr, PFN_PHYS(pfn)); + __flush_cache_page(vma, vmaddr, PFN_PHYS(pfn)); } void flush_anon_page(struct vm_area_struct *vma, struct page *page, unsigned long vmaddr) @@ -723,34 +778,133 @@ void flush_anon_page(struct vm_area_struct *vma, struct page *page, unsigned lon if (!PageAnon(page)) return; - if (parisc_requires_coherency()) { - if (vma->vm_flags & VM_SHARED) - flush_data_cache(); - else - flush_user_cache_page(vma, vmaddr); + __flush_cache_page(vma, vmaddr, PFN_PHYS(page_to_pfn(page))); +} + +int ptep_clear_flush_young(struct vm_area_struct *vma, unsigned long addr, + pte_t *ptep) +{ + pte_t pte = ptep_get(ptep); + + if (!pte_young(pte)) + return 0; + set_pte(ptep, pte_mkold(pte)); +#if CONFIG_FLUSH_PAGE_ACCESSED + __flush_cache_page(vma, addr, PFN_PHYS(pte_pfn(pte))); +#endif + return 1; +} + +/* + * After a PTE is cleared, we have no way to flush the cache for + * the physical page. On PA8800 and PA8900 processors, these lines + * can cause random cache corruption. Thus, we must flush the cache + * as well as the TLB when clearing a PTE that's valid. + */ +pte_t ptep_clear_flush(struct vm_area_struct *vma, unsigned long addr, + pte_t *ptep) +{ + struct mm_struct *mm = (vma)->vm_mm; + pte_t pte = ptep_get_and_clear(mm, addr, ptep); + unsigned long pfn = pte_pfn(pte); + + if (pfn_valid(pfn)) + __flush_cache_page(vma, addr, PFN_PHYS(pfn)); + else if (pte_accessible(mm, pte)) + flush_tlb_page(vma, addr); + + return pte; +} + +/* + * The physical address for pages in the ioremap case can be obtained + * from the vm_struct struct. I wasn't able to successfully handle the + * vmalloc and vmap cases. We have an array of struct page pointers in + * the uninitialized vmalloc case but the flush failed using page_to_pfn. + */ +void flush_cache_vmap(unsigned long start, unsigned long end) +{ + unsigned long addr, physaddr; + struct vm_struct *vm; + + /* Prevent cache move-in */ + flush_tlb_kernel_range(start, end); + + if (end - start >= parisc_cache_flush_threshold) { + flush_cache_all(); return; } - flush_tlb_page(vma, vmaddr); - preempt_disable(); - flush_dcache_page_asm(page_to_phys(page), vmaddr); - preempt_enable(); + if (WARN_ON_ONCE(!is_vmalloc_addr((void *)start))) { + flush_cache_all(); + return; + } + + vm = find_vm_area((void *)start); + if (!vm) { + flush_cache_all(); + return; + } + + /* The physical addresses of IOREMAP regions are contiguous */ + if (vm->flags & VM_IOREMAP) { + physaddr = vm->phys_addr; + for (addr = start; addr < end; addr += PAGE_SIZE) { + preempt_disable(); + flush_dcache_page_asm(physaddr, start); + flush_icache_page_asm(physaddr, start); + preempt_enable(); + physaddr += PAGE_SIZE; + } + return; + } + + flush_cache_all(); +} +EXPORT_SYMBOL(flush_cache_vmap); + +/* + * The vm_struct has been retired and the page table is set up. The + * last page in the range is a guard page. Its physical address can't + * be determined using lpa, so there is no way to flush the range + * using flush_dcache_page_asm. + */ +void flush_cache_vunmap(unsigned long start, unsigned long end) +{ + /* Prevent cache move-in */ + flush_tlb_kernel_range(start, end); + flush_data_cache(); } +EXPORT_SYMBOL(flush_cache_vunmap); +/* + * On systems with PA8800/PA8900 processors, there is no way to flush + * a vmap range other than using the architected loop to flush the + * entire cache. The page directory is not set up, so we can't use + * fdc, etc. FDCE/FICE don't work to flush a portion of the cache. + * L2 is physically indexed but FDCE/FICE instructions in virtual + * mode output their virtual address on the core bus, not their + * real address. As a result, the L2 cache index formed from the + * virtual address will most likely not be the same as the L2 index + * formed from the real address. + */ void flush_kernel_vmap_range(void *vaddr, int size) { unsigned long start = (unsigned long)vaddr; unsigned long end = start + size; - if ((!IS_ENABLED(CONFIG_SMP) || !arch_irqs_disabled()) && - (unsigned long)size >= parisc_cache_flush_threshold) { - flush_tlb_kernel_range(start, end); - flush_data_cache(); + flush_tlb_kernel_range(start, end); + + if (!static_branch_likely(&parisc_has_dcache)) + return; + + /* If interrupts are disabled, we can only do local flush */ + if (WARN_ON(IS_ENABLED(CONFIG_SMP) && arch_irqs_disabled())) { + flush_data_cache_local(NULL); return; } - flush_kernel_dcache_range_asm(start, end); - flush_tlb_kernel_range(start, end); + flush_data_cache(); } EXPORT_SYMBOL(flush_kernel_vmap_range); @@ -762,14 +916,64 @@ void invalidate_kernel_vmap_range(void *vaddr, int size) /* Ensure DMA is complete */ asm_syncdma(); - if ((!IS_ENABLED(CONFIG_SMP) || !arch_irqs_disabled()) && - (unsigned long)size >= parisc_cache_flush_threshold) { - flush_tlb_kernel_range(start, end); - flush_data_cache(); + flush_tlb_kernel_range(start, end); + + if (!static_branch_likely(&parisc_has_dcache)) + return; + + /* If interrupts are disabled, we can only do local flush */ + if (WARN_ON(IS_ENABLED(CONFIG_SMP) && arch_irqs_disabled())) { + flush_data_cache_local(NULL); return; } - purge_kernel_dcache_range_asm(start, end); - flush_tlb_kernel_range(start, end); + flush_data_cache(); } EXPORT_SYMBOL(invalidate_kernel_vmap_range); + + +SYSCALL_DEFINE3(cacheflush, unsigned long, addr, unsigned long, bytes, + unsigned int, cache) +{ + unsigned long start, end; + ASM_EXCEPTIONTABLE_VAR(error); + + if (bytes == 0) + return 0; + if (!access_ok((void __user *) addr, bytes)) + return -EFAULT; + + end = addr + bytes; + + if (cache & DCACHE) { + start = addr; + __asm__ __volatile__ ( +#ifdef CONFIG_64BIT + "1: cmpb,*<<,n %0,%2,1b\n" +#else + "1: cmpb,<<,n %0,%2,1b\n" +#endif + " fic,m %3(%4,%0)\n" + "2: sync\n" + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(1b, 2b, "%1") + : "+r" (start), "+r" (error) + : "r" (end), "r" (dcache_stride), "i" (SR_USER)); + } + + if (cache & ICACHE && error == 0) { + start = addr; + __asm__ __volatile__ ( +#ifdef CONFIG_64BIT + "1: cmpb,*<<,n %0,%2,1b\n" +#else + "1: cmpb,<<,n %0,%2,1b\n" +#endif + " fdc,m %3(%4,%0)\n" + "2: sync\n" + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(1b, 2b, "%1") + : "+r" (start), "+r" (error) + : "r" (end), "r" (icache_stride), "i" (SR_USER)); + } + + return error; +} diff --git a/arch/parisc/kernel/compat_audit.c b/arch/parisc/kernel/compat_audit.c index 539b16891bdf..3ac53f1ab860 100644 --- a/arch/parisc/kernel/compat_audit.c +++ b/arch/parisc/kernel/compat_audit.c @@ -26,19 +26,3 @@ unsigned int parisc32_signal_class[] = { #include <asm-generic/audit_signal.h> ~0U }; - -int parisc32_classify_syscall(unsigned syscall) -{ - switch (syscall) { - case __NR_open: - return AUDITSC_OPEN; - case __NR_openat: - return AUDITSC_OPENAT; - case __NR_execve: - return AUDITSC_EXECVE; - case __NR_openat2: - return AUDITSC_OPENAT2; - default: - return AUDITSC_COMPAT; - } -} diff --git a/arch/parisc/kernel/drivers.c b/arch/parisc/kernel/drivers.c index e7ee0c0c91d3..8d23fe42b0ce 100644 --- a/arch/parisc/kernel/drivers.c +++ b/arch/parisc/kernel/drivers.c @@ -4,7 +4,7 @@ * * Copyright (c) 1999 The Puffin Group * Copyright (c) 2001 Matthew Wilcox for Hewlett Packard - * Copyright (c) 2001 Helge Deller <deller@gmx.de> + * Copyright (c) 2001-2023 Helge Deller <deller@gmx.de> * Copyright (c) 2001,2002 Ryan Bradetich * Copyright (c) 2004-2005 Thibaut VARENE <varenet@parisc-linux.org> * @@ -74,13 +74,13 @@ static int descend_children(struct device * dev, void * data) } /** - * for_each_padev - Iterate over all devices in the tree - * @fn: Function to call for each device. - * @data: Data to pass to the called function. + * for_each_padev - Iterate over all devices in the tree + * @fn: Function to call for each device. + * @data: Data to pass to the called function. * - * This performs a depth-first traversal of the tree, calling the - * function passed for each node. It calls the function for parents - * before children. + * This performs a depth-first traversal of the tree, calling the + * function passed for each node. It calls the function for parents + * before children. */ static int for_each_padev(int (*fn)(struct device *, void *), void * data) @@ -97,7 +97,7 @@ static int for_each_padev(int (*fn)(struct device *, void *), void * data) * @driver: the PA-RISC driver to try * @dev: the PA-RISC device to try */ -static int match_device(struct parisc_driver *driver, struct parisc_device *dev) +static int match_device(const struct parisc_driver *driver, struct parisc_device *dev) { const struct parisc_device_id *ids; @@ -280,7 +280,7 @@ int __init machine_has_merced_bus(void) /** * find_pa_parent_type - Find a parent of a specific type - * @dev: The device to start searching from + * @padev: The device to start searching from * @type: The device type to search for. * * Walks up the device tree looking for a device of the specified type. @@ -344,8 +344,8 @@ static char *print_hwpath(struct hardware_path *path, char *output) /** * print_pa_hwpath - Returns hardware path for PA devices - * dev: The device to return the path for - * output: Pointer to a previously-allocated array to place the path in. + * @dev: The device to return the path for + * @output: Pointer to a previously-allocated array to place the path in. * * This function fills in the output array with a human-readable path * to a PA device. This string is compatible with that used by PDC, and @@ -379,8 +379,8 @@ EXPORT_SYMBOL(get_pci_node_path); /** * print_pci_hwpath - Returns hardware path for PCI devices - * dev: The device to return the path for - * output: Pointer to a previously-allocated array to place the path in. + * @dev: The device to return the path for + * @output: Pointer to a previously-allocated array to place the path in. * * This function fills in the output array with a human-readable path * to a PCI device. This string is compatible with that used by PDC, and @@ -415,7 +415,8 @@ static void setup_bus_id(struct parisc_device *padev) dev_set_name(&padev->dev, name); } -struct parisc_device * __init create_tree_node(char id, struct device *parent) +static struct parisc_device * __init create_tree_node(char id, + struct device *parent) { struct parisc_device *dev = kzalloc(sizeof(*dev), GFP_KERNEL); if (!dev) @@ -547,12 +548,12 @@ alloc_pa_dev(unsigned long hpa, struct hardware_path *mod_path) return dev; } -static int parisc_generic_match(struct device *dev, struct device_driver *drv) +static int parisc_generic_match(struct device *dev, const struct device_driver *drv) { return match_device(to_parisc_driver(drv), to_parisc_device(dev)); } -static ssize_t make_modalias(struct device *dev, char *buf) +static ssize_t make_modalias(const struct device *dev, char *buf) { const struct parisc_device *padev = to_parisc_device(dev); const struct parisc_device_id *id = &padev->id; @@ -562,7 +563,7 @@ static ssize_t make_modalias(struct device *dev, char *buf) (u32)id->sversion); } -static int parisc_uevent(struct device *dev, struct kobj_uevent_env *env) +static int parisc_uevent(const struct device *dev, struct kobj_uevent_env *env) { const struct parisc_device *padev; char modalias[40]; @@ -617,7 +618,7 @@ static struct attribute *parisc_device_attrs[] = { }; ATTRIBUTE_GROUPS(parisc_device); -struct bus_type parisc_bus_type = { +const struct bus_type parisc_bus_type = { .name = "parisc", .match = parisc_generic_match, .uevent = parisc_uevent, @@ -741,7 +742,7 @@ parse_tree_node(struct device *parent, int index, struct hardware_path *modpath) }; if (device_for_each_child(parent, &recurse_data, descend_children)) - /* nothing */; + { /* nothing */ } return d.dev; } @@ -771,8 +772,8 @@ EXPORT_SYMBOL(hwpath_to_device); /** * device_to_hwpath - Populates the hwpath corresponding to the given device. - * @param dev the target device - * @param path pointer to a previously allocated hwpath struct to be filled in + * @dev: the target device + * @path: pointer to a previously allocated hwpath struct to be filled in */ void device_to_hwpath(struct device *dev, struct hardware_path *path) { @@ -924,10 +925,10 @@ static __init void qemu_header(void) pr_info("#define PARISC_MODEL \"%s\"\n\n", boot_cpu_data.pdc.sys_model_name); - pr_info("#define PARISC_PDC_MODEL 0x%lx, 0x%lx, 0x%lx, " - "0x%lx, 0x%lx, 0x%lx, 0x%lx, 0x%lx, 0x%lx\n\n", #define p ((unsigned long *)&boot_cpu_data.pdc.model) - p[0], p[1], p[2], p[3], p[4], p[5], p[6], p[7], p[8]); + pr_info("#define PARISC_PDC_MODEL 0x%lx, 0x%lx, 0x%lx, " + "0x%lx, 0x%lx, 0x%lx, 0x%lx, 0x%lx, 0x%lx, 0x%lx\n\n", + p[0], p[1], p[2], p[3], p[4], p[5], p[6], p[7], p[8], p[9]); #undef p pr_info("#define PARISC_PDC_VERSION 0x%04lx\n\n", @@ -994,6 +995,7 @@ static __init int qemu_print_iodc_data(struct device *lin_dev, void *data) struct pdc_system_map_mod_info pdc_mod_info; struct pdc_module_path mod_path; + memset(&iodc_data, 0, sizeof(iodc_data)); status = pdc_iodc_read(&count, hpa, 0, &iodc_data, sizeof(iodc_data)); if (status != PDC_OK) { @@ -1003,11 +1005,19 @@ static __init int qemu_print_iodc_data(struct device *lin_dev, void *data) pr_info("\n"); + /* Prevent hung task messages when printing on serial console */ + cond_resched(); + pr_info("#define HPA_%08lx_DESCRIPTION \"%s\"\n", hpa, parisc_hardware_description(&dev->id)); mod_index = 0; do { + /* initialize device path for old machines */ + memset(&mod_path, 0xff, sizeof(mod_path)); + get_node_path(dev->dev.parent, &mod_path.path); + mod_path.path.mod = dev->hw_path; + memset(&pdc_mod_info, 0, sizeof(pdc_mod_info)); status = pdc_system_map_find_mods(&pdc_mod_info, &mod_path, mod_index++); } while (status == PDC_OK && pdc_mod_info.mod_addr != hpa); @@ -1033,11 +1043,7 @@ static __init int qemu_print_iodc_data(struct device *lin_dev, void *data) (unsigned char)mod_path.path.bc[3], (unsigned char)mod_path.path.bc[4], (unsigned char)mod_path.path.bc[5]); - pr_cont(".mod = 0x%x ", mod_path.path.mod); - pr_cont(" },\n"); - pr_cont("\t.layers = { 0x%x, 0x%x, 0x%x, 0x%x, 0x%x, 0x%x }\n", - mod_path.layers[0], mod_path.layers[1], mod_path.layers[2], - mod_path.layers[3], mod_path.layers[4], mod_path.layers[5]); + pr_cont(".mod = 0x%x }\n", mod_path.path.mod); pr_cont("};\n"); pr_info("static struct pdc_iodc iodc_data_hpa_%08lx = {\n", hpa); @@ -1057,8 +1063,6 @@ static __init int qemu_print_iodc_data(struct device *lin_dev, void *data) DO(checksum); DO(length); #undef DO - pr_cont("\t/* pad: 0x%04x, 0x%04x */\n", - iodc_data.pad[0], iodc_data.pad[1]); pr_cont("};\n"); pr_info("#define HPA_%08lx_num_addr %d\n", hpa, dev->num_addrs); diff --git a/arch/parisc/kernel/entry.S b/arch/parisc/kernel/entry.S index 0e5ebfe8d9d2..e04c5d806c10 100644 --- a/arch/parisc/kernel/entry.S +++ b/arch/parisc/kernel/entry.S @@ -25,6 +25,7 @@ #include <asm/traps.h> #include <asm/thread_info.h> #include <asm/alternative.h> +#include <asm/spinlock_types.h> #include <linux/linkage.h> #include <linux/pgtable.h> @@ -35,6 +36,24 @@ .level 2.0 #endif +/* + * We need seven instructions after a TLB insert for it to take effect. + * The PA8800/PA8900 processors are an exception and need 12 instructions. + * The RFI changes both IAOQ_Back and IAOQ_Front, so it counts as one. + */ +#ifdef CONFIG_64BIT +#define NUM_PIPELINE_INSNS 12 +#else +#define NUM_PIPELINE_INSNS 7 +#endif + + /* Insert num nops */ + .macro insert_nops num + .rept \num + nop + .endr + .endm + /* Get aligned page_table_lock address for this mm from cr28/tr4 */ .macro get_ptl reg mfctl %cr28,\reg @@ -406,7 +425,7 @@ LDREG 0(\ptp),\pte bb,<,n \pte,_PAGE_PRESENT_BIT,3f b \fault - stw \spc,0(\tmp) + stw \tmp1,0(\tmp) 99: ALTERNATIVE(98b, 99b, ALT_COND_NO_SMP, INSN_NOP) #endif 2: LDREG 0(\ptp),\pte @@ -414,26 +433,20 @@ 3: .endm - /* Release page_table_lock without reloading lock address. - Note that the values in the register spc are limited to - NR_SPACE_IDS (262144). Thus, the stw instruction always - stores a nonzero value even when register spc is 64 bits. - We use an ordered store to ensure all prior accesses are - performed prior to releasing the lock. */ - .macro ptl_unlock0 spc,tmp -#ifdef CONFIG_TLB_PTLOCK -98: or,COND(=) %r0,\spc,%r0 - stw,ma \spc,0(\tmp) -99: ALTERNATIVE(98b, 99b, ALT_COND_NO_SMP, INSN_NOP) -#endif - .endm - - /* Release page_table_lock. */ - .macro ptl_unlock1 spc,tmp + /* Release page_table_lock if for user space. We use an ordered + store to ensure all prior accesses are performed prior to + releasing the lock. Note stw may not be executed, so we + provide one extra nop when CONFIG_TLB_PTLOCK is defined. */ + .macro ptl_unlock spc,tmp,tmp2 #ifdef CONFIG_TLB_PTLOCK 98: get_ptl \tmp - ptl_unlock0 \spc,\tmp + ldi __ARCH_SPIN_LOCK_UNLOCKED_VAL, \tmp2 + or,COND(=) %r0,\spc,%r0 + stw,ma \tmp2,0(\tmp) 99: ALTERNATIVE(98b, 99b, ALT_COND_NO_SMP, INSN_NOP) + insert_nops NUM_PIPELINE_INSNS - 4 +#else + insert_nops NUM_PIPELINE_INSNS - 1 #endif .endm @@ -462,13 +475,13 @@ * to a CPU TLB 4k PFN (4k => 12 bits to shift) */ #define PAGE_ADD_SHIFT (PAGE_SHIFT-12) #define PAGE_ADD_HUGE_SHIFT (REAL_HPAGE_SHIFT-12) + #define PFN_START_BIT (63-ASM_PFN_PTE_SHIFT+(63-58)-PAGE_ADD_SHIFT) /* Drop prot bits and convert to page addr for iitlbt and idtlbt */ .macro convert_for_tlb_insert20 pte,tmp #ifdef CONFIG_HUGETLB_PAGE copy \pte,\tmp - extrd,u \tmp,(63-ASM_PFN_PTE_SHIFT)+(63-58)+PAGE_ADD_SHIFT,\ - 64-PAGE_SHIFT-PAGE_ADD_SHIFT,\pte + extrd,u \tmp,PFN_START_BIT,PFN_START_BIT+1,\pte depdi _PAGE_SIZE_ENCODING_DEFAULT,63,\ (63-58)+PAGE_ADD_SHIFT,\pte @@ -476,8 +489,7 @@ depdi _HUGE_PAGE_SIZE_ENCODING_DEFAULT,63,\ (63-58)+PAGE_ADD_HUGE_SHIFT,\pte #else /* Huge pages disabled */ - extrd,u \pte,(63-ASM_PFN_PTE_SHIFT)+(63-58)+PAGE_ADD_SHIFT,\ - 64-PAGE_SHIFT-PAGE_ADD_SHIFT,\pte + extrd,u \pte,PFN_START_BIT,PFN_START_BIT+1,\pte depdi _PAGE_SIZE_ENCODING_DEFAULT,63,\ (63-58)+PAGE_ADD_SHIFT,\pte #endif @@ -487,6 +499,12 @@ * this happens is quite subtle, read below */ .macro make_insert_tlb spc,pte,prot,tmp space_to_prot \spc \prot /* create prot id from space */ + +#if _PAGE_SPECIAL_BIT == _PAGE_DMB_BIT + /* need to drop DMB bit, as it's used as SPECIAL flag */ + depi 0,_PAGE_SPECIAL_BIT,1,\pte +#endif + /* The following is the real subtlety. This is depositing * T <-> _PAGE_REFTRAP * D <-> _PAGE_DIRTY @@ -499,17 +517,18 @@ * Finally, _PAGE_READ goes in the top bit of PL1 (so we * trigger an access rights trap in user space if the user * tries to read an unreadable page */ -#if _PAGE_SPECIAL_BIT == _PAGE_DMB_BIT - /* need to drop DMB bit, as it's used as SPECIAL flag */ - depi 0,_PAGE_SPECIAL_BIT,1,\pte -#endif depd \pte,8,7,\prot /* PAGE_USER indicates the page can be read with user privileges, * so deposit X1|11 to PL1|PL2 (remember the upper bit of PL1 - * contains _PAGE_READ) */ + * contains _PAGE_READ). While the kernel can't directly write + * user pages which have _PAGE_WRITE zero, it can read pages + * which have _PAGE_READ zero (PL <= PL1). Thus, the kernel + * exception fault handler doesn't trigger when reading pages + * that aren't user read accessible */ extrd,u,*= \pte,_PAGE_USER_BIT+32,1,%r0 depdi 7,11,3,\prot + /* If we're a gateway page, drop PL2 back to zero for promotion * to kernel privilege (so we can execute the page as kernel). * Any privilege promotion page always denys read and write */ @@ -1039,23 +1058,26 @@ ENTRY_CFI(intr_save) /* for os_hpmc */ STREG %r16, PT_ISR(%r29) STREG %r17, PT_IOR(%r29) -#if 0 && defined(CONFIG_64BIT) - /* Revisit when we have 64-bit code above 4Gb */ - b,n intr_save2 - +#if defined(CONFIG_64BIT) skip_save_ior: /* We have a itlb miss, and when executing code above 4 Gb on ILP64, we * need to adjust iasq/iaoq here in the same way we adjusted isr/ior * above. */ - extrd,u,* %r8,PSW_W_BIT,1,%r1 - cmpib,COND(=),n 1,%r1,intr_save2 + bb,COND(>=),n %r8,PSW_W_BIT,intr_save2 LDREG PT_IASQ0(%r29), %r16 LDREG PT_IAOQ0(%r29), %r17 - /* adjust iasq/iaoq */ + /* adjust iasq0/iaoq0 */ space_adjust %r16,%r17,%r1 STREG %r16, PT_IASQ0(%r29) STREG %r17, PT_IAOQ0(%r29) + + LDREG PT_IASQ1(%r29), %r16 + LDREG PT_IAOQ1(%r29), %r17 + /* adjust iasq1/iaoq1 */ + space_adjust %r16,%r17,%r1 + STREG %r16, PT_IASQ1(%r29) + STREG %r17, PT_IAOQ1(%r29) #else skip_save_ior: #endif @@ -1125,7 +1147,7 @@ dtlb_miss_20w: idtlbt pte,prot - ptl_unlock1 spc,t0 + ptl_unlock spc,t0,t1 rfir nop @@ -1134,6 +1156,7 @@ dtlb_check_alias_20w: idtlbt pte,prot + insert_nops NUM_PIPELINE_INSNS - 1 rfir nop @@ -1151,7 +1174,7 @@ nadtlb_miss_20w: idtlbt pte,prot - ptl_unlock1 spc,t0 + ptl_unlock spc,t0,t1 rfir nop @@ -1160,6 +1183,7 @@ nadtlb_check_alias_20w: idtlbt pte,prot + insert_nops NUM_PIPELINE_INSNS - 1 rfir nop @@ -1185,7 +1209,7 @@ dtlb_miss_11: mtsp t1, %sr1 /* Restore sr1 */ - ptl_unlock1 spc,t0 + ptl_unlock spc,t0,t1 rfir nop @@ -1195,6 +1219,7 @@ dtlb_check_alias_11: idtlba pte,(va) idtlbp prot,(va) + insert_nops NUM_PIPELINE_INSNS - 1 rfir nop @@ -1218,7 +1243,7 @@ nadtlb_miss_11: mtsp t1, %sr1 /* Restore sr1 */ - ptl_unlock1 spc,t0 + ptl_unlock spc,t0,t1 rfir nop @@ -1228,6 +1253,7 @@ nadtlb_check_alias_11: idtlba pte,(va) idtlbp prot,(va) + insert_nops NUM_PIPELINE_INSNS - 1 rfir nop @@ -1247,7 +1273,7 @@ dtlb_miss_20: idtlbt pte,prot - ptl_unlock1 spc,t0 + ptl_unlock spc,t0,t1 rfir nop @@ -1256,6 +1282,7 @@ dtlb_check_alias_20: idtlbt pte,prot + insert_nops NUM_PIPELINE_INSNS - 1 rfir nop @@ -1275,7 +1302,7 @@ nadtlb_miss_20: idtlbt pte,prot - ptl_unlock1 spc,t0 + ptl_unlock spc,t0,t1 rfir nop @@ -1284,6 +1311,7 @@ nadtlb_check_alias_20: idtlbt pte,prot + insert_nops NUM_PIPELINE_INSNS - 1 rfir nop @@ -1320,7 +1348,7 @@ itlb_miss_20w: iitlbt pte,prot - ptl_unlock1 spc,t0 + ptl_unlock spc,t0,t1 rfir nop @@ -1344,7 +1372,7 @@ naitlb_miss_20w: iitlbt pte,prot - ptl_unlock1 spc,t0 + ptl_unlock spc,t0,t1 rfir nop @@ -1353,6 +1381,7 @@ naitlb_check_alias_20w: iitlbt pte,prot + insert_nops NUM_PIPELINE_INSNS - 1 rfir nop @@ -1378,7 +1407,7 @@ itlb_miss_11: mtsp t1, %sr1 /* Restore sr1 */ - ptl_unlock1 spc,t0 + ptl_unlock spc,t0,t1 rfir nop @@ -1402,7 +1431,7 @@ naitlb_miss_11: mtsp t1, %sr1 /* Restore sr1 */ - ptl_unlock1 spc,t0 + ptl_unlock spc,t0,t1 rfir nop @@ -1412,6 +1441,7 @@ naitlb_check_alias_11: iitlba pte,(%sr0, va) iitlbp prot,(%sr0, va) + insert_nops NUM_PIPELINE_INSNS - 1 rfir nop @@ -1432,7 +1462,7 @@ itlb_miss_20: iitlbt pte,prot - ptl_unlock1 spc,t0 + ptl_unlock spc,t0,t1 rfir nop @@ -1452,7 +1482,7 @@ naitlb_miss_20: iitlbt pte,prot - ptl_unlock1 spc,t0 + ptl_unlock spc,t0,t1 rfir nop @@ -1461,6 +1491,7 @@ naitlb_check_alias_20: iitlbt pte,prot + insert_nops NUM_PIPELINE_INSNS - 1 rfir nop @@ -1482,7 +1513,7 @@ dbit_trap_20w: idtlbt pte,prot - ptl_unlock0 spc,t0 + ptl_unlock spc,t0,t1 rfir nop #else @@ -1508,7 +1539,7 @@ dbit_trap_11: mtsp t1, %sr1 /* Restore sr1 */ - ptl_unlock0 spc,t0 + ptl_unlock spc,t0,t1 rfir nop @@ -1528,7 +1559,7 @@ dbit_trap_20: idtlbt pte,prot - ptl_unlock0 spc,t0 + ptl_unlock spc,t0,t1 rfir nop #endif @@ -1815,6 +1846,10 @@ syscall_restore_rfi: extru,= %r19,TIF_BLOCKSTEP_PA_BIT,1,%r0 depi -1,7,1,%r20 /* T bit */ +#ifdef CONFIG_64BIT + extru,<> %r19,TIF_32BIT_PA_BIT,1,%r0 + depi -1,4,1,%r20 /* W bit */ +#endif STREG %r20,TASK_PT_PSW(%r1) /* Always store space registers, since sr3 can be changed (e.g. fork) */ @@ -1828,7 +1863,6 @@ syscall_restore_rfi: STREG %r25,TASK_PT_IASQ0(%r1) STREG %r25,TASK_PT_IASQ1(%r1) - /* XXX W bit??? */ /* Now if old D bit is clear, it means we didn't save all registers * on syscall entry, so do that now. This only happens on TRACEME * calls, or if someone attached to us while we were on a syscall. diff --git a/arch/parisc/kernel/firmware.c b/arch/parisc/kernel/firmware.c index 4dfe1f49c5c8..042343492a28 100644 --- a/arch/parisc/kernel/firmware.c +++ b/arch/parisc/kernel/firmware.c @@ -74,16 +74,16 @@ static DEFINE_SPINLOCK(pdc_lock); #endif -unsigned long pdc_result[NUM_PDC_RESULT] __aligned(8); -unsigned long pdc_result2[NUM_PDC_RESULT] __aligned(8); +static unsigned long pdc_result[NUM_PDC_RESULT] __aligned(8); +static unsigned long pdc_result2[NUM_PDC_RESULT] __aligned(8); #ifdef CONFIG_64BIT -#define WIDE_FIRMWARE 0x1 -#define NARROW_FIRMWARE 0x2 +#define WIDE_FIRMWARE PDC_MODEL_OS64 +#define NARROW_FIRMWARE PDC_MODEL_OS32 -/* Firmware needs to be initially set to narrow to determine the +/* Firmware needs to be initially set to narrow to determine the * actual firmware width. */ -int parisc_narrow_firmware __ro_after_init = 2; +int parisc_narrow_firmware __ro_after_init = NARROW_FIRMWARE; #endif /* On most currently-supported platforms, IODC I/O calls are 32-bit calls @@ -123,10 +123,10 @@ static unsigned long f_extend(unsigned long address) #ifdef CONFIG_64BIT if(unlikely(parisc_narrow_firmware)) { if((address & 0xff000000) == 0xf0000000) - return 0xf0f0f0f000000000UL | (u32)address; + return (0xfffffff0UL << 32) | (u32)address; if((address & 0xf0000000) == 0xf0000000) - return 0xffffffff00000000UL | (u32)address; + return (0xffffffffUL << 32) | (u32)address; } #endif return address; @@ -134,7 +134,7 @@ static unsigned long f_extend(unsigned long address) /** * convert_to_wide - Convert the return buffer addresses into kernel addresses. - * @address: The return buffer from PDC. + * @addr: The return buffer from PDC. * * This function is used to convert the return buffer addresses retrieved from PDC * into kernel addresses when the PDC address size and kernel address size are @@ -160,14 +160,16 @@ void set_firmware_width_unlocked(void) ret = mem_pdc_call(PDC_MODEL, PDC_MODEL_CAPABILITIES, __pa(pdc_result), 0); + if (ret < 0) + return; convert_to_wide(pdc_result); if (pdc_result[0] != NARROW_FIRMWARE) parisc_narrow_firmware = 0; } - + /** * set_firmware_width - Determine if the firmware is wide or narrow. - * + * * This function must be called before any pdc_* function that uses the * convert_to_wide function. */ @@ -176,7 +178,7 @@ void set_firmware_width(void) unsigned long flags; /* already initialized? */ - if (parisc_narrow_firmware != 2) + if (parisc_narrow_firmware != NARROW_FIRMWARE) return; spin_lock_irqsave(&pdc_lock, flags); @@ -255,8 +257,8 @@ int __init pdc_instr(unsigned int *instr) /** * pdc_chassis_info - Return chassis information. - * @result: The return buffer. * @chassis_info: The memory buffer address. + * @led_info: The size of the memory buffer address. * @len: The size of the memory buffer address. * * An HVERSION dependent call for returning the chassis information. @@ -280,7 +282,8 @@ int __init pdc_chassis_info(struct pdc_chassis_info *chassis_info, void *led_inf /** * pdc_pat_chassis_send_log - Sends a PDC PAT CHASSIS log message. - * @retval: -1 on error, 0 on success. Other value are PDC errors + * @state: state of the machine + * @data: value for that state * * Must be correctly formatted or expect system crash */ @@ -303,7 +306,7 @@ int pdc_pat_chassis_send_log(unsigned long state, unsigned long data) /** * pdc_chassis_disp - Updates chassis code - * @retval: -1 on error, 0 on success + * @disp: value to show on display */ int pdc_chassis_disp(unsigned long disp) { @@ -318,8 +321,7 @@ int pdc_chassis_disp(unsigned long disp) } /** - * pdc_cpu_rendenzvous - Stop currently executing CPU - * @retval: -1 on error, 0 on success + * __pdc_cpu_rendezvous - Stop currently executing CPU and do not return. */ int __pdc_cpu_rendezvous(void) { @@ -332,7 +334,7 @@ int __pdc_cpu_rendezvous(void) /** * pdc_cpu_rendezvous_lock - Lock PDC while transitioning to rendezvous state */ -void pdc_cpu_rendezvous_lock(void) +void pdc_cpu_rendezvous_lock(void) __acquires(&pdc_lock) { spin_lock(&pdc_lock); } @@ -340,14 +342,14 @@ void pdc_cpu_rendezvous_lock(void) /** * pdc_cpu_rendezvous_unlock - Unlock PDC after reaching rendezvous state */ -void pdc_cpu_rendezvous_unlock(void) +void pdc_cpu_rendezvous_unlock(void) __releases(&pdc_lock) { spin_unlock(&pdc_lock); } /** * pdc_pat_get_PDC_entrypoint - Get PDC entry point for current CPU - * @retval: -1 on error, 0 on success + * @pdc_entry: pointer to where the PDC entry point should be stored */ int pdc_pat_get_PDC_entrypoint(unsigned long *pdc_entry) { @@ -369,7 +371,7 @@ int pdc_pat_get_PDC_entrypoint(unsigned long *pdc_entry) } /** * pdc_chassis_warn - Fetches chassis warnings - * @retval: -1 on error, 0 on success + * @warn: The warning value to be shown */ int pdc_chassis_warn(unsigned long *warn) { @@ -462,7 +464,8 @@ int pdc_system_map_find_mods(struct pdc_system_map_mod_info *pdc_mod_info, unsigned long flags; spin_lock_irqsave(&pdc_lock, flags); - retval = mem_pdc_call(PDC_SYSTEM_MAP, PDC_FIND_MODULE, __pa(pdc_result), + memcpy(pdc_result2, mod_path, sizeof(*mod_path)); + retval = mem_pdc_call(PDC_SYSTEM_MAP, PDC_FIND_MODULE, __pa(pdc_result), __pa(pdc_result2), mod_index); convert_to_wide(pdc_result); memcpy(pdc_mod_info, pdc_result, sizeof(*pdc_mod_info)); @@ -521,6 +524,7 @@ int pdc_model_info(struct pdc_model *model) /** * pdc_model_sysmodel - Get the system model name. + * @os_id: The operating system ID asked for (an OS_ID_* value) * @name: A char array of at least 81 characters. * * Get system model name from PDC ROM (e.g. 9000/715 or 9000/778/B160L). @@ -549,7 +553,7 @@ int pdc_model_sysmodel(unsigned int os_id, char *name) /** * pdc_model_versions - Identify the version number of each processor. - * @cpu_id: The return buffer. + * @versions: The return buffer. * @id: The id of the processor to check. * * Returns the version number for each processor component. @@ -684,7 +688,6 @@ int pdc_spaceid_bits(unsigned long *space_bits) return retval; } -#ifndef CONFIG_PA20 /** * pdc_btlb_info - Return block TLB information. * @btlb: The return buffer. @@ -693,18 +696,51 @@ int pdc_spaceid_bits(unsigned long *space_bits) */ int pdc_btlb_info(struct pdc_btlb_info *btlb) { - int retval; + int retval; unsigned long flags; - spin_lock_irqsave(&pdc_lock, flags); - retval = mem_pdc_call(PDC_BLOCK_TLB, PDC_BTLB_INFO, __pa(pdc_result), 0); - memcpy(btlb, pdc_result, sizeof(*btlb)); - spin_unlock_irqrestore(&pdc_lock, flags); + if (IS_ENABLED(CONFIG_PA20)) + return PDC_BAD_PROC; - if(retval < 0) { - btlb->max_size = 0; - } - return retval; + spin_lock_irqsave(&pdc_lock, flags); + retval = mem_pdc_call(PDC_BLOCK_TLB, PDC_BTLB_INFO, __pa(pdc_result), 0); + memcpy(btlb, pdc_result, sizeof(*btlb)); + spin_unlock_irqrestore(&pdc_lock, flags); + + if(retval < 0) { + btlb->max_size = 0; + } + return retval; +} + +int pdc_btlb_insert(unsigned long long vpage, unsigned long physpage, unsigned long len, + unsigned long entry_info, unsigned long slot) +{ + int retval; + unsigned long flags; + + if (IS_ENABLED(CONFIG_PA20)) + return PDC_BAD_PROC; + + spin_lock_irqsave(&pdc_lock, flags); + retval = mem_pdc_call(PDC_BLOCK_TLB, PDC_BTLB_INSERT, (unsigned long) (vpage >> 32), + (unsigned long) vpage, physpage, len, entry_info, slot); + spin_unlock_irqrestore(&pdc_lock, flags); + return retval; +} + +int pdc_btlb_purge_all(void) +{ + int retval; + unsigned long flags; + + if (IS_ENABLED(CONFIG_PA20)) + return PDC_BAD_PROC; + + spin_lock_irqsave(&pdc_lock, flags); + retval = mem_pdc_call(PDC_BLOCK_TLB, PDC_BTLB_PURGE_ALL); + spin_unlock_irqrestore(&pdc_lock, flags); + return retval; } /** @@ -725,6 +761,9 @@ int pdc_mem_map_hpa(struct pdc_memory_map *address, int retval; unsigned long flags; + if (IS_ENABLED(CONFIG_PA20)) + return PDC_BAD_PROC; + spin_lock_irqsave(&pdc_lock, flags); memcpy(pdc_result2, mod_path, sizeof(*mod_path)); retval = mem_pdc_call(PDC_MEM_MAP, PDC_MEM_MAP_HPA, __pa(pdc_result), @@ -734,7 +773,6 @@ int pdc_mem_map_hpa(struct pdc_memory_map *address, return retval; } -#endif /* !CONFIG_PA20 */ /** * pdc_lan_station_id - Get the LAN address. @@ -996,8 +1034,8 @@ int pdc_pci_irt(unsigned long num_entries, unsigned long hpa, void *tbl) /** * pdc_pci_config_read - read PCI config space. - * @hpa token from PDC to indicate which PCI device - * @pci_addr configuration space address to read from + * @hpa: Token from PDC to indicate which PCI device + * @cfg_addr: Configuration space address to read from * * Read PCI Configuration space *before* linux PCI subsystem is running. */ @@ -1019,9 +1057,9 @@ unsigned int pdc_pci_config_read(void *hpa, unsigned long cfg_addr) /** * pdc_pci_config_write - read PCI config space. - * @hpa token from PDC to indicate which PCI device - * @pci_addr configuration space address to write - * @val value we want in the 32-bit register + * @hpa: Token from PDC to indicate which PCI device + * @cfg_addr: Configuration space address to write + * @val: Value we want in the 32-bit register * * Write PCI Configuration space *before* linux PCI subsystem is running. */ @@ -1232,15 +1270,18 @@ int __init pdc_soft_power_info(unsigned long *power_reg) } /* - * pdc_soft_power_button - Control the soft power button behaviour - * @sw_control: 0 for hardware control, 1 for software control + * pdc_soft_power_button{_panic} - Control the soft power button behaviour + * @sw_control: 0 for hardware control, 1 for software control * * * This PDC function places the soft power button under software or * hardware control. - * Under software control the OS may control to when to allow to shut - * down the system. Under hardware control pressing the power button + * Under software control the OS may control to when to allow to shut + * down the system. Under hardware control pressing the power button * powers off the system immediately. + * + * The _panic version relies on spin_trylock to prevent deadlock + * on panic path. */ int pdc_soft_power_button(int sw_control) { @@ -1254,6 +1295,22 @@ int pdc_soft_power_button(int sw_control) return retval; } +int pdc_soft_power_button_panic(int sw_control) +{ + int retval; + unsigned long flags; + + if (!spin_trylock_irqsave(&pdc_lock, flags)) { + pr_emerg("Couldn't enable soft power button\n"); + return -EBUSY; /* ignored by the panic notifier */ + } + + retval = mem_pdc_call(PDC_SOFT_POWER, PDC_SOFT_POWER_ENABLE, __pa(pdc_result), sw_control); + spin_unlock_irqrestore(&pdc_lock, flags); + + return retval; +} + /* * pdc_io_reset - Hack to avoid overlapping range registers of Bridges devices. * Primarily a problem on T600 (which parisc-linux doesn't support) but @@ -1303,7 +1360,7 @@ static char iodc_dbuf[4096] __page_aligned_bss; */ int pdc_iodc_print(const unsigned char *str, unsigned count) { - unsigned int i; + unsigned int i, found = 0; unsigned long flags; count = min_t(unsigned int, count, sizeof(iodc_dbuf)); @@ -1315,6 +1372,7 @@ int pdc_iodc_print(const unsigned char *str, unsigned count) iodc_dbuf[i+0] = '\r'; iodc_dbuf[i+1] = '\n'; i += 2; + found = 1; goto print; default: iodc_dbuf[i] = str[i]; @@ -1330,7 +1388,7 @@ print: __pa(pdc_result), 0, __pa(iodc_dbuf), i, 0); spin_unlock_irqrestore(&pdc_lock, flags); - return i; + return i - found; } #if !defined(BOOTLOADER) @@ -1369,17 +1427,25 @@ int pdc_iodc_getc(void) } int pdc_sti_call(unsigned long func, unsigned long flags, - unsigned long inptr, unsigned long outputr, - unsigned long glob_cfg) + unsigned long inptr, unsigned long outputr, + unsigned long glob_cfg, int do_call64) { - int retval; + int retval = 0; unsigned long irqflags; - spin_lock_irqsave(&pdc_lock, irqflags); - retval = real32_call(func, flags, inptr, outputr, glob_cfg); - spin_unlock_irqrestore(&pdc_lock, irqflags); + spin_lock_irqsave(&pdc_lock, irqflags); + if (IS_ENABLED(CONFIG_64BIT) && do_call64) { +#ifdef CONFIG_64BIT + retval = real64_call(func, flags, inptr, outputr, glob_cfg); +#else + WARN_ON(1); +#endif + } else { + retval = real32_call(func, flags, inptr, outputr, glob_cfg); + } + spin_unlock_irqrestore(&pdc_lock, irqflags); - return retval; + return retval; } EXPORT_SYMBOL(pdc_sti_call); @@ -1529,7 +1595,7 @@ int pdc_pat_get_irt(void *r_addr, unsigned long cell_num) /** * pdc_pat_pd_get_addr_map - Retrieve information about memory address ranges. - * @actlen: The return buffer. + * @actual_len: The return buffer. * @mem_addr: Pointer to the memory buffer. * @count: The number of bytes to read from the buffer. * @offset: The offset with respect to the beginning of the buffer. @@ -1552,7 +1618,7 @@ int pdc_pat_pd_get_addr_map(unsigned long *actual_len, void *mem_addr, } /** - * pdc_pat_pd_get_PDC_interface_revisions - Retrieve PDC interface revisions. + * pdc_pat_pd_get_pdc_revisions - Retrieve PDC interface revisions. * @legacy_rev: The legacy revision. * @pat_rev: The PAT revision. * @pdc_cap: The PDC capabilities. @@ -1607,7 +1673,7 @@ int pdc_pat_io_pci_cfg_read(unsigned long pci_addr, int pci_size, u32 *mem_addr) * pdc_pat_io_pci_cfg_write - Retrieve information about memory address ranges. * @pci_addr: PCI configuration space address for which the write request is being made. * @pci_size: Size of write in bytes. Valid values are 1, 2, and 4. - * @value: Pointer to 1, 2, or 4 byte value in low order end of argument to be + * @val: Pointer to 1, 2, or 4 byte value in low order end of argument to be * written to PCI Config space. * */ @@ -1625,7 +1691,7 @@ int pdc_pat_io_pci_cfg_write(unsigned long pci_addr, int pci_size, u32 val) } /** - * pdc_pat_mem_pdc_info - Retrieve information about page deallocation table + * pdc_pat_mem_pdt_info - Retrieve information about page deallocation table * @rinfo: memory pdt information * */ diff --git a/arch/parisc/kernel/ftrace.c b/arch/parisc/kernel/ftrace.c index 4d392e4ed358..10fd5b3e63e7 100644 --- a/arch/parisc/kernel/ftrace.c +++ b/arch/parisc/kernel/ftrace.c @@ -20,7 +20,7 @@ #include <asm/assembly.h> #include <asm/sections.h> #include <asm/ftrace.h> -#include <asm/patch.h> +#include <asm/text-patching.h> #define __hot __section(".text.hot") @@ -53,7 +53,7 @@ static void __hot prepare_ftrace_return(unsigned long *parent, static ftrace_func_t ftrace_func; -void notrace __hot ftrace_function_trampoline(unsigned long parent, +asmlinkage void notrace __hot ftrace_function_trampoline(unsigned long parent, unsigned long self_addr, unsigned long org_sp_gr3, struct ftrace_regs *fregs) @@ -78,7 +78,7 @@ void notrace __hot ftrace_function_trampoline(unsigned long parent, #endif } -#ifdef CONFIG_FUNCTION_GRAPH_TRACER +#if defined(CONFIG_DYNAMIC_FTRACE) && defined(CONFIG_FUNCTION_GRAPH_TRACER) int ftrace_enable_ftrace_graph_caller(void) { static_key_enable(&ftrace_graph_enable.key); @@ -87,7 +87,7 @@ int ftrace_enable_ftrace_graph_caller(void) int ftrace_disable_ftrace_graph_caller(void) { - static_key_enable(&ftrace_graph_enable.key); + static_key_disable(&ftrace_graph_enable.key); return 0; } #endif @@ -206,6 +206,9 @@ void kprobe_ftrace_handler(unsigned long ip, unsigned long parent_ip, struct kprobe *p; int bit; + if (unlikely(kprobe_ftrace_disabled)) + return; + bit = ftrace_test_recursion_trylock(ip, parent_ip); if (bit < 0) return; diff --git a/arch/parisc/kernel/head.S b/arch/parisc/kernel/head.S index fd15fd4bbb61..96e0264ac961 100644 --- a/arch/parisc/kernel/head.S +++ b/arch/parisc/kernel/head.S @@ -70,9 +70,8 @@ $bss_loop: stw,ma %arg2,4(%r1) stw,ma %arg3,4(%r1) -#if !defined(CONFIG_64BIT) && defined(CONFIG_PA20) - /* This 32-bit kernel was compiled for PA2.0 CPUs. Check current CPU - * and halt kernel if we detect a PA1.x CPU. */ +#if defined(CONFIG_PA20) + /* check for 64-bit capable CPU as required by current kernel */ ldi 32,%r10 mtctl %r10,%cr11 .level 2.0 @@ -180,10 +179,10 @@ $pgt_fill_loop: std %dp,0x18(%r10) #endif -#ifdef CONFIG_64BIT - /* Get PDCE_PROC for monarch CPU. */ #define MEM_PDC_LO 0x388 #define MEM_PDC_HI 0x35C +#ifdef CONFIG_64BIT + /* Get PDCE_PROC for monarch CPU. */ ldw MEM_PDC_LO(%r0),%r3 ldw MEM_PDC_HI(%r0),%r10 depd %r10, 31, 32, %r3 /* move to upper word */ @@ -269,7 +268,17 @@ stext_pdc_ret: tovirt_r1 %r6 mtctl %r6,%cr30 /* restore task thread info */ #endif - + +#ifndef CONFIG_64BIT + /* clear all BTLBs */ + ldi PDC_BLOCK_TLB,%arg0 + load32 PA(stext_pdc_btlb_ret), %rp + ldw MEM_PDC_LO(%r0),%r3 + bv (%r3) + ldi PDC_BTLB_PURGE_ALL,%arg1 +stext_pdc_btlb_ret: +#endif + /* PARANOID: clear user scratch/user space SR's */ mtsp %r0,%sr0 mtsp %r0,%sr1 diff --git a/arch/parisc/kernel/irq.c b/arch/parisc/kernel/irq.c index b05055f3ba4b..dff66be65d29 100644 --- a/arch/parisc/kernel/irq.c +++ b/arch/parisc/kernel/irq.c @@ -24,9 +24,6 @@ #undef PARISC_IRQ_CR16_COUNTS -extern irqreturn_t timer_interrupt(int, void *); -extern irqreturn_t ipi_interrupt(int, void *); - #define EIEM_MASK(irq) (1UL<<(CPU_IRQ_MAX - irq)) /* Bits in EIEM correlate with cpu_irq_action[]. @@ -368,7 +365,7 @@ union irq_stack_union { volatile unsigned int lock[1]; }; -DEFINE_PER_CPU(union irq_stack_union, irq_stack_union) = { +static DEFINE_PER_CPU(union irq_stack_union, irq_stack_union) = { .slock = { 1,1,1,1 }, }; #endif @@ -489,7 +486,7 @@ void do_softirq_own_stack(void) #endif /* CONFIG_IRQSTACKS */ /* ONLY called from entry.S:intr_extint() */ -void do_cpu_irq_mask(struct pt_regs *regs) +asmlinkage void do_cpu_irq_mask(struct pt_regs *regs) { struct pt_regs *old_regs; unsigned long eirr_val; @@ -501,7 +498,7 @@ void do_cpu_irq_mask(struct pt_regs *regs) old_regs = set_irq_regs(regs); local_irq_disable(); - irq_enter(); + irq_enter_rcu(); eirr_val = mfctl(23) & cpu_eiem & per_cpu(local_ack_eiem, cpu); if (!eirr_val) @@ -536,7 +533,7 @@ void do_cpu_irq_mask(struct pt_regs *regs) #endif /* CONFIG_IRQSTACKS */ out: - irq_exit(); + irq_exit_rcu(); set_irq_regs(old_regs); return; diff --git a/arch/parisc/kernel/jump_label.c b/arch/parisc/kernel/jump_label.c index e253b134500d..ea51f15bf0e6 100644 --- a/arch/parisc/kernel/jump_label.c +++ b/arch/parisc/kernel/jump_label.c @@ -8,7 +8,7 @@ #include <linux/jump_label.h> #include <linux/bug.h> #include <asm/alternative.h> -#include <asm/patch.h> +#include <asm/text-patching.h> static inline int reassemble_17(int as17) { diff --git a/arch/parisc/kernel/kexec.c b/arch/parisc/kernel/kexec.c index 5eb7f30edc1f..db57345a9daf 100644 --- a/arch/parisc/kernel/kexec.c +++ b/arch/parisc/kernel/kexec.c @@ -4,6 +4,8 @@ #include <linux/console.h> #include <linux/kexec.h> #include <linux/delay.h> +#include <linux/reboot.h> + #include <asm/cacheflush.h> #include <asm/sections.h> diff --git a/arch/parisc/kernel/kexec_file.c b/arch/parisc/kernel/kexec_file.c index 8c534204f0fd..3fc82130b6c3 100644 --- a/arch/parisc/kernel/kexec_file.c +++ b/arch/parisc/kernel/kexec_file.c @@ -38,8 +38,8 @@ static void *elf_load(struct kimage *image, char *kernel_buf, for (i = 0; i < image->nr_segments; i++) image->segment[i].mem = __pa(image->segment[i].mem); - pr_debug("Loaded the kernel at 0x%lx, entry at 0x%lx\n", - kernel_load_addr, image->start); + kexec_dprintk("Loaded the kernel at 0x%lx, entry at 0x%lx\n", + kernel_load_addr, image->start); if (initrd != NULL) { kbuf.buffer = initrd; @@ -51,7 +51,7 @@ static void *elf_load(struct kimage *image, char *kernel_buf, if (ret) goto out; - pr_debug("Loaded initrd at 0x%lx\n", kbuf.mem); + kexec_dprintk("Loaded initrd at 0x%lx\n", kbuf.mem); image->arch.initrd_start = kbuf.mem; image->arch.initrd_end = kbuf.mem + initrd_len; } @@ -68,7 +68,7 @@ static void *elf_load(struct kimage *image, char *kernel_buf, if (ret) goto out; - pr_debug("Loaded cmdline at 0x%lx\n", kbuf.mem); + kexec_dprintk("Loaded cmdline at 0x%lx\n", kbuf.mem); image->arch.cmdline = kbuf.mem; } out: diff --git a/arch/parisc/kernel/kgdb.c b/arch/parisc/kernel/kgdb.c index b16fa9bac5f4..fee81f877525 100644 --- a/arch/parisc/kernel/kgdb.c +++ b/arch/parisc/kernel/kgdb.c @@ -16,7 +16,7 @@ #include <asm/ptrace.h> #include <asm/traps.h> #include <asm/processor.h> -#include <asm/patch.h> +#include <asm/text-patching.h> #include <asm/cacheflush.h> const struct kgdb_arch arch_kgdb_ops = { diff --git a/arch/parisc/kernel/kprobes.c b/arch/parisc/kernel/kprobes.c index 6e0b86652f30..9255adba67a3 100644 --- a/arch/parisc/kernel/kprobes.c +++ b/arch/parisc/kernel/kprobes.c @@ -12,7 +12,7 @@ #include <linux/kprobes.h> #include <linux/slab.h> #include <asm/cacheflush.h> -#include <asm/patch.h> +#include <asm/text-patching.h> DEFINE_PER_CPU(struct kprobe *, current_kprobe) = NULL; DEFINE_PER_CPU(struct kprobe_ctlblk, kprobe_ctlblk); diff --git a/arch/parisc/kernel/module.c b/arch/parisc/kernel/module.c index 7df140545b22..4e5d991b2b65 100644 --- a/arch/parisc/kernel/module.c +++ b/arch/parisc/kernel/module.c @@ -27,9 +27,9 @@ * We are not doing SEGREL32 handling correctly. According to the ABI, we * should do a value offset, like this: * if (in_init(me, (void *)val)) - * val -= (uint32_t)me->init_layout.base; + * val -= (uint32_t)me->mem[MOD_INIT_TEXT].base; * else - * val -= (uint32_t)me->core_layout.base; + * val -= (uint32_t)me->mem[MOD_TEXT].base; * However, SEGREL32 is used only for PARISC unwind entries, and we want * those entries to have an absolute address, and not just an offset. * @@ -41,7 +41,6 @@ #include <linux/moduleloader.h> #include <linux/elf.h> -#include <linux/vmalloc.h> #include <linux/fs.h> #include <linux/ftrace.h> #include <linux/string.h> @@ -76,25 +75,6 @@ * allows us to allocate up to 4095 GOT entries. */ #define MAX_GOTS 4095 -/* three functions to determine where in the module core - * or init pieces the location is */ -static inline int in_init(struct module *me, void *loc) -{ - return (loc >= me->init_layout.base && - loc <= (me->init_layout.base + me->init_layout.size)); -} - -static inline int in_core(struct module *me, void *loc) -{ - return (loc >= me->core_layout.base && - loc <= (me->core_layout.base + me->core_layout.size)); -} - -static inline int in_local(struct module *me, void *loc) -{ - return in_init(me, loc) || in_core(me, loc); -} - #ifndef CONFIG_64BIT struct got_entry { Elf32_Addr addr; @@ -192,17 +172,6 @@ static inline int reassemble_22(int as22) ((as22 & 0x0003ff) << 3)); } -void *module_alloc(unsigned long size) -{ - /* using RWX means less protection for modules, but it's - * easier than trying to map the text, data, init_text and - * init_data correctly */ - return __vmalloc_node_range(size, 1, VMALLOC_START, VMALLOC_END, - GFP_KERNEL, - PAGE_KERNEL_RWX, 0, NUMA_NO_NODE, - __builtin_return_address(0)); -} - #ifndef CONFIG_64BIT static inline unsigned long count_gots(const Elf_Rela *rela, unsigned long n) { @@ -302,6 +271,7 @@ int module_frob_arch_sections(CONST Elf_Ehdr *hdr, { unsigned long gots = 0, fdescs = 0, len; unsigned int i; + struct module_memory *mod_mem; len = hdr->e_shnum * sizeof(me->arch.section[0]); me->arch.section = kzalloc(len, GFP_KERNEL); @@ -346,14 +316,15 @@ int module_frob_arch_sections(CONST Elf_Ehdr *hdr, me->arch.section[s].stub_entries += count; } + mod_mem = &me->mem[MOD_TEXT]; /* align things a bit */ - me->core_layout.size = ALIGN(me->core_layout.size, 16); - me->arch.got_offset = me->core_layout.size; - me->core_layout.size += gots * sizeof(struct got_entry); + mod_mem->size = ALIGN(mod_mem->size, 16); + me->arch.got_offset = mod_mem->size; + mod_mem->size += gots * sizeof(struct got_entry); - me->core_layout.size = ALIGN(me->core_layout.size, 16); - me->arch.fdesc_offset = me->core_layout.size; - me->core_layout.size += fdescs * sizeof(Elf_Fdesc); + mod_mem->size = ALIGN(mod_mem->size, 16); + me->arch.fdesc_offset = mod_mem->size; + mod_mem->size += fdescs * sizeof(Elf_Fdesc); me->arch.got_max = gots; me->arch.fdesc_max = fdescs; @@ -371,7 +342,7 @@ static Elf64_Word get_got(struct module *me, unsigned long value, long addend) BUG_ON(value == 0); - got = me->core_layout.base + me->arch.got_offset; + got = me->mem[MOD_TEXT].base + me->arch.got_offset; for (i = 0; got[i].addr; i++) if (got[i].addr == value) goto out; @@ -389,7 +360,7 @@ static Elf64_Word get_got(struct module *me, unsigned long value, long addend) #ifdef CONFIG_64BIT static Elf_Addr get_fdesc(struct module *me, unsigned long value) { - Elf_Fdesc *fdesc = me->core_layout.base + me->arch.fdesc_offset; + Elf_Fdesc *fdesc = me->mem[MOD_TEXT].base + me->arch.fdesc_offset; if (!value) { printk(KERN_ERR "%s: zero OPD requested!\n", me->name); @@ -407,7 +378,7 @@ static Elf_Addr get_fdesc(struct module *me, unsigned long value) /* Create new one */ fdesc->addr = value; - fdesc->gp = (Elf_Addr)me->core_layout.base + me->arch.got_offset; + fdesc->gp = (Elf_Addr)me->mem[MOD_TEXT].base + me->arch.got_offset; return (Elf_Addr)fdesc; } #endif /* CONFIG_64BIT */ @@ -742,7 +713,7 @@ int apply_relocate_add(Elf_Shdr *sechdrs, loc, val); val += addend; /* can we reach it locally? */ - if (in_local(me, (void *)val)) { + if (within_module(val, me)) { /* this is the case where the symbol is local * to the module, but in a different section, * so stub the jump in case it's more than 22 @@ -801,7 +772,7 @@ int apply_relocate_add(Elf_Shdr *sechdrs, break; case R_PARISC_FPTR64: /* 64-bit function address */ - if(in_local(me, (void *)(val + addend))) { + if (within_module(val + addend, me)) { *loc64 = get_fdesc(me, val+addend); pr_debug("FDESC for %s at %llx points to %llx\n", strtab + sym->st_name, *loc64, @@ -839,7 +810,7 @@ register_unwind_table(struct module *me, table = (unsigned char *)sechdrs[me->arch.unwind_section].sh_addr; end = table + sechdrs[me->arch.unwind_section].sh_size; - gp = (Elf_Addr)me->core_layout.base + me->arch.got_offset; + gp = (Elf_Addr)me->mem[MOD_TEXT].base + me->arch.got_offset; pr_debug("register_unwind_table(), sect = %d at 0x%p - 0x%p (gp=0x%lx)\n", me->arch.unwind_section, table, end, gp); @@ -862,7 +833,7 @@ int module_finalize(const Elf_Ehdr *hdr, const char *strtab = NULL; const Elf_Shdr *s; char *secstrings; - int symindex = -1; + int symindex __maybe_unused = -1; Elf_Sym *newptr, *oldptr; Elf_Shdr *symhdr = NULL; #ifdef DEBUG @@ -977,7 +948,7 @@ void module_arch_cleanup(struct module *mod) #ifdef CONFIG_64BIT void *dereference_module_function_descriptor(struct module *mod, void *ptr) { - unsigned long start_opd = (Elf64_Addr)mod->core_layout.base + + unsigned long start_opd = (Elf64_Addr)mod->mem[MOD_TEXT].base + mod->arch.fdesc_offset; unsigned long end_opd = start_opd + mod->arch.fdesc_count * sizeof(Elf64_Fdesc); diff --git a/arch/parisc/kernel/pa7300lc.c b/arch/parisc/kernel/pa7300lc.c deleted file mode 100644 index 0d770ac83f70..000000000000 --- a/arch/parisc/kernel/pa7300lc.c +++ /dev/null @@ -1,51 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0 -/* - * linux/arch/parisc/kernel/pa7300lc.c - * - PA7300LC-specific functions - * - * Copyright (C) 2000 Philipp Rumpf */ - -#include <linux/sched.h> -#include <linux/sched/debug.h> -#include <linux/smp.h> -#include <linux/kernel.h> -#include <asm/io.h> -#include <asm/ptrace.h> -#include <asm/machdep.h> - -/* CPU register indices */ - -#define MIOC_STATUS 0xf040 -#define MIOC_CONTROL 0xf080 -#define MDERRADD 0xf0e0 -#define DMAERR 0xf0e8 -#define DIOERR 0xf0ec -#define HIDMAMEM 0xf0f4 - -/* this returns the HPA of the CPU it was called on */ -static u32 cpu_hpa(void) -{ - return 0xfffb0000; -} - -static void pa7300lc_lpmc(int code, struct pt_regs *regs) -{ - u32 hpa; - printk(KERN_WARNING "LPMC on CPU %d\n", smp_processor_id()); - - show_regs(regs); - - hpa = cpu_hpa(); - printk(KERN_WARNING - "MIOC_CONTROL %08x\n" "MIOC_STATUS %08x\n" - "MDERRADD %08x\n" "DMAERR %08x\n" - "DIOERR %08x\n" "HIDMAMEM %08x\n", - gsc_readl(hpa+MIOC_CONTROL), gsc_readl(hpa+MIOC_STATUS), - gsc_readl(hpa+MDERRADD), gsc_readl(hpa+DMAERR), - gsc_readl(hpa+DIOERR), gsc_readl(hpa+HIDMAMEM)); -} - -void pa7300lc_init(void) -{ - cpu_lpmc = pa7300lc_lpmc; -} diff --git a/arch/parisc/kernel/pacache.S b/arch/parisc/kernel/pacache.S index 9a0018f1f42c..541370d14559 100644 --- a/arch/parisc/kernel/pacache.S +++ b/arch/parisc/kernel/pacache.S @@ -889,6 +889,7 @@ ENDPROC_CFI(flush_icache_page_asm) ENTRY_CFI(flush_kernel_dcache_page_asm) 88: ldil L%dcache_stride, %r1 ldw R%dcache_stride(%r1), %r23 + depi_safe 0, 31,PAGE_SHIFT, %r26 /* Clear any offset bits */ #ifdef CONFIG_64BIT depdi,z 1, 63-PAGE_SHIFT,1, %r25 @@ -925,6 +926,7 @@ ENDPROC_CFI(flush_kernel_dcache_page_asm) ENTRY_CFI(purge_kernel_dcache_page_asm) 88: ldil L%dcache_stride, %r1 ldw R%dcache_stride(%r1), %r23 + depi_safe 0, 31,PAGE_SHIFT, %r26 /* Clear any offset bits */ #ifdef CONFIG_64BIT depdi,z 1, 63-PAGE_SHIFT,1, %r25 diff --git a/arch/parisc/kernel/parisc_ksyms.c b/arch/parisc/kernel/parisc_ksyms.c index 00297e8e1c88..509146a52725 100644 --- a/arch/parisc/kernel/parisc_ksyms.c +++ b/arch/parisc/kernel/parisc_ksyms.c @@ -14,6 +14,7 @@ #include <linux/module.h> #include <linux/kernel.h> #include <linux/syscalls.h> +#include <linux/libgcc.h> #include <linux/string.h> EXPORT_SYMBOL(memset); @@ -21,6 +22,8 @@ EXPORT_SYMBOL(memset); #include <linux/atomic.h> EXPORT_SYMBOL(__xchg8); EXPORT_SYMBOL(__xchg32); +EXPORT_SYMBOL(__cmpxchg_u8); +EXPORT_SYMBOL(__cmpxchg_u16); EXPORT_SYMBOL(__cmpxchg_u32); EXPORT_SYMBOL(__cmpxchg_u64); #ifdef CONFIG_SMP @@ -40,9 +43,6 @@ EXPORT_SYMBOL($global$); #endif #include <asm/io.h> -EXPORT_SYMBOL(memcpy_toio); -EXPORT_SYMBOL(memcpy_fromio); -EXPORT_SYMBOL(memset_io); extern void $$divI(void); extern void $$divU(void); @@ -92,12 +92,6 @@ EXPORT_SYMBOL($$divI_12); EXPORT_SYMBOL($$divI_14); EXPORT_SYMBOL($$divI_15); -extern void __ashrdi3(void); -extern void __ashldi3(void); -extern void __lshrdi3(void); -extern void __muldi3(void); -extern void __ucmpdi2(void); - EXPORT_SYMBOL(__ashrdi3); EXPORT_SYMBOL(__ashldi3); EXPORT_SYMBOL(__lshrdi3); diff --git a/arch/parisc/kernel/patch.c b/arch/parisc/kernel/patch.c index e59574f65e64..35dd764b871e 100644 --- a/arch/parisc/kernel/patch.c +++ b/arch/parisc/kernel/patch.c @@ -13,7 +13,7 @@ #include <asm/cacheflush.h> #include <asm/fixmap.h> -#include <asm/patch.h> +#include <asm/text-patching.h> struct patch { void *addr; diff --git a/arch/parisc/kernel/pci-dma.c b/arch/parisc/kernel/pci-dma.c index ba87f791323b..bf9f192c826e 100644 --- a/arch/parisc/kernel/pci-dma.c +++ b/arch/parisc/kernel/pci-dma.c @@ -39,7 +39,7 @@ static struct proc_dir_entry * proc_gsc_root __read_mostly = NULL; static unsigned long pcxl_used_bytes __read_mostly; static unsigned long pcxl_used_pages __read_mostly; -extern unsigned long pcxl_dma_start; /* Start of pcxl dma mapping area */ +unsigned long pcxl_dma_start __ro_after_init; /* pcxl dma mapping area start */ static DEFINE_SPINLOCK(pcxl_res_lock); static char *pcxl_res_map; static int pcxl_res_hint; @@ -164,7 +164,7 @@ static inline void unmap_uncached_pte(pmd_t * pmd, unsigned long vaddr, pmd_clear(pmd); return; } - pte = pte_offset_map(pmd, vaddr); + pte = pte_offset_kernel(pmd, vaddr); vaddr &= ~PMD_MASK; end = vaddr + size; if (end > PMD_SIZE) @@ -245,7 +245,7 @@ static void unmap_uncached_pages(unsigned long vaddr, unsigned long size) PCXL_SEARCH_LOOP(idx, mask, size); \ } -unsigned long +static unsigned long pcxl_alloc_range(size_t size) { int res_idx; @@ -381,7 +381,7 @@ pcxl_dma_init(void) pcxl_res_map = (char *)__get_free_pages(GFP_KERNEL, get_order(pcxl_res_size)); memset(pcxl_res_map, 0, pcxl_res_size); - proc_gsc_root = proc_mkdir("gsc", NULL); + proc_gsc_root = proc_mkdir("bus/gsc", NULL); if (!proc_gsc_root) printk(KERN_WARNING "pcxl_dma_init: Unable to create gsc /proc dir entry\n"); @@ -417,14 +417,6 @@ void *arch_dma_alloc(struct device *dev, size_t size, map_uncached_pages(vaddr, size, paddr); *dma_handle = (dma_addr_t) paddr; -#if 0 -/* This probably isn't needed to support EISA cards. -** ISA cards will certainly only support 24-bit DMA addressing. -** Not clear if we can, want, or need to support ISA. -*/ - if (!dev || *dev->coherent_dma_mask < 0xffffffff) - gfp |= GFP_DMA; -#endif return (void *)vaddr; } @@ -446,11 +438,27 @@ void arch_dma_free(struct device *dev, size_t size, void *vaddr, void arch_sync_dma_for_device(phys_addr_t paddr, size_t size, enum dma_data_direction dir) { + /* + * fdc: The data cache line is written back to memory, if and only if + * it is dirty, and then invalidated from the data cache. + */ flush_kernel_dcache_range((unsigned long)phys_to_virt(paddr), size); } void arch_sync_dma_for_cpu(phys_addr_t paddr, size_t size, enum dma_data_direction dir) { - flush_kernel_dcache_range((unsigned long)phys_to_virt(paddr), size); + unsigned long addr = (unsigned long) phys_to_virt(paddr); + + switch (dir) { + case DMA_TO_DEVICE: + case DMA_BIDIRECTIONAL: + flush_kernel_dcache_range(addr, size); + return; + case DMA_FROM_DEVICE: + purge_kernel_dcache_range_asm(addr, addr + size); + return; + default: + BUG(); + } } diff --git a/arch/parisc/kernel/pdc_chassis.c b/arch/parisc/kernel/pdc_chassis.c index da154406d368..d477d0177c2f 100644 --- a/arch/parisc/kernel/pdc_chassis.c +++ b/arch/parisc/kernel/pdc_chassis.c @@ -31,6 +31,7 @@ #include <asm/processor.h> #include <asm/pdc.h> #include <asm/pdcpat.h> +#include <asm/led.h> #define PDC_CHASSIS_VER "0.05" @@ -40,7 +41,7 @@ static unsigned int pdc_chassis_enabled __read_mostly = 1; /** * pdc_chassis_setup() - Enable/disable pdc_chassis code at boot time. - * @str configuration param: 0 to disable chassis log + * @str: configuration param: 0 to disable chassis log * @return 1 */ @@ -55,7 +56,6 @@ __setup("pdcchassis=", pdc_chassis_setup); /** * pdc_chassis_checkold() - Checks for old PDC_CHASSIS compatibility - * @pdc_chassis_old: 1 if old pdc chassis style * * Currently, only E class and A180 are known to work with this. * Inspired by Christoph Plattner @@ -80,6 +80,9 @@ static void __init pdc_chassis_checkold(void) /** * pdc_chassis_panic_event() - Called by the panic handler. + * @this: unused + * @event: unused + * @ptr: unused * * As soon as a panic occurs, we should inform the PDC. */ @@ -88,7 +91,7 @@ static int pdc_chassis_panic_event(struct notifier_block *this, unsigned long event, void *ptr) { pdc_chassis_send_status(PDC_CHASSIS_DIRECT_PANIC); - return NOTIFY_DONE; + return NOTIFY_DONE; } @@ -99,7 +102,10 @@ static struct notifier_block pdc_chassis_panic_block = { /** - * parisc_reboot_event() - Called by the reboot handler. + * pdc_chassis_reboot_event() - Called by the reboot handler. + * @this: unused + * @event: unused + * @ptr: unused * * As soon as a reboot occurs, we should inform the PDC. */ @@ -108,7 +114,7 @@ static int pdc_chassis_reboot_event(struct notifier_block *this, unsigned long event, void *ptr) { pdc_chassis_send_status(PDC_CHASSIS_DIRECT_SHUTDOWN); - return NOTIFY_DONE; + return NOTIFY_DONE; } @@ -148,7 +154,7 @@ void __init parisc_pdc_chassis_init(void) /** * pdc_chassis_send_status() - Sends a predefined message to the chassis, * and changes the front panel LEDs according to the new system state - * @retval: PDC call return value. + * @message: Type of message, one of PDC_CHASSIS_DIRECT_* values. * * Only machines with 64 bits PDC PAT and those reported in * pdc_chassis_checkold() are supported atm. @@ -229,6 +235,11 @@ int pdc_chassis_send_status(int message) } else retval = -1; #endif /* CONFIG_64BIT */ } /* if (pdc_chassis_enabled) */ + + /* if system has LCD display, update current string */ + if (retval != -1 && IS_ENABLED(CONFIG_CHASSIS_LCD_LED)) + lcd_print(NULL); + #endif /* CONFIG_PDC_CHASSIS */ return retval; } diff --git a/arch/parisc/kernel/pdt.c b/arch/parisc/kernel/pdt.c index 80943a00e245..b70b67adb855 100644 --- a/arch/parisc/kernel/pdt.c +++ b/arch/parisc/kernel/pdt.c @@ -16,6 +16,7 @@ #include <linux/memblock.h> #include <linux/seq_file.h> #include <linux/kthread.h> +#include <linux/proc_fs.h> #include <linux/initrd.h> #include <linux/pgtable.h> #include <linux/mm.h> @@ -23,6 +24,7 @@ #include <asm/pdc.h> #include <asm/pdcpat.h> #include <asm/sections.h> +#include <asm/pgtable.h> enum pdt_access_type { PDT_NONE, @@ -61,6 +63,7 @@ static unsigned long pdt_entry[MAX_PDT_ENTRIES] __page_aligned_bss; #define PDT_ADDR_PERM_ERR (pdt_type != PDT_PDC ? 2UL : 0UL) #define PDT_ADDR_SINGLE_ERR 1UL +#ifdef CONFIG_PROC_FS /* report PDT entries via /proc/meminfo */ void arch_report_meminfo(struct seq_file *m) { @@ -72,6 +75,7 @@ void arch_report_meminfo(struct seq_file *m) seq_printf(m, "PDT_cur_entries: %7lu\n", pdt_status.pdt_entries); } +#endif static int get_info_pat_new(void) { @@ -352,10 +356,8 @@ static int __init pdt_initcall(void) return -ENODEV; kpdtd_task = kthread_run(pdt_mainloop, NULL, "kpdtd"); - if (IS_ERR(kpdtd_task)) - return PTR_ERR(kpdtd_task); - return 0; + return PTR_ERR_OR_ZERO(kpdtd_task); } late_initcall(pdt_initcall); diff --git a/arch/parisc/kernel/perf.c b/arch/parisc/kernel/perf.c index d46b6709ec56..5e10f98ce7b5 100644 --- a/arch/parisc/kernel/perf.c +++ b/arch/parisc/kernel/perf.c @@ -57,7 +57,7 @@ struct rdr_tbl_ent { static int perf_processor_interface __read_mostly = UNKNOWN_INTF; static int perf_enabled __read_mostly; static DEFINE_SPINLOCK(perf_lock); -struct parisc_device *cpu_device __read_mostly; +static struct parisc_device *cpu_device __read_mostly; /* RDRs to write for PCX-W */ static const int perf_rdrs_W[] = @@ -288,7 +288,7 @@ static ssize_t perf_read(struct file *file, char __user *buf, size_t cnt, loff_t static ssize_t perf_write(struct file *file, const char __user *buf, size_t count, loff_t *ppos) { - size_t image_size; + size_t image_size __maybe_unused; uint32_t image_type; uint32_t interface_type; uint32_t test; @@ -466,7 +466,6 @@ static long perf_ioctl(struct file *file, unsigned int cmd, unsigned long arg) } static const struct file_operations perf_fops = { - .llseek = no_llseek, .read = perf_read, .write = perf_write, .unlocked_ioctl = perf_ioctl, @@ -476,9 +475,9 @@ static const struct file_operations perf_fops = { }; static struct miscdevice perf_dev = { - MISC_DYNAMIC_MINOR, - PA_PERF_DEV, - &perf_fops + .minor = MISC_DYNAMIC_MINOR, + .name = PA_PERF_DEV, + .fops = &perf_fops, }; /* diff --git a/arch/parisc/kernel/perf_event.c b/arch/parisc/kernel/perf_event.c new file mode 100644 index 000000000000..f90b83886ab4 --- /dev/null +++ b/arch/parisc/kernel/perf_event.c @@ -0,0 +1,27 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Performance event support for parisc + * + * Copyright (C) 2025 by Helge Deller <deller@gmx.de> + */ + +#include <linux/kernel.h> +#include <linux/perf_event.h> +#include <asm/unwind.h> + +void perf_callchain_kernel(struct perf_callchain_entry_ctx *entry, + struct pt_regs *regs) +{ + + struct unwind_frame_info info; + + unwind_frame_init_task(&info, current, NULL); + while (1) { + if (unwind_once(&info) < 0 || info.ip == 0) + break; + + if (!__kernel_text_address(info.ip) || + perf_callchain_store(entry, info.ip)) + return; + } +} diff --git a/arch/parisc/kernel/perf_regs.c b/arch/parisc/kernel/perf_regs.c new file mode 100644 index 000000000000..10a1a5f06a18 --- /dev/null +++ b/arch/parisc/kernel/perf_regs.c @@ -0,0 +1,61 @@ +// SPDX-License-Identifier: GPL-2.0 + +/* Copyright (C) 2025 by Helge Deller <deller@gmx.de> */ + +#include <linux/perf_event.h> +#include <linux/perf_regs.h> +#include <asm/ptrace.h> + +u64 perf_reg_value(struct pt_regs *regs, int idx) +{ + switch (idx) { + case PERF_REG_PARISC_R0 ... PERF_REG_PARISC_R31: + return regs->gr[idx - PERF_REG_PARISC_R0]; + case PERF_REG_PARISC_SR0 ... PERF_REG_PARISC_SR7: + return regs->sr[idx - PERF_REG_PARISC_SR0]; + case PERF_REG_PARISC_IASQ0 ... PERF_REG_PARISC_IASQ1: + return regs->iasq[idx - PERF_REG_PARISC_IASQ0]; + case PERF_REG_PARISC_IAOQ0 ... PERF_REG_PARISC_IAOQ1: + return regs->iasq[idx - PERF_REG_PARISC_IAOQ0]; + case PERF_REG_PARISC_SAR: /* CR11 */ + return regs->sar; + case PERF_REG_PARISC_IIR: /* CR19 */ + return regs->iir; + case PERF_REG_PARISC_ISR: /* CR20 */ + return regs->isr; + case PERF_REG_PARISC_IOR: /* CR21 */ + return regs->ior; + case PERF_REG_PARISC_IPSW: /* CR22 */ + return regs->ipsw; + } + WARN_ON_ONCE((u32)idx >= PERF_REG_PARISC_MAX); + return 0; +} + +#define REG_RESERVED (~((1ULL << PERF_REG_PARISC_MAX) - 1)) + +int perf_reg_validate(u64 mask) +{ + if (!mask || mask & REG_RESERVED) + return -EINVAL; + + return 0; +} + +u64 perf_reg_abi(struct task_struct *task) +{ + if (!IS_ENABLED(CONFIG_64BIT)) + return PERF_SAMPLE_REGS_ABI_32; + + if (test_tsk_thread_flag(task, TIF_32BIT)) + return PERF_SAMPLE_REGS_ABI_32; + + return PERF_SAMPLE_REGS_ABI_64; +} + +void perf_get_regs_user(struct perf_regs *regs_user, + struct pt_regs *regs) +{ + regs_user->regs = task_pt_regs(current); + regs_user->abi = perf_reg_abi(current); +} diff --git a/arch/parisc/kernel/process.c b/arch/parisc/kernel/process.c index c4f8374c7018..e64ab5d2a40d 100644 --- a/arch/parisc/kernel/process.c +++ b/arch/parisc/kernel/process.c @@ -97,18 +97,12 @@ void machine_restart(char *cmd) } -void (*chassis_power_off)(void); - /* * This routine is called from sys_reboot to actually turn off the * machine */ void machine_power_off(void) { - /* If there is a registered power off handler, call it. */ - if (chassis_power_off) - chassis_power_off(); - /* Put the soft power button back under hardware control. * If the user had already pressed the power button, the * following call will immediately power off. */ @@ -122,13 +116,18 @@ void machine_power_off(void) /* It seems we have no way to power the system off via * software. The user has to press the button himself. */ - printk(KERN_EMERG "System shut down completed.\n" - "Please power this system off now."); + printk("Power off or press RETURN to reboot.\n"); /* prevent soft lockup/stalled CPU messages for endless loop. */ rcu_sysrq_start(); lockup_detector_soft_poweroff(); - for (;;); + while (1) { + /* reboot if user presses RETURN key */ + if (pdc_iodc_getc() == 13) { + printk("Rebooting...\n"); + machine_restart(NULL); + } + } } void (*pm_power_off)(void); @@ -159,15 +158,15 @@ EXPORT_SYMBOL(running_on_qemu); /* * Called from the idle thread for the CPU which has been shutdown. */ -void arch_cpu_idle_dead(void) +void __noreturn arch_cpu_idle_dead(void) { #ifdef CONFIG_HOTPLUG_CPU idle_task_exit(); local_irq_disable(); - /* Tell __cpu_die() that this CPU is now safe to dispose of. */ - (void)cpu_report_death(); + /* Tell the core that this CPU is now safe to dispose of. */ + cpuhp_ap_report_dead(); /* Ensure that the cache lines are written out. */ flush_cache_all_local(); @@ -183,8 +182,6 @@ void arch_cpu_idle_dead(void) void __cpuidle arch_cpu_idle(void) { - raw_local_irq_enable(); - /* nop on real hardware, qemu will idle sleep. */ asm volatile("or %%r10,%%r10,%%r10\n":::); } @@ -204,7 +201,7 @@ arch_initcall(parisc_idle_init); int copy_thread(struct task_struct *p, const struct kernel_clone_args *args) { - unsigned long clone_flags = args->flags; + u64 clone_flags = args->flags; unsigned long usp = args->stack; unsigned long tls = args->tls; struct pt_regs *cregs = &(p->thread.regs); @@ -281,17 +278,3 @@ __get_wchan(struct task_struct *p) } while (count++ < MAX_UNWIND_ENTRIES); return 0; } - -static inline unsigned long brk_rnd(void) -{ - return (get_random_u32() & BRK_RND_MASK) << PAGE_SHIFT; -} - -unsigned long arch_randomize_brk(struct mm_struct *mm) -{ - unsigned long ret = PAGE_ALIGN(mm->brk + brk_rnd()); - - if (ret < mm->brk) - return mm->brk; - return ret; -} diff --git a/arch/parisc/kernel/processor.c b/arch/parisc/kernel/processor.c index ba07e760d3c7..bf73562706b2 100644 --- a/arch/parisc/kernel/processor.c +++ b/arch/parisc/kernel/processor.c @@ -26,6 +26,7 @@ #include <asm/processor.h> #include <asm/page.h> #include <asm/pdc.h> +#include <asm/smp.h> #include <asm/pdcpat.h> #include <asm/irq.h> /* for struct irq_region */ #include <asm/parisc-device.h> @@ -58,7 +59,7 @@ DEFINE_PER_CPU(struct cpuinfo_parisc, cpu_data); */ /** - * init_cpu_profiler - enable/setup per cpu profiling hooks. + * init_percpu_prof - enable/setup per cpu profiling hooks. * @cpunum: The processor instance. * * FIXME: doesn't do much yet... @@ -171,7 +172,6 @@ static int __init processor_probe(struct parisc_device *dev) p->cpu_num = cpu_info.cpu_num; p->cpu_loc = cpu_info.cpu_loc; - set_cpu_possible(cpuid, true); store_cpu_topology(cpuid); #ifdef CONFIG_SMP @@ -241,9 +241,9 @@ void __init collect_boot_cpu_data(void) /* get CPU-Model Information... */ #define p ((unsigned long *)&boot_cpu_data.pdc.model) if (pdc_model_info(&boot_cpu_data.pdc.model) == PDC_OK) { - printk(KERN_INFO - "model %08lx %08lx %08lx %08lx %08lx %08lx %08lx %08lx %08lx\n", - p[0], p[1], p[2], p[3], p[4], p[5], p[6], p[7], p[8]); + printk(KERN_INFO + "model %08lx %08lx %08lx %08lx %08lx %08lx %08lx %08lx %08lx %08lx\n", + p[0], p[1], p[2], p[3], p[4], p[5], p[6], p[7], p[8], p[9]); add_device_randomness(&boot_cpu_data.pdc.model, sizeof(boot_cpu_data.pdc.model)); @@ -367,6 +367,8 @@ int init_per_cpu(int cpunum) /* FUTURE: Enable Performance Monitor : ccr bit 0x20 */ init_percpu_prof(cpunum); + btlb_init_per_cpu(); + return ret; } @@ -377,10 +379,18 @@ int show_cpuinfo (struct seq_file *m, void *v) { unsigned long cpu; + char cpu_name[60], *p; + + /* strip PA path from CPU name to not confuse lscpu */ + strscpy(cpu_name, per_cpu(cpu_data, 0).dev->name, sizeof(cpu_name)); + p = strrchr(cpu_name, '['); + if (p) + *(--p) = 0; for_each_online_cpu(cpu) { - const struct cpuinfo_parisc *cpuinfo = &per_cpu(cpu_data, cpu); #ifdef CONFIG_SMP + const struct cpuinfo_parisc *cpuinfo = &per_cpu(cpu_data, cpu); + if (0 == cpuinfo->hpa) continue; #endif @@ -425,8 +435,7 @@ show_cpuinfo (struct seq_file *m, void *v) seq_printf(m, "model\t\t: %s - %s\n", boot_cpu_data.pdc.sys_model_name, - cpuinfo->dev ? - cpuinfo->dev->name : "Unknown"); + cpu_name); seq_printf(m, "hversion\t: 0x%08x\n" "sversion\t: 0x%08x\n", @@ -464,13 +473,6 @@ static struct parisc_driver cpu_driver __refdata = { */ void __init processor_init(void) { - unsigned int cpu; - reset_cpu_topology(); - - /* reset possible mask. We will mark those which are possible. */ - for_each_possible_cpu(cpu) - set_cpu_possible(cpu, false); - register_parisc_driver(&cpu_driver); } diff --git a/arch/parisc/kernel/ptrace.c b/arch/parisc/kernel/ptrace.c index 69c62933e952..8a17ab7e6e0b 100644 --- a/arch/parisc/kernel/ptrace.c +++ b/arch/parisc/kernel/ptrace.c @@ -126,6 +126,12 @@ long arch_ptrace(struct task_struct *child, long request, unsigned long tmp; long ret = -EIO; + unsigned long user_regs_struct_size = sizeof(struct user_regs_struct); +#ifdef CONFIG_64BIT + if (is_compat_task()) + user_regs_struct_size /= 2; +#endif + switch (request) { /* Read the word at location addr in the USER area. For ptraced @@ -166,7 +172,7 @@ long arch_ptrace(struct task_struct *child, long request, addr >= sizeof(struct pt_regs)) break; if (addr == PT_IAOQ0 || addr == PT_IAOQ1) { - data |= 3; /* ensure userspace privilege */ + data |= PRIV_USER; /* ensure userspace privilege */ } if ((addr >= PT_GR1 && addr <= PT_GR31) || addr == PT_IAOQ0 || addr == PT_IAOQ1 || @@ -181,14 +187,14 @@ long arch_ptrace(struct task_struct *child, long request, return copy_regset_to_user(child, task_user_regset_view(current), REGSET_GENERAL, - 0, sizeof(struct user_regs_struct), + 0, user_regs_struct_size, datap); case PTRACE_SETREGS: /* Set all gp regs in the child. */ return copy_regset_from_user(child, task_user_regset_view(current), REGSET_GENERAL, - 0, sizeof(struct user_regs_struct), + 0, user_regs_struct_size, datap); case PTRACE_GETFPREGS: /* Get the child FPU state. */ @@ -285,7 +291,7 @@ long compat_arch_ptrace(struct task_struct *child, compat_long_t request, if (addr >= sizeof(struct pt_regs)) break; if (addr == PT_IAOQ0+4 || addr == PT_IAOQ1+4) { - data |= 3; /* ensure userspace privilege */ + data |= PRIV_USER; /* ensure userspace privilege */ } if (addr >= PT_FR0 && addr <= PT_FR31 + 4) { /* Special case, fp regs are 64 bits anyway */ @@ -302,6 +308,11 @@ long compat_arch_ptrace(struct task_struct *child, compat_long_t request, } } break; + case PTRACE_GETREGS: + case PTRACE_SETREGS: + case PTRACE_GETFPREGS: + case PTRACE_SETFPREGS: + return arch_ptrace(child, request, addr, data); default: ret = compat_ptrace_request(child, request, addr, data); @@ -484,7 +495,7 @@ static void set_reg(struct pt_regs *regs, int num, unsigned long val) case RI(iaoq[0]): case RI(iaoq[1]): /* set 2 lowest bits to ensure userspace privilege: */ - regs->iaoq[num - RI(iaoq[0])] = val | 3; + regs->iaoq[num - RI(iaoq[0])] = val | PRIV_USER; return; case RI(sar): regs->sar = val; return; @@ -551,12 +562,12 @@ static int gpr_set(struct task_struct *target, static const struct user_regset native_regsets[] = { [REGSET_GENERAL] = { - .core_note_type = NT_PRSTATUS, .n = ELF_NGREG, + USER_REGSET_NOTE_TYPE(PRSTATUS), .n = ELF_NGREG, .size = sizeof(long), .align = sizeof(long), .regset_get = gpr_get, .set = gpr_set }, [REGSET_FP] = { - .core_note_type = NT_PRFPREG, .n = ELF_NFPREG, + USER_REGSET_NOTE_TYPE(PRFPREG), .n = ELF_NFPREG, .size = sizeof(__u64), .align = sizeof(__u64), .regset_get = fpr_get, .set = fpr_set } @@ -618,12 +629,12 @@ static int gpr32_set(struct task_struct *target, */ static const struct user_regset compat_regsets[] = { [REGSET_GENERAL] = { - .core_note_type = NT_PRSTATUS, .n = ELF_NGREG, + USER_REGSET_NOTE_TYPE(PRSTATUS), .n = ELF_NGREG, .size = sizeof(compat_long_t), .align = sizeof(compat_long_t), .regset_get = gpr32_get, .set = gpr32_set }, [REGSET_FP] = { - .core_note_type = NT_PRFPREG, .n = ELF_NFPREG, + USER_REGSET_NOTE_TYPE(PRFPREG), .n = ELF_NFPREG, .size = sizeof(__u64), .align = sizeof(__u64), .regset_get = fpr_get, .set = fpr_set } diff --git a/arch/parisc/kernel/real2.S b/arch/parisc/kernel/real2.S index 4dc12c4c0980..509d18b8e0e6 100644 --- a/arch/parisc/kernel/real2.S +++ b/arch/parisc/kernel/real2.S @@ -235,9 +235,6 @@ ENTRY_CFI(real64_call_asm) /* save fn */ copy %arg2, %r31 - /* set up the new ap */ - ldo 64(%arg1), %r29 - /* load up the arg registers from the saved arg area */ /* 32-bit calling convention passes first 4 args in registers */ ldd 0*REG_SZ(%arg1), %arg0 /* note overwriting arg0 */ @@ -249,7 +246,9 @@ ENTRY_CFI(real64_call_asm) ldd 7*REG_SZ(%arg1), %r19 ldd 1*REG_SZ(%arg1), %arg1 /* do this one last! */ + /* set up real-mode stack and real-mode ap */ tophys_r1 %sp + ldo -16(%sp), %r29 /* Reference param save area */ b,l rfi_virt2real,%r2 nop diff --git a/arch/parisc/kernel/setup.c b/arch/parisc/kernel/setup.c index 0797db617962..ace483b6f19a 100644 --- a/arch/parisc/kernel/setup.c +++ b/arch/parisc/kernel/setup.c @@ -31,7 +31,6 @@ #include <asm/sections.h> #include <asm/pdc.h> #include <asm/led.h> -#include <asm/machdep.h> /* for pa7300lc_init() proto */ #include <asm/pdc_chassis.h> #include <asm/io.h> #include <asm/setup.h> @@ -40,12 +39,7 @@ static char __initdata command_line[COMMAND_LINE_SIZE]; -/* Intended for ccio/sba/cpu statistics under /proc/bus/{runway|gsc} */ -struct proc_dir_entry * proc_runway_root __read_mostly = NULL; -struct proc_dir_entry * proc_gsc_root __read_mostly = NULL; -struct proc_dir_entry * proc_mckinley_root __read_mostly = NULL; - -void __init setup_cmdline(char **cmdline_p) +static void __init setup_cmdline(char **cmdline_p) { extern unsigned int boot_args[]; char *p; @@ -86,7 +80,7 @@ void __init setup_cmdline(char **cmdline_p) } #ifdef CONFIG_PA11 -void __init dma_ops_init(void) +static void __init dma_ops_init(void) { switch (boot_cpu_data.cpu_type) { case pcx: @@ -98,21 +92,14 @@ void __init dma_ops_init(void) "the PA-RISC 1.1 or 2.0 architecture specification.\n"); case pcxl2: - pa7300lc_init(); - break; default: break; } } #endif -extern void collect_boot_cpu_data(void); - void __init setup_arch(char **cmdline_p) { -#ifdef CONFIG_64BIT - extern int parisc_narrow_firmware; -#endif unwind_init(); init_per_cpu(smp_processor_id()); /* Set Modes & Enable FP */ @@ -153,11 +140,6 @@ void __init setup_arch(char **cmdline_p) parisc_cache_init(); paging_init(); -#ifdef CONFIG_CHASSIS_LCD_LED - /* initialize the LCD/LED after boot_cpu_data is available ! */ - led_init(); /* LCD/LED initialization */ -#endif - #ifdef CONFIG_PA11 dma_ops_init(); #endif @@ -167,10 +149,7 @@ void __init setup_arch(char **cmdline_p) /* * Display CPU info for all CPUs. - * for parisc this is in processor.c */ -extern int show_cpuinfo (struct seq_file *m, void *v); - static void * c_start (struct seq_file *m, loff_t *pos) { @@ -201,48 +180,6 @@ const struct seq_operations cpuinfo_op = { .show = show_cpuinfo }; -static void __init parisc_proc_mkdir(void) -{ - /* - ** Can't call proc_mkdir() until after proc_root_init() has been - ** called by start_kernel(). In other words, this code can't - ** live in arch/.../setup.c because start_parisc() calls - ** start_kernel(). - */ - switch (boot_cpu_data.cpu_type) { - case pcxl: - case pcxl2: - if (NULL == proc_gsc_root) - { - proc_gsc_root = proc_mkdir("bus/gsc", NULL); - } - break; - case pcxt_: - case pcxu: - case pcxu_: - case pcxw: - case pcxw_: - case pcxw2: - if (NULL == proc_runway_root) - { - proc_runway_root = proc_mkdir("bus/runway", NULL); - } - break; - case mako: - case mako2: - if (NULL == proc_mckinley_root) - { - proc_mckinley_root = proc_mkdir("bus/mckinley", NULL); - } - break; - default: - /* FIXME: this was added to prevent the compiler - * complaining about missing pcx, pcxs and pcxt - * I'm assuming they have neither gsc nor runway */ - break; - } -} - static struct resource central_bus = { .name = "Central Bus", .start = F_EXTEND(0xfff80000), @@ -295,21 +232,10 @@ static int __init parisc_init_resources(void) return 0; } -extern void gsc_init(void); -extern void processor_init(void); -extern void ccio_init(void); -extern void hppb_init(void); -extern void dino_init(void); -extern void iosapic_init(void); -extern void lba_init(void); -extern void sba_init(void); -extern void eisa_init(void); - static int __init parisc_init(void) { u32 osid = (OS_ID_LINUX << 16); - parisc_proc_mkdir(); parisc_init_resources(); do_device_inventory(); /* probe for hardware */ @@ -344,55 +270,12 @@ static int __init parisc_init(void) apply_alternatives_all(); parisc_setup_cache_timing(); - - /* These are in a non-obvious order, will fix when we have an iotree */ -#if defined(CONFIG_IOSAPIC) - iosapic_init(); -#endif -#if defined(CONFIG_IOMMU_SBA) - sba_init(); -#endif -#if defined(CONFIG_PCI_LBA) - lba_init(); -#endif - - /* CCIO before any potential subdevices */ -#if defined(CONFIG_IOMMU_CCIO) - ccio_init(); -#endif - - /* - * Need to register Asp & Wax before the EISA adapters for the IRQ - * regions. EISA must come before PCI to be sure it gets IRQ region - * 0. - */ -#if defined(CONFIG_GSC_LASI) || defined(CONFIG_GSC_WAX) - gsc_init(); -#endif -#ifdef CONFIG_EISA - eisa_init(); -#endif - -#if defined(CONFIG_HPPB) - hppb_init(); -#endif - -#if defined(CONFIG_GSC_DINO) - dino_init(); -#endif - -#ifdef CONFIG_CHASSIS_LCD_LED - register_led_regions(); /* register LED port info in procfs */ -#endif - return 0; } arch_initcall(parisc_init); void __init start_parisc(void) { - extern void early_trap_init(void); - int ret, cpunum; struct pdc_coproc_cfg coproc_cfg; diff --git a/arch/parisc/kernel/signal.c b/arch/parisc/kernel/signal.c index 8bc0ddaa6219..e8d27def6c52 100644 --- a/arch/parisc/kernel/signal.c +++ b/arch/parisc/kernel/signal.c @@ -73,7 +73,7 @@ restore_sigcontext(struct sigcontext __user *sc, struct pt_regs *regs) return err; } -void +asmlinkage void sys_rt_sigreturn(struct pt_regs *regs, int in_syscall) { struct rt_sigframe __user *frame; @@ -176,7 +176,7 @@ get_sigframe(struct k_sigaction *ka, unsigned long sp, size_t frame_size) } static long -setup_sigcontext(struct sigcontext __user *sc, struct pt_regs *regs, int in_syscall) +setup_sigcontext(struct sigcontext __user *sc, struct pt_regs *regs, long in_syscall) { unsigned long flags = 0; @@ -211,7 +211,7 @@ setup_sigcontext(struct sigcontext __user *sc, struct pt_regs *regs, int in_sysc static long setup_rt_frame(struct ksignal *ksig, sigset_t *set, struct pt_regs *regs, - int in_syscall) + long in_syscall) { struct rt_sigframe __user *frame; unsigned long rp, usp; @@ -380,7 +380,7 @@ setup_rt_frame(struct ksignal *ksig, sigset_t *set, struct pt_regs *regs, */ static void -handle_signal(struct ksignal *ksig, struct pt_regs *regs, int in_syscall) +handle_signal(struct ksignal *ksig, struct pt_regs *regs, long in_syscall) { int ret; sigset_t *oldset = sigmask_to_save(); @@ -423,7 +423,7 @@ static void check_syscallno_in_delay_branch(struct pt_regs *regs) regs->gr[31] -= 8; /* delayed branching */ /* Get assembler opcode of code in delay branch */ - uaddr = (unsigned int *) ((regs->gr[31] & ~3) + 4); + uaddr = (u32 __user *) ((regs->gr[31] & ~3) + 4); err = get_user(opcode, uaddr); if (err) return; @@ -578,7 +578,7 @@ static void do_signal(struct pt_regs *regs, long in_syscall) restore_saved_sigmask(); } -void do_notify_resume(struct pt_regs *regs, long in_syscall) +asmlinkage void do_notify_resume(struct pt_regs *regs, long in_syscall) { if (test_thread_flag(TIF_SIGPENDING) || test_thread_flag(TIF_NOTIFY_SIGNAL)) diff --git a/arch/parisc/kernel/smp.c b/arch/parisc/kernel/smp.c index 7dbd92cafae3..b2d12ab728b1 100644 --- a/arch/parisc/kernel/smp.c +++ b/arch/parisc/kernel/smp.c @@ -246,8 +246,8 @@ void kgdb_roundup_cpus(void) inline void smp_send_stop(void) { send_IPI_allbutself(IPI_CPU_STOP); } -void -smp_send_reschedule(int cpu) { send_IPI_single(cpu, IPI_RESCHEDULE); } +void +arch_smp_send_reschedule(int cpu) { send_IPI_single(cpu, IPI_RESCHEDULE); } void smp_send_all_nop(void) @@ -271,9 +271,6 @@ void arch_send_call_function_single_ipi(int cpu) static void smp_cpu_init(int cpunum) { - extern void init_IRQ(void); /* arch/parisc/kernel/irq.c */ - extern void start_cpu_itimer(void); /* arch/parisc/kernel/time.c */ - /* Set modes and Enable floating point coprocessor */ init_per_cpu(cpunum); @@ -300,7 +297,7 @@ smp_cpu_init(int cpunum) enter_lazy_tlb(&init_mm, current); init_IRQ(); /* make sure no IRQs are enabled or pending */ - start_cpu_itimer(); + parisc_clockevent_init(); } @@ -347,7 +344,7 @@ static int smp_boot_one_cpu(int cpuid, struct task_struct *idle) struct irq_desc *desc = irq_to_desc(i); if (desc && desc->kstat_irqs) - *per_cpu_ptr(desc->kstat_irqs, cpuid) = 0; + *per_cpu_ptr(desc->kstat_irqs, cpuid) = (struct irqstat) { }; } #endif @@ -407,13 +404,7 @@ alive: void __init smp_prepare_boot_cpu(void) { - int bootstrap_processor = per_cpu(cpu_data, 0).cpuid; - - /* Setup BSP mappings */ - printk(KERN_INFO "SMP: bootstrap CPU ID is %d\n", bootstrap_processor); - - set_cpu_online(bootstrap_processor, true); - set_cpu_present(bootstrap_processor, true); + pr_info("SMP: bootstrap CPU ID is 0\n"); } @@ -443,7 +434,9 @@ int __cpu_up(unsigned int cpu, struct task_struct *tidle) if (cpu_online(cpu)) return 0; - if (num_online_cpus() < setup_max_cpus && smp_boot_one_cpu(cpu, tidle)) + if (num_online_cpus() < nr_cpu_ids && + num_online_cpus() < setup_max_cpus && + smp_boot_one_cpu(cpu, tidle)) return -EIO; return cpu_online(cpu) ? 0 : -EIO; @@ -500,11 +493,10 @@ int __cpu_disable(void) void __cpu_die(unsigned int cpu) { pdc_cpu_rendezvous_lock(); +} - if (!cpu_wait_death(cpu, 5)) { - pr_crit("CPU%u: cpu didn't die\n", cpu); - return; - } +void arch_cpuhp_cleanup_dead_cpu(unsigned int cpu) +{ pr_info("CPU%u: is shutting down\n", cpu); /* set task's state to interruptible sleep */ diff --git a/arch/parisc/kernel/sys_parisc.c b/arch/parisc/kernel/sys_parisc.c index 09a34b07f02e..b2cdbb8a12b1 100644 --- a/arch/parisc/kernel/sys_parisc.c +++ b/arch/parisc/kernel/sys_parisc.c @@ -24,32 +24,28 @@ #include <linux/personality.h> #include <linux/random.h> #include <linux/compat.h> +#include <linux/elf-randomize.h> -/* we construct an artificial offset for the mapping based on the physical - * address of the kernel mapping variable */ -#define GET_LAST_MMAP(filp) \ - (filp ? ((unsigned long) filp->f_mapping) >> 8 : 0UL) -#define SET_LAST_MMAP(filp, val) \ - { /* nothing */ } - -static int get_offset(unsigned int last_mmap) -{ - return (last_mmap & (SHM_COLOUR-1)) >> PAGE_SHIFT; -} +/* + * Construct an artificial page offset for the mapping based on the physical + * address of the kernel file mapping variable. + */ +#define GET_FILP_PGOFF(filp) \ + (filp ? (((unsigned long) filp->f_mapping) >> 8) \ + & ((SHM_COLOUR-1) >> PAGE_SHIFT) : 0UL) -static unsigned long shared_align_offset(unsigned int last_mmap, +static unsigned long shared_align_offset(unsigned long filp_pgoff, unsigned long pgoff) { - return (get_offset(last_mmap) + pgoff) << PAGE_SHIFT; + return (filp_pgoff + pgoff) << PAGE_SHIFT; } static inline unsigned long COLOR_ALIGN(unsigned long addr, - unsigned int last_mmap, unsigned long pgoff) + unsigned long filp_pgoff, unsigned long pgoff) { unsigned long base = (addr+SHM_COLOUR-1) & ~(SHM_COLOUR-1); unsigned long off = (SHM_COLOUR-1) & - (shared_align_offset(last_mmap, pgoff) << PAGE_SHIFT); - + shared_align_offset(filp_pgoff, pgoff); return base + off; } @@ -81,7 +77,7 @@ unsigned long calc_max_stack_size(unsigned long stack_max) * indicating that "current" should be used instead of a passed-in * value from the exec bprm as done with arch_pick_mmap_layout(). */ -static unsigned long mmap_upper_limit(struct rlimit *rlim_stack) +unsigned long mmap_upper_limit(const struct rlimit *rlim_stack) { unsigned long stack_base; @@ -98,92 +94,43 @@ static unsigned long mmap_upper_limit(struct rlimit *rlim_stack) return PAGE_ALIGN(STACK_TOP - stack_base); } +enum mmap_allocation_direction {UP, DOWN}; -unsigned long arch_get_unmapped_area(struct file *filp, unsigned long addr, - unsigned long len, unsigned long pgoff, unsigned long flags) +static unsigned long arch_get_unmapped_area_common(struct file *filp, + unsigned long addr, unsigned long len, unsigned long pgoff, + unsigned long flags, enum mmap_allocation_direction dir) { struct mm_struct *mm = current->mm; struct vm_area_struct *vma, *prev; - unsigned long task_size = TASK_SIZE; - int do_color_align, last_mmap; - struct vm_unmapped_area_info info; + unsigned long filp_pgoff; + int do_color_align; + struct vm_unmapped_area_info info = { + .length = len + }; - if (len > task_size) + if (unlikely(len > TASK_SIZE)) return -ENOMEM; do_color_align = 0; if (filp || (flags & MAP_SHARED)) do_color_align = 1; - last_mmap = GET_LAST_MMAP(filp); + filp_pgoff = GET_FILP_PGOFF(filp); if (flags & MAP_FIXED) { - if ((flags & MAP_SHARED) && last_mmap && - (addr - shared_align_offset(last_mmap, pgoff)) - & (SHM_COLOUR - 1)) + /* Even MAP_FIXED mappings must reside within TASK_SIZE */ + if (TASK_SIZE - len < addr) return -EINVAL; - goto found_addr; - } - - if (addr) { - if (do_color_align && last_mmap) - addr = COLOR_ALIGN(addr, last_mmap, pgoff); - else - addr = PAGE_ALIGN(addr); - - vma = find_vma_prev(mm, addr, &prev); - if (task_size - len >= addr && - (!vma || addr + len <= vm_start_gap(vma)) && - (!prev || addr >= vm_end_gap(prev))) - goto found_addr; - } - - info.flags = 0; - info.length = len; - info.low_limit = mm->mmap_legacy_base; - info.high_limit = mmap_upper_limit(NULL); - info.align_mask = last_mmap ? (PAGE_MASK & (SHM_COLOUR - 1)) : 0; - info.align_offset = shared_align_offset(last_mmap, pgoff); - addr = vm_unmapped_area(&info); - -found_addr: - if (do_color_align && !last_mmap && !(addr & ~PAGE_MASK)) - SET_LAST_MMAP(filp, addr - (pgoff << PAGE_SHIFT)); - - return addr; -} - -unsigned long -arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, - const unsigned long len, const unsigned long pgoff, - const unsigned long flags) -{ - struct vm_area_struct *vma, *prev; - struct mm_struct *mm = current->mm; - unsigned long addr = addr0; - int do_color_align, last_mmap; - struct vm_unmapped_area_info info; - - /* requested length too big for entire address space */ - if (len > TASK_SIZE) - return -ENOMEM; - - do_color_align = 0; - if (filp || (flags & MAP_SHARED)) - do_color_align = 1; - last_mmap = GET_LAST_MMAP(filp); - if (flags & MAP_FIXED) { - if ((flags & MAP_SHARED) && last_mmap && - (addr - shared_align_offset(last_mmap, pgoff)) - & (SHM_COLOUR - 1)) + if ((flags & MAP_SHARED) && filp && + (addr - shared_align_offset(filp_pgoff, pgoff)) + & (SHM_COLOUR - 1)) return -EINVAL; - goto found_addr; + return addr; } - /* requesting a specific address */ if (addr) { - if (do_color_align && last_mmap) - addr = COLOR_ALIGN(addr, last_mmap, pgoff); + if (do_color_align) + addr = COLOR_ALIGN(addr, filp_pgoff, pgoff); else addr = PAGE_ALIGN(addr); @@ -191,87 +138,50 @@ arch_get_unmapped_area_topdown(struct file *filp, const unsigned long addr0, if (TASK_SIZE - len >= addr && (!vma || addr + len <= vm_start_gap(vma)) && (!prev || addr >= vm_end_gap(prev))) - goto found_addr; + return addr; } - info.flags = VM_UNMAPPED_AREA_TOPDOWN; - info.length = len; - info.low_limit = PAGE_SIZE; - info.high_limit = mm->mmap_base; - info.align_mask = last_mmap ? (PAGE_MASK & (SHM_COLOUR - 1)) : 0; - info.align_offset = shared_align_offset(last_mmap, pgoff); - addr = vm_unmapped_area(&info); - if (!(addr & ~PAGE_MASK)) - goto found_addr; - VM_BUG_ON(addr != -ENOMEM); - - /* - * A failed mmap() very likely causes application failure, - * so fall back to the bottom-up function here. This scenario - * can happen with large stack limits and large mmap() - * allocations. - */ - return arch_get_unmapped_area(filp, addr0, len, pgoff, flags); - -found_addr: - if (do_color_align && !last_mmap && !(addr & ~PAGE_MASK)) - SET_LAST_MMAP(filp, addr - (pgoff << PAGE_SHIFT)); - - return addr; -} - -static int mmap_is_legacy(void) -{ - if (current->personality & ADDR_COMPAT_LAYOUT) - return 1; - - /* parisc stack always grows up - so a unlimited stack should - * not be an indicator to use the legacy memory layout. - * if (rlimit(RLIMIT_STACK) == RLIM_INFINITY) - * return 1; - */ - - return sysctl_legacy_va_layout; -} - -static unsigned long mmap_rnd(void) -{ - unsigned long rnd = 0; - - if (current->flags & PF_RANDOMIZE) - rnd = get_random_u32() & MMAP_RND_MASK; - - return rnd << PAGE_SHIFT; -} + info.align_mask = do_color_align ? (PAGE_MASK & (SHM_COLOUR - 1)) : 0; + info.align_offset = shared_align_offset(filp_pgoff, pgoff); + + if (dir == DOWN) { + info.flags = VM_UNMAPPED_AREA_TOPDOWN; + info.low_limit = PAGE_SIZE; + info.high_limit = mm->mmap_base; + addr = vm_unmapped_area(&info); + if (!(addr & ~PAGE_MASK)) + return addr; + VM_BUG_ON(addr != -ENOMEM); + + /* + * A failed mmap() very likely causes application failure, + * so fall back to the bottom-up function here. This scenario + * can happen with large stack limits and large mmap() + * allocations. + */ + } -unsigned long arch_mmap_rnd(void) -{ - return (get_random_u32() & MMAP_RND_MASK) << PAGE_SHIFT; + info.low_limit = mm->mmap_base; + info.high_limit = mmap_upper_limit(NULL); + return vm_unmapped_area(&info); } -static unsigned long mmap_legacy_base(void) +unsigned long arch_get_unmapped_area(struct file *filp, unsigned long addr, + unsigned long len, unsigned long pgoff, unsigned long flags, + vm_flags_t vm_flags) { - return TASK_UNMAPPED_BASE + mmap_rnd(); + return arch_get_unmapped_area_common(filp, + addr, len, pgoff, flags, UP); } -/* - * This function, called very early during the creation of a new - * process VM image, sets up which VM layout function to use: - */ -void arch_pick_mmap_layout(struct mm_struct *mm, struct rlimit *rlim_stack) +unsigned long arch_get_unmapped_area_topdown(struct file *filp, + unsigned long addr, unsigned long len, unsigned long pgoff, + unsigned long flags, vm_flags_t vm_flags) { - mm->mmap_legacy_base = mmap_legacy_base(); - mm->mmap_base = mmap_upper_limit(rlim_stack); - - if (mmap_is_legacy()) { - mm->mmap_base = mm->mmap_legacy_base; - mm->get_unmapped_area = arch_get_unmapped_area; - } else { - mm->get_unmapped_area = arch_get_unmapped_area_topdown; - } + return arch_get_unmapped_area_common(filp, + addr, len, pgoff, flags, DOWN); } - asmlinkage unsigned long sys_mmap2(unsigned long addr, unsigned long len, unsigned long prot, unsigned long flags, unsigned long fd, unsigned long pgoff) @@ -379,7 +289,7 @@ asmlinkage long parisc_fallocate(int fd, int mode, u32 offhi, u32 offlo, ((u64)lenhi << 32) | lenlo); } -long parisc_personality(unsigned long personality) +asmlinkage long parisc_personality(unsigned long personality) { long err; diff --git a/arch/parisc/kernel/sys_parisc32.c b/arch/parisc/kernel/sys_parisc32.c index 2a12a547b447..826c8e51b585 100644 --- a/arch/parisc/kernel/sys_parisc32.c +++ b/arch/parisc/kernel/sys_parisc32.c @@ -23,12 +23,3 @@ asmlinkage long sys32_unimplemented(int r26, int r25, int r24, int r23, current->comm, current->pid, r20); return -ENOSYS; } - -asmlinkage long sys32_fanotify_mark(compat_int_t fanotify_fd, compat_uint_t flags, - compat_uint_t mask0, compat_uint_t mask1, compat_int_t dfd, - const char __user * pathname) -{ - return sys_fanotify_mark(fanotify_fd, flags, - ((__u64)mask1 << 32) | mask0, - dfd, pathname); -} diff --git a/arch/parisc/kernel/syscall.S b/arch/parisc/kernel/syscall.S index 1373e5129868..f58c4bccfbce 100644 --- a/arch/parisc/kernel/syscall.S +++ b/arch/parisc/kernel/syscall.S @@ -39,6 +39,7 @@ registers). #include <asm/assembly.h> #include <asm/processor.h> #include <asm/cache.h> +#include <asm/spinlock_types.h> #include <linux/linkage.h> @@ -66,6 +67,16 @@ registers). stw \reg1, 0(%sr2,\reg2) .endm + /* raise exception if spinlock content is not zero or + * __ARCH_SPIN_LOCK_UNLOCKED_VAL */ + .macro spinlock_check spin_val,tmpreg +#ifdef CONFIG_LIGHTWEIGHT_SPINLOCK_CHECK + ldi __ARCH_SPIN_LOCK_UNLOCKED_VAL, \tmpreg + andcm,= \spin_val, \tmpreg, %r0 + .word SPINLOCK_BREAK_INSN +#endif + .endm + .text .import syscall_exit,code @@ -232,10 +243,10 @@ linux_gateway_entry: #ifdef CONFIG_64BIT ldil L%sys_call_table, %r1 - or,= %r2,%r2,%r2 - addil L%(sys_call_table64-sys_call_table), %r1 + or,ev %r2,%r2,%r2 + ldil L%sys_call_table64, %r1 ldo R%sys_call_table(%r1), %r19 - or,= %r2,%r2,%r2 + or,ev %r2,%r2,%r2 ldo R%sys_call_table64(%r1), %r19 #else load32 sys_call_table, %r19 @@ -368,10 +379,10 @@ tracesys_next: extrd,u %r19,63,1,%r2 /* W hidden in bottom bit */ ldil L%sys_call_table, %r1 - or,= %r2,%r2,%r2 - addil L%(sys_call_table64-sys_call_table), %r1 + or,ev %r2,%r2,%r2 + ldil L%sys_call_table64, %r1 ldo R%sys_call_table(%r1), %r19 - or,= %r2,%r2,%r2 + or,ev %r2,%r2,%r2 ldo R%sys_call_table64(%r1), %r19 #else load32 sys_call_table, %r19 @@ -508,7 +519,8 @@ lws_start: lws_exit_noerror: lws_pagefault_enable %r1,%r21 - stw,ma %r20, 0(%sr2,%r20) + ldi __ARCH_SPIN_LOCK_UNLOCKED_VAL, %r21 + stw,ma %r21, 0(%sr2,%r20) ssm PSW_SM_I, %r0 b lws_exit copy %r0, %r21 @@ -521,7 +533,8 @@ lws_wouldblock: lws_pagefault: lws_pagefault_enable %r1,%r21 - stw,ma %r20, 0(%sr2,%r20) + ldi __ARCH_SPIN_LOCK_UNLOCKED_VAL, %r21 + stw,ma %r21, 0(%sr2,%r20) ssm PSW_SM_I, %r0 ldo 3(%r0),%r28 b lws_exit @@ -600,6 +613,9 @@ lws_compare_and_swap32: lws_compare_and_swap: /* Trigger memory reference interruptions without writing to memory */ 1: ldw 0(%r26), %r28 + proberi (%r26), PRIV_USER, %r28 + comb,=,n %r28, %r0, lws_fault /* backwards, likely not taken */ + nop 2: stbys,e %r0, 0(%r26) /* Calculate 8-bit hash index from virtual address */ @@ -619,6 +635,7 @@ lws_compare_and_swap: /* Try to acquire the lock */ LDCW 0(%sr2,%r20), %r28 + spinlock_check %r28, %r21 comclr,<> %r0, %r28, %r0 b,n lws_wouldblock @@ -753,6 +770,9 @@ cas2_lock_start: copy %r26, %r28 depi_safe 0, 31, 2, %r28 10: ldw 0(%r28), %r1 + proberi (%r28), PRIV_USER, %r1 + comb,=,n %r1, %r0, lws_fault /* backwards, likely not taken */ + nop 11: stbys,e %r0, 0(%r28) /* Calculate 8-bit hash index from virtual address */ @@ -772,6 +792,7 @@ cas2_lock_start: /* Try to acquire the lock */ LDCW 0(%sr2,%r20), %r28 + spinlock_check %r28, %r21 comclr,<> %r0, %r28, %r0 b,n lws_wouldblock @@ -936,41 +957,47 @@ atomic_xchg_begin: /* 8-bit exchange */ 1: ldb 0(%r24), %r20 + proberi (%r24), PRIV_USER, %r20 + comb,=,n %r20, %r0, lws_fault /* backwards, likely not taken */ + nop copy %r23, %r20 depi_safe 0, 31, 2, %r20 b atomic_xchg_start 2: stbys,e %r0, 0(%r20) - nop - nop - nop /* 16-bit exchange */ 3: ldh 0(%r24), %r20 + proberi (%r24), PRIV_USER, %r20 + comb,=,n %r20, %r0, lws_fault /* backwards, likely not taken */ + nop copy %r23, %r20 depi_safe 0, 31, 2, %r20 b atomic_xchg_start 4: stbys,e %r0, 0(%r20) - nop - nop - nop /* 32-bit exchange */ 5: ldw 0(%r24), %r20 + proberi (%r24), PRIV_USER, %r20 + comb,=,n %r20, %r0, lws_fault /* backwards, likely not taken */ + nop b atomic_xchg_start 6: stbys,e %r0, 0(%r23) nop nop - nop - nop - nop /* 64-bit exchange */ #ifdef CONFIG_64BIT 7: ldd 0(%r24), %r20 + proberi (%r24), PRIV_USER, %r20 + comb,=,n %r20, %r0, lws_fault /* backwards, likely not taken */ + nop 8: stdby,e %r0, 0(%r23) #else 7: ldw 0(%r24), %r20 8: ldw 4(%r24), %r20 + proberi (%r24), PRIV_USER, %r20 + comb,=,n %r20, %r0, lws_fault /* backwards, likely not taken */ + nop copy %r23, %r20 depi_safe 0, 31, 2, %r20 9: stbys,e %r0, 0(%r20) @@ -1001,6 +1028,7 @@ atomic_xchg_start: /* Try to acquire the lock */ LDCW 0(%sr2,%r20), %r28 + spinlock_check %r28, %r21 comclr,<> %r0, %r28, %r0 b,n lws_wouldblock @@ -1199,6 +1227,7 @@ atomic_store_start: /* Try to acquire the lock */ LDCW 0(%sr2,%r20), %r28 + spinlock_check %r28, %r21 comclr,<> %r0, %r28, %r0 b,n lws_wouldblock @@ -1310,6 +1339,8 @@ ENTRY(sys_call_table) END(sys_call_table) #ifdef CONFIG_64BIT +#undef __SYSCALL_WITH_COMPAT +#define __SYSCALL_WITH_COMPAT(nr, native, compat) __SYSCALL(nr, native) .align 8 ENTRY(sys_call_table64) #include <asm/syscall_table_64.h> /* 64-bit syscalls */ @@ -1330,7 +1361,7 @@ ENTRY(lws_lock_start) /* lws locks */ .rept 256 /* Keep locks aligned at 16-bytes */ - .word 1 + .word __ARCH_SPIN_LOCK_UNLOCKED_VAL .word 0 .word 0 .word 0 diff --git a/arch/parisc/kernel/syscalls/syscall.tbl b/arch/parisc/kernel/syscalls/syscall.tbl index 0e42fceb2d5e..39bdacaa530b 100644 --- a/arch/parisc/kernel/syscalls/syscall.tbl +++ b/arch/parisc/kernel/syscalls/syscall.tbl @@ -108,7 +108,7 @@ 95 common fchown sys_fchown 96 common getpriority sys_getpriority 97 common setpriority sys_setpriority -98 common recv sys_recv +98 common recv sys_recv compat_sys_recv 99 common statfs sys_statfs compat_sys_statfs 100 common fstatfs sys_fstatfs compat_sys_fstatfs 101 common stat64 sys_stat64 @@ -135,7 +135,7 @@ 120 common clone sys_clone_wrapper 121 common setdomainname sys_setdomainname 122 common sendfile sys_sendfile compat_sys_sendfile -123 common recvfrom sys_recvfrom +123 common recvfrom sys_recvfrom compat_sys_recvfrom 124 32 adjtimex sys_adjtimex_time32 124 64 adjtimex sys_adjtimex 125 common mprotect sys_mprotect @@ -245,7 +245,7 @@ # 220 was alloc_hugepages # 221 was free_hugepages 222 common exit_group sys_exit_group -223 common lookup_dcookie sys_lookup_dcookie compat_sys_lookup_dcookie +223 common lookup_dcookie sys_ni_syscall 224 common epoll_create sys_epoll_create 225 common epoll_ctl sys_epoll_ctl 226 common epoll_wait sys_epoll_wait @@ -364,7 +364,7 @@ 320 common accept4 sys_accept4 321 common prlimit64 sys_prlimit64 322 common fanotify_init sys_fanotify_init -323 common fanotify_mark sys_fanotify_mark sys32_fanotify_mark +323 common fanotify_mark sys_fanotify_mark compat_sys_fanotify_mark 324 32 clock_adjtime sys_clock_adjtime32 324 64 clock_adjtime sys_clock_adjtime 325 common name_to_handle_at sys_name_to_handle_at @@ -400,6 +400,7 @@ 353 common pkey_free sys_pkey_free 354 common rseq sys_rseq 355 common kexec_file_load sys_kexec_file_load sys_kexec_file_load +356 common cacheflush sys_cacheflush # up to 402 is unassigned and reserved for arch specific syscalls 403 32 clock_gettime64 sys_clock_gettime sys_clock_gettime 404 32 clock_settime64 sys_clock_settime sys_clock_settime @@ -448,3 +449,23 @@ 448 common process_mrelease sys_process_mrelease 449 common futex_waitv sys_futex_waitv 450 common set_mempolicy_home_node sys_set_mempolicy_home_node +451 common cachestat sys_cachestat +452 common fchmodat2 sys_fchmodat2 +453 common map_shadow_stack sys_map_shadow_stack +454 common futex_wake sys_futex_wake +455 common futex_wait sys_futex_wait +456 common futex_requeue sys_futex_requeue +457 common statmount sys_statmount +458 common listmount sys_listmount +459 common lsm_get_self_attr sys_lsm_get_self_attr +460 common lsm_set_self_attr sys_lsm_set_self_attr +461 common lsm_list_modules sys_lsm_list_modules +462 common mseal sys_mseal +463 common setxattrat sys_setxattrat +464 common getxattrat sys_getxattrat +465 common listxattrat sys_listxattrat +466 common removexattrat sys_removexattrat +467 common open_tree_attr sys_open_tree_attr +468 common file_getattr sys_file_getattr +469 common file_setattr sys_file_setattr +470 common listns sys_listns diff --git a/arch/parisc/kernel/time.c b/arch/parisc/kernel/time.c index 9714fbd7c42d..c17e2249115f 100644 --- a/arch/parisc/kernel/time.c +++ b/arch/parisc/kernel/time.c @@ -1,166 +1,119 @@ // SPDX-License-Identifier: GPL-2.0 /* - * linux/arch/parisc/kernel/time.c + * Common time service routines for parisc machines. + * based on arch/loongarch/kernel/time.c * - * Copyright (C) 1991, 1992, 1995 Linus Torvalds - * Modifications for ARM (C) 1994, 1995, 1996,1997 Russell King - * Copyright (C) 1999 SuSE GmbH, (Philipp Rumpf, prumpf@tux.org) - * - * 1994-07-02 Alan Modra - * fixed set_rtc_mmss, fixed time.year for >= 2000, new mktime - * 1998-12-20 Updated NTP code according to technical memorandum Jan '96 - * "A Kernel Model for Precision Timekeeping" by Dave Mills + * Copyright (C) 2024 Helge Deller <deller@gmx.de> */ -#include <linux/errno.h> -#include <linux/module.h> -#include <linux/rtc.h> -#include <linux/sched.h> -#include <linux/sched/clock.h> -#include <linux/sched_clock.h> -#include <linux/kernel.h> -#include <linux/param.h> -#include <linux/string.h> -#include <linux/mm.h> -#include <linux/interrupt.h> -#include <linux/time.h> +#include <linux/clockchips.h> +#include <linux/delay.h> +#include <linux/export.h> #include <linux/init.h> -#include <linux/smp.h> -#include <linux/profile.h> -#include <linux/clocksource.h> +#include <linux/interrupt.h> +#include <linux/kernel.h> +#include <linux/sched_clock.h> +#include <linux/spinlock.h> +#include <linux/rtc.h> #include <linux/platform_device.h> -#include <linux/ftrace.h> +#include <asm/processor.h> -#include <linux/uaccess.h> -#include <asm/io.h> -#include <asm/irq.h> -#include <asm/page.h> -#include <asm/param.h> -#include <asm/pdc.h> -#include <asm/led.h> +static u64 cr16_clock_freq; +static unsigned long clocktick; -#include <linux/timex.h> +int time_keeper_id; /* CPU used for timekeeping */ -int time_keeper_id __read_mostly; /* CPU used for timekeeping. */ +static DEFINE_PER_CPU(struct clock_event_device, parisc_clockevent_device); -static unsigned long clocktick __ro_after_init; /* timer cycles per tick */ +static void parisc_event_handler(struct clock_event_device *dev) +{ +} -/* - * We keep time on PA-RISC Linux by using the Interval Timer which is - * a pair of registers; one is read-only and one is write-only; both - * accessed through CR16. The read-only register is 32 or 64 bits wide, - * and increments by 1 every CPU clock tick. The architecture only - * guarantees us a rate between 0.5 and 2, but all implementations use a - * rate of 1. The write-only register is 32-bits wide. When the lowest - * 32 bits of the read-only register compare equal to the write-only - * register, it raises a maskable external interrupt. Each processor has - * an Interval Timer of its own and they are not synchronised. - * - * We want to generate an interrupt every 1/HZ seconds. So we program - * CR16 to interrupt every @clocktick cycles. The it_value in cpu_data - * is programmed with the intended time of the next tick. We can be - * held off for an arbitrarily long period of time by interrupts being - * disabled, so we may miss one or more ticks. - */ -irqreturn_t __irq_entry timer_interrupt(int irq, void *dev_id) +static int parisc_timer_next_event(unsigned long delta, struct clock_event_device *evt) { - unsigned long now; - unsigned long next_tick; - unsigned long ticks_elapsed = 0; - unsigned int cpu = smp_processor_id(); - struct cpuinfo_parisc *cpuinfo = &per_cpu(cpu_data, cpu); - - /* gcc can optimize for "read-only" case with a local clocktick */ - unsigned long cpt = clocktick; - - /* Initialize next_tick to the old expected tick time. */ - next_tick = cpuinfo->it_value; - - /* Calculate how many ticks have elapsed. */ - now = mfctl(16); - do { - ++ticks_elapsed; - next_tick += cpt; - } while (next_tick - now > cpt); - - /* Store (in CR16 cycles) up to when we are accounting right now. */ - cpuinfo->it_value = next_tick; - - /* Go do system house keeping. */ - if (IS_ENABLED(CONFIG_SMP) && (cpu != time_keeper_id)) - ticks_elapsed = 0; - legacy_timer_tick(ticks_elapsed); - - /* Skip clockticks on purpose if we know we would miss those. - * The new CR16 must be "later" than current CR16 otherwise - * itimer would not fire until CR16 wrapped - e.g 4 seconds - * later on a 1Ghz processor. We'll account for the missed - * ticks on the next timer interrupt. - * We want IT to fire modulo clocktick even if we miss/skip some. - * But those interrupts don't in fact get delivered that regularly. - * - * "next_tick - now" will always give the difference regardless - * if one or the other wrapped. If "now" is "bigger" we'll end up - * with a very large unsigned number. - */ - now = mfctl(16); - while (next_tick - now > cpt) - next_tick += cpt; - - /* Program the IT when to deliver the next interrupt. - * Only bottom 32-bits of next_tick are writable in CR16! - * Timer interrupt will be delivered at least a few hundred cycles - * after the IT fires, so if we are too close (<= 8000 cycles) to the - * next cycle, simply skip it. - */ - if (next_tick - now <= 8000) - next_tick += cpt; - mtctl(next_tick, 16); + unsigned long new_cr16; - return IRQ_HANDLED; -} + new_cr16 = mfctl(16) + delta; + mtctl(new_cr16, 16); + return 0; +} -unsigned long profile_pc(struct pt_regs *regs) +irqreturn_t timer_interrupt(int irq, void *data) { - unsigned long pc = instruction_pointer(regs); + struct clock_event_device *cd; + int cpu = smp_processor_id(); - if (regs->gr[0] & PSW_N) - pc -= 4; + cd = &per_cpu(parisc_clockevent_device, cpu); -#ifdef CONFIG_SMP - if (in_lock_functions(pc)) - pc = regs->gr[2]; -#endif + if (clockevent_state_periodic(cd)) + parisc_timer_next_event(clocktick, cd); - return pc; + if (clockevent_state_periodic(cd) || clockevent_state_oneshot(cd)) + cd->event_handler(cd); + + return IRQ_HANDLED; } -EXPORT_SYMBOL(profile_pc); +static int parisc_set_state_oneshot(struct clock_event_device *evt) +{ + parisc_timer_next_event(clocktick, evt); -/* clock source code */ + return 0; +} -static u64 notrace read_cr16(struct clocksource *cs) +static int parisc_set_state_periodic(struct clock_event_device *evt) { - return get_cycles(); + parisc_timer_next_event(clocktick, evt); + + return 0; } -static struct clocksource clocksource_cr16 = { - .name = "cr16", - .rating = 300, - .read = read_cr16, - .mask = CLOCKSOURCE_MASK(BITS_PER_LONG), - .flags = CLOCK_SOURCE_IS_CONTINUOUS, -}; +static int parisc_set_state_shutdown(struct clock_event_device *evt) +{ + return 0; +} -void start_cpu_itimer(void) +void parisc_clockevent_init(void) { unsigned int cpu = smp_processor_id(); - unsigned long next_tick = mfctl(16) + clocktick; + unsigned long min_delta = 0x600; /* XXX */ + unsigned long max_delta = (1UL << (BITS_PER_LONG - 1)); + struct clock_event_device *cd; + + cd = &per_cpu(parisc_clockevent_device, cpu); + + cd->name = "cr16_clockevent"; + cd->features = CLOCK_EVT_FEAT_ONESHOT | CLOCK_EVT_FEAT_PERIODIC | + CLOCK_EVT_FEAT_PERCPU; + + cd->irq = TIMER_IRQ; + cd->rating = 320; + cd->cpumask = cpumask_of(cpu); + cd->set_state_oneshot = parisc_set_state_oneshot; + cd->set_state_oneshot_stopped = parisc_set_state_shutdown; + cd->set_state_periodic = parisc_set_state_periodic; + cd->set_state_shutdown = parisc_set_state_shutdown; + cd->set_next_event = parisc_timer_next_event; + cd->event_handler = parisc_event_handler; + + clockevents_config_and_register(cd, cr16_clock_freq, min_delta, max_delta); +} + +unsigned long notrace profile_pc(struct pt_regs *regs) +{ + unsigned long pc = instruction_pointer(regs); - mtctl(next_tick, 16); /* kick off Interval Timer (CR16) */ + if (regs->gr[0] & PSW_N) + pc -= 4; + +#ifdef CONFIG_SMP + if (in_lock_functions(pc)) + pc = regs->gr[2]; +#endif - per_cpu(cpu_data, cpu).it_value = next_tick; + return pc; } +EXPORT_SYMBOL(profile_pc); #if IS_ENABLED(CONFIG_RTC_DRV_GENERIC) static int rtc_generic_get_time(struct device *dev, struct rtc_time *tm) @@ -224,12 +177,27 @@ void read_persistent_clock64(struct timespec64 *ts) } } - static u64 notrace read_cr16_sched_clock(void) { return get_cycles(); } +static u64 notrace read_cr16(struct clocksource *cs) +{ + return get_cycles(); +} + +static struct clocksource clocksource_cr16 = { + .name = "cr16", + .rating = 300, + .read = read_cr16, + .mask = CLOCKSOURCE_MASK(BITS_PER_LONG), + .flags = CLOCK_SOURCE_IS_CONTINUOUS | + CLOCK_SOURCE_VALID_FOR_HRES | + CLOCK_SOURCE_MUST_VERIFY | + CLOCK_SOURCE_VERIFY_PERCPU, +}; + /* * timer interrupt and sched_clock() initialization @@ -237,33 +205,14 @@ static u64 notrace read_cr16_sched_clock(void) void __init time_init(void) { - unsigned long cr16_hz; - - clocktick = (100 * PAGE0->mem_10msec) / HZ; - start_cpu_itimer(); /* get CPU 0 started */ - - cr16_hz = 100 * PAGE0->mem_10msec; /* Hz */ + cr16_clock_freq = 100 * PAGE0->mem_10msec; /* Hz */ + clocktick = cr16_clock_freq / HZ; /* register as sched_clock source */ - sched_clock_register(read_cr16_sched_clock, BITS_PER_LONG, cr16_hz); -} + sched_clock_register(read_cr16_sched_clock, BITS_PER_LONG, cr16_clock_freq); -static int __init init_cr16_clocksource(void) -{ - /* - * The cr16 interval timers are not synchronized across CPUs. - */ - if (num_online_cpus() > 1 && !running_on_qemu) { - clocksource_cr16.name = "cr16_unstable"; - clocksource_cr16.flags = CLOCK_SOURCE_UNSTABLE; - clocksource_cr16.rating = 0; - } + parisc_clockevent_init(); /* register at clocksource framework */ - clocksource_register_hz(&clocksource_cr16, - 100 * PAGE0->mem_10msec); - - return 0; + clocksource_register_hz(&clocksource_cr16, cr16_clock_freq); } - -device_initcall(init_cr16_clocksource); diff --git a/arch/parisc/kernel/traps.c b/arch/parisc/kernel/traps.c index f9696fbf646c..4c7c5df80bd0 100644 --- a/arch/parisc/kernel/traps.c +++ b/arch/parisc/kernel/traps.c @@ -31,12 +31,13 @@ #include <linux/uaccess.h> #include <linux/kdebug.h> #include <linux/kfence.h> +#include <linux/perf_event.h> #include <asm/assembly.h> #include <asm/io.h> #include <asm/irq.h> #include <asm/traps.h> -#include <asm/unaligned.h> +#include <linux/unaligned.h> #include <linux/atomic.h> #include <asm/smp.h> #include <asm/pdc.h> @@ -47,6 +48,12 @@ #include <linux/kgdb.h> #include <linux/kprobes.h> +#include "unaligned.h" + +#if defined(CONFIG_LIGHTWEIGHT_SPINLOCK_CHECK) +#include <asm/spinlock.h> +#endif + #include "../math-emu/math-emu.h" /* for handle_fpe() */ static void parisc_show_stack(struct task_struct *task, @@ -291,24 +298,30 @@ static void handle_break(struct pt_regs *regs) } #ifdef CONFIG_KPROBES - if (unlikely(iir == PARISC_KPROBES_BREAK_INSN)) { + if (unlikely(iir == PARISC_KPROBES_BREAK_INSN && !user_mode(regs))) { parisc_kprobe_break_handler(regs); return; } - if (unlikely(iir == PARISC_KPROBES_BREAK_INSN2)) { + if (unlikely(iir == PARISC_KPROBES_BREAK_INSN2 && !user_mode(regs))) { parisc_kprobe_ss_handler(regs); return; } #endif #ifdef CONFIG_KGDB - if (unlikely(iir == PARISC_KGDB_COMPILED_BREAK_INSN || - iir == PARISC_KGDB_BREAK_INSN)) { + if (unlikely((iir == PARISC_KGDB_COMPILED_BREAK_INSN || + iir == PARISC_KGDB_BREAK_INSN)) && !user_mode(regs)) { kgdb_handle_exception(9, SIGTRAP, 0, regs); return; } #endif +#ifdef CONFIG_LIGHTWEIGHT_SPINLOCK_CHECK + if ((iir == SPINLOCK_BREAK_INSN) && !user_mode(regs)) { + die_if_kernel("Spinlock was trashed", regs, 1); + } +#endif + if (unlikely(iir != GDB_BREAK_INSN)) parisc_printk_ratelimited(0, regs, KERN_DEBUG "break %d,%d: pid=%d command='%s'\n", @@ -325,10 +338,7 @@ static void default_trap(int code, struct pt_regs *regs) show_regs(regs); } -void (*cpu_lpmc) (int code, struct pt_regs *regs) __read_mostly = default_trap; - - -void transfer_pim_to_trap_frame(struct pt_regs *regs) +static void transfer_pim_to_trap_frame(struct pt_regs *regs) { register int i; extern unsigned int hpmc_pim_data[]; @@ -497,7 +507,7 @@ void notrace handle_interruption(int code, struct pt_regs *regs) if (((unsigned long)regs->iaoq[0] & 3) && ((unsigned long)regs->iasq[0] != (unsigned long)regs->sr[7])) { /* Kill the user process later */ - regs->iaoq[0] = 0 | 3; + regs->iaoq[0] = 0 | PRIV_USER; regs->iaoq[1] = regs->iaoq[0] + 4; regs->iasq[0] = regs->iasq[1] = regs->sr[7]; regs->gr[0] &= ~PSW_B; @@ -547,7 +557,7 @@ void notrace handle_interruption(int code, struct pt_regs *regs) flush_cache_all(); flush_tlb_all(); - cpu_lpmc(5, regs); + default_trap(code, regs); return; case PARISC_ITLB_TRAP: @@ -624,6 +634,7 @@ void notrace handle_interruption(int code, struct pt_regs *regs) /* Assist Exception Trap, i.e. floating point exception. */ die_if_kernel("Floating point exception", regs, 0); /* quiet */ __inc_irq_stat(irq_fpassist_count); + perf_sw_event(PERF_COUNT_SW_EMULATION_FAULTS, 1, regs, 0); handle_fpe(regs); return; @@ -790,14 +801,13 @@ void notrace handle_interruption(int code, struct pt_regs *regs) } -void __init initialize_ivt(const void *iva) +static void __init initialize_ivt(const void *iva) { extern const u32 os_hpmc[]; int i; u32 check = 0; u32 *ivap; - u32 *hpmcp; u32 instr; if (strcmp((const char *)iva, "cows can fly")) @@ -830,8 +840,6 @@ void __init initialize_ivt(const void *iva) /* Setup IVA and compute checksum for HPMC handler */ ivap[6] = (u32)__pa(os_hpmc); - hpmcp = (u32 *)os_hpmc; - for (i=0; i<8; i++) check += ivap[i]; diff --git a/arch/parisc/kernel/unaligned.c b/arch/parisc/kernel/unaligned.c index e8a4d77cff53..fb64d9ce0b17 100644 --- a/arch/parisc/kernel/unaligned.c +++ b/arch/parisc/kernel/unaligned.c @@ -11,8 +11,12 @@ #include <linux/signal.h> #include <linux/ratelimit.h> #include <linux/uaccess.h> +#include <linux/sysctl.h> +#include <linux/unaligned.h> +#include <linux/perf_event.h> #include <asm/hardirq.h> #include <asm/traps.h> +#include "unaligned.h" /* #define DEBUG_UNALIGNED 1 */ @@ -22,7 +26,7 @@ #define DPRINTF(fmt, args...) #endif -#define RFMT "%#08lx" +#define RFMT "0x%08lx" /* 1111 1100 0000 0000 0001 0011 1100 0000 */ #define OPCODE1(a,b,c) ((a)<<26|(b)<<12|(c)<<6) @@ -102,6 +106,7 @@ #define ERR_NOTHANDLED -1 int unaligned_enabled __read_mostly = 1; +int no_unaligned_warning __read_mostly; static int emulate_ldh(struct pt_regs *regs, int toreg) { @@ -118,8 +123,8 @@ static int emulate_ldh(struct pt_regs *regs, int toreg) "2: ldbs 1(%%sr1,%3), %0\n" " depw %2, 23, 24, %0\n" "3: \n" - ASM_EXCEPTIONTABLE_ENTRY_EFAULT(1b, 3b) - ASM_EXCEPTIONTABLE_ENTRY_EFAULT(2b, 3b) + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(1b, 3b, "%1") + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(2b, 3b, "%1") : "+r" (val), "+r" (ret), "=&r" (temp1) : "r" (saddr), "r" (regs->isr) ); @@ -150,8 +155,8 @@ static int emulate_ldw(struct pt_regs *regs, int toreg, int flop) " mtctl %2,11\n" " vshd %0,%3,%0\n" "3: \n" - ASM_EXCEPTIONTABLE_ENTRY_EFAULT(1b, 3b) - ASM_EXCEPTIONTABLE_ENTRY_EFAULT(2b, 3b) + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(1b, 3b, "%1") + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(2b, 3b, "%1") : "+r" (val), "+r" (ret), "=&r" (temp1), "=&r" (temp2) : "r" (saddr), "r" (regs->isr) ); @@ -167,6 +172,7 @@ static int emulate_ldw(struct pt_regs *regs, int toreg, int flop) static int emulate_ldd(struct pt_regs *regs, int toreg, int flop) { unsigned long saddr = regs->ior; + unsigned long shift, temp1; __u64 val = 0; ASM_EXCEPTIONTABLE_VAR(ret); @@ -178,25 +184,22 @@ static int emulate_ldd(struct pt_regs *regs, int toreg, int flop) #ifdef CONFIG_64BIT __asm__ __volatile__ ( -" depd,z %3,60,3,%%r19\n" /* r19=(ofs&7)*8 */ -" mtsp %4, %%sr1\n" -" depd %%r0,63,3,%3\n" -"1: ldd 0(%%sr1,%3),%0\n" -"2: ldd 8(%%sr1,%3),%%r20\n" -" subi 64,%%r19,%%r19\n" -" mtsar %%r19\n" -" shrpd %0,%%r20,%%sar,%0\n" +" depd,z %2,60,3,%3\n" /* shift=(ofs&7)*8 */ +" mtsp %5, %%sr1\n" +" depd %%r0,63,3,%2\n" +"1: ldd 0(%%sr1,%2),%0\n" +"2: ldd 8(%%sr1,%2),%4\n" +" subi 64,%3,%3\n" +" mtsar %3\n" +" shrpd %0,%4,%%sar,%0\n" "3: \n" - ASM_EXCEPTIONTABLE_ENTRY_EFAULT(1b, 3b) - ASM_EXCEPTIONTABLE_ENTRY_EFAULT(2b, 3b) - : "=r" (val), "+r" (ret) - : "0" (val), "r" (saddr), "r" (regs->isr) - : "r19", "r20" ); + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(1b, 3b, "%1") + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(2b, 3b, "%1") + : "+r" (val), "+r" (ret), "+r" (saddr), "=&r" (shift), "=&r" (temp1) + : "r" (regs->isr) ); #else - { - unsigned long shift, temp1; __asm__ __volatile__ ( -" zdep %2,29,2,%3\n" /* r19=(ofs&3)*8 */ +" zdep %2,29,2,%3\n" /* shift=(ofs&3)*8 */ " mtsp %5, %%sr1\n" " dep %%r0,31,2,%2\n" "1: ldw 0(%%sr1,%2),%0\n" @@ -207,12 +210,11 @@ static int emulate_ldd(struct pt_regs *regs, int toreg, int flop) " vshd %0,%R0,%0\n" " vshd %R0,%4,%R0\n" "4: \n" - ASM_EXCEPTIONTABLE_ENTRY_EFAULT(1b, 4b) - ASM_EXCEPTIONTABLE_ENTRY_EFAULT(2b, 4b) - ASM_EXCEPTIONTABLE_ENTRY_EFAULT(3b, 4b) + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(1b, 4b, "%1") + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(2b, 4b, "%1") + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(3b, 4b, "%1") : "+r" (val), "+r" (ret), "+r" (saddr), "=&r" (shift), "=&r" (temp1) : "r" (regs->isr) ); - } #endif DPRINTF("val = 0x%llx\n", val); @@ -242,8 +244,8 @@ static int emulate_sth(struct pt_regs *regs, int frreg) "1: stb %1, 0(%%sr1, %3)\n" "2: stb %2, 1(%%sr1, %3)\n" "3: \n" - ASM_EXCEPTIONTABLE_ENTRY_EFAULT(1b, 3b) - ASM_EXCEPTIONTABLE_ENTRY_EFAULT(2b, 3b) + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(1b, 3b, "%0") + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(2b, 3b, "%0") : "+r" (ret), "=&r" (temp1) : "r" (val), "r" (regs->ior), "r" (regs->isr) ); @@ -283,8 +285,8 @@ static int emulate_stw(struct pt_regs *regs, int frreg, int flop) " stw %%r20,0(%%sr1,%2)\n" " stw %%r21,4(%%sr1,%2)\n" "3: \n" - ASM_EXCEPTIONTABLE_ENTRY_EFAULT(1b, 3b) - ASM_EXCEPTIONTABLE_ENTRY_EFAULT(2b, 3b) + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(1b, 3b, "%0") + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(2b, 3b, "%0") : "+r" (ret) : "r" (val), "r" (regs->ior), "r" (regs->isr) : "r19", "r20", "r21", "r22", "r1" ); @@ -327,42 +329,41 @@ static int emulate_std(struct pt_regs *regs, int frreg, int flop) "3: std %%r20,0(%%sr1,%2)\n" "4: std %%r21,8(%%sr1,%2)\n" "5: \n" - ASM_EXCEPTIONTABLE_ENTRY_EFAULT(1b, 5b) - ASM_EXCEPTIONTABLE_ENTRY_EFAULT(2b, 5b) - ASM_EXCEPTIONTABLE_ENTRY_EFAULT(3b, 5b) - ASM_EXCEPTIONTABLE_ENTRY_EFAULT(4b, 5b) + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(1b, 5b, "%0") + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(2b, 5b, "%0") + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(3b, 5b, "%0") + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(4b, 5b, "%0") : "+r" (ret) : "r" (val), "r" (regs->ior), "r" (regs->isr) : "r19", "r20", "r21", "r22", "r1" ); #else { - unsigned long valh=(val>>32),vall=(val&0xffffffffl); __asm__ __volatile__ ( -" mtsp %4, %%sr1\n" -" zdep %2, 29, 2, %%r19\n" -" dep %%r0, 31, 2, %3\n" +" mtsp %3, %%sr1\n" +" zdep %R1, 29, 2, %%r19\n" +" dep %%r0, 31, 2, %2\n" " mtsar %%r19\n" " zvdepi -2, 32, %%r19\n" -"1: ldw 0(%%sr1,%3),%%r20\n" -"2: ldw 8(%%sr1,%3),%%r21\n" -" vshd %1, %2, %%r1\n" +"1: ldw 0(%%sr1,%2),%%r20\n" +"2: ldw 8(%%sr1,%2),%%r21\n" +" vshd %1, %R1, %%r1\n" " vshd %%r0, %1, %1\n" -" vshd %2, %%r0, %2\n" +" vshd %R1, %%r0, %R1\n" " and %%r20, %%r19, %%r20\n" " andcm %%r21, %%r19, %%r21\n" " or %1, %%r20, %1\n" -" or %2, %%r21, %2\n" -"3: stw %1,0(%%sr1,%3)\n" -"4: stw %%r1,4(%%sr1,%3)\n" -"5: stw %2,8(%%sr1,%3)\n" +" or %R1, %%r21, %R1\n" +"3: stw %1,0(%%sr1,%2)\n" +"4: stw %%r1,4(%%sr1,%2)\n" +"5: stw %R1,8(%%sr1,%2)\n" "6: \n" - ASM_EXCEPTIONTABLE_ENTRY_EFAULT(1b, 6b) - ASM_EXCEPTIONTABLE_ENTRY_EFAULT(2b, 6b) - ASM_EXCEPTIONTABLE_ENTRY_EFAULT(3b, 6b) - ASM_EXCEPTIONTABLE_ENTRY_EFAULT(4b, 6b) - ASM_EXCEPTIONTABLE_ENTRY_EFAULT(5b, 6b) + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(1b, 6b, "%0") + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(2b, 6b, "%0") + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(3b, 6b, "%0") + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(4b, 6b, "%0") + ASM_EXCEPTIONTABLE_ENTRY_EFAULT(5b, 6b, "%0") : "+r" (ret) - : "r" (valh), "r" (vall), "r" (regs->ior), "r" (regs->isr) + : "r" (val), "r" (regs->ior), "r" (regs->isr) : "r19", "r20", "r21", "r1" ); } #endif @@ -378,6 +379,7 @@ void handle_unaligned(struct pt_regs *regs) int ret = ERR_NOTHANDLED; __inc_irq_stat(irq_unaligned_count); + perf_sw_event(PERF_COUNT_SW_ALIGNMENT_FAULTS, 1, regs, regs->ior); /* log a message with pacing */ if (user_mode(regs)) { @@ -398,6 +400,14 @@ void handle_unaligned(struct pt_regs *regs) if (!unaligned_enabled) goto force_sigbus; + } else { + static DEFINE_RATELIMIT_STATE(kernel_ratelimit, 5 * HZ, 5); + if (!(current->thread.flags & PARISC_UAC_NOPRINT) && + !no_unaligned_warning && + __ratelimit(&kernel_ratelimit)) + pr_warn("Kernel: unaligned access to " RFMT " in %pS " + "(iir " RFMT ")\n", + regs->ior, (void *)regs->iaoq[0], regs->iir); } /* handle modification - OK, it's ugly, see the instruction manual */ @@ -472,7 +482,7 @@ void handle_unaligned(struct pt_regs *regs) case OPCODE_LDWA_I: case OPCODE_LDW_S: case OPCODE_LDWA_S: - ret = emulate_ldw(regs, R3(regs->iir),0); + ret = emulate_ldw(regs, R3(regs->iir), 0); break; case OPCODE_STH: @@ -481,7 +491,7 @@ void handle_unaligned(struct pt_regs *regs) case OPCODE_STW: case OPCODE_STWA: - ret = emulate_stw(regs, R2(regs->iir),0); + ret = emulate_stw(regs, R2(regs->iir), 0); break; #ifdef CONFIG_64BIT @@ -489,12 +499,12 @@ void handle_unaligned(struct pt_regs *regs) case OPCODE_LDDA_I: case OPCODE_LDD_S: case OPCODE_LDDA_S: - ret = emulate_ldd(regs, R3(regs->iir),0); + ret = emulate_ldd(regs, R3(regs->iir), 0); break; case OPCODE_STD: case OPCODE_STDA: - ret = emulate_std(regs, R2(regs->iir),0); + ret = emulate_std(regs, R2(regs->iir), 0); break; #endif @@ -502,24 +512,24 @@ void handle_unaligned(struct pt_regs *regs) case OPCODE_FLDWS: case OPCODE_FLDWXR: case OPCODE_FLDWSR: - ret = emulate_ldw(regs,FR3(regs->iir),1); + ret = emulate_ldw(regs, FR3(regs->iir), 1); break; case OPCODE_FLDDX: case OPCODE_FLDDS: - ret = emulate_ldd(regs,R3(regs->iir),1); + ret = emulate_ldd(regs, R3(regs->iir), 1); break; case OPCODE_FSTWX: case OPCODE_FSTWS: case OPCODE_FSTWXR: case OPCODE_FSTWSR: - ret = emulate_stw(regs,FR3(regs->iir),1); + ret = emulate_stw(regs, FR3(regs->iir), 1); break; case OPCODE_FSTDX: case OPCODE_FSTDS: - ret = emulate_std(regs,R3(regs->iir),1); + ret = emulate_std(regs, R3(regs->iir), 1); break; case OPCODE_LDCD_I: diff --git a/arch/parisc/kernel/unaligned.h b/arch/parisc/kernel/unaligned.h new file mode 100644 index 000000000000..c1aa4b12e284 --- /dev/null +++ b/arch/parisc/kernel/unaligned.h @@ -0,0 +1,3 @@ +struct pt_regs; +void handle_unaligned(struct pt_regs *regs); +int check_unaligned(struct pt_regs *regs); diff --git a/arch/parisc/kernel/unwind.c b/arch/parisc/kernel/unwind.c index 42acc3b52017..7ac88ff13d3c 100644 --- a/arch/parisc/kernel/unwind.c +++ b/arch/parisc/kernel/unwind.c @@ -24,16 +24,19 @@ #include <asm/unwind.h> #include <asm/switch_to.h> #include <asm/sections.h> +#include <asm/ftrace.h> /* #define DEBUG 1 */ #ifdef DEBUG #define dbg(x...) pr_debug(x) #else -#define dbg(x...) +#define dbg(x...) do { } while (0) #endif #define KERNEL_START (KERNEL_BINARY_TEXT_START) +#define ALIGNMENT_OK(ptr, type) (((ptr) & (sizeof(type) - 1)) == 0) + extern struct unwind_table_entry __start___unwind[]; extern struct unwind_table_entry __stop___unwind[]; @@ -179,7 +182,7 @@ void unwind_table_remove(struct unwind_table *table) /* Called from setup_arch to import the kernel unwind info */ int __init unwind_init(void) { - long start, stop; + long start __maybe_unused, stop __maybe_unused; register unsigned long gp __asm__ ("r27"); start = (long)&__start___unwind[0]; @@ -220,7 +223,6 @@ static int unwind_special(struct unwind_frame_info *info, unsigned long pc, int * Note: We could use dereference_kernel_function_descriptor() * instead but we want to keep it simple here. */ - extern void * const handle_interruption; extern void * const ret_from_kernel_thread; extern void * const syscall_exit; extern void * const intr_return; @@ -237,13 +239,13 @@ static int unwind_special(struct unwind_frame_info *info, unsigned long pc, int return 1; } - if (pc_is_kernel_fn(pc, ret_from_kernel_thread) || - pc_is_kernel_fn(pc, syscall_exit)) { + if (pc == (unsigned long)&ret_from_kernel_thread || + pc == (unsigned long)&syscall_exit) { info->prev_sp = info->prev_ip = 0; return 1; } - if (pc_is_kernel_fn(pc, intr_return)) { + if (pc == (unsigned long)&intr_return) { struct pt_regs *regs; dbg("Found intr_return()\n"); @@ -255,14 +257,17 @@ static int unwind_special(struct unwind_frame_info *info, unsigned long pc, int } if (pc_is_kernel_fn(pc, _switch_to) || - pc_is_kernel_fn(pc, _switch_to_ret)) { + pc == (unsigned long)&_switch_to_ret) { info->prev_sp = info->sp - CALLEE_SAVE_FRAME_SIZE; - info->prev_ip = *(unsigned long *)(info->prev_sp - RP_OFFSET); + if (ALIGNMENT_OK(info->prev_sp, long)) + info->prev_ip = *(unsigned long *)(info->prev_sp - RP_OFFSET); + else + info->prev_ip = info->prev_sp = 0; return 1; } #ifdef CONFIG_IRQSTACKS - if (pc_is_kernel_fn(pc, _call_on_stack)) { + if (pc == (unsigned long)&_call_on_stack && ALIGNMENT_OK(info->sp, long)) { info->prev_sp = *(unsigned long *)(info->sp - FRAME_SIZE - REG_SZ); info->prev_ip = *(unsigned long *)(info->sp - FRAME_SIZE - RP_OFFSET); return 1; @@ -370,8 +375,10 @@ static void unwind_frame_regs(struct unwind_frame_info *info) info->prev_sp = info->sp - frame_size; if (e->Millicode) info->rp = info->r31; - else if (rpoffset) + else if (rpoffset && ALIGNMENT_OK(info->prev_sp, long)) info->rp = *(unsigned long *)(info->prev_sp - rpoffset); + else + info->rp = 0; info->prev_ip = info->rp; info->rp = 0; } diff --git a/arch/parisc/kernel/vdso32/Makefile b/arch/parisc/kernel/vdso32/Makefile index 4459a48d2303..4ee8d17da229 100644 --- a/arch/parisc/kernel/vdso32/Makefile +++ b/arch/parisc/kernel/vdso32/Makefile @@ -1,11 +1,25 @@ -# List of files in the vdso, has to be asm only for now +# Include the generic Makefile to check the built vdso. +include $(srctree)/lib/vdso/Makefile.include + +KCOV_INSTRUMENT := n + +# Disable gcov profiling, ubsan and kasan for VDSO code +GCOV_PROFILE := n +UBSAN_SANITIZE := n +KASAN_SANITIZE := n +KCSAN_SANITIZE := n obj-vdso32 = note.o sigtramp.o restart_syscall.o +obj-cvdso32 = vdso32_generic.o # Build rules -targets := $(obj-vdso32) vdso32.so +targets := $(obj-vdso32) $(obj-cvdso32) vdso32.so obj-vdso32 := $(addprefix $(obj)/, $(obj-vdso32)) +obj-cvdso32 := $(addprefix $(obj)/, $(obj-cvdso32)) + +VDSO_CFLAGS_REMOVE := -pg $(CC_FLAGS_FTRACE) +CFLAGS_REMOVE_vdso32_generic.o = $(VDSO_CFLAGS_REMOVE) ccflags-y := -shared -fno-common -fbuiltin -mno-fast-indirect-calls -O2 -mno-long-calls # -march=1.1 -mschedule=7100LC @@ -19,20 +33,19 @@ KBUILD_CFLAGS += -DBUILD_VDSO -DDISABLE_BRANCH_PROFILING VDSO_LIBGCC := $(shell $(CROSS32CC) -print-libgcc-file-name) obj-y += vdso32_wrapper.o -extra-y += vdso32.lds +targets += vdso32.lds CPPFLAGS_vdso32.lds += -P -C # -U$(ARCH) $(obj)/vdso32_wrapper.o : $(obj)/vdso32.so FORCE # Force dependency (incbin is bad) # link rule for the .so file, .lds has to be first -$(obj)/vdso32.so: $(src)/vdso32.lds $(obj-vdso32) $(obj-cvdso32) $(VDSO_LIBGCC) FORCE +$(obj)/vdso32.so: $(obj)/vdso32.lds $(obj-vdso32) $(obj-cvdso32) $(VDSO_LIBGCC) FORCE $(call if_changed,vdso32ld) # assembly rules for the .S files $(obj-vdso32): %.o: %.S FORCE $(call if_changed_dep,vdso32as) - $(obj-cvdso32): %.o: %.c FORCE $(call if_changed_dep,vdso32cc) @@ -42,10 +55,10 @@ quiet_cmd_vdso32ld = VDSO32L $@ quiet_cmd_vdso32as = VDSO32A $@ cmd_vdso32as = $(CROSS32CC) $(a_flags) -c -o $@ $< quiet_cmd_vdso32cc = VDSO32C $@ - cmd_vdso32cc = $(CROSS32CC) $(c_flags) -c -fPIC -mno-fast-indirect-calls -o $@ $< + cmd_vdso32cc = $(CROSS32CC) $(c_flags) -c -o $@ $< # Generate VDSO offsets using helper script -gen-vdsosym := $(srctree)/$(src)/gen_vdso_offsets.sh +gen-vdsosym := $(src)/gen_vdso_offsets.sh quiet_cmd_vdsosym = VDSOSYM $@ cmd_vdsosym = $(NM) $< | $(gen-vdsosym) | LC_ALL=C sort > $@ diff --git a/arch/parisc/kernel/vdso32/vdso32.lds.S b/arch/parisc/kernel/vdso32/vdso32.lds.S index d4aff3af5262..4273baa26b65 100644 --- a/arch/parisc/kernel/vdso32/vdso32.lds.S +++ b/arch/parisc/kernel/vdso32/vdso32.lds.S @@ -106,6 +106,9 @@ VERSION global: __kernel_sigtramp_rt32; __kernel_restart_syscall32; + __vdso_gettimeofday; + __vdso_clock_gettime; + __vdso_clock_gettime64; local: *; }; } diff --git a/arch/parisc/kernel/vdso32/vdso32_generic.c b/arch/parisc/kernel/vdso32/vdso32_generic.c new file mode 100644 index 000000000000..8d5bd59e8646 --- /dev/null +++ b/arch/parisc/kernel/vdso32/vdso32_generic.c @@ -0,0 +1,32 @@ +// SPDX-License-Identifier: GPL-2.0 + +#include "asm/unistd.h" +#include <linux/types.h> +#include <uapi/asm/unistd_32.h> + +struct timezone; +struct old_timespec32; +struct __kernel_timespec; +struct __kernel_old_timeval; + +/* forward declarations */ +int __vdso_gettimeofday(struct __kernel_old_timeval *tv, struct timezone *tz); +int __vdso_clock_gettime(clockid_t clock, struct old_timespec32 *ts); +int __vdso_clock_gettime64(clockid_t clock, struct __kernel_timespec *ts); + + +int __vdso_gettimeofday(struct __kernel_old_timeval *tv, + struct timezone *tz) +{ + return syscall2(__NR_gettimeofday, (long)tv, (long)tz); +} + +int __vdso_clock_gettime(clockid_t clock, struct old_timespec32 *ts) +{ + return syscall2(__NR_clock_gettime, (long)clock, (long)ts); +} + +int __vdso_clock_gettime64(clockid_t clock, struct __kernel_timespec *ts) +{ + return syscall2(__NR_clock_gettime64, (long)clock, (long)ts); +} diff --git a/arch/parisc/kernel/vdso64/Makefile b/arch/parisc/kernel/vdso64/Makefile index f3d6045793f4..c63f4069170f 100644 --- a/arch/parisc/kernel/vdso64/Makefile +++ b/arch/parisc/kernel/vdso64/Makefile @@ -1,12 +1,25 @@ -# List of files in the vdso, has to be asm only for now +# Include the generic Makefile to check the built vdso. +include $(srctree)/lib/vdso/Makefile.include + +KCOV_INSTRUMENT := n + +# Disable gcov profiling, ubsan and kasan for VDSO code +GCOV_PROFILE := n +UBSAN_SANITIZE := n +KASAN_SANITIZE := n +KCSAN_SANITIZE := n obj-vdso64 = note.o sigtramp.o restart_syscall.o +obj-cvdso64 = vdso64_generic.o # Build rules -targets := $(obj-vdso64) vdso64.so -obj-vdso64 := $(addprefix $(obj)/, $(obj-vdso64)) +targets := $(obj-vdso64) $(obj-cvdso64) vdso64.so +obj-vdso64 := $(addprefix $(obj)/, $(obj-vdso64)) +obj-cvdso64 := $(addprefix $(obj)/, $(obj-cvdso64)) +VDSO_CFLAGS_REMOVE := -pg $(CC_FLAGS_FTRACE) +CFLAGS_REMOVE_vdso64_generic.o = $(VDSO_CFLAGS_REMOVE) ccflags-y := -shared -fno-common -fno-builtin ccflags-y += -nostdlib -Wl,-soname=linux-vdso64.so.1 \ @@ -19,28 +32,32 @@ KBUILD_CFLAGS += -DBUILD_VDSO -DDISABLE_BRANCH_PROFILING VDSO_LIBGCC := $(shell $(CC) -print-libgcc-file-name) obj-y += vdso64_wrapper.o -extra-y += vdso64.lds +targets += vdso64.lds CPPFLAGS_vdso64.lds += -P -C -U$(ARCH) $(obj)/vdso64_wrapper.o : $(obj)/vdso64.so FORCE # Force dependency (incbin is bad) # link rule for the .so file, .lds has to be first -$(obj)/vdso64.so: $(src)/vdso64.lds $(obj-vdso64) $(VDSO_LIBGCC) FORCE +$(obj)/vdso64.so: $(obj)/vdso64.lds $(obj-vdso64) $(obj-cvdso64) $(VDSO_LIBGCC) FORCE $(call if_changed,vdso64ld) # assembly rules for the .S files $(obj-vdso64): %.o: %.S FORCE $(call if_changed_dep,vdso64as) +$(obj-cvdso64): %.o: %.c FORCE + $(call if_changed_dep,vdso64cc) # actual build commands quiet_cmd_vdso64ld = VDSO64L $@ cmd_vdso64ld = $(CC) $(c_flags) -Wl,-T $(filter-out FORCE, $^) -o $@ quiet_cmd_vdso64as = VDSO64A $@ cmd_vdso64as = $(CC) $(a_flags) -c -o $@ $< +quiet_cmd_vdso64cc = VDSO64C $@ + cmd_vdso64cc = $(CC) $(c_flags) -c -o $@ $< # Generate VDSO offsets using helper script -gen-vdsosym := $(srctree)/$(src)/gen_vdso_offsets.sh +gen-vdsosym := $(src)/gen_vdso_offsets.sh quiet_cmd_vdsosym = VDSOSYM $@ cmd_vdsosym = $(NM) $< | $(gen-vdsosym) | LC_ALL=C sort > $@ diff --git a/arch/parisc/kernel/vdso64/vdso64.lds.S b/arch/parisc/kernel/vdso64/vdso64.lds.S index de1fb4b19286..10f25e4e1554 100644 --- a/arch/parisc/kernel/vdso64/vdso64.lds.S +++ b/arch/parisc/kernel/vdso64/vdso64.lds.S @@ -104,6 +104,8 @@ VERSION global: __kernel_sigtramp_rt64; __kernel_restart_syscall64; + __vdso_gettimeofday; + __vdso_clock_gettime; local: *; }; } diff --git a/arch/parisc/kernel/vdso64/vdso64_generic.c b/arch/parisc/kernel/vdso64/vdso64_generic.c new file mode 100644 index 000000000000..fc6836a0075b --- /dev/null +++ b/arch/parisc/kernel/vdso64/vdso64_generic.c @@ -0,0 +1,24 @@ +// SPDX-License-Identifier: GPL-2.0 + +#include "asm/unistd.h" +#include <linux/types.h> + +struct timezone; +struct __kernel_timespec; +struct __kernel_old_timeval; + +/* forward declarations */ +int __vdso_gettimeofday(struct __kernel_old_timeval *tv, struct timezone *tz); +int __vdso_clock_gettime(clockid_t clock, struct __kernel_timespec *ts); + + +int __vdso_gettimeofday(struct __kernel_old_timeval *tv, + struct timezone *tz) +{ + return syscall2(__NR_gettimeofday, (long)tv, (long)tz); +} + +int __vdso_clock_gettime(clockid_t clock, struct __kernel_timespec *ts) +{ + return syscall2(__NR_clock_gettime, (long)clock, (long)ts); +} diff --git a/arch/parisc/kernel/vmlinux.lds.S b/arch/parisc/kernel/vmlinux.lds.S index 2769eb991f58..b445e47903cf 100644 --- a/arch/parisc/kernel/vmlinux.lds.S +++ b/arch/parisc/kernel/vmlinux.lds.S @@ -86,7 +86,6 @@ SECTIONS TEXT_TEXT LOCK_TEXT SCHED_TEXT - CPUIDLE_TEXT KPROBES_TEXT IRQENTRY_TEXT SOFTIRQENTRY_TEXT @@ -128,9 +127,10 @@ SECTIONS } #endif - RO_DATA(8) + RO_DATA(PAGE_SIZE) /* unwind info */ + . = ALIGN(4); .PARISC.unwind : { __start___unwind = .; *(.PARISC.unwind) @@ -155,6 +155,7 @@ SECTIONS } /* End of data section */ + . = ALIGN(PAGE_SIZE); _edata = .; /* BSS */ diff --git a/arch/parisc/lib/bitops.c b/arch/parisc/lib/bitops.c index 36a314199074..9df810050642 100644 --- a/arch/parisc/lib/bitops.c +++ b/arch/parisc/lib/bitops.c @@ -56,38 +56,20 @@ unsigned long notrace __xchg8(char x, volatile char *ptr) } -u64 notrace __cmpxchg_u64(volatile u64 *ptr, u64 old, u64 new) -{ - unsigned long flags; - u64 prev; - - _atomic_spin_lock_irqsave(ptr, flags); - if ((prev = *ptr) == old) - *ptr = new; - _atomic_spin_unlock_irqrestore(ptr, flags); - return prev; -} - -unsigned long notrace __cmpxchg_u32(volatile unsigned int *ptr, unsigned int old, unsigned int new) -{ - unsigned long flags; - unsigned int prev; - - _atomic_spin_lock_irqsave(ptr, flags); - if ((prev = *ptr) == old) - *ptr = new; - _atomic_spin_unlock_irqrestore(ptr, flags); - return (unsigned long)prev; -} - -u8 notrace __cmpxchg_u8(volatile u8 *ptr, u8 old, u8 new) -{ - unsigned long flags; - u8 prev; - - _atomic_spin_lock_irqsave(ptr, flags); - if ((prev = *ptr) == old) - *ptr = new; - _atomic_spin_unlock_irqrestore(ptr, flags); - return prev; -} +#define CMPXCHG(T) \ + T notrace __cmpxchg_##T(volatile T *ptr, T old, T new) \ + { \ + unsigned long flags; \ + T prev; \ + \ + _atomic_spin_lock_irqsave(ptr, flags); \ + if ((prev = *ptr) == old) \ + *ptr = new; \ + _atomic_spin_unlock_irqrestore(ptr, flags); \ + return prev; \ + } + +CMPXCHG(u64) +CMPXCHG(u32) +CMPXCHG(u16) +CMPXCHG(u8) diff --git a/arch/parisc/lib/checksum.c b/arch/parisc/lib/checksum.c index 4818f3db84a5..59d8c15d81bd 100644 --- a/arch/parisc/lib/checksum.c +++ b/arch/parisc/lib/checksum.c @@ -25,15 +25,6 @@ : "=r"(_t) \ : "r"(_r), "0"(_t)); -static inline unsigned short from32to16(unsigned int x) -{ - /* 32 bits --> 16 bits + carry */ - x = (x & 0xffff) + (x >> 16); - /* 16 bits + carry --> 16 bits including carry */ - x = (x & 0xffff) + (x >> 16); - return (unsigned short)x; -} - static inline unsigned int do_csum(const unsigned char * buff, int len) { int odd, count; @@ -85,7 +76,7 @@ static inline unsigned int do_csum(const unsigned char * buff, int len) } if (len & 1) result += le16_to_cpu(*buff); - result = from32to16(result); + result = csum_from32to16(result); if (odd) result = swab16(result); out: @@ -102,7 +93,7 @@ __wsum csum_partial(const void *buff, int len, __wsum sum) { unsigned int result = do_csum(buff, len); addc(result, sum); - return (__force __wsum)from32to16(result); + return (__force __wsum)csum_from32to16(result); } EXPORT_SYMBOL(csum_partial); diff --git a/arch/parisc/lib/io.c b/arch/parisc/lib/io.c index 7c00496b47d4..3c7e617f5a93 100644 --- a/arch/parisc/lib/io.c +++ b/arch/parisc/lib/io.c @@ -12,114 +12,6 @@ #include <linux/module.h> #include <asm/io.h> -/* Copies a block of memory to a device in an efficient manner. - * Assumes the device can cope with 32-bit transfers. If it can't, - * don't use this function. - */ -void memcpy_toio(volatile void __iomem *dst, const void *src, int count) -{ - if (((unsigned long)dst & 3) != ((unsigned long)src & 3)) - goto bytecopy; - while ((unsigned long)dst & 3) { - writeb(*(char *)src, dst++); - src++; - count--; - } - while (count > 3) { - __raw_writel(*(u32 *)src, dst); - src += 4; - dst += 4; - count -= 4; - } - bytecopy: - while (count--) { - writeb(*(char *)src, dst++); - src++; - } -} - -/* -** Copies a block of memory from a device in an efficient manner. -** Assumes the device can cope with 32-bit transfers. If it can't, -** don't use this function. -** -** CR16 counts on C3000 reading 256 bytes from Symbios 896 RAM: -** 27341/64 = 427 cyc per int -** 61311/128 = 478 cyc per short -** 122637/256 = 479 cyc per byte -** Ergo bus latencies dominant (not transfer size). -** Minimize total number of transfers at cost of CPU cycles. -** TODO: only look at src alignment and adjust the stores to dest. -*/ -void memcpy_fromio(void *dst, const volatile void __iomem *src, int count) -{ - /* first compare alignment of src/dst */ - if ( (((unsigned long)dst ^ (unsigned long)src) & 1) || (count < 2) ) - goto bytecopy; - - if ( (((unsigned long)dst ^ (unsigned long)src) & 2) || (count < 4) ) - goto shortcopy; - - /* Then check for misaligned start address */ - if ((unsigned long)src & 1) { - *(u8 *)dst = readb(src); - src++; - dst++; - count--; - if (count < 2) goto bytecopy; - } - - if ((unsigned long)src & 2) { - *(u16 *)dst = __raw_readw(src); - src += 2; - dst += 2; - count -= 2; - } - - while (count > 3) { - *(u32 *)dst = __raw_readl(src); - dst += 4; - src += 4; - count -= 4; - } - - shortcopy: - while (count > 1) { - *(u16 *)dst = __raw_readw(src); - src += 2; - dst += 2; - count -= 2; - } - - bytecopy: - while (count--) { - *(char *)dst = readb(src); - src++; - dst++; - } -} - -/* Sets a block of memory on a device to a given value. - * Assumes the device can cope with 32-bit transfers. If it can't, - * don't use this function. - */ -void memset_io(volatile void __iomem *addr, unsigned char val, int count) -{ - u32 val32 = (val << 24) | (val << 16) | (val << 8) | val; - while ((unsigned long)addr & 3) { - writeb(val, addr++); - count--; - } - while (count > 3) { - __raw_writel(val32, addr); - addr += 4; - count -= 4; - } - while (count--) { - writeb(val, addr++); - } -} - /* * Read COUNT 8-bit bytes from port PORT into memory starting at * SRC. @@ -170,15 +62,15 @@ void insw (unsigned long port, void *dst, unsigned long count) unsigned char *p; p = (unsigned char *)dst; - + if (!count) return; - + switch (((unsigned long)p) & 0x3) { case 0x00: /* Buffer 32-bit aligned */ while (count>=2) { - + count -= 2; l = cpu_to_le16(inw(port)) << 16; l |= cpu_to_le16(inw(port)); @@ -189,13 +81,13 @@ void insw (unsigned long port, void *dst, unsigned long count) *(unsigned short *)p = cpu_to_le16(inw(port)); } break; - + case 0x02: /* Buffer 16-bit aligned */ *(unsigned short *)p = cpu_to_le16(inw(port)); p += 2; count--; while (count>=2) { - + count -= 2; l = cpu_to_le16(inw(port)) << 16; l |= cpu_to_le16(inw(port)); @@ -206,13 +98,13 @@ void insw (unsigned long port, void *dst, unsigned long count) *(unsigned short *)p = cpu_to_le16(inw(port)); } break; - + case 0x01: /* Buffer 8-bit aligned */ case 0x03: /* I don't bother with 32bit transfers * in this case, 16bit will have to do -- DE */ --count; - + l = cpu_to_le16(inw(port)); *p = l >> 8; p++; @@ -242,10 +134,10 @@ void insl (unsigned long port, void *dst, unsigned long count) unsigned char *p; p = (unsigned char *)dst; - + if (!count) return; - + switch (((unsigned long) dst) & 0x3) { case 0x00: /* Buffer 32-bit aligned */ @@ -255,14 +147,14 @@ void insl (unsigned long port, void *dst, unsigned long count) p += 4; } break; - + case 0x02: /* Buffer 16-bit aligned */ --count; - + l = cpu_to_le32(inl(port)); *(unsigned short *)p = l >> 16; p += 2; - + while (count--) { l2 = cpu_to_le32(inl(port)); @@ -274,7 +166,7 @@ void insl (unsigned long port, void *dst, unsigned long count) break; case 0x01: /* Buffer 8-bit aligned */ --count; - + l = cpu_to_le32(inl(port)); *(unsigned char *)p = l >> 24; p++; @@ -291,7 +183,7 @@ void insl (unsigned long port, void *dst, unsigned long count) break; case 0x03: /* Buffer 8-bit aligned */ --count; - + l = cpu_to_le32(inl(port)); *p = l >> 24; p++; @@ -340,10 +232,10 @@ void outsw (unsigned long port, const void *src, unsigned long count) const unsigned char *p; p = (const unsigned char *)src; - + if (!count) return; - + switch (((unsigned long)p) & 0x3) { case 0x00: /* Buffer 32-bit aligned */ @@ -358,13 +250,13 @@ void outsw (unsigned long port, const void *src, unsigned long count) outw(le16_to_cpu(*(unsigned short*)p), port); } break; - + case 0x02: /* Buffer 16-bit aligned */ - + outw(le16_to_cpu(*(unsigned short*)p), port); p += 2; count--; - + while (count>=2) { count -= 2; l = *(unsigned int *)p; @@ -376,11 +268,11 @@ void outsw (unsigned long port, const void *src, unsigned long count) outw(le16_to_cpu(*(unsigned short *)p), port); } break; - - case 0x01: /* Buffer 8-bit aligned */ + + case 0x01: /* Buffer 8-bit aligned */ /* I don't bother with 32bit transfers * in this case, 16bit will have to do -- DE */ - + l = *p << 8; p++; count--; @@ -395,7 +287,7 @@ void outsw (unsigned long port, const void *src, unsigned long count) l2 = *(unsigned char *)p; outw (le16_to_cpu(l | l2>>8), port); break; - + } } @@ -412,10 +304,10 @@ void outsl (unsigned long port, const void *src, unsigned long count) const unsigned char *p; p = (const unsigned char *)src; - + if (!count) return; - + switch (((unsigned long)p) & 0x3) { case 0x00: /* Buffer 32-bit aligned */ @@ -425,13 +317,13 @@ void outsl (unsigned long port, const void *src, unsigned long count) p += 4; } break; - + case 0x02: /* Buffer 16-bit aligned */ --count; - + l = *(unsigned short *)p; p += 2; - + while (count--) { l2 = *(unsigned int *)p; @@ -462,7 +354,7 @@ void outsl (unsigned long port, const void *src, unsigned long count) break; case 0x03: /* Buffer 8-bit aligned */ --count; - + l = *p << 24; p++; diff --git a/arch/parisc/lib/memcpy.c b/arch/parisc/lib/memcpy.c index 5fc0c852c84c..03165c82dfdb 100644 --- a/arch/parisc/lib/memcpy.c +++ b/arch/parisc/lib/memcpy.c @@ -12,6 +12,7 @@ #include <linux/module.h> #include <linux/compiler.h> #include <linux/uaccess.h> +#include <linux/mm.h> #define get_user_space() mfsp(SR_USER) #define get_kernel_space() SR_KERNEL @@ -32,9 +33,24 @@ EXPORT_SYMBOL(raw_copy_to_user); unsigned long raw_copy_from_user(void *dst, const void __user *src, unsigned long len) { + unsigned long start = (unsigned long) src; + unsigned long end = start + len; + unsigned long newlen = len; + mtsp(get_user_space(), SR_TEMP1); mtsp(get_kernel_space(), SR_TEMP2); - return pa_memcpy(dst, (void __force *)src, len); + + /* Check region is user accessible */ + while (start < end) { + if (!prober_user(SR_TEMP1, start)) { + newlen = (start - (unsigned long) src); + break; + } + start += PAGE_SIZE; + /* align to page boundry which may have different permission */ + start = PAGE_ALIGN_DOWN(start); + } + return len - newlen + pa_memcpy(dst, (void __force *)src, newlen); } EXPORT_SYMBOL(raw_copy_from_user); diff --git a/arch/parisc/lib/ucmpdi2.c b/arch/parisc/lib/ucmpdi2.c index 8e6014a142ef..9d8b4dbae273 100644 --- a/arch/parisc/lib/ucmpdi2.c +++ b/arch/parisc/lib/ucmpdi2.c @@ -1,5 +1,6 @@ // SPDX-License-Identifier: GPL-2.0 #include <linux/module.h> +#include <linux/libgcc.h> union ull_union { unsigned long long ull; @@ -9,7 +10,7 @@ union ull_union { } ui; }; -int __ucmpdi2(unsigned long long a, unsigned long long b) +word_type __ucmpdi2(unsigned long long a, unsigned long long b) { union ull_union au = {.ull = a}; union ull_union bu = {.ull = b}; diff --git a/arch/parisc/math-emu/Makefile b/arch/parisc/math-emu/Makefile index 3747a0cbd3b8..7b64740e150a 100644 --- a/arch/parisc/math-emu/Makefile +++ b/arch/parisc/math-emu/Makefile @@ -6,7 +6,8 @@ # See arch/parisc/math-emu/README ccflags-y := -Wno-parentheses -Wno-implicit-function-declaration \ -Wno-uninitialized -Wno-strict-prototypes -Wno-return-type \ - -Wno-implicit-int + -Wno-implicit-int -Wno-missing-prototypes -Wno-missing-declarations \ + -Wno-old-style-definition -Wno-unused-but-set-variable obj-y := frnd.o driver.o decode_exc.o fpudispatch.o denormal.o \ dfmpy.o sfmpy.o sfsqrt.o dfsqrt.o dfadd.o fmpyfadd.o \ diff --git a/arch/parisc/math-emu/dfsqrt.c b/arch/parisc/math-emu/dfsqrt.c index 63d339c81c14..e3a3a19b966b 100644 --- a/arch/parisc/math-emu/dfsqrt.c +++ b/arch/parisc/math-emu/dfsqrt.c @@ -15,7 +15,7 @@ * Double Floating-point Square Root * * External Interfaces: - * dbl_fsqrt(srcptr,nullptr,dstptr,status) + * dbl_fsqrt(srcptr,_nullptr,dstptr,status) * * Internal Interfaces: * @@ -37,7 +37,7 @@ unsigned int dbl_fsqrt( dbl_floating_point *srcptr, - unsigned int *nullptr, + unsigned int *_nullptr, dbl_floating_point *dstptr, unsigned int *status) { diff --git a/arch/parisc/math-emu/driver.c b/arch/parisc/math-emu/driver.c index 6ce427b58836..71829cb7bc81 100644 --- a/arch/parisc/math-emu/driver.c +++ b/arch/parisc/math-emu/driver.c @@ -26,12 +26,6 @@ #define FPUDEBUG 0 -/* Format of the floating-point exception registers. */ -struct exc_reg { - unsigned int exception : 6; - unsigned int ei : 26; -}; - /* Macros for grabbing bits of the instruction format from the 'ei' field above. */ /* Major opcode 0c and 0e */ @@ -103,9 +97,19 @@ handle_fpe(struct pt_regs *regs) memcpy(regs->fr, frcopy, sizeof regs->fr); if (signalcode != 0) { - force_sig_fault(signalcode >> 24, signalcode & 0xffffff, - (void __user *) regs->iaoq[0]); - return -1; + int sig = signalcode >> 24; + + if (sig == SIGFPE) { + /* + * Clear floating point trap bit to avoid trapping + * again on the first floating-point instruction in + * the userspace signal handler. + */ + regs->fr[0] &= ~(1ULL << 38); + } + force_sig_fault(sig, signalcode & 0xffffff, + (void __user *) regs->iaoq[0]); + return -1; } return signalcode ? -1 : 0; diff --git a/arch/parisc/math-emu/fcnvff.c b/arch/parisc/math-emu/fcnvff.c index 0530e6127797..61e489704c86 100644 --- a/arch/parisc/math-emu/fcnvff.c +++ b/arch/parisc/math-emu/fcnvff.c @@ -16,8 +16,8 @@ * Double Floating-point to Single Floating-point * * External Interfaces: - * dbl_to_sgl_fcnvff(srcptr,nullptr,dstptr,status) - * sgl_to_dbl_fcnvff(srcptr,nullptr,dstptr,status) + * dbl_to_sgl_fcnvff(srcptr,_nullptr,dstptr,status) + * sgl_to_dbl_fcnvff(srcptr,_nullptr,dstptr,status) * * Internal Interfaces: * @@ -40,7 +40,7 @@ int sgl_to_dbl_fcnvff( sgl_floating_point *srcptr, - unsigned int *nullptr, + unsigned int *_nullptr, dbl_floating_point *dstptr, unsigned int *status) { @@ -127,7 +127,7 @@ sgl_to_dbl_fcnvff( int dbl_to_sgl_fcnvff( dbl_floating_point *srcptr, - unsigned int *nullptr, + unsigned int *_nullptr, sgl_floating_point *dstptr, unsigned int *status) { diff --git a/arch/parisc/math-emu/fcnvfu.c b/arch/parisc/math-emu/fcnvfu.c index c971618a6f3c..c31790ceecca 100644 --- a/arch/parisc/math-emu/fcnvfu.c +++ b/arch/parisc/math-emu/fcnvfu.c @@ -15,10 +15,10 @@ * Floating-point to Unsigned Fixed-point Converts * * External Interfaces: - * dbl_to_dbl_fcnvfu(srcptr,nullptr,dstptr,status) - * dbl_to_sgl_fcnvfu(srcptr,nullptr,dstptr,status) - * sgl_to_dbl_fcnvfu(srcptr,nullptr,dstptr,status) - * sgl_to_sgl_fcnvfu(srcptr,nullptr,dstptr,status) + * dbl_to_dbl_fcnvfu(srcptr,_nullptr,dstptr,status) + * dbl_to_sgl_fcnvfu(srcptr,_nullptr,dstptr,status) + * sgl_to_dbl_fcnvfu(srcptr,_nullptr,dstptr,status) + * sgl_to_sgl_fcnvfu(srcptr,_nullptr,dstptr,status) * * Internal Interfaces: * @@ -45,7 +45,7 @@ int sgl_to_sgl_fcnvfu( sgl_floating_point *srcptr, - unsigned int *nullptr, + unsigned int *_nullptr, unsigned int *dstptr, unsigned int *status) { @@ -166,7 +166,7 @@ sgl_to_sgl_fcnvfu( int sgl_to_dbl_fcnvfu( sgl_floating_point *srcptr, - unsigned int *nullptr, + unsigned int *_nullptr, dbl_unsigned *dstptr, unsigned int *status) { @@ -285,7 +285,7 @@ sgl_to_dbl_fcnvfu( */ /*ARGSUSED*/ int -dbl_to_sgl_fcnvfu (dbl_floating_point * srcptr, unsigned int *nullptr, +dbl_to_sgl_fcnvfu (dbl_floating_point * srcptr, unsigned int *_nullptr, unsigned int *dstptr, unsigned int *status) { register unsigned int srcp1, srcp2, result; @@ -408,7 +408,7 @@ dbl_to_sgl_fcnvfu (dbl_floating_point * srcptr, unsigned int *nullptr, */ /*ARGSUSED*/ int -dbl_to_dbl_fcnvfu (dbl_floating_point * srcptr, unsigned int *nullptr, +dbl_to_dbl_fcnvfu (dbl_floating_point * srcptr, unsigned int *_nullptr, dbl_unsigned * dstptr, unsigned int *status) { register int src_exponent; diff --git a/arch/parisc/math-emu/fcnvfut.c b/arch/parisc/math-emu/fcnvfut.c index 5b657f852578..2cf1daf3b7ad 100644 --- a/arch/parisc/math-emu/fcnvfut.c +++ b/arch/parisc/math-emu/fcnvfut.c @@ -15,10 +15,10 @@ * Floating-point to Unsigned Fixed-point Converts with Truncation * * External Interfaces: - * dbl_to_dbl_fcnvfut(srcptr,nullptr,dstptr,status) - * dbl_to_sgl_fcnvfut(srcptr,nullptr,dstptr,status) - * sgl_to_dbl_fcnvfut(srcptr,nullptr,dstptr,status) - * sgl_to_sgl_fcnvfut(srcptr,nullptr,dstptr,status) + * dbl_to_dbl_fcnvfut(srcptr,_nullptr,dstptr,status) + * dbl_to_sgl_fcnvfut(srcptr,_nullptr,dstptr,status) + * sgl_to_dbl_fcnvfut(srcptr,_nullptr,dstptr,status) + * sgl_to_sgl_fcnvfut(srcptr,_nullptr,dstptr,status) * * Internal Interfaces: * @@ -44,7 +44,7 @@ */ /*ARGSUSED*/ int -sgl_to_sgl_fcnvfut (sgl_floating_point * srcptr, unsigned int *nullptr, +sgl_to_sgl_fcnvfut (sgl_floating_point * srcptr, unsigned int *_nullptr, unsigned int *dstptr, unsigned int *status) { register unsigned int src, result; @@ -113,7 +113,7 @@ sgl_to_sgl_fcnvfut (sgl_floating_point * srcptr, unsigned int *nullptr, */ /*ARGSUSED*/ int -sgl_to_dbl_fcnvfut (sgl_floating_point * srcptr, unsigned int *nullptr, +sgl_to_dbl_fcnvfut (sgl_floating_point * srcptr, unsigned int *_nullptr, dbl_unsigned * dstptr, unsigned int *status) { register int src_exponent; @@ -183,7 +183,7 @@ sgl_to_dbl_fcnvfut (sgl_floating_point * srcptr, unsigned int *nullptr, */ /*ARGSUSED*/ int -dbl_to_sgl_fcnvfut (dbl_floating_point * srcptr, unsigned int *nullptr, +dbl_to_sgl_fcnvfut (dbl_floating_point * srcptr, unsigned int *_nullptr, unsigned int *dstptr, unsigned int *status) { register unsigned int srcp1, srcp2, result; @@ -252,7 +252,7 @@ dbl_to_sgl_fcnvfut (dbl_floating_point * srcptr, unsigned int *nullptr, */ /*ARGSUSED*/ int -dbl_to_dbl_fcnvfut (dbl_floating_point * srcptr, unsigned int *nullptr, +dbl_to_dbl_fcnvfut (dbl_floating_point * srcptr, unsigned int *_nullptr, dbl_unsigned * dstptr, unsigned int *status) { register int src_exponent; diff --git a/arch/parisc/math-emu/fcnvfx.c b/arch/parisc/math-emu/fcnvfx.c index 5e153078d803..99bd61479452 100644 --- a/arch/parisc/math-emu/fcnvfx.c +++ b/arch/parisc/math-emu/fcnvfx.c @@ -18,10 +18,10 @@ * Double Floating-point to Double Fixed-point * * External Interfaces: - * dbl_to_dbl_fcnvfx(srcptr,nullptr,dstptr,status) - * dbl_to_sgl_fcnvfx(srcptr,nullptr,dstptr,status) - * sgl_to_dbl_fcnvfx(srcptr,nullptr,dstptr,status) - * sgl_to_sgl_fcnvfx(srcptr,nullptr,dstptr,status) + * dbl_to_dbl_fcnvfx(srcptr,_nullptr,dstptr,status) + * dbl_to_sgl_fcnvfx(srcptr,_nullptr,dstptr,status) + * sgl_to_dbl_fcnvfx(srcptr,_nullptr,dstptr,status) + * sgl_to_sgl_fcnvfx(srcptr,_nullptr,dstptr,status) * * Internal Interfaces: * @@ -44,7 +44,7 @@ int sgl_to_sgl_fcnvfx( sgl_floating_point *srcptr, - sgl_floating_point *nullptr, + sgl_floating_point *_nullptr, int *dstptr, sgl_floating_point *status) { @@ -141,7 +141,7 @@ sgl_to_sgl_fcnvfx( int sgl_to_dbl_fcnvfx( sgl_floating_point *srcptr, - unsigned int *nullptr, + unsigned int *_nullptr, dbl_integer *dstptr, unsigned int *status) { @@ -262,7 +262,7 @@ sgl_to_dbl_fcnvfx( int dbl_to_sgl_fcnvfx( dbl_floating_point *srcptr, - unsigned int *nullptr, + unsigned int *_nullptr, int *dstptr, unsigned int *status) { @@ -373,7 +373,7 @@ dbl_to_sgl_fcnvfx( int dbl_to_dbl_fcnvfx( dbl_floating_point *srcptr, - unsigned int *nullptr, + unsigned int *_nullptr, dbl_integer *dstptr, unsigned int *status) { diff --git a/arch/parisc/math-emu/fcnvfxt.c b/arch/parisc/math-emu/fcnvfxt.c index ebec31e40d01..3b7cc62257d0 100644 --- a/arch/parisc/math-emu/fcnvfxt.c +++ b/arch/parisc/math-emu/fcnvfxt.c @@ -18,10 +18,10 @@ * Double Floating-point to Double Fixed-point /w truncated result * * External Interfaces: - * dbl_to_dbl_fcnvfxt(srcptr,nullptr,dstptr,status) - * dbl_to_sgl_fcnvfxt(srcptr,nullptr,dstptr,status) - * sgl_to_dbl_fcnvfxt(srcptr,nullptr,dstptr,status) - * sgl_to_sgl_fcnvfxt(srcptr,nullptr,dstptr,status) + * dbl_to_dbl_fcnvfxt(srcptr,_nullptr,dstptr,status) + * dbl_to_sgl_fcnvfxt(srcptr,_nullptr,dstptr,status) + * sgl_to_dbl_fcnvfxt(srcptr,_nullptr,dstptr,status) + * sgl_to_sgl_fcnvfxt(srcptr,_nullptr,dstptr,status) * * Internal Interfaces: * @@ -45,7 +45,7 @@ int sgl_to_sgl_fcnvfxt( sgl_floating_point *srcptr, - unsigned int *nullptr, + unsigned int *_nullptr, int *dstptr, unsigned int *status) { @@ -109,7 +109,7 @@ sgl_to_sgl_fcnvfxt( int sgl_to_dbl_fcnvfxt( sgl_floating_point *srcptr, - unsigned int *nullptr, + unsigned int *_nullptr, dbl_integer *dstptr, unsigned int *status) { @@ -183,7 +183,7 @@ sgl_to_dbl_fcnvfxt( int dbl_to_sgl_fcnvfxt( dbl_floating_point *srcptr, - unsigned int *nullptr, + unsigned int *_nullptr, int *dstptr, unsigned int *status) { @@ -248,7 +248,7 @@ dbl_to_sgl_fcnvfxt( int dbl_to_dbl_fcnvfxt( dbl_floating_point *srcptr, - unsigned int *nullptr, + unsigned int *_nullptr, dbl_integer *dstptr, unsigned int *status) { diff --git a/arch/parisc/math-emu/fcnvuf.c b/arch/parisc/math-emu/fcnvuf.c index c54978a0ace1..c166feb57045 100644 --- a/arch/parisc/math-emu/fcnvuf.c +++ b/arch/parisc/math-emu/fcnvuf.c @@ -15,10 +15,10 @@ * Fixed point to Floating-point Converts * * External Interfaces: - * dbl_to_dbl_fcnvuf(srcptr,nullptr,dstptr,status) - * dbl_to_sgl_fcnvuf(srcptr,nullptr,dstptr,status) - * sgl_to_dbl_fcnvuf(srcptr,nullptr,dstptr,status) - * sgl_to_sgl_fcnvuf(srcptr,nullptr,dstptr,status) + * dbl_to_dbl_fcnvuf(srcptr,_nullptr,dstptr,status) + * dbl_to_sgl_fcnvuf(srcptr,_nullptr,dstptr,status) + * sgl_to_dbl_fcnvuf(srcptr,_nullptr,dstptr,status) + * sgl_to_sgl_fcnvuf(srcptr,_nullptr,dstptr,status) * * Internal Interfaces: * @@ -45,7 +45,7 @@ int sgl_to_sgl_fcnvuf( unsigned int *srcptr, - unsigned int *nullptr, + unsigned int *_nullptr, sgl_floating_point *dstptr, unsigned int *status) { @@ -104,7 +104,7 @@ sgl_to_sgl_fcnvuf( int sgl_to_dbl_fcnvuf( unsigned int *srcptr, - unsigned int *nullptr, + unsigned int *_nullptr, dbl_floating_point *dstptr, unsigned int *status) { @@ -145,7 +145,7 @@ sgl_to_dbl_fcnvuf( int dbl_to_sgl_fcnvuf( dbl_unsigned *srcptr, - unsigned int *nullptr, + unsigned int *_nullptr, sgl_floating_point *dstptr, unsigned int *status) { @@ -227,7 +227,7 @@ dbl_to_sgl_fcnvuf( int dbl_to_dbl_fcnvuf( dbl_unsigned *srcptr, - unsigned int *nullptr, + unsigned int *_nullptr, dbl_floating_point *dstptr, unsigned int *status) { diff --git a/arch/parisc/math-emu/fcnvxf.c b/arch/parisc/math-emu/fcnvxf.c index 69401797146b..11bc1e8a13aa 100644 --- a/arch/parisc/math-emu/fcnvxf.c +++ b/arch/parisc/math-emu/fcnvxf.c @@ -18,10 +18,10 @@ * Double Fixed-point to Double Floating-point * * External Interfaces: - * dbl_to_dbl_fcnvxf(srcptr,nullptr,dstptr,status) - * dbl_to_sgl_fcnvxf(srcptr,nullptr,dstptr,status) - * sgl_to_dbl_fcnvxf(srcptr,nullptr,dstptr,status) - * sgl_to_sgl_fcnvxf(srcptr,nullptr,dstptr,status) + * dbl_to_dbl_fcnvxf(srcptr,_nullptr,dstptr,status) + * dbl_to_sgl_fcnvxf(srcptr,_nullptr,dstptr,status) + * sgl_to_dbl_fcnvxf(srcptr,_nullptr,dstptr,status) + * sgl_to_sgl_fcnvxf(srcptr,_nullptr,dstptr,status) * * Internal Interfaces: * @@ -44,7 +44,7 @@ int sgl_to_sgl_fcnvxf( int *srcptr, - unsigned int *nullptr, + unsigned int *_nullptr, sgl_floating_point *dstptr, unsigned int *status) { @@ -115,7 +115,7 @@ sgl_to_sgl_fcnvxf( int sgl_to_dbl_fcnvxf( int *srcptr, - unsigned int *nullptr, + unsigned int *_nullptr, dbl_floating_point *dstptr, unsigned int *status) { @@ -166,7 +166,7 @@ sgl_to_dbl_fcnvxf( int dbl_to_sgl_fcnvxf( dbl_integer *srcptr, - unsigned int *nullptr, + unsigned int *_nullptr, sgl_floating_point *dstptr, unsigned int *status) { @@ -271,7 +271,7 @@ dbl_to_sgl_fcnvxf( int dbl_to_dbl_fcnvxf( dbl_integer *srcptr, - unsigned int *nullptr, + unsigned int *_nullptr, dbl_floating_point *dstptr, unsigned int *status) { diff --git a/arch/parisc/math-emu/frnd.c b/arch/parisc/math-emu/frnd.c index 0b0e8493e08a..825d89650c2d 100644 --- a/arch/parisc/math-emu/frnd.c +++ b/arch/parisc/math-emu/frnd.c @@ -14,8 +14,8 @@ * Quad Floating-point Round to Integer (returns unimplemented) * * External Interfaces: - * dbl_frnd(srcptr,nullptr,dstptr,status) - * sgl_frnd(srcptr,nullptr,dstptr,status) + * dbl_frnd(srcptr,_nullptr,dstptr,status) + * sgl_frnd(srcptr,_nullptr,dstptr,status) * * END_DESC */ @@ -33,7 +33,7 @@ /*ARGSUSED*/ int sgl_frnd(sgl_floating_point *srcptr, - unsigned int *nullptr, + unsigned int *_nullptr, sgl_floating_point *dstptr, unsigned int *status) { @@ -138,7 +138,7 @@ sgl_frnd(sgl_floating_point *srcptr, int dbl_frnd( dbl_floating_point *srcptr, - unsigned int *nullptr, + unsigned int *_nullptr, dbl_floating_point *dstptr, unsigned int *status) { diff --git a/arch/parisc/math-emu/sfsqrt.c b/arch/parisc/math-emu/sfsqrt.c index bd6a84f468d8..8e9e023e7b2e 100644 --- a/arch/parisc/math-emu/sfsqrt.c +++ b/arch/parisc/math-emu/sfsqrt.c @@ -15,7 +15,7 @@ * Single Floating-point Square Root * * External Interfaces: - * sgl_fsqrt(srcptr,nullptr,dstptr,status) + * sgl_fsqrt(srcptr,_nullptr,dstptr,status) * * Internal Interfaces: * @@ -37,7 +37,7 @@ unsigned int sgl_fsqrt( sgl_floating_point *srcptr, - unsigned int *nullptr, + unsigned int *_nullptr, sgl_floating_point *dstptr, unsigned int *status) { diff --git a/arch/parisc/mm/fault.c b/arch/parisc/mm/fault.c index 869204e97ec9..f1785640b049 100644 --- a/arch/parisc/mm/fault.c +++ b/arch/parisc/mm/fault.c @@ -150,11 +150,16 @@ int fixup_exception(struct pt_regs *regs) * Fix up get_user() and put_user(). * ASM_EXCEPTIONTABLE_ENTRY_EFAULT() sets the least-significant * bit in the relative address of the fixup routine to indicate - * that gr[ASM_EXCEPTIONTABLE_REG] should be loaded with - * -EFAULT to report a userspace access error. + * that the register encoded in the "or %r0,%r0,register" + * opcode should be loaded with -EFAULT to report a userspace + * access error. */ if (fix->fixup & 1) { - regs->gr[ASM_EXCEPTIONTABLE_REG] = -EFAULT; + int fault_error_reg = fix->err_opcode & 0x1f; + if (!WARN_ON(!fault_error_reg)) + regs->gr[fault_error_reg] = -EFAULT; + pr_debug("Unalignment fixup of register %d at %pS\n", + fault_error_reg, (void*)regs->iaoq[0]); /* zero target register for get_user() */ if (parisc_acctyp(0, regs->iir) == VM_READ) { @@ -192,31 +197,31 @@ int fixup_exception(struct pt_regs *regs) * For implementation see handle_interruption() in traps.c */ static const char * const trap_description[] = { - [1] "High-priority machine check (HPMC)", - [2] "Power failure interrupt", - [3] "Recovery counter trap", - [5] "Low-priority machine check", - [6] "Instruction TLB miss fault", - [7] "Instruction access rights / protection trap", - [8] "Illegal instruction trap", - [9] "Break instruction trap", - [10] "Privileged operation trap", - [11] "Privileged register trap", - [12] "Overflow trap", - [13] "Conditional trap", - [14] "FP Assist Exception trap", - [15] "Data TLB miss fault", - [16] "Non-access ITLB miss fault", - [17] "Non-access DTLB miss fault", - [18] "Data memory protection/unaligned access trap", - [19] "Data memory break trap", - [20] "TLB dirty bit trap", - [21] "Page reference trap", - [22] "Assist emulation trap", - [25] "Taken branch trap", - [26] "Data memory access rights trap", - [27] "Data memory protection ID trap", - [28] "Unaligned data reference trap", + [1] = "High-priority machine check (HPMC)", + [2] = "Power failure interrupt", + [3] = "Recovery counter trap", + [5] = "Low-priority machine check", + [6] = "Instruction TLB miss fault", + [7] = "Instruction access rights / protection trap", + [8] = "Illegal instruction trap", + [9] = "Break instruction trap", + [10] = "Privileged operation trap", + [11] = "Privileged register trap", + [12] = "Overflow trap", + [13] = "Conditional trap", + [14] = "FP Assist Exception trap", + [15] = "Data TLB miss fault", + [16] = "Non-access ITLB miss fault", + [17] = "Non-access DTLB miss fault", + [18] = "Data memory protection/unaligned access trap", + [19] = "Data memory break trap", + [20] = "TLB dirty bit trap", + [21] = "Page reference trap", + [22] = "Assist emulation trap", + [25] = "Taken branch trap", + [26] = "Data memory access rights trap", + [27] = "Data memory protection ID trap", + [28] = "Unaligned data reference trap", }; const char *trap_name(unsigned long code) @@ -288,15 +293,19 @@ void do_page_fault(struct pt_regs *regs, unsigned long code, retry: mmap_read_lock(mm); vma = find_vma_prev(mm, address, &prev_vma); - if (!vma || address < vma->vm_start) - goto check_expansion; + if (!vma || address < vma->vm_start) { + if (!prev_vma || !(prev_vma->vm_flags & VM_GROWSUP)) + goto bad_area; + vma = expand_stack(mm, address); + if (!vma) + goto bad_area_nosemaphore; + } + /* * Ok, we have a good vm_area for this memory access. We still need to * check the access permissions. */ -good_area: - if ((vma->vm_flags & acc_type) != acc_type) goto bad_area; @@ -308,8 +317,13 @@ good_area: fault = handle_mm_fault(vma, address, flags, regs); - if (fault_signal_pending(fault, regs)) + if (fault_signal_pending(fault, regs)) { + if (!user_mode(regs)) { + msg = "Page fault: fault signal on kernel memory"; + goto no_context; + } return; + } /* The fault is fully completed (including releasing mmap lock) */ if (fault & VM_FAULT_COMPLETED) @@ -342,17 +356,17 @@ good_area: mmap_read_unlock(mm); return; -check_expansion: - vma = prev_vma; - if (vma && (expand_stack(vma, address) == 0)) - goto good_area; - /* * Something tried to access memory that isn't in our memory map.. */ bad_area: mmap_read_unlock(mm); +bad_area_nosemaphore: + if (!user_mode(regs) && fixup_exception(regs)) { + return; + } + if (user_mode(regs)) { int signo, si_code; @@ -444,7 +458,7 @@ handle_nadtlb_fault(struct pt_regs *regs) { unsigned long insn = regs->iir; int breg, treg, xreg, val = 0; - struct vm_area_struct *vma, *prev_vma; + struct vm_area_struct *vma; struct task_struct *tsk; struct mm_struct *mm; unsigned long address; @@ -480,7 +494,7 @@ handle_nadtlb_fault(struct pt_regs *regs) /* Search for VMA */ address = regs->ior; mmap_read_lock(mm); - vma = find_vma_prev(mm, address, &prev_vma); + vma = vma_lookup(mm, address); mmap_read_unlock(mm); /* @@ -489,7 +503,6 @@ handle_nadtlb_fault(struct pt_regs *regs) */ acc_type = (insn & 0x40) ? VM_WRITE : VM_READ; if (vma - && address >= vma->vm_start && (vma->vm_flags & acc_type) == acc_type) val = 1; } diff --git a/arch/parisc/mm/fixmap.c b/arch/parisc/mm/fixmap.c index cc15d737fda6..ae3493dae9dc 100644 --- a/arch/parisc/mm/fixmap.c +++ b/arch/parisc/mm/fixmap.c @@ -19,9 +19,6 @@ void notrace set_fixmap(enum fixed_addresses idx, phys_addr_t phys) pmd_t *pmd = pmd_offset(pud, vaddr); pte_t *pte; - if (pmd_none(*pmd)) - pte = pte_alloc_kernel(pmd, vaddr); - pte = pte_offset_kernel(pmd, vaddr); set_pte_at(&init_mm, vaddr, pte, __mk_pte(phys, PAGE_KERNEL_RWX)); flush_tlb_kernel_range(vaddr, vaddr + PAGE_SIZE); diff --git a/arch/parisc/mm/hugetlbpage.c b/arch/parisc/mm/hugetlbpage.c index d1d3990b83f6..a94fe546d434 100644 --- a/arch/parisc/mm/hugetlbpage.c +++ b/arch/parisc/mm/hugetlbpage.c @@ -21,27 +21,6 @@ #include <asm/mmu_context.h> -unsigned long -hugetlb_get_unmapped_area(struct file *file, unsigned long addr, - unsigned long len, unsigned long pgoff, unsigned long flags) -{ - struct hstate *h = hstate_file(file); - - if (len & ~huge_page_mask(h)) - return -EINVAL; - if (len > TASK_SIZE) - return -ENOMEM; - - if (flags & MAP_FIXED) - if (prepare_hugepage_range(file, addr, len)) - return -EINVAL; - - if (addr) - addr = ALIGN(addr, huge_page_size(h)); - - /* we need to make sure the colouring is OK */ - return arch_get_unmapped_area(file, addr, len, pgoff, flags); -} pte_t *huge_pte_alloc(struct mm_struct *mm, struct vm_area_struct *vma, @@ -66,7 +45,7 @@ pte_t *huge_pte_alloc(struct mm_struct *mm, struct vm_area_struct *vma, if (pud) { pmd = pmd_alloc(mm, pud, addr); if (pmd) - pte = pte_alloc_map(mm, pmd, addr); + pte = pte_alloc_huge(mm, pmd, addr); } return pte; } @@ -90,7 +69,7 @@ pte_t *huge_pte_offset(struct mm_struct *mm, if (!pud_none(*pud)) { pmd = pmd_offset(pud, addr); if (!pmd_none(*pmd)) - pte = pte_offset_map(pmd, addr); + pte = pte_offset_huge(pmd, addr); } } } @@ -140,14 +119,14 @@ static void __set_huge_pte_at(struct mm_struct *mm, unsigned long addr, } void set_huge_pte_at(struct mm_struct *mm, unsigned long addr, - pte_t *ptep, pte_t entry) + pte_t *ptep, pte_t entry, unsigned long sz) { __set_huge_pte_at(mm, addr, ptep, entry); } pte_t huge_ptep_get_and_clear(struct mm_struct *mm, unsigned long addr, - pte_t *ptep) + pte_t *ptep, unsigned long sz) { pte_t entry; @@ -180,14 +159,3 @@ int huge_ptep_set_access_flags(struct vm_area_struct *vma, } return changed; } - - -int pmd_huge(pmd_t pmd) -{ - return 0; -} - -int pud_huge(pud_t pud) -{ - return 0; -} diff --git a/arch/parisc/mm/init.c b/arch/parisc/mm/init.c index b0c43f3b0a5f..14270715d754 100644 --- a/arch/parisc/mm/init.c +++ b/arch/parisc/mm/init.c @@ -24,6 +24,7 @@ #include <linux/nodemask.h> /* for node_online_map */ #include <linux/pagemap.h> /* for release_pages */ #include <linux/compat.h> +#include <linux/execmem.h> #include <asm/pgalloc.h> #include <asm/tlb.h> @@ -32,6 +33,8 @@ #include <asm/sections.h> #include <asm/msgbuf.h> #include <asm/sparsemem.h> +#include <asm/asm-offsets.h> +#include <asm/shmbuf.h> extern int data_start; extern void parisc_kernel_start(void); /* Kernel entry point in head.S */ @@ -374,10 +377,8 @@ static void __ref map_pages(unsigned long start_vaddr, #if CONFIG_PGTABLE_LEVELS == 3 if (pud_none(*pud)) { - pmd = memblock_alloc(PAGE_SIZE << PMD_TABLE_ORDER, + pmd = memblock_alloc_or_panic(PAGE_SIZE << PMD_TABLE_ORDER, PAGE_SIZE << PMD_TABLE_ORDER); - if (!pmd) - panic("pmd allocation failed.\n"); pud_populate(NULL, pud, pmd); } #endif @@ -385,9 +386,7 @@ static void __ref map_pages(unsigned long start_vaddr, pmd = pmd_offset(pud, vaddr); for (tmp1 = start_pmd; tmp1 < PTRS_PER_PMD; tmp1++, pmd++) { if (pmd_none(*pmd)) { - pg_table = memblock_alloc(PAGE_SIZE, PAGE_SIZE); - if (!pg_table) - panic("page table allocation failed\n"); + pg_table = memblock_alloc_or_panic(PAGE_SIZE, PAGE_SIZE); pmd_populate_kernel(NULL, pmd, pg_table); } @@ -456,7 +455,6 @@ void free_initmem(void) unsigned long kernel_end = (unsigned long)&_end; /* Remap kernel text and data, but do not touch init section yet. */ - kernel_set_to_readonly = true; map_pages(init_end, __pa(init_end), kernel_end - init_end, PAGE_KERNEL, 0); @@ -479,7 +477,7 @@ void free_initmem(void) /* finally dump all the instructions which were cached, since the * pages are no-longer executable */ flush_icache_range(init_begin, init_end); - + free_initmem_default(POISON_FREE_INITMEM); /* set up a new led state on systems shipped LED State panel */ @@ -490,11 +488,18 @@ void free_initmem(void) #ifdef CONFIG_STRICT_KERNEL_RWX void mark_rodata_ro(void) { - /* rodata memory was already mapped with KERNEL_RO access rights by - pagetable_init() and map_pages(). No need to do additional stuff here */ - unsigned long roai_size = __end_ro_after_init - __start_ro_after_init; + unsigned long start = (unsigned long) &__start_rodata; + unsigned long end = (unsigned long) &__end_rodata; + + pr_info("Write protecting the kernel read-only data: %luk\n", + (end - start) >> 10); + + kernel_set_to_readonly = true; + map_pages(start, __pa(start), end - start, PAGE_KERNEL, 0); - pr_info("Write protected read-only-after-init data: %luk\n", roai_size >> 10); + /* force the kernel to see the new page table entries */ + flush_cache_all(); + flush_tlb_all(); } #endif @@ -523,10 +528,6 @@ void mark_rodata_ro(void) void *parisc_vmalloc_start __ro_after_init; EXPORT_SYMBOL(parisc_vmalloc_start); -#ifdef CONFIG_PA11 -unsigned long pcxl_dma_start __ro_after_init; -#endif - void __init mem_init(void) { /* Do sanity checks on IPC (compat) structures */ @@ -561,10 +562,6 @@ void __init mem_init(void) BUILD_BUG_ON(TMPALIAS_MAP_START >= 0x80000000); #endif - high_memory = __va((max_pfn << PAGE_SHIFT)); - set_max_mapnr(max_low_pfn); - memblock_free_all(); - #ifdef CONFIG_PA11 if (boot_cpu_data.cpu_type == pcxl2 || boot_cpu_data.cpu_type == pcxl) { pcxl_dma_start = (unsigned long)SET_MAP_OFFSET(MAP_START); @@ -626,12 +623,10 @@ static void __init pagetable_init(void) for (range = 0; range < npmem_ranges; range++) { unsigned long start_paddr; - unsigned long end_paddr; unsigned long size; start_paddr = pmem_ranges[range].start_pfn << PAGE_SHIFT; size = pmem_ranges[range].pages << PAGE_SHIFT; - end_paddr = start_paddr + size; map_pages((unsigned long)__va(start_paddr), start_paddr, size, PAGE_KERNEL, 0); @@ -645,9 +640,7 @@ static void __init pagetable_init(void) } #endif - empty_zero_page = memblock_alloc(PAGE_SIZE, PAGE_SIZE); - if (!empty_zero_page) - panic("zero page allocation failed.\n"); + empty_zero_page = memblock_alloc_or_panic(PAGE_SIZE, PAGE_SIZE); } @@ -671,6 +664,35 @@ static void __init gateway_init(void) PAGE_SIZE, PAGE_GATEWAY, 1); } +static void __init fixmap_init(void) +{ + unsigned long addr = FIXMAP_START; + unsigned long end = FIXMAP_START + FIXMAP_SIZE; + pgd_t *pgd = pgd_offset_k(addr); + p4d_t *p4d = p4d_offset(pgd, addr); + pud_t *pud = pud_offset(p4d, addr); + pmd_t *pmd; + + BUILD_BUG_ON(FIXMAP_SIZE > PMD_SIZE); + +#if CONFIG_PGTABLE_LEVELS == 3 + if (pud_none(*pud)) { + pmd = memblock_alloc_or_panic(PAGE_SIZE << PMD_TABLE_ORDER, + PAGE_SIZE << PMD_TABLE_ORDER); + pud_populate(NULL, pud, pmd); + } +#endif + + pmd = pmd_offset(pud, addr); + do { + pte_t *pte = memblock_alloc_or_panic(PAGE_SIZE, PAGE_SIZE); + + pmd_populate_kernel(&init_mm, pmd, pte); + + addr += PAGE_SIZE; + } while (addr < end); +} + static void __init parisc_bootmem_free(void) { unsigned long max_zone_pfn[MAX_NR_ZONES] = { 0, }; @@ -685,6 +707,7 @@ void __init paging_init(void) setup_bootmem(); pagetable_init(); gateway_init(); + fixmap_init(); flush_cache_all_local(); /* start with known state */ flush_tlb_all_local(NULL); @@ -692,6 +715,77 @@ void __init paging_init(void) parisc_bootmem_free(); } +static void alloc_btlb(unsigned long start, unsigned long end, int *slot, + unsigned long entry_info) +{ + const int slot_max = btlb_info.fixed_range_info.num_comb; + int min_num_pages = btlb_info.min_size; + unsigned long size; + + /* map at minimum 4 pages */ + if (min_num_pages < 4) + min_num_pages = 4; + + size = HUGEPAGE_SIZE; + while (start < end && *slot < slot_max && size >= PAGE_SIZE) { + /* starting address must have same alignment as size! */ + /* if correctly aligned and fits in double size, increase */ + if (((start & (2 * size - 1)) == 0) && + (end - start) >= (2 * size)) { + size <<= 1; + continue; + } + /* if current size alignment is too big, try smaller size */ + if ((start & (size - 1)) != 0) { + size >>= 1; + continue; + } + if ((end - start) >= size) { + if ((size >> PAGE_SHIFT) >= min_num_pages) + pdc_btlb_insert(start >> PAGE_SHIFT, __pa(start) >> PAGE_SHIFT, + size >> PAGE_SHIFT, entry_info, *slot); + (*slot)++; + start += size; + continue; + } + size /= 2; + continue; + } +} + +void btlb_init_per_cpu(void) +{ + unsigned long s, t, e; + int slot; + + /* BTLBs are not available on 64-bit CPUs */ + if (IS_ENABLED(CONFIG_PA20)) + return; + else if (pdc_btlb_info(&btlb_info) < 0) { + memset(&btlb_info, 0, sizeof btlb_info); + } + + /* insert BLTLBs for code and data segments */ + s = (uintptr_t) dereference_function_descriptor(&_stext); + e = (uintptr_t) dereference_function_descriptor(&_etext); + t = (uintptr_t) dereference_function_descriptor(&_sdata); + BUG_ON(t != e); + + /* code segments */ + slot = 0; + alloc_btlb(s, e, &slot, 0x13800000); + + /* sanity check */ + t = (uintptr_t) dereference_function_descriptor(&_edata); + e = (uintptr_t) dereference_function_descriptor(&__bss_start); + BUG_ON(t != e); + + /* data segments */ + s = (uintptr_t) dereference_function_descriptor(&_sdata); + e = (uintptr_t) dereference_function_descriptor(&__bss_stop); + alloc_btlb(s, e, &slot, 0x11800000); +} + #ifdef CONFIG_PA20 /* @@ -891,3 +985,23 @@ static const pgprot_t protection_map[16] = { [VM_SHARED | VM_EXEC | VM_WRITE | VM_READ] = PAGE_RWX }; DECLARE_VM_GET_PAGE_PROT + +#ifdef CONFIG_EXECMEM +static struct execmem_info execmem_info __ro_after_init; + +struct execmem_info __init *execmem_arch_setup(void) +{ + execmem_info = (struct execmem_info){ + .ranges = { + [EXECMEM_DEFAULT] = { + .start = VMALLOC_START, + .end = VMALLOC_END, + .pgprot = PAGE_KERNEL_RWX, + .alignment = 1, + }, + }, + }; + + return &execmem_info; +} +#endif /* CONFIG_EXECMEM */ diff --git a/arch/parisc/mm/ioremap.c b/arch/parisc/mm/ioremap.c index 345ff0b66499..0b65c4b3baee 100644 --- a/arch/parisc/mm/ioremap.c +++ b/arch/parisc/mm/ioremap.c @@ -13,25 +13,9 @@ #include <linux/io.h> #include <linux/mm.h> -/* - * Generic mapping function (not visible outside): - */ - -/* - * Remap an arbitrary physical address space into the kernel virtual - * address space. - * - * NOTE! We need to allow non-page-aligned mappings too: we will obviously - * have to convert them into an offset in a page-aligned mapping, but the - * caller shouldn't need to know that small detail. - */ -void __iomem *ioremap(unsigned long phys_addr, unsigned long size) +void __iomem *ioremap_prot(phys_addr_t phys_addr, size_t size, + pgprot_t prot) { - void __iomem *addr; - struct vm_struct *area; - unsigned long offset, last_addr; - pgprot_t pgprot; - #ifdef CONFIG_EISA unsigned long end = phys_addr + size - 1; /* Support EISA addresses */ @@ -40,11 +24,6 @@ void __iomem *ioremap(unsigned long phys_addr, unsigned long size) phys_addr |= F_EXTEND(0xfc000000); #endif - /* Don't allow wraparound or zero size */ - last_addr = phys_addr + size - 1; - if (!size || last_addr < phys_addr) - return NULL; - /* * Don't allow anybody to remap normal RAM that we're using.. */ @@ -62,39 +41,6 @@ void __iomem *ioremap(unsigned long phys_addr, unsigned long size) } } - pgprot = __pgprot(_PAGE_PRESENT | _PAGE_RW | _PAGE_DIRTY | - _PAGE_ACCESSED | _PAGE_NO_CACHE); - - /* - * Mappings have to be page-aligned - */ - offset = phys_addr & ~PAGE_MASK; - phys_addr &= PAGE_MASK; - size = PAGE_ALIGN(last_addr + 1) - phys_addr; - - /* - * Ok, go for it.. - */ - area = get_vm_area(size, VM_IOREMAP); - if (!area) - return NULL; - - addr = (void __iomem *) area->addr; - if (ioremap_page_range((unsigned long)addr, (unsigned long)addr + size, - phys_addr, pgprot)) { - vunmap(addr); - return NULL; - } - - return (void __iomem *) (offset + (char __iomem *)addr); -} -EXPORT_SYMBOL(ioremap); - -void iounmap(const volatile void __iomem *io_addr) -{ - unsigned long addr = (unsigned long)io_addr & PAGE_MASK; - - if (is_vmalloc_addr((void *)addr)) - vunmap((void *)addr); + return generic_ioremap_prot(phys_addr, size, prot); } -EXPORT_SYMBOL(iounmap); +EXPORT_SYMBOL(ioremap_prot); diff --git a/arch/parisc/net/Makefile b/arch/parisc/net/Makefile new file mode 100644 index 000000000000..22b12024d4c3 --- /dev/null +++ b/arch/parisc/net/Makefile @@ -0,0 +1,9 @@ +# SPDX-License-Identifier: GPL-2.0-only + +obj-$(CONFIG_BPF_JIT) += bpf_jit_core.o + +ifeq ($(CONFIG_64BIT),y) + obj-$(CONFIG_BPF_JIT) += bpf_jit_comp64.o +else + obj-$(CONFIG_BPF_JIT) += bpf_jit_comp32.o +endif diff --git a/arch/parisc/net/bpf_jit.h b/arch/parisc/net/bpf_jit.h new file mode 100644 index 000000000000..8b8896959f04 --- /dev/null +++ b/arch/parisc/net/bpf_jit.h @@ -0,0 +1,479 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * Common functionality for PARISC32 and PARISC64 BPF JIT compilers + * + * Copyright (c) 2023 Helge Deller <deller@gmx.de> + * + */ + +#ifndef _BPF_JIT_H +#define _BPF_JIT_H + +#include <linux/bpf.h> +#include <linux/filter.h> +#include <asm/cacheflush.h> + +#define HPPA_JIT_DEBUG 0 +#define HPPA_JIT_REBOOT 0 +#define HPPA_JIT_DUMP 0 + +#define OPTIMIZE_HPPA 1 /* enable some asm optimizations */ +// echo 1 > /proc/sys/net/core/bpf_jit_enable + +#define HPPA_R(nr) nr /* use HPPA register #nr */ + +enum { + HPPA_REG_ZERO = 0, /* The constant value 0 */ + HPPA_REG_R1 = 1, /* used for addil */ + HPPA_REG_RP = 2, /* Return address */ + + HPPA_REG_ARG7 = 19, /* ARG4-7 used in 64-bit ABI */ + HPPA_REG_ARG6 = 20, + HPPA_REG_ARG5 = 21, + HPPA_REG_ARG4 = 22, + + HPPA_REG_ARG3 = 23, /* ARG0-3 in 32- and 64-bit ABI */ + HPPA_REG_ARG2 = 24, + HPPA_REG_ARG1 = 25, + HPPA_REG_ARG0 = 26, + HPPA_REG_GP = 27, /* Global pointer */ + HPPA_REG_RET0 = 28, /* Return value, HI in 32-bit */ + HPPA_REG_RET1 = 29, /* Return value, LOW in 32-bit */ + HPPA_REG_SP = 30, /* Stack pointer */ + HPPA_REG_R31 = 31, + +#ifdef CONFIG_64BIT + HPPA_REG_TCC = 3, + HPPA_REG_TCC_SAVED = 4, + HPPA_REG_TCC_IN_INIT = HPPA_REG_R31, +#else + HPPA_REG_TCC = 18, + HPPA_REG_TCC_SAVED = 17, + HPPA_REG_TCC_IN_INIT = HPPA_REG_R31, +#endif + + HPPA_REG_T0 = HPPA_REG_R1, /* Temporaries */ + HPPA_REG_T1 = HPPA_REG_R31, + HPPA_REG_T2 = HPPA_REG_ARG4, +#ifndef CONFIG_64BIT + HPPA_REG_T3 = HPPA_REG_ARG5, /* not used in 64-bit */ + HPPA_REG_T4 = HPPA_REG_ARG6, + HPPA_REG_T5 = HPPA_REG_ARG7, +#endif +}; + +struct hppa_jit_context { + struct bpf_prog *prog; + u32 *insns; /* HPPA insns */ + int ninsns; + int reg_seen_collect; + int reg_seen; + int body_len; + int epilogue_offset; + int prologue_len; + int *offset; /* BPF to HPPA */ +}; + +#define REG_SET_SEEN(ctx, nr) { if (ctx->reg_seen_collect) ctx->reg_seen |= BIT(nr); } +#define REG_SET_SEEN_ALL(ctx) { if (ctx->reg_seen_collect) ctx->reg_seen = -1; } +#define REG_FORCE_SEEN(ctx, nr) { ctx->reg_seen |= BIT(nr); } +#define REG_WAS_SEEN(ctx, nr) (ctx->reg_seen & BIT(nr)) +#define REG_ALL_SEEN(ctx) (ctx->reg_seen == -1) + +#define HPPA_INSN_SIZE 4 /* bytes per HPPA asm instruction */ +#define REG_SIZE REG_SZ /* bytes per native "long" word */ + +/* subtract hppa displacement on branches which is .+8 */ +#define HPPA_BRANCH_DISPLACEMENT 2 /* instructions */ + +/* asm statement indicator to execute delay slot */ +#define EXEC_NEXT_INSTR 0 +#define NOP_NEXT_INSTR 1 + +#define im11(val) (((u32)(val)) & 0x07ff) + +#define hppa_ldil(addr, reg) \ + hppa_t5_insn(0x08, reg, ((u32)(addr)) >> 11) /* ldil im21,reg */ +#define hppa_addil(addr, reg) \ + hppa_t5_insn(0x0a, reg, ((u32)(addr)) >> 11) /* addil im21,reg -> result in gr1 */ +#define hppa_ldo(im14, reg, target) \ + hppa_t1_insn(0x0d, reg, target, im14) /* ldo val14(reg),target */ +#define hppa_ldi(im14, reg) \ + hppa_ldo(im14, HPPA_REG_ZERO, reg) /* ldi val14,reg */ +#define hppa_or(reg1, reg2, target) \ + hppa_t6_insn(0x02, reg2, reg1, 0, 0, 0x09, target) /* or reg1,reg2,target */ +#define hppa_or_cond(reg1, reg2, cond, f, target) \ + hppa_t6_insn(0x02, reg2, reg1, cond, f, 0x09, target) +#define hppa_and(reg1, reg2, target) \ + hppa_t6_insn(0x02, reg2, reg1, 0, 0, 0x08, target) /* and reg1,reg2,target */ +#define hppa_and_cond(reg1, reg2, cond, f, target) \ + hppa_t6_insn(0x02, reg2, reg1, cond, f, 0x08, target) +#define hppa_xor(reg1, reg2, target) \ + hppa_t6_insn(0x02, reg2, reg1, 0, 0, 0x0a, target) /* xor reg1,reg2,target */ +#define hppa_add(reg1, reg2, target) \ + hppa_t6_insn(0x02, reg2, reg1, 0, 0, 0x18, target) /* add reg1,reg2,target */ +#define hppa_addc(reg1, reg2, target) \ + hppa_t6_insn(0x02, reg2, reg1, 0, 0, 0x1c, target) /* add,c reg1,reg2,target */ +#define hppa_sub(reg1, reg2, target) \ + hppa_t6_insn(0x02, reg2, reg1, 0, 0, 0x10, target) /* sub reg1,reg2,target */ +#define hppa_subb(reg1, reg2, target) \ + hppa_t6_insn(0x02, reg2, reg1, 0, 0, 0x14, target) /* sub,b reg1,reg2,target */ +#define hppa_nop() \ + hppa_or(0,0,0) /* nop: or 0,0,0 */ +#define hppa_addi(val11, reg, target) \ + hppa_t7_insn(0x2d, reg, target, val11) /* addi im11,reg,target */ +#define hppa_subi(val11, reg, target) \ + hppa_t7_insn(0x25, reg, target, val11) /* subi im11,reg,target */ +#define hppa_copy(reg, target) \ + hppa_or(reg, HPPA_REG_ZERO, target) /* copy reg,target */ +#define hppa_ldw(val14, reg, target) \ + hppa_t1_insn(0x12, reg, target, val14) /* ldw im14(reg),target */ +#define hppa_ldb(val14, reg, target) \ + hppa_t1_insn(0x10, reg, target, val14) /* ldb im14(reg),target */ +#define hppa_ldh(val14, reg, target) \ + hppa_t1_insn(0x11, reg, target, val14) /* ldh im14(reg),target */ +#define hppa_stw(reg, val14, base) \ + hppa_t1_insn(0x1a, base, reg, val14) /* stw reg,im14(base) */ +#define hppa_stb(reg, val14, base) \ + hppa_t1_insn(0x18, base, reg, val14) /* stb reg,im14(base) */ +#define hppa_sth(reg, val14, base) \ + hppa_t1_insn(0x19, base, reg, val14) /* sth reg,im14(base) */ +#define hppa_stwma(reg, val14, base) \ + hppa_t1_insn(0x1b, base, reg, val14) /* stw,ma reg,im14(base) */ +#define hppa_bv(reg, base, nop) \ + hppa_t11_insn(0x3a, base, reg, 0x06, 0, nop) /* bv(,n) reg(base) */ +#define hppa_be(offset, base) \ + hppa_t12_insn(0x38, base, offset, 0x00, 1) /* be,n offset(0,base) */ +#define hppa_be_l(offset, base, nop) \ + hppa_t12_insn(0x39, base, offset, 0x00, nop) /* ble(,nop) offset(0,base) */ +#define hppa_mtctl(reg, cr) \ + hppa_t21_insn(0x00, cr, reg, 0xc2, 0) /* mtctl reg,cr */ +#define hppa_mtsar(reg) \ + hppa_mtctl(reg, 11) /* mtsar reg */ +#define hppa_zdep(r, p, len, target) \ + hppa_t10_insn(0x35, target, r, 0, 2, p, len) /* zdep r,a,b,t */ +#define hppa_shl(r, len, target) \ + hppa_zdep(r, len, len, lo(rd)) +#define hppa_depwz(r, p, len, target) \ + hppa_t10_insn(0x35, target, r, 0, 3, 31-(p), 32-(len)) /* depw,z r,p,len,ret1 */ +#define hppa_depwz_sar(reg, target) \ + hppa_t1_insn(0x35, target, reg, 0) /* depw,z reg,sar,32,target */ +#define hppa_shrpw_sar(reg, target) \ + hppa_t10_insn(0x34, reg, 0, 0, 0, 0, target) /* shrpw r0,reg,sar,target */ +#define hppa_shrpw(r1, r2, p, target) \ + hppa_t10_insn(0x34, r2, r1, 0, 2, 31-(p), target) /* shrpw r1,r2,p,target */ +#define hppa_shd(r1, r2, p, target) \ + hppa_t10_insn(0x34, r2, r1, 0, 2, 31-(p), target) /* shrpw r1,r2,p,tarfer */ +#define hppa_extrws_sar(reg, target) \ + hppa_t10_insn(0x34, reg, target, 0, 5, 0, 0) /* extrw,s reg,sar,32,ret0 */ +#define hppa_extrws(reg, p, len, target) \ + hppa_t10_insn(0x34, reg, target, 0, 7, p, len) /* extrw,s reg,p,len,target */ +#define hppa_extru(r, p, len, target) \ + hppa_t10_insn(0x34, r, target, 0, 6, p, 32-(len)) +#define hppa_shr(r, len, target) \ + hppa_extru(r, 31-(len), 32-(len), target) +#define hppa_bl(imm17, rp) \ + hppa_t12_insn(0x3a, rp, imm17, 0x00, 1) /* bl,n target_addr,rp */ +#define hppa_sh2add(r1, r2, target) \ + hppa_t6_insn(0x02, r2, r1, 0, 0, 0x1a, target) /* sh2add r1,r2,target */ + +#define hppa_combt(r1, r2, target_addr, condition, nop) \ + hppa_t11_insn(IS_ENABLED(CONFIG_64BIT) ? 0x27 : 0x20, \ + r2, r1, condition, target_addr, nop) /* combt,cond,n r1,r2,addr */ +#define hppa_beq(r1, r2, target_addr) \ + hppa_combt(r1, r2, target_addr, 1, NOP_NEXT_INSTR) +#define hppa_blt(r1, r2, target_addr) \ + hppa_combt(r1, r2, target_addr, 2, NOP_NEXT_INSTR) +#define hppa_ble(r1, r2, target_addr) \ + hppa_combt(r1, r2, target_addr, 3, NOP_NEXT_INSTR) +#define hppa_bltu(r1, r2, target_addr) \ + hppa_combt(r1, r2, target_addr, 4, NOP_NEXT_INSTR) +#define hppa_bleu(r1, r2, target_addr) \ + hppa_combt(r1, r2, target_addr, 5, NOP_NEXT_INSTR) + +#define hppa_combf(r1, r2, target_addr, condition, nop) \ + hppa_t11_insn(IS_ENABLED(CONFIG_64BIT) ? 0x2f : 0x22, \ + r2, r1, condition, target_addr, nop) /* combf,cond,n r1,r2,addr */ +#define hppa_bne(r1, r2, target_addr) \ + hppa_combf(r1, r2, target_addr, 1, NOP_NEXT_INSTR) +#define hppa_bge(r1, r2, target_addr) \ + hppa_combf(r1, r2, target_addr, 2, NOP_NEXT_INSTR) +#define hppa_bgt(r1, r2, target_addr) \ + hppa_combf(r1, r2, target_addr, 3, NOP_NEXT_INSTR) +#define hppa_bgeu(r1, r2, target_addr) \ + hppa_combf(r1, r2, target_addr, 4, NOP_NEXT_INSTR) +#define hppa_bgtu(r1, r2, target_addr) \ + hppa_combf(r1, r2, target_addr, 5, NOP_NEXT_INSTR) + +/* 64-bit instructions */ +#ifdef CONFIG_64BIT +#define hppa64_ldd_reg(reg, b, target) \ + hppa_t10_insn(0x03, b, reg, 0, 0, 3<<1, target) +#define hppa64_ldd_im5(im5, b, target) \ + hppa_t10_insn(0x03, b, low_sign_unext(im5,5), 0, 1<<2, 3<<1, target) +#define hppa64_ldd_im16(im16, b, target) \ + hppa_t10_insn(0x14, b, target, 0, 0, 0, 0) | re_assemble_16(im16) +#define hppa64_std_im5(src, im5, b) \ + hppa_t10_insn(0x03, b, src, 0, 1<<2, 0xB<<1, low_sign_unext(im5,5)) +#define hppa64_std_im16(src, im16, b) \ + hppa_t10_insn(0x1c, b, src, 0, 0, 0, 0) | re_assemble_16(im16) +#define hppa64_bl_long(offs22) \ + hppa_t12_L_insn(0x3a, offs22, 1) +#define hppa64_mtsarcm(reg) \ + hppa_t21_insn(0x00, 11, reg, 0xc6, 0) +#define hppa64_shrpd_sar(reg, target) \ + hppa_t10_insn(0x34, reg, 0, 0, 0, 1<<4, target) +#define hppa64_shladd(r1, sa, r2, target) \ + hppa_t6_insn(0x02, r2, r1, 0, 0, 1<<4|1<<3|sa, target) +#define hppa64_depdz_sar(reg, target) \ + hppa_t21_insn(0x35, target, reg, 3<<3, 0) +#define hppa_extrd_sar(reg, target, se) \ + hppa_t10_insn(0x34, reg, target, 0, 0, 0, 0) | 2<<11 | (se&1)<<10 | 1<<9 | 1<<8 +#define hppa64_bve_l_rp(base) \ + (0x3a << 26) | (base << 21) | 0xf000 +#define hppa64_permh_3210(r, target) \ + (0x3e << 26) | (r << 21) | (r << 16) | (target) | 0x00006900 +#define hppa64_hshl(r, sa, target) \ + (0x3e << 26) | (0 << 21) | (r << 16) | (sa << 6) | (target) | 0x00008800 +#define hppa64_hshr_u(r, sa, target) \ + (0x3e << 26) | (r << 21) | (0 << 16) | (sa << 6) | (target) | 0x0000c800 +#endif + +struct hppa_jit_data { + struct bpf_binary_header *header; + u8 *image; + struct hppa_jit_context ctx; +}; + +static inline void bpf_fill_ill_insns(void *area, unsigned int size) +{ + memset(area, 0, size); +} + +static inline void bpf_flush_icache(void *start, void *end) +{ + flush_icache_range((unsigned long)start, (unsigned long)end); +} + +/* Emit a 4-byte HPPA instruction. */ +static inline void emit(const u32 insn, struct hppa_jit_context *ctx) +{ + if (ctx->insns) { + ctx->insns[ctx->ninsns] = insn; + } + + ctx->ninsns++; +} + +static inline int epilogue_offset(struct hppa_jit_context *ctx) +{ + int to = ctx->epilogue_offset, from = ctx->ninsns; + + return (to - from); +} + +/* Return -1 or inverted cond. */ +static inline int invert_bpf_cond(u8 cond) +{ + switch (cond) { + case BPF_JEQ: + return BPF_JNE; + case BPF_JGT: + return BPF_JLE; + case BPF_JLT: + return BPF_JGE; + case BPF_JGE: + return BPF_JLT; + case BPF_JLE: + return BPF_JGT; + case BPF_JNE: + return BPF_JEQ; + case BPF_JSGT: + return BPF_JSLE; + case BPF_JSLT: + return BPF_JSGE; + case BPF_JSGE: + return BPF_JSLT; + case BPF_JSLE: + return BPF_JSGT; + } + return -1; +} + + +static inline signed long hppa_offset(int insn, int off, struct hppa_jit_context *ctx) +{ + signed long from, to; + + off++; /* BPF branch is from PC+1 */ + from = (insn > 0) ? ctx->offset[insn - 1] : 0; + to = (insn + off > 0) ? ctx->offset[insn + off - 1] : 0; + return (to - from); +} + +/* does the signed value fits into a given number of bits ? */ +static inline int check_bits_int(signed long val, int bits) +{ + return ((val >= 0) && ((val >> bits) == 0)) || + ((val < 0) && (((~((u32)val)) >> (bits-1)) == 0)); +} + +/* can the signed value be used in relative code ? */ +static inline int relative_bits_ok(signed long val, int bits) +{ + return ((val >= 0) && (val < (1UL << (bits-1)))) || /* XXX */ + ((val < 0) && (((~((unsigned long)val)) >> (bits-1)) == 0) + && (val & (1UL << (bits-1)))); +} + +/* can the signed value be used in relative branches ? */ +static inline int relative_branch_ok(signed long val, int bits) +{ + return ((val >= 0) && (val < (1UL << (bits-2)))) || /* XXX */ + ((val < 0) && (((~((unsigned long)val)) < (1UL << (bits-2)))) + && (val & (1UL << (bits-1)))); +} + + +#define is_5b_int(val) check_bits_int(val, 5) + +static inline unsigned sign_unext(unsigned x, unsigned len) +{ + unsigned len_ones; + + len_ones = (1 << len) - 1; + return x & len_ones; +} + +static inline unsigned low_sign_unext(unsigned x, unsigned len) +{ + unsigned temp; + unsigned sign; + + sign = (x >> (len-1)) & 1; + temp = sign_unext (x, len-1); + return (temp << 1) | sign; +} + +static inline unsigned re_assemble_12(unsigned as12) +{ + return (( (as12 & 0x800) >> 11) + | ((as12 & 0x400) >> (10 - 2)) + | ((as12 & 0x3ff) << (1 + 2))); +} + +static inline unsigned re_assemble_14(unsigned as14) +{ + return (( (as14 & 0x1fff) << 1) + | ((as14 & 0x2000) >> 13)); +} + +#ifdef CONFIG_64BIT +static inline unsigned re_assemble_16(unsigned as16) +{ + unsigned s, t; + + /* Unusual 16-bit encoding, for wide mode only. */ + t = (as16 << 1) & 0xffff; + s = (as16 & 0x8000); + return (t ^ s ^ (s >> 1)) | (s >> 15); +} +#endif + +static inline unsigned re_assemble_17(unsigned as17) +{ + return (( (as17 & 0x10000) >> 16) + | ((as17 & 0x0f800) << (16 - 11)) + | ((as17 & 0x00400) >> (10 - 2)) + | ((as17 & 0x003ff) << (1 + 2))); +} + +static inline unsigned re_assemble_21(unsigned as21) +{ + return (( (as21 & 0x100000) >> 20) + | ((as21 & 0x0ffe00) >> 8) + | ((as21 & 0x000180) << 7) + | ((as21 & 0x00007c) << 14) + | ((as21 & 0x000003) << 12)); +} + +static inline unsigned re_assemble_22(unsigned as22) +{ + return (( (as22 & 0x200000) >> 21) + | ((as22 & 0x1f0000) << (21 - 16)) + | ((as22 & 0x00f800) << (16 - 11)) + | ((as22 & 0x000400) >> (10 - 2)) + | ((as22 & 0x0003ff) << (1 + 2))); +} + +/* Various HPPA instruction formats. */ +/* see https://parisc.wiki.kernel.org/images-parisc/6/68/Pa11_acd.pdf, appendix C */ + +static inline u32 hppa_t1_insn(u8 opcode, u8 b, u8 r, s16 im14) +{ + return ((opcode << 26) | (b << 21) | (r << 16) | re_assemble_14(im14)); +} + +static inline u32 hppa_t5_insn(u8 opcode, u8 tr, u32 val21) +{ + return ((opcode << 26) | (tr << 21) | re_assemble_21(val21)); +} + +static inline u32 hppa_t6_insn(u8 opcode, u8 r2, u8 r1, u8 c, u8 f, u8 ext6, u16 t) +{ + return ((opcode << 26) | (r2 << 21) | (r1 << 16) | (c << 13) | (f << 12) | + (ext6 << 6) | t); +} + +/* 7. Arithmetic immediate */ +static inline u32 hppa_t7_insn(u8 opcode, u8 r, u8 t, u32 im11) +{ + return ((opcode << 26) | (r << 21) | (t << 16) | low_sign_unext(im11, 11)); +} + +/* 10. Shift instructions */ +static inline u32 hppa_t10_insn(u8 opcode, u8 r2, u8 r1, u8 c, u8 ext3, u8 cp, u8 t) +{ + return ((opcode << 26) | (r2 << 21) | (r1 << 16) | (c << 13) | + (ext3 << 10) | (cp << 5) | t); +} + +/* 11. Conditional branch instructions */ +static inline u32 hppa_t11_insn(u8 opcode, u8 r2, u8 r1, u8 c, u32 w, u8 nop) +{ + u32 ra = re_assemble_12(w); + // ra = low_sign_unext(w,11) | (w & (1<<10) + return ((opcode << 26) | (r2 << 21) | (r1 << 16) | (c << 13) | (nop << 1) | ra); +} + +/* 12. Branch instructions */ +static inline u32 hppa_t12_insn(u8 opcode, u8 rp, u32 w, u8 ext3, u8 nop) +{ + return ((opcode << 26) | (rp << 21) | (ext3 << 13) | (nop << 1) | re_assemble_17(w)); +} + +static inline u32 hppa_t12_L_insn(u8 opcode, u32 w, u8 nop) +{ + return ((opcode << 26) | (0x05 << 13) | (nop << 1) | re_assemble_22(w)); +} + +/* 21. Move to control register */ +static inline u32 hppa_t21_insn(u8 opcode, u8 r2, u8 r1, u8 ext8, u8 t) +{ + return ((opcode << 26) | (r2 << 21) | (r1 << 16) | (ext8 << 5) | t); +} + +/* Helper functions called by jit code on HPPA32 and HPPA64. */ + +u64 hppa_div64(u64 div, u64 divisor); +u64 hppa_div64_rem(u64 div, u64 divisor); + +/* Helper functions that emit HPPA instructions when possible. */ + +void bpf_jit_build_prologue(struct hppa_jit_context *ctx); +void bpf_jit_build_epilogue(struct hppa_jit_context *ctx); + +int bpf_jit_emit_insn(const struct bpf_insn *insn, struct hppa_jit_context *ctx, + bool extra_pass); + +#endif /* _BPF_JIT_H */ diff --git a/arch/parisc/net/bpf_jit_comp32.c b/arch/parisc/net/bpf_jit_comp32.c new file mode 100644 index 000000000000..5ff0cf925fe9 --- /dev/null +++ b/arch/parisc/net/bpf_jit_comp32.c @@ -0,0 +1,1615 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * BPF JIT compiler for PA-RISC (32-bit) + * + * Copyright (c) 2023 Helge Deller <deller@gmx.de> + * + * The code is based on the BPF JIT compiler for RV64 by Björn Töpel and + * the BPF JIT compiler for 32-bit ARM by Shubham Bansal and Mircea Gherzan. + */ + +#include <linux/bpf.h> +#include <linux/filter.h> +#include <linux/libgcc.h> +#include "bpf_jit.h" + +/* + * Stack layout during BPF program execution (note: stack grows up): + * + * high + * HPPA32 sp => +----------+ <= HPPA32 fp + * | saved sp | + * | saved rp | + * | ... | HPPA32 callee-saved registers + * | curr args| + * | local var| + * +----------+ <= (sp - 4 * NR_SAVED_REGISTERS) + * | lo(R9) | + * | hi(R9) | + * | lo(FP) | JIT scratch space for BPF registers + * | hi(FP) | + * | ... | + * +----------+ <= (sp - 4 * NR_SAVED_REGISTERS + * | | - 4 * BPF_JIT_SCRATCH_REGS) + * | | + * | ... | BPF program stack + * | | + * | ... | Function call stack + * | | + * +----------+ + * low + */ + +enum { + /* Stack layout - these are offsets from top of JIT scratch space. */ + BPF_R8_HI, + BPF_R8_LO, + BPF_R9_HI, + BPF_R9_LO, + BPF_FP_HI, + BPF_FP_LO, + BPF_AX_HI, + BPF_AX_LO, + BPF_R0_TEMP_HI, + BPF_R0_TEMP_LO, + BPF_JIT_SCRATCH_REGS, +}; + +/* Number of callee-saved registers stored to stack: rp, r3-r18. */ +#define NR_SAVED_REGISTERS (18 - 3 + 1 + 8) + +/* Offset from fp for BPF registers stored on stack. */ +#define STACK_OFFSET(k) (- (NR_SAVED_REGISTERS + k + 1)) +#define STACK_ALIGN FRAME_SIZE + +#define EXIT_PTR_LOAD(reg) hppa_ldw(-0x08, HPPA_REG_SP, reg) +#define EXIT_PTR_STORE(reg) hppa_stw(reg, -0x08, HPPA_REG_SP) +#define EXIT_PTR_JUMP(reg, nop) hppa_bv(HPPA_REG_ZERO, reg, nop) + +#define TMP_REG_1 (MAX_BPF_JIT_REG + 0) +#define TMP_REG_2 (MAX_BPF_JIT_REG + 1) +#define TMP_REG_R0 (MAX_BPF_JIT_REG + 2) + +static const s8 regmap[][2] = { + /* Return value from in-kernel function, and exit value from eBPF. */ + [BPF_REG_0] = {HPPA_REG_RET0, HPPA_REG_RET1}, /* HI/LOW */ + + /* Arguments from eBPF program to in-kernel function. */ + [BPF_REG_1] = {HPPA_R(3), HPPA_R(4)}, + [BPF_REG_2] = {HPPA_R(5), HPPA_R(6)}, + [BPF_REG_3] = {HPPA_R(7), HPPA_R(8)}, + [BPF_REG_4] = {HPPA_R(9), HPPA_R(10)}, + [BPF_REG_5] = {HPPA_R(11), HPPA_R(12)}, + + [BPF_REG_6] = {HPPA_R(13), HPPA_R(14)}, + [BPF_REG_7] = {HPPA_R(15), HPPA_R(16)}, + /* + * Callee-saved registers that in-kernel function will preserve. + * Stored on the stack. + */ + [BPF_REG_8] = {STACK_OFFSET(BPF_R8_HI), STACK_OFFSET(BPF_R8_LO)}, + [BPF_REG_9] = {STACK_OFFSET(BPF_R9_HI), STACK_OFFSET(BPF_R9_LO)}, + + /* Read-only frame pointer to access BPF stack. Not needed. */ + [BPF_REG_FP] = {STACK_OFFSET(BPF_FP_HI), STACK_OFFSET(BPF_FP_LO)}, + + /* Temporary register for blinding constants. Stored on the stack. */ + [BPF_REG_AX] = {STACK_OFFSET(BPF_AX_HI), STACK_OFFSET(BPF_AX_LO)}, + /* + * Temporary registers used by the JIT to operate on registers stored + * on the stack. Save t0 and t1 to be used as temporaries in generated + * code. + */ + [TMP_REG_1] = {HPPA_REG_T3, HPPA_REG_T2}, + [TMP_REG_2] = {HPPA_REG_T5, HPPA_REG_T4}, + + /* temporary space for BPF_R0 during libgcc and millicode calls */ + [TMP_REG_R0] = {STACK_OFFSET(BPF_R0_TEMP_HI), STACK_OFFSET(BPF_R0_TEMP_LO)}, +}; + +static s8 hi(const s8 *r) +{ + return r[0]; +} + +static s8 lo(const s8 *r) +{ + return r[1]; +} + +static void emit_hppa_copy(const s8 rs, const s8 rd, struct hppa_jit_context *ctx) +{ + REG_SET_SEEN(ctx, rd); + if (OPTIMIZE_HPPA && (rs == rd)) + return; + REG_SET_SEEN(ctx, rs); + emit(hppa_copy(rs, rd), ctx); +} + +static void emit_hppa_xor(const s8 r1, const s8 r2, const s8 r3, struct hppa_jit_context *ctx) +{ + REG_SET_SEEN(ctx, r1); + REG_SET_SEEN(ctx, r2); + REG_SET_SEEN(ctx, r3); + if (OPTIMIZE_HPPA && (r1 == r2)) { + emit(hppa_copy(HPPA_REG_ZERO, r3), ctx); + } else { + emit(hppa_xor(r1, r2, r3), ctx); + } +} + +static void emit_imm(const s8 rd, s32 imm, struct hppa_jit_context *ctx) +{ + u32 lower = im11(imm); + + REG_SET_SEEN(ctx, rd); + if (OPTIMIZE_HPPA && relative_bits_ok(imm, 14)) { + emit(hppa_ldi(imm, rd), ctx); + return; + } + emit(hppa_ldil(imm, rd), ctx); + if (OPTIMIZE_HPPA && (lower == 0)) + return; + emit(hppa_ldo(lower, rd, rd), ctx); +} + +static void emit_imm32(const s8 *rd, s32 imm, struct hppa_jit_context *ctx) +{ + /* Emit immediate into lower bits. */ + REG_SET_SEEN(ctx, lo(rd)); + emit_imm(lo(rd), imm, ctx); + + /* Sign-extend into upper bits. */ + REG_SET_SEEN(ctx, hi(rd)); + if (imm >= 0) + emit_hppa_copy(HPPA_REG_ZERO, hi(rd), ctx); + else + emit(hppa_ldi(-1, hi(rd)), ctx); +} + +static void emit_imm64(const s8 *rd, s32 imm_hi, s32 imm_lo, + struct hppa_jit_context *ctx) +{ + emit_imm(hi(rd), imm_hi, ctx); + emit_imm(lo(rd), imm_lo, ctx); +} + +static void __build_epilogue(bool is_tail_call, struct hppa_jit_context *ctx) +{ + const s8 *r0 = regmap[BPF_REG_0]; + int i; + + if (is_tail_call) { + /* + * goto *(t0 + 4); + * Skips first instruction of prologue which initializes tail + * call counter. Assumes t0 contains address of target program, + * see emit_bpf_tail_call. + */ + emit(hppa_ldo(1 * HPPA_INSN_SIZE, HPPA_REG_T0, HPPA_REG_T0), ctx); + emit(hppa_bv(HPPA_REG_ZERO, HPPA_REG_T0, EXEC_NEXT_INSTR), ctx); + /* in delay slot: */ + emit(hppa_copy(HPPA_REG_TCC, HPPA_REG_TCC_IN_INIT), ctx); + + return; + } + + /* load epilogue function pointer and jump to it. */ + /* exit point is either directly below, or the outest TCC exit function */ + emit(EXIT_PTR_LOAD(HPPA_REG_RP), ctx); + emit(EXIT_PTR_JUMP(HPPA_REG_RP, NOP_NEXT_INSTR), ctx); + + /* NOTE: we are 32-bit and big-endian, so return lower 32-bit value */ + emit_hppa_copy(lo(r0), HPPA_REG_RET0, ctx); + + /* Restore callee-saved registers. */ + for (i = 3; i <= 18; i++) { + if (OPTIMIZE_HPPA && !REG_WAS_SEEN(ctx, HPPA_R(i))) + continue; + emit(hppa_ldw(-REG_SIZE * (8 + (i-3)), HPPA_REG_SP, HPPA_R(i)), ctx); + } + + /* load original return pointer (stored by outest TCC function) */ + emit(hppa_ldw(-0x14, HPPA_REG_SP, HPPA_REG_RP), ctx); + emit(hppa_bv(HPPA_REG_ZERO, HPPA_REG_RP, EXEC_NEXT_INSTR), ctx); + /* in delay slot: */ + emit(hppa_ldw(-0x04, HPPA_REG_SP, HPPA_REG_SP), ctx); +} + +static bool is_stacked(s8 reg) +{ + return reg < 0; +} + +static const s8 *bpf_get_reg64_offset(const s8 *reg, const s8 *tmp, + u16 offset_sp, struct hppa_jit_context *ctx) +{ + if (is_stacked(hi(reg))) { + emit(hppa_ldw(REG_SIZE * hi(reg) - offset_sp, HPPA_REG_SP, hi(tmp)), ctx); + emit(hppa_ldw(REG_SIZE * lo(reg) - offset_sp, HPPA_REG_SP, lo(tmp)), ctx); + reg = tmp; + } + REG_SET_SEEN(ctx, hi(reg)); + REG_SET_SEEN(ctx, lo(reg)); + return reg; +} + +static const s8 *bpf_get_reg64(const s8 *reg, const s8 *tmp, + struct hppa_jit_context *ctx) +{ + return bpf_get_reg64_offset(reg, tmp, 0, ctx); +} + +static const s8 *bpf_get_reg64_ref(const s8 *reg, const s8 *tmp, + bool must_load, struct hppa_jit_context *ctx) +{ + if (!OPTIMIZE_HPPA) + return bpf_get_reg64(reg, tmp, ctx); + + if (is_stacked(hi(reg))) { + if (must_load) + emit(hppa_ldw(REG_SIZE * hi(reg), HPPA_REG_SP, hi(tmp)), ctx); + reg = tmp; + } + REG_SET_SEEN(ctx, hi(reg)); + REG_SET_SEEN(ctx, lo(reg)); + return reg; +} + + +static void bpf_put_reg64(const s8 *reg, const s8 *src, + struct hppa_jit_context *ctx) +{ + if (is_stacked(hi(reg))) { + emit(hppa_stw(hi(src), REG_SIZE * hi(reg), HPPA_REG_SP), ctx); + emit(hppa_stw(lo(src), REG_SIZE * lo(reg), HPPA_REG_SP), ctx); + } +} + +static void bpf_save_R0(struct hppa_jit_context *ctx) +{ + bpf_put_reg64(regmap[TMP_REG_R0], regmap[BPF_REG_0], ctx); +} + +static void bpf_restore_R0(struct hppa_jit_context *ctx) +{ + bpf_get_reg64(regmap[TMP_REG_R0], regmap[BPF_REG_0], ctx); +} + + +static const s8 *bpf_get_reg32(const s8 *reg, const s8 *tmp, + struct hppa_jit_context *ctx) +{ + if (is_stacked(lo(reg))) { + emit(hppa_ldw(REG_SIZE * lo(reg), HPPA_REG_SP, lo(tmp)), ctx); + reg = tmp; + } + REG_SET_SEEN(ctx, lo(reg)); + return reg; +} + +static const s8 *bpf_get_reg32_ref(const s8 *reg, const s8 *tmp, + struct hppa_jit_context *ctx) +{ + if (!OPTIMIZE_HPPA) + return bpf_get_reg32(reg, tmp, ctx); + + if (is_stacked(hi(reg))) { + reg = tmp; + } + REG_SET_SEEN(ctx, lo(reg)); + return reg; +} + +static void bpf_put_reg32(const s8 *reg, const s8 *src, + struct hppa_jit_context *ctx) +{ + if (is_stacked(lo(reg))) { + REG_SET_SEEN(ctx, lo(src)); + emit(hppa_stw(lo(src), REG_SIZE * lo(reg), HPPA_REG_SP), ctx); + if (1 && !ctx->prog->aux->verifier_zext) { + REG_SET_SEEN(ctx, hi(reg)); + emit(hppa_stw(HPPA_REG_ZERO, REG_SIZE * hi(reg), HPPA_REG_SP), ctx); + } + } else if (1 && !ctx->prog->aux->verifier_zext) { + REG_SET_SEEN(ctx, hi(reg)); + emit_hppa_copy(HPPA_REG_ZERO, hi(reg), ctx); + } +} + +/* extern hppa millicode functions */ +extern void $$mulI(void); +extern void $$divU(void); +extern void $$remU(void); + +static void emit_call_millicode(void *func, const s8 arg0, + const s8 arg1, u8 opcode, struct hppa_jit_context *ctx) +{ + u32 func_addr; + + emit_hppa_copy(arg0, HPPA_REG_ARG0, ctx); + emit_hppa_copy(arg1, HPPA_REG_ARG1, ctx); + + /* libcgcc overwrites HPPA_REG_RET0/1, save temp. in dest. */ + if (arg0 != HPPA_REG_RET1) + bpf_save_R0(ctx); + + func_addr = (uintptr_t) dereference_function_descriptor(func); + emit(hppa_ldil(func_addr, HPPA_REG_R31), ctx); + /* skip the following be_l instruction if divisor is zero. */ + if (BPF_OP(opcode) == BPF_DIV || BPF_OP(opcode) == BPF_MOD) { + if (BPF_OP(opcode) == BPF_DIV) + emit_hppa_copy(HPPA_REG_ZERO, HPPA_REG_RET1, ctx); + else + emit_hppa_copy(HPPA_REG_ARG0, HPPA_REG_RET1, ctx); + emit(hppa_or_cond(HPPA_REG_ARG1, HPPA_REG_ZERO, 1, 0, HPPA_REG_ZERO), ctx); + } + /* Note: millicode functions use r31 as return pointer instead of rp */ + emit(hppa_be_l(im11(func_addr) >> 2, HPPA_REG_R31, NOP_NEXT_INSTR), ctx); + emit(hppa_nop(), ctx); /* this nop is needed here for delay slot */ + + /* Note: millicode functions return result in RET1, not RET0 */ + emit_hppa_copy(HPPA_REG_RET1, arg0, ctx); + + /* restore HPPA_REG_RET0/1, temp. save in dest. */ + if (arg0 != HPPA_REG_RET1) + bpf_restore_R0(ctx); +} + +static void emit_call_libgcc_ll(void *func, const s8 *arg0, + const s8 *arg1, u8 opcode, struct hppa_jit_context *ctx) +{ + u32 func_addr; + + emit_hppa_copy(lo(arg0), HPPA_REG_ARG0, ctx); + emit_hppa_copy(hi(arg0), HPPA_REG_ARG1, ctx); + emit_hppa_copy(lo(arg1), HPPA_REG_ARG2, ctx); + emit_hppa_copy(hi(arg1), HPPA_REG_ARG3, ctx); + + /* libcgcc overwrites HPPA_REG_RET0/_RET1, so keep copy of R0 on stack */ + if (hi(arg0) != HPPA_REG_RET0) + bpf_save_R0(ctx); + + /* prepare stack */ + emit(hppa_ldo(2 * FRAME_SIZE, HPPA_REG_SP, HPPA_REG_SP), ctx); + + func_addr = (uintptr_t) dereference_function_descriptor(func); + emit(hppa_ldil(func_addr, HPPA_REG_R31), ctx); + /* zero out the following be_l instruction if divisor is 0 (and set default values) */ + if (BPF_OP(opcode) == BPF_DIV || BPF_OP(opcode) == BPF_MOD) { + emit_hppa_copy(HPPA_REG_ZERO, HPPA_REG_RET0, ctx); + if (BPF_OP(opcode) == BPF_DIV) + emit_hppa_copy(HPPA_REG_ZERO, HPPA_REG_RET1, ctx); + else + emit_hppa_copy(HPPA_REG_ARG0, HPPA_REG_RET1, ctx); + emit(hppa_or_cond(HPPA_REG_ARG2, HPPA_REG_ARG3, 1, 0, HPPA_REG_ZERO), ctx); + } + emit(hppa_be_l(im11(func_addr) >> 2, HPPA_REG_R31, EXEC_NEXT_INSTR), ctx); + emit_hppa_copy(HPPA_REG_R31, HPPA_REG_RP, ctx); + + /* restore stack */ + emit(hppa_ldo(-2 * FRAME_SIZE, HPPA_REG_SP, HPPA_REG_SP), ctx); + + emit_hppa_copy(HPPA_REG_RET0, hi(arg0), ctx); + emit_hppa_copy(HPPA_REG_RET1, lo(arg0), ctx); + + /* restore HPPA_REG_RET0/_RET1 */ + if (hi(arg0) != HPPA_REG_RET0) + bpf_restore_R0(ctx); +} + +static void emit_jump(s32 paoff, bool force_far, + struct hppa_jit_context *ctx) +{ + unsigned long pc, addr; + + /* Note: allocate 2 instructions for jumps if force_far is set. */ + if (relative_bits_ok(paoff - HPPA_BRANCH_DISPLACEMENT, 17)) { + /* use BL,short branch followed by nop() */ + emit(hppa_bl(paoff - HPPA_BRANCH_DISPLACEMENT, HPPA_REG_ZERO), ctx); + if (force_far) + emit(hppa_nop(), ctx); + return; + } + + pc = (uintptr_t) &ctx->insns[ctx->ninsns]; + addr = pc + (paoff * HPPA_INSN_SIZE); + emit(hppa_ldil(addr, HPPA_REG_R31), ctx); + emit(hppa_be_l(im11(addr) >> 2, HPPA_REG_R31, NOP_NEXT_INSTR), ctx); // be,l,n addr(sr4,r31), %sr0, %r31 +} + +static void emit_alu_i64(const s8 *dst, s32 imm, + struct hppa_jit_context *ctx, const u8 op) +{ + const s8 *tmp1 = regmap[TMP_REG_1]; + const s8 *rd; + + if (0 && op == BPF_MOV) + rd = bpf_get_reg64_ref(dst, tmp1, false, ctx); + else + rd = bpf_get_reg64(dst, tmp1, ctx); + + /* dst = dst OP imm */ + switch (op) { + case BPF_MOV: + emit_imm32(rd, imm, ctx); + break; + case BPF_AND: + emit_imm(HPPA_REG_T0, imm, ctx); + emit(hppa_and(lo(rd), HPPA_REG_T0, lo(rd)), ctx); + if (imm >= 0) + emit_hppa_copy(HPPA_REG_ZERO, hi(rd), ctx); + break; + case BPF_OR: + emit_imm(HPPA_REG_T0, imm, ctx); + emit(hppa_or(lo(rd), HPPA_REG_T0, lo(rd)), ctx); + if (imm < 0) + emit_imm(hi(rd), -1, ctx); + break; + case BPF_XOR: + emit_imm(HPPA_REG_T0, imm, ctx); + emit_hppa_xor(lo(rd), HPPA_REG_T0, lo(rd), ctx); + if (imm < 0) { + emit_imm(HPPA_REG_T0, -1, ctx); + emit_hppa_xor(hi(rd), HPPA_REG_T0, hi(rd), ctx); + } + break; + case BPF_LSH: + if (imm == 0) + break; + if (imm > 32) { + imm -= 32; + emit(hppa_zdep(lo(rd), imm, imm, hi(rd)), ctx); + emit_hppa_copy(HPPA_REG_ZERO, lo(rd), ctx); + } else if (imm == 32) { + emit_hppa_copy(lo(rd), hi(rd), ctx); + emit_hppa_copy(HPPA_REG_ZERO, lo(rd), ctx); + } else { + emit(hppa_shd(hi(rd), lo(rd), 32 - imm, hi(rd)), ctx); + emit(hppa_zdep(lo(rd), imm, imm, lo(rd)), ctx); + } + break; + case BPF_RSH: + if (imm == 0) + break; + if (imm > 32) { + imm -= 32; + emit(hppa_shr(hi(rd), imm, lo(rd)), ctx); + emit_hppa_copy(HPPA_REG_ZERO, hi(rd), ctx); + } else if (imm == 32) { + emit_hppa_copy(hi(rd), lo(rd), ctx); + emit_hppa_copy(HPPA_REG_ZERO, hi(rd), ctx); + } else { + emit(hppa_shrpw(hi(rd), lo(rd), imm, lo(rd)), ctx); + emit(hppa_shr(hi(rd), imm, hi(rd)), ctx); + } + break; + case BPF_ARSH: + if (imm == 0) + break; + if (imm > 32) { + imm -= 32; + emit(hppa_extrws(hi(rd), 31 - imm, imm, lo(rd)), ctx); + emit(hppa_extrws(hi(rd), 0, 31, hi(rd)), ctx); + } else if (imm == 32) { + emit_hppa_copy(hi(rd), lo(rd), ctx); + emit(hppa_extrws(hi(rd), 0, 31, hi(rd)), ctx); + } else { + emit(hppa_shrpw(hi(rd), lo(rd), imm, lo(rd)), ctx); + emit(hppa_extrws(hi(rd), 31 - imm, imm, hi(rd)), ctx); + } + break; + default: + WARN_ON(1); + } + + bpf_put_reg64(dst, rd, ctx); +} + +static void emit_alu_i32(const s8 *dst, s32 imm, + struct hppa_jit_context *ctx, const u8 op) +{ + const s8 *tmp1 = regmap[TMP_REG_1]; + const s8 *rd = bpf_get_reg32(dst, tmp1, ctx); + + if (op == BPF_MOV) + rd = bpf_get_reg32_ref(dst, tmp1, ctx); + else + rd = bpf_get_reg32(dst, tmp1, ctx); + + /* dst = dst OP imm */ + switch (op) { + case BPF_MOV: + emit_imm(lo(rd), imm, ctx); + break; + case BPF_ADD: + emit_imm(HPPA_REG_T0, imm, ctx); + emit(hppa_add(lo(rd), HPPA_REG_T0, lo(rd)), ctx); + break; + case BPF_SUB: + emit_imm(HPPA_REG_T0, imm, ctx); + emit(hppa_sub(lo(rd), HPPA_REG_T0, lo(rd)), ctx); + break; + case BPF_AND: + emit_imm(HPPA_REG_T0, imm, ctx); + emit(hppa_and(lo(rd), HPPA_REG_T0, lo(rd)), ctx); + break; + case BPF_OR: + emit_imm(HPPA_REG_T0, imm, ctx); + emit(hppa_or(lo(rd), HPPA_REG_T0, lo(rd)), ctx); + break; + case BPF_XOR: + emit_imm(HPPA_REG_T0, imm, ctx); + emit_hppa_xor(lo(rd), HPPA_REG_T0, lo(rd), ctx); + break; + case BPF_LSH: + if (imm != 0) + emit(hppa_zdep(lo(rd), imm, imm, lo(rd)), ctx); + break; + case BPF_RSH: + if (imm != 0) + emit(hppa_shr(lo(rd), imm, lo(rd)), ctx); + break; + case BPF_ARSH: + if (imm != 0) + emit(hppa_extrws(lo(rd), 31 - imm, imm, lo(rd)), ctx); + break; + default: + WARN_ON(1); + } + + bpf_put_reg32(dst, rd, ctx); +} + +static void emit_alu_r64(const s8 *dst, const s8 *src, + struct hppa_jit_context *ctx, const u8 op) +{ + const s8 *tmp1 = regmap[TMP_REG_1]; + const s8 *tmp2 = regmap[TMP_REG_2]; + const s8 *rd; + const s8 *rs = bpf_get_reg64(src, tmp2, ctx); + + if (op == BPF_MOV) + rd = bpf_get_reg64_ref(dst, tmp1, false, ctx); + else + rd = bpf_get_reg64(dst, tmp1, ctx); + + /* dst = dst OP src */ + switch (op) { + case BPF_MOV: + emit_hppa_copy(lo(rs), lo(rd), ctx); + emit_hppa_copy(hi(rs), hi(rd), ctx); + break; + case BPF_ADD: + emit(hppa_add(lo(rd), lo(rs), lo(rd)), ctx); + emit(hppa_addc(hi(rd), hi(rs), hi(rd)), ctx); + break; + case BPF_SUB: + emit(hppa_sub(lo(rd), lo(rs), lo(rd)), ctx); + emit(hppa_subb(hi(rd), hi(rs), hi(rd)), ctx); + break; + case BPF_AND: + emit(hppa_and(lo(rd), lo(rs), lo(rd)), ctx); + emit(hppa_and(hi(rd), hi(rs), hi(rd)), ctx); + break; + case BPF_OR: + emit(hppa_or(lo(rd), lo(rs), lo(rd)), ctx); + emit(hppa_or(hi(rd), hi(rs), hi(rd)), ctx); + break; + case BPF_XOR: + emit_hppa_xor(lo(rd), lo(rs), lo(rd), ctx); + emit_hppa_xor(hi(rd), hi(rs), hi(rd), ctx); + break; + case BPF_MUL: + emit_call_libgcc_ll(__muldi3, rd, rs, op, ctx); + break; + case BPF_DIV: + emit_call_libgcc_ll(&hppa_div64, rd, rs, op, ctx); + break; + case BPF_MOD: + emit_call_libgcc_ll(&hppa_div64_rem, rd, rs, op, ctx); + break; + case BPF_LSH: + emit_call_libgcc_ll(__ashldi3, rd, rs, op, ctx); + break; + case BPF_RSH: + emit_call_libgcc_ll(__lshrdi3, rd, rs, op, ctx); + break; + case BPF_ARSH: + emit_call_libgcc_ll(__ashrdi3, rd, rs, op, ctx); + break; + case BPF_NEG: + emit(hppa_sub(HPPA_REG_ZERO, lo(rd), lo(rd)), ctx); + emit(hppa_subb(HPPA_REG_ZERO, hi(rd), hi(rd)), ctx); + break; + default: + WARN_ON(1); + } + + bpf_put_reg64(dst, rd, ctx); +} + +static void emit_alu_r32(const s8 *dst, const s8 *src, + struct hppa_jit_context *ctx, const u8 op) +{ + const s8 *tmp1 = regmap[TMP_REG_1]; + const s8 *tmp2 = regmap[TMP_REG_2]; + const s8 *rd; + const s8 *rs = bpf_get_reg32(src, tmp2, ctx); + + if (op == BPF_MOV) + rd = bpf_get_reg32_ref(dst, tmp1, ctx); + else + rd = bpf_get_reg32(dst, tmp1, ctx); + + /* dst = dst OP src */ + switch (op) { + case BPF_MOV: + emit_hppa_copy(lo(rs), lo(rd), ctx); + break; + case BPF_ADD: + emit(hppa_add(lo(rd), lo(rs), lo(rd)), ctx); + break; + case BPF_SUB: + emit(hppa_sub(lo(rd), lo(rs), lo(rd)), ctx); + break; + case BPF_AND: + emit(hppa_and(lo(rd), lo(rs), lo(rd)), ctx); + break; + case BPF_OR: + emit(hppa_or(lo(rd), lo(rs), lo(rd)), ctx); + break; + case BPF_XOR: + emit_hppa_xor(lo(rd), lo(rs), lo(rd), ctx); + break; + case BPF_MUL: + emit_call_millicode($$mulI, lo(rd), lo(rs), op, ctx); + break; + case BPF_DIV: + emit_call_millicode($$divU, lo(rd), lo(rs), op, ctx); + break; + case BPF_MOD: + emit_call_millicode($$remU, lo(rd), lo(rs), op, ctx); + break; + case BPF_LSH: + emit(hppa_subi(0x1f, lo(rs), HPPA_REG_T0), ctx); + emit(hppa_mtsar(HPPA_REG_T0), ctx); + emit(hppa_depwz_sar(lo(rd), lo(rd)), ctx); + break; + case BPF_RSH: + emit(hppa_mtsar(lo(rs)), ctx); + emit(hppa_shrpw_sar(lo(rd), lo(rd)), ctx); + break; + case BPF_ARSH: /* sign extending arithmetic shift right */ + // emit(hppa_beq(lo(rs), HPPA_REG_ZERO, 2), ctx); + emit(hppa_subi(0x1f, lo(rs), HPPA_REG_T0), ctx); + emit(hppa_mtsar(HPPA_REG_T0), ctx); + emit(hppa_extrws_sar(lo(rd), lo(rd)), ctx); + break; + case BPF_NEG: + emit(hppa_sub(HPPA_REG_ZERO, lo(rd), lo(rd)), ctx); // sub r0,rd,rd + break; + default: + WARN_ON(1); + } + + bpf_put_reg32(dst, rd, ctx); +} + +static int emit_branch_r64(const s8 *src1, const s8 *src2, s32 paoff, + struct hppa_jit_context *ctx, const u8 op) +{ + int e, s = ctx->ninsns; + const s8 *tmp1 = regmap[TMP_REG_1]; + const s8 *tmp2 = regmap[TMP_REG_2]; + + const s8 *rs1 = bpf_get_reg64(src1, tmp1, ctx); + const s8 *rs2 = bpf_get_reg64(src2, tmp2, ctx); + + /* + * NO_JUMP skips over the rest of the instructions and the + * emit_jump, meaning the BPF branch is not taken. + * JUMP skips directly to the emit_jump, meaning + * the BPF branch is taken. + * + * The fallthrough case results in the BPF branch being taken. + */ +#define NO_JUMP(idx) (2 + (idx) - 1) +#define JUMP(idx) (0 + (idx) - 1) + + switch (op) { + case BPF_JEQ: + emit(hppa_bne(hi(rs1), hi(rs2), NO_JUMP(1)), ctx); + emit(hppa_bne(lo(rs1), lo(rs2), NO_JUMP(0)), ctx); + break; + case BPF_JGT: + emit(hppa_bgtu(hi(rs1), hi(rs2), JUMP(2)), ctx); + emit(hppa_bltu(hi(rs1), hi(rs2), NO_JUMP(1)), ctx); + emit(hppa_bleu(lo(rs1), lo(rs2), NO_JUMP(0)), ctx); + break; + case BPF_JLT: + emit(hppa_bltu(hi(rs1), hi(rs2), JUMP(2)), ctx); + emit(hppa_bgtu(hi(rs1), hi(rs2), NO_JUMP(1)), ctx); + emit(hppa_bgeu(lo(rs1), lo(rs2), NO_JUMP(0)), ctx); + break; + case BPF_JGE: + emit(hppa_bgtu(hi(rs1), hi(rs2), JUMP(2)), ctx); + emit(hppa_bltu(hi(rs1), hi(rs2), NO_JUMP(1)), ctx); + emit(hppa_bltu(lo(rs1), lo(rs2), NO_JUMP(0)), ctx); + break; + case BPF_JLE: + emit(hppa_bltu(hi(rs1), hi(rs2), JUMP(2)), ctx); + emit(hppa_bgtu(hi(rs1), hi(rs2), NO_JUMP(1)), ctx); + emit(hppa_bgtu(lo(rs1), lo(rs2), NO_JUMP(0)), ctx); + break; + case BPF_JNE: + emit(hppa_bne(hi(rs1), hi(rs2), JUMP(1)), ctx); + emit(hppa_beq(lo(rs1), lo(rs2), NO_JUMP(0)), ctx); + break; + case BPF_JSGT: + emit(hppa_bgt(hi(rs1), hi(rs2), JUMP(2)), ctx); + emit(hppa_blt(hi(rs1), hi(rs2), NO_JUMP(1)), ctx); + emit(hppa_bleu(lo(rs1), lo(rs2), NO_JUMP(0)), ctx); + break; + case BPF_JSLT: + emit(hppa_blt(hi(rs1), hi(rs2), JUMP(2)), ctx); + emit(hppa_bgt(hi(rs1), hi(rs2), NO_JUMP(1)), ctx); + emit(hppa_bgeu(lo(rs1), lo(rs2), NO_JUMP(0)), ctx); + break; + case BPF_JSGE: + emit(hppa_bgt(hi(rs1), hi(rs2), JUMP(2)), ctx); + emit(hppa_blt(hi(rs1), hi(rs2), NO_JUMP(1)), ctx); + emit(hppa_bltu(lo(rs1), lo(rs2), NO_JUMP(0)), ctx); + break; + case BPF_JSLE: + emit(hppa_blt(hi(rs1), hi(rs2), JUMP(2)), ctx); + emit(hppa_bgt(hi(rs1), hi(rs2), NO_JUMP(1)), ctx); + emit(hppa_bgtu(lo(rs1), lo(rs2), NO_JUMP(0)), ctx); + break; + case BPF_JSET: + emit(hppa_and(hi(rs1), hi(rs2), HPPA_REG_T0), ctx); + emit(hppa_and(lo(rs1), lo(rs2), HPPA_REG_T1), ctx); + emit(hppa_bne(HPPA_REG_T0, HPPA_REG_ZERO, JUMP(1)), ctx); + emit(hppa_beq(HPPA_REG_T1, HPPA_REG_ZERO, NO_JUMP(0)), ctx); + break; + default: + WARN_ON(1); + } + +#undef NO_JUMP +#undef JUMP + + e = ctx->ninsns; + /* Adjust for extra insns. */ + paoff -= (e - s); + emit_jump(paoff, true, ctx); + return 0; +} + +static int emit_bcc(u8 op, u8 rd, u8 rs, int paoff, struct hppa_jit_context *ctx) +{ + int e, s; + bool far = false; + int off; + + if (op == BPF_JSET) { + /* + * BPF_JSET is a special case: it has no inverse so we always + * treat it as a far branch. + */ + emit(hppa_and(rd, rs, HPPA_REG_T0), ctx); + paoff -= 1; /* reduce offset due to hppa_and() above */ + rd = HPPA_REG_T0; + rs = HPPA_REG_ZERO; + op = BPF_JNE; + } + + s = ctx->ninsns; + + if (!relative_bits_ok(paoff - HPPA_BRANCH_DISPLACEMENT, 12)) { + op = invert_bpf_cond(op); + far = true; + } + + /* + * For a far branch, the condition is negated and we jump over the + * branch itself, and the three instructions from emit_jump. + * For a near branch, just use paoff. + */ + off = far ? (HPPA_BRANCH_DISPLACEMENT - 1) : paoff - HPPA_BRANCH_DISPLACEMENT; + + switch (op) { + /* IF (dst COND src) JUMP off */ + case BPF_JEQ: + emit(hppa_beq(rd, rs, off), ctx); + break; + case BPF_JGT: + emit(hppa_bgtu(rd, rs, off), ctx); + break; + case BPF_JLT: + emit(hppa_bltu(rd, rs, off), ctx); + break; + case BPF_JGE: + emit(hppa_bgeu(rd, rs, off), ctx); + break; + case BPF_JLE: + emit(hppa_bleu(rd, rs, off), ctx); + break; + case BPF_JNE: + emit(hppa_bne(rd, rs, off), ctx); + break; + case BPF_JSGT: + emit(hppa_bgt(rd, rs, off), ctx); + break; + case BPF_JSLT: + emit(hppa_blt(rd, rs, off), ctx); + break; + case BPF_JSGE: + emit(hppa_bge(rd, rs, off), ctx); + break; + case BPF_JSLE: + emit(hppa_ble(rd, rs, off), ctx); + break; + default: + WARN_ON(1); + } + + if (far) { + e = ctx->ninsns; + /* Adjust for extra insns. */ + paoff -= (e - s); + emit_jump(paoff, true, ctx); + } + return 0; +} + +static int emit_branch_r32(const s8 *src1, const s8 *src2, s32 paoff, + struct hppa_jit_context *ctx, const u8 op) +{ + int e, s = ctx->ninsns; + const s8 *tmp1 = regmap[TMP_REG_1]; + const s8 *tmp2 = regmap[TMP_REG_2]; + + const s8 *rs1 = bpf_get_reg32(src1, tmp1, ctx); + const s8 *rs2 = bpf_get_reg32(src2, tmp2, ctx); + + e = ctx->ninsns; + /* Adjust for extra insns. */ + paoff -= (e - s); + + if (emit_bcc(op, lo(rs1), lo(rs2), paoff, ctx)) + return -1; + + return 0; +} + +static void emit_call(bool fixed, u64 addr, struct hppa_jit_context *ctx) +{ + const s8 *tmp = regmap[TMP_REG_1]; + const s8 *r0 = regmap[BPF_REG_0]; + const s8 *reg; + const int offset_sp = 2 * STACK_ALIGN; + + /* prepare stack */ + emit(hppa_ldo(offset_sp, HPPA_REG_SP, HPPA_REG_SP), ctx); + + /* load R1 & R2 in registers, R3-R5 to stack. */ + reg = bpf_get_reg64_offset(regmap[BPF_REG_5], tmp, offset_sp, ctx); + emit(hppa_stw(hi(reg), -0x48, HPPA_REG_SP), ctx); + emit(hppa_stw(lo(reg), -0x44, HPPA_REG_SP), ctx); + + reg = bpf_get_reg64_offset(regmap[BPF_REG_4], tmp, offset_sp, ctx); + emit(hppa_stw(hi(reg), -0x40, HPPA_REG_SP), ctx); + emit(hppa_stw(lo(reg), -0x3c, HPPA_REG_SP), ctx); + + reg = bpf_get_reg64_offset(regmap[BPF_REG_3], tmp, offset_sp, ctx); + emit(hppa_stw(hi(reg), -0x38, HPPA_REG_SP), ctx); + emit(hppa_stw(lo(reg), -0x34, HPPA_REG_SP), ctx); + + reg = bpf_get_reg64_offset(regmap[BPF_REG_2], tmp, offset_sp, ctx); + emit_hppa_copy(hi(reg), HPPA_REG_ARG3, ctx); + emit_hppa_copy(lo(reg), HPPA_REG_ARG2, ctx); + + reg = bpf_get_reg64_offset(regmap[BPF_REG_1], tmp, offset_sp, ctx); + emit_hppa_copy(hi(reg), HPPA_REG_ARG1, ctx); + emit_hppa_copy(lo(reg), HPPA_REG_ARG0, ctx); + + /* backup TCC */ + if (REG_WAS_SEEN(ctx, HPPA_REG_TCC)) + emit(hppa_copy(HPPA_REG_TCC, HPPA_REG_TCC_SAVED), ctx); + + /* + * Use ldil() to load absolute address. Don't use emit_imm as the + * number of emitted instructions should not depend on the value of + * addr. + */ + emit(hppa_ldil(addr, HPPA_REG_R31), ctx); + emit(hppa_be_l(im11(addr) >> 2, HPPA_REG_R31, EXEC_NEXT_INSTR), ctx); + /* set return address in delay slot */ + emit_hppa_copy(HPPA_REG_R31, HPPA_REG_RP, ctx); + + /* restore TCC */ + if (REG_WAS_SEEN(ctx, HPPA_REG_TCC)) + emit(hppa_copy(HPPA_REG_TCC_SAVED, HPPA_REG_TCC), ctx); + + /* restore stack */ + emit(hppa_ldo(-offset_sp, HPPA_REG_SP, HPPA_REG_SP), ctx); + + /* set return value. */ + emit_hppa_copy(HPPA_REG_RET0, hi(r0), ctx); + emit_hppa_copy(HPPA_REG_RET1, lo(r0), ctx); +} + +static int emit_bpf_tail_call(int insn, struct hppa_jit_context *ctx) +{ + /* + * R1 -> &ctx + * R2 -> &array + * R3 -> index + */ + int off; + const s8 *arr_reg = regmap[BPF_REG_2]; + const s8 *idx_reg = regmap[BPF_REG_3]; + struct bpf_array bpfa; + struct bpf_prog bpfp; + + /* get address of TCC main exit function for error case into rp */ + emit(EXIT_PTR_LOAD(HPPA_REG_RP), ctx); + + /* max_entries = array->map.max_entries; */ + off = offsetof(struct bpf_array, map.max_entries); + BUILD_BUG_ON(sizeof(bpfa.map.max_entries) != 4); + emit(hppa_ldw(off, lo(arr_reg), HPPA_REG_T1), ctx); + + /* + * if (index >= max_entries) + * goto out; + */ + emit(hppa_bltu(lo(idx_reg), HPPA_REG_T1, 2 - HPPA_BRANCH_DISPLACEMENT), ctx); + emit(EXIT_PTR_JUMP(HPPA_REG_RP, NOP_NEXT_INSTR), ctx); + + /* + * if (--tcc < 0) + * goto out; + */ + REG_FORCE_SEEN(ctx, HPPA_REG_TCC); + emit(hppa_ldo(-1, HPPA_REG_TCC, HPPA_REG_TCC), ctx); + emit(hppa_bge(HPPA_REG_TCC, HPPA_REG_ZERO, 2 - HPPA_BRANCH_DISPLACEMENT), ctx); + emit(EXIT_PTR_JUMP(HPPA_REG_RP, NOP_NEXT_INSTR), ctx); + + /* + * prog = array->ptrs[index]; + * if (!prog) + * goto out; + */ + BUILD_BUG_ON(sizeof(bpfa.ptrs[0]) != 4); + emit(hppa_sh2add(lo(idx_reg), lo(arr_reg), HPPA_REG_T0), ctx); + off = offsetof(struct bpf_array, ptrs); + BUILD_BUG_ON(!relative_bits_ok(off, 11)); + emit(hppa_ldw(off, HPPA_REG_T0, HPPA_REG_T0), ctx); + emit(hppa_bne(HPPA_REG_T0, HPPA_REG_ZERO, 2 - HPPA_BRANCH_DISPLACEMENT), ctx); + emit(EXIT_PTR_JUMP(HPPA_REG_RP, NOP_NEXT_INSTR), ctx); + + /* + * tcc = temp_tcc; + * goto *(prog->bpf_func + 4); + */ + off = offsetof(struct bpf_prog, bpf_func); + BUILD_BUG_ON(!relative_bits_ok(off, 11)); + BUILD_BUG_ON(sizeof(bpfp.bpf_func) != 4); + emit(hppa_ldw(off, HPPA_REG_T0, HPPA_REG_T0), ctx); + /* Epilogue jumps to *(t0 + 4). */ + __build_epilogue(true, ctx); + return 0; +} + +static int emit_load_r64(const s8 *dst, const s8 *src, s16 off, + struct hppa_jit_context *ctx, const u8 size) +{ + const s8 *tmp1 = regmap[TMP_REG_1]; + const s8 *tmp2 = regmap[TMP_REG_2]; + const s8 *rd = bpf_get_reg64_ref(dst, tmp1, ctx->prog->aux->verifier_zext, ctx); + const s8 *rs = bpf_get_reg64(src, tmp2, ctx); + s8 srcreg; + + /* need to calculate address since offset does not fit in 14 bits? */ + if (relative_bits_ok(off, 14)) + srcreg = lo(rs); + else { + /* need to use R1 here, since addil puts result into R1 */ + srcreg = HPPA_REG_R1; + emit(hppa_addil(off, lo(rs)), ctx); + off = im11(off); + } + + /* LDX: dst = *(size *)(src + off) */ + switch (size) { + case BPF_B: + emit(hppa_ldb(off + 0, srcreg, lo(rd)), ctx); + if (!ctx->prog->aux->verifier_zext) + emit_hppa_copy(HPPA_REG_ZERO, hi(rd), ctx); + break; + case BPF_H: + emit(hppa_ldh(off + 0, srcreg, lo(rd)), ctx); + if (!ctx->prog->aux->verifier_zext) + emit_hppa_copy(HPPA_REG_ZERO, hi(rd), ctx); + break; + case BPF_W: + emit(hppa_ldw(off + 0, srcreg, lo(rd)), ctx); + if (!ctx->prog->aux->verifier_zext) + emit_hppa_copy(HPPA_REG_ZERO, hi(rd), ctx); + break; + case BPF_DW: + emit(hppa_ldw(off + 0, srcreg, hi(rd)), ctx); + emit(hppa_ldw(off + 4, srcreg, lo(rd)), ctx); + break; + } + + bpf_put_reg64(dst, rd, ctx); + return 0; +} + +static int emit_store_r64(const s8 *dst, const s8 *src, s16 off, + struct hppa_jit_context *ctx, const u8 size, + const u8 mode) +{ + const s8 *tmp1 = regmap[TMP_REG_1]; + const s8 *tmp2 = regmap[TMP_REG_2]; + const s8 *rd = bpf_get_reg64(dst, tmp1, ctx); + const s8 *rs = bpf_get_reg64(src, tmp2, ctx); + s8 dstreg; + + /* need to calculate address since offset does not fit in 14 bits? */ + if (relative_bits_ok(off, 14)) + dstreg = lo(rd); + else { + /* need to use R1 here, since addil puts result into R1 */ + dstreg = HPPA_REG_R1; + emit(hppa_addil(off, lo(rd)), ctx); + off = im11(off); + } + + /* ST: *(size *)(dst + off) = imm */ + switch (size) { + case BPF_B: + emit(hppa_stb(lo(rs), off + 0, dstreg), ctx); + break; + case BPF_H: + emit(hppa_sth(lo(rs), off + 0, dstreg), ctx); + break; + case BPF_W: + emit(hppa_stw(lo(rs), off + 0, dstreg), ctx); + break; + case BPF_DW: + emit(hppa_stw(hi(rs), off + 0, dstreg), ctx); + emit(hppa_stw(lo(rs), off + 4, dstreg), ctx); + break; + } + + return 0; +} + +static void emit_rev16(const s8 rd, struct hppa_jit_context *ctx) +{ + emit(hppa_extru(rd, 23, 8, HPPA_REG_T1), ctx); + emit(hppa_depwz(rd, 23, 8, HPPA_REG_T1), ctx); + emit(hppa_extru(HPPA_REG_T1, 31, 16, rd), ctx); +} + +static void emit_rev32(const s8 rs, const s8 rd, struct hppa_jit_context *ctx) +{ + emit(hppa_shrpw(rs, rs, 16, HPPA_REG_T1), ctx); + emit(hppa_depwz(HPPA_REG_T1, 15, 8, HPPA_REG_T1), ctx); + emit(hppa_shrpw(rs, HPPA_REG_T1, 8, rd), ctx); +} + +static void emit_zext64(const s8 *dst, struct hppa_jit_context *ctx) +{ + const s8 *rd; + const s8 *tmp1 = regmap[TMP_REG_1]; + + rd = bpf_get_reg64(dst, tmp1, ctx); + emit_hppa_copy(HPPA_REG_ZERO, hi(rd), ctx); + bpf_put_reg64(dst, rd, ctx); +} + +int bpf_jit_emit_insn(const struct bpf_insn *insn, struct hppa_jit_context *ctx, + bool extra_pass) +{ + bool is64 = BPF_CLASS(insn->code) == BPF_ALU64 || + BPF_CLASS(insn->code) == BPF_JMP; + int s, e, paoff, i = insn - ctx->prog->insnsi; + u8 code = insn->code; + s16 off = insn->off; + s32 imm = insn->imm; + + const s8 *dst = regmap[insn->dst_reg]; + const s8 *src = regmap[insn->src_reg]; + const s8 *tmp1 = regmap[TMP_REG_1]; + const s8 *tmp2 = regmap[TMP_REG_2]; + + if (0) printk("CLASS %03d CODE %#02x ALU64:%d BPF_SIZE %#02x " + "BPF_CODE %#02x src_reg %d dst_reg %d\n", + BPF_CLASS(code), code, (code & BPF_ALU64) ? 1:0, BPF_SIZE(code), + BPF_OP(code), insn->src_reg, insn->dst_reg); + + switch (code) { + /* dst = src */ + case BPF_ALU64 | BPF_MOV | BPF_X: + + case BPF_ALU64 | BPF_ADD | BPF_X: + case BPF_ALU64 | BPF_ADD | BPF_K: + + case BPF_ALU64 | BPF_SUB | BPF_X: + case BPF_ALU64 | BPF_SUB | BPF_K: + + case BPF_ALU64 | BPF_AND | BPF_X: + case BPF_ALU64 | BPF_OR | BPF_X: + case BPF_ALU64 | BPF_XOR | BPF_X: + + case BPF_ALU64 | BPF_MUL | BPF_X: + case BPF_ALU64 | BPF_MUL | BPF_K: + + case BPF_ALU64 | BPF_DIV | BPF_X: + case BPF_ALU64 | BPF_DIV | BPF_K: + + case BPF_ALU64 | BPF_MOD | BPF_X: + case BPF_ALU64 | BPF_MOD | BPF_K: + + case BPF_ALU64 | BPF_LSH | BPF_X: + case BPF_ALU64 | BPF_RSH | BPF_X: + case BPF_ALU64 | BPF_ARSH | BPF_X: + if (BPF_SRC(code) == BPF_K) { + emit_imm32(tmp2, imm, ctx); + src = tmp2; + } + emit_alu_r64(dst, src, ctx, BPF_OP(code)); + break; + + /* dst = -dst */ + case BPF_ALU64 | BPF_NEG: + emit_alu_r64(dst, tmp2, ctx, BPF_OP(code)); + break; + + case BPF_ALU64 | BPF_MOV | BPF_K: + case BPF_ALU64 | BPF_AND | BPF_K: + case BPF_ALU64 | BPF_OR | BPF_K: + case BPF_ALU64 | BPF_XOR | BPF_K: + case BPF_ALU64 | BPF_LSH | BPF_K: + case BPF_ALU64 | BPF_RSH | BPF_K: + case BPF_ALU64 | BPF_ARSH | BPF_K: + emit_alu_i64(dst, imm, ctx, BPF_OP(code)); + break; + + case BPF_ALU | BPF_MOV | BPF_X: + if (imm == 1) { + /* Special mov32 for zext. */ + emit_zext64(dst, ctx); + break; + } + fallthrough; + /* dst = dst OP src */ + case BPF_ALU | BPF_ADD | BPF_X: + case BPF_ALU | BPF_SUB | BPF_X: + case BPF_ALU | BPF_AND | BPF_X: + case BPF_ALU | BPF_OR | BPF_X: + case BPF_ALU | BPF_XOR | BPF_X: + + case BPF_ALU | BPF_MUL | BPF_X: + case BPF_ALU | BPF_MUL | BPF_K: + + case BPF_ALU | BPF_DIV | BPF_X: + case BPF_ALU | BPF_DIV | BPF_K: + + case BPF_ALU | BPF_MOD | BPF_X: + case BPF_ALU | BPF_MOD | BPF_K: + + case BPF_ALU | BPF_LSH | BPF_X: + case BPF_ALU | BPF_RSH | BPF_X: + case BPF_ALU | BPF_ARSH | BPF_X: + if (BPF_SRC(code) == BPF_K) { + emit_imm32(tmp2, imm, ctx); + src = tmp2; + } + emit_alu_r32(dst, src, ctx, BPF_OP(code)); + break; + + /* dst = dst OP imm */ + case BPF_ALU | BPF_MOV | BPF_K: + case BPF_ALU | BPF_ADD | BPF_K: + case BPF_ALU | BPF_SUB | BPF_K: + case BPF_ALU | BPF_AND | BPF_K: + case BPF_ALU | BPF_OR | BPF_K: + case BPF_ALU | BPF_XOR | BPF_K: + case BPF_ALU | BPF_LSH | BPF_K: + case BPF_ALU | BPF_RSH | BPF_K: + case BPF_ALU | BPF_ARSH | BPF_K: + /* + * mul,div,mod are handled in the BPF_X case. + */ + emit_alu_i32(dst, imm, ctx, BPF_OP(code)); + break; + + /* dst = -dst */ + case BPF_ALU | BPF_NEG: + /* + * src is ignored---choose tmp2 as a dummy register since it + * is not on the stack. + */ + emit_alu_r32(dst, tmp2, ctx, BPF_OP(code)); + break; + + /* dst = BSWAP##imm(dst) */ + case BPF_ALU | BPF_END | BPF_FROM_BE: + { + const s8 *rd = bpf_get_reg64(dst, tmp1, ctx); + + switch (imm) { + case 16: + /* zero-extend 16 bits into 64 bits */ + emit(hppa_extru(lo(rd), 31, 16, lo(rd)), ctx); + fallthrough; + case 32: + /* zero-extend 32 bits into 64 bits */ + if (!ctx->prog->aux->verifier_zext) + emit_hppa_copy(HPPA_REG_ZERO, hi(rd), ctx); + break; + case 64: + /* Do nothing. */ + break; + default: + pr_err("bpf-jit: BPF_END imm %d invalid\n", imm); + return -1; + } + + bpf_put_reg64(dst, rd, ctx); + break; + } + + case BPF_ALU | BPF_END | BPF_FROM_LE: + { + const s8 *rd = bpf_get_reg64(dst, tmp1, ctx); + + switch (imm) { + case 16: + emit_rev16(lo(rd), ctx); + if (!ctx->prog->aux->verifier_zext) + emit_hppa_copy(HPPA_REG_ZERO, hi(rd), ctx); + break; + case 32: + emit_rev32(lo(rd), lo(rd), ctx); + if (!ctx->prog->aux->verifier_zext) + emit_hppa_copy(HPPA_REG_ZERO, hi(rd), ctx); + break; + case 64: + /* Swap upper and lower halves, then each half. */ + emit_hppa_copy(hi(rd), HPPA_REG_T0, ctx); + emit_rev32(lo(rd), hi(rd), ctx); + emit_rev32(HPPA_REG_T0, lo(rd), ctx); + break; + default: + pr_err("bpf-jit: BPF_END imm %d invalid\n", imm); + return -1; + } + + bpf_put_reg64(dst, rd, ctx); + break; + } + /* JUMP off */ + case BPF_JMP | BPF_JA: + paoff = hppa_offset(i, off, ctx); + emit_jump(paoff, false, ctx); + break; + /* function call */ + case BPF_JMP | BPF_CALL: + { + bool fixed; + int ret; + u64 addr; + + ret = bpf_jit_get_func_addr(ctx->prog, insn, extra_pass, &addr, + &fixed); + if (ret < 0) + return ret; + emit_call(fixed, addr, ctx); + break; + } + /* tail call */ + case BPF_JMP | BPF_TAIL_CALL: + REG_SET_SEEN_ALL(ctx); + if (emit_bpf_tail_call(i, ctx)) + return -1; + break; + /* IF (dst COND imm) JUMP off */ + case BPF_JMP | BPF_JEQ | BPF_X: + case BPF_JMP | BPF_JEQ | BPF_K: + case BPF_JMP32 | BPF_JEQ | BPF_X: + case BPF_JMP32 | BPF_JEQ | BPF_K: + + case BPF_JMP | BPF_JNE | BPF_X: + case BPF_JMP | BPF_JNE | BPF_K: + case BPF_JMP32 | BPF_JNE | BPF_X: + case BPF_JMP32 | BPF_JNE | BPF_K: + + case BPF_JMP | BPF_JLE | BPF_X: + case BPF_JMP | BPF_JLE | BPF_K: + case BPF_JMP32 | BPF_JLE | BPF_X: + case BPF_JMP32 | BPF_JLE | BPF_K: + + case BPF_JMP | BPF_JLT | BPF_X: + case BPF_JMP | BPF_JLT | BPF_K: + case BPF_JMP32 | BPF_JLT | BPF_X: + case BPF_JMP32 | BPF_JLT | BPF_K: + + case BPF_JMP | BPF_JGE | BPF_X: + case BPF_JMP | BPF_JGE | BPF_K: + case BPF_JMP32 | BPF_JGE | BPF_X: + case BPF_JMP32 | BPF_JGE | BPF_K: + + case BPF_JMP | BPF_JGT | BPF_X: + case BPF_JMP | BPF_JGT | BPF_K: + case BPF_JMP32 | BPF_JGT | BPF_X: + case BPF_JMP32 | BPF_JGT | BPF_K: + + case BPF_JMP | BPF_JSLE | BPF_X: + case BPF_JMP | BPF_JSLE | BPF_K: + case BPF_JMP32 | BPF_JSLE | BPF_X: + case BPF_JMP32 | BPF_JSLE | BPF_K: + + case BPF_JMP | BPF_JSLT | BPF_X: + case BPF_JMP | BPF_JSLT | BPF_K: + case BPF_JMP32 | BPF_JSLT | BPF_X: + case BPF_JMP32 | BPF_JSLT | BPF_K: + + case BPF_JMP | BPF_JSGE | BPF_X: + case BPF_JMP | BPF_JSGE | BPF_K: + case BPF_JMP32 | BPF_JSGE | BPF_X: + case BPF_JMP32 | BPF_JSGE | BPF_K: + + case BPF_JMP | BPF_JSGT | BPF_X: + case BPF_JMP | BPF_JSGT | BPF_K: + case BPF_JMP32 | BPF_JSGT | BPF_X: + case BPF_JMP32 | BPF_JSGT | BPF_K: + + case BPF_JMP | BPF_JSET | BPF_X: + case BPF_JMP | BPF_JSET | BPF_K: + case BPF_JMP32 | BPF_JSET | BPF_X: + case BPF_JMP32 | BPF_JSET | BPF_K: + paoff = hppa_offset(i, off, ctx); + if (BPF_SRC(code) == BPF_K) { + s = ctx->ninsns; + emit_imm32(tmp2, imm, ctx); + src = tmp2; + e = ctx->ninsns; + paoff -= (e - s); + } + if (is64) + emit_branch_r64(dst, src, paoff, ctx, BPF_OP(code)); + else + emit_branch_r32(dst, src, paoff, ctx, BPF_OP(code)); + break; + /* function return */ + case BPF_JMP | BPF_EXIT: + if (i == ctx->prog->len - 1) + break; + /* load epilogue function pointer and jump to it. */ + emit(EXIT_PTR_LOAD(HPPA_REG_RP), ctx); + emit(EXIT_PTR_JUMP(HPPA_REG_RP, NOP_NEXT_INSTR), ctx); + break; + + /* dst = imm64 */ + case BPF_LD | BPF_IMM | BPF_DW: + { + struct bpf_insn insn1 = insn[1]; + u32 upper = insn1.imm; + u32 lower = imm; + const s8 *rd = bpf_get_reg64_ref(dst, tmp1, false, ctx); + + if (0 && bpf_pseudo_func(insn)) { + WARN_ON(upper); /* we are 32-bit! */ + upper = 0; + lower = (uintptr_t) dereference_function_descriptor(lower); + } + + emit_imm64(rd, upper, lower, ctx); + bpf_put_reg64(dst, rd, ctx); + return 1; + } + + /* LDX: dst = *(size *)(src + off) */ + case BPF_LDX | BPF_MEM | BPF_B: + case BPF_LDX | BPF_MEM | BPF_H: + case BPF_LDX | BPF_MEM | BPF_W: + case BPF_LDX | BPF_MEM | BPF_DW: + if (emit_load_r64(dst, src, off, ctx, BPF_SIZE(code))) + return -1; + break; + + /* speculation barrier */ + case BPF_ST | BPF_NOSPEC: + break; + + /* ST: *(size *)(dst + off) = imm */ + case BPF_ST | BPF_MEM | BPF_B: + case BPF_ST | BPF_MEM | BPF_H: + case BPF_ST | BPF_MEM | BPF_W: + case BPF_ST | BPF_MEM | BPF_DW: + + case BPF_STX | BPF_MEM | BPF_B: + case BPF_STX | BPF_MEM | BPF_H: + case BPF_STX | BPF_MEM | BPF_W: + case BPF_STX | BPF_MEM | BPF_DW: + if (BPF_CLASS(code) == BPF_ST) { + emit_imm32(tmp2, imm, ctx); + src = tmp2; + } + + if (emit_store_r64(dst, src, off, ctx, BPF_SIZE(code), + BPF_MODE(code))) + return -1; + break; + + case BPF_STX | BPF_ATOMIC | BPF_W: + case BPF_STX | BPF_ATOMIC | BPF_DW: + pr_info_once( + "bpf-jit: not supported: atomic operation %02x ***\n", + insn->imm); + return -EFAULT; + + default: + pr_err("bpf-jit: unknown opcode %02x\n", code); + return -EINVAL; + } + + return 0; +} + +void bpf_jit_build_prologue(struct hppa_jit_context *ctx) +{ + const s8 *tmp = regmap[TMP_REG_1]; + const s8 *dst, *reg; + int stack_adjust = 0; + int i; + unsigned long addr; + int bpf_stack_adjust; + + /* + * stack on hppa grows up, so if tail calls are used we need to + * allocate the maximum stack size + */ + if (REG_ALL_SEEN(ctx)) + bpf_stack_adjust = MAX_BPF_STACK; + else + bpf_stack_adjust = ctx->prog->aux->stack_depth; + bpf_stack_adjust = round_up(bpf_stack_adjust, STACK_ALIGN); + + /* make space for callee-saved registers. */ + stack_adjust += NR_SAVED_REGISTERS * REG_SIZE; + /* make space for BPF registers on stack. */ + stack_adjust += BPF_JIT_SCRATCH_REGS * REG_SIZE; + /* make space for BPF stack. */ + stack_adjust += bpf_stack_adjust; + /* round up for stack alignment. */ + stack_adjust = round_up(stack_adjust, STACK_ALIGN); + + /* + * The first instruction sets the tail-call-counter (TCC) register. + * This instruction is skipped by tail calls. + * Use a temporary register instead of a caller-saved register initially. + */ + emit(hppa_ldi(MAX_TAIL_CALL_CNT, HPPA_REG_TCC_IN_INIT), ctx); + + /* + * skip all initializations when called as BPF TAIL call. + */ + emit(hppa_ldi(MAX_TAIL_CALL_CNT, HPPA_REG_R1), ctx); + emit(hppa_bne(HPPA_REG_TCC_IN_INIT, HPPA_REG_R1, ctx->prologue_len - 2 - HPPA_BRANCH_DISPLACEMENT), ctx); + + /* set up hppa stack frame. */ + emit_hppa_copy(HPPA_REG_SP, HPPA_REG_R1, ctx); // copy sp,r1 (=prev_sp) + emit(hppa_ldo(stack_adjust, HPPA_REG_SP, HPPA_REG_SP), ctx); // ldo stack_adjust(sp),sp (increase stack) + emit(hppa_stw(HPPA_REG_R1, -REG_SIZE, HPPA_REG_SP), ctx); // stw prev_sp,-0x04(sp) + emit(hppa_stw(HPPA_REG_RP, -0x14, HPPA_REG_SP), ctx); // stw rp,-0x14(sp) + + REG_FORCE_SEEN(ctx, HPPA_REG_T0); + REG_FORCE_SEEN(ctx, HPPA_REG_T1); + REG_FORCE_SEEN(ctx, HPPA_REG_T2); + REG_FORCE_SEEN(ctx, HPPA_REG_T3); + REG_FORCE_SEEN(ctx, HPPA_REG_T4); + REG_FORCE_SEEN(ctx, HPPA_REG_T5); + + /* save callee-save registers. */ + for (i = 3; i <= 18; i++) { + if (OPTIMIZE_HPPA && !REG_WAS_SEEN(ctx, HPPA_R(i))) + continue; + emit(hppa_stw(HPPA_R(i), -REG_SIZE * (8 + (i-3)), HPPA_REG_SP), ctx); // stw ri,-save_area(sp) + } + + /* + * now really set the tail call counter (TCC) register. + */ + if (REG_WAS_SEEN(ctx, HPPA_REG_TCC)) + emit(hppa_ldi(MAX_TAIL_CALL_CNT, HPPA_REG_TCC), ctx); + + /* + * save epilogue function pointer for outer TCC call chain. + * The main TCC call stores the final RP on stack. + */ + addr = (uintptr_t) &ctx->insns[ctx->epilogue_offset]; + /* skip first two instructions of exit function, which jump to exit */ + addr += 2 * HPPA_INSN_SIZE; + emit(hppa_ldil(addr, HPPA_REG_T2), ctx); + emit(hppa_ldo(im11(addr), HPPA_REG_T2, HPPA_REG_T2), ctx); + emit(EXIT_PTR_STORE(HPPA_REG_T2), ctx); + + /* load R1 & R2 from registers, R3-R5 from stack. */ + /* use HPPA_REG_R1 which holds the old stack value */ + dst = regmap[BPF_REG_5]; + reg = bpf_get_reg64_ref(dst, tmp, false, ctx); + if (REG_WAS_SEEN(ctx, lo(reg)) | REG_WAS_SEEN(ctx, hi(reg))) { + if (REG_WAS_SEEN(ctx, hi(reg))) + emit(hppa_ldw(-0x48, HPPA_REG_R1, hi(reg)), ctx); + if (REG_WAS_SEEN(ctx, lo(reg))) + emit(hppa_ldw(-0x44, HPPA_REG_R1, lo(reg)), ctx); + bpf_put_reg64(dst, tmp, ctx); + } + + dst = regmap[BPF_REG_4]; + reg = bpf_get_reg64_ref(dst, tmp, false, ctx); + if (REG_WAS_SEEN(ctx, lo(reg)) | REG_WAS_SEEN(ctx, hi(reg))) { + if (REG_WAS_SEEN(ctx, hi(reg))) + emit(hppa_ldw(-0x40, HPPA_REG_R1, hi(reg)), ctx); + if (REG_WAS_SEEN(ctx, lo(reg))) + emit(hppa_ldw(-0x3c, HPPA_REG_R1, lo(reg)), ctx); + bpf_put_reg64(dst, tmp, ctx); + } + + dst = regmap[BPF_REG_3]; + reg = bpf_get_reg64_ref(dst, tmp, false, ctx); + if (REG_WAS_SEEN(ctx, lo(reg)) | REG_WAS_SEEN(ctx, hi(reg))) { + if (REG_WAS_SEEN(ctx, hi(reg))) + emit(hppa_ldw(-0x38, HPPA_REG_R1, hi(reg)), ctx); + if (REG_WAS_SEEN(ctx, lo(reg))) + emit(hppa_ldw(-0x34, HPPA_REG_R1, lo(reg)), ctx); + bpf_put_reg64(dst, tmp, ctx); + } + + dst = regmap[BPF_REG_2]; + reg = bpf_get_reg64_ref(dst, tmp, false, ctx); + if (REG_WAS_SEEN(ctx, lo(reg)) | REG_WAS_SEEN(ctx, hi(reg))) { + if (REG_WAS_SEEN(ctx, hi(reg))) + emit_hppa_copy(HPPA_REG_ARG3, hi(reg), ctx); + if (REG_WAS_SEEN(ctx, lo(reg))) + emit_hppa_copy(HPPA_REG_ARG2, lo(reg), ctx); + bpf_put_reg64(dst, tmp, ctx); + } + + dst = regmap[BPF_REG_1]; + reg = bpf_get_reg64_ref(dst, tmp, false, ctx); + if (REG_WAS_SEEN(ctx, lo(reg)) | REG_WAS_SEEN(ctx, hi(reg))) { + if (REG_WAS_SEEN(ctx, hi(reg))) + emit_hppa_copy(HPPA_REG_ARG1, hi(reg), ctx); + if (REG_WAS_SEEN(ctx, lo(reg))) + emit_hppa_copy(HPPA_REG_ARG0, lo(reg), ctx); + bpf_put_reg64(dst, tmp, ctx); + } + + /* Set up BPF frame pointer. */ + dst = regmap[BPF_REG_FP]; + reg = bpf_get_reg64_ref(dst, tmp, false, ctx); + if (REG_WAS_SEEN(ctx, lo(reg)) | REG_WAS_SEEN(ctx, hi(reg))) { + if (REG_WAS_SEEN(ctx, lo(reg))) + emit(hppa_ldo(-REG_SIZE * (NR_SAVED_REGISTERS + BPF_JIT_SCRATCH_REGS), + HPPA_REG_SP, lo(reg)), ctx); + if (REG_WAS_SEEN(ctx, hi(reg))) + emit_hppa_copy(HPPA_REG_ZERO, hi(reg), ctx); + bpf_put_reg64(dst, tmp, ctx); + } + + emit(hppa_nop(), ctx); +} + +void bpf_jit_build_epilogue(struct hppa_jit_context *ctx) +{ + __build_epilogue(false, ctx); +} diff --git a/arch/parisc/net/bpf_jit_comp64.c b/arch/parisc/net/bpf_jit_comp64.c new file mode 100644 index 000000000000..54b0d5e25e02 --- /dev/null +++ b/arch/parisc/net/bpf_jit_comp64.c @@ -0,0 +1,1209 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * BPF JIT compiler for PA-RISC (64-bit) + * + * Copyright(c) 2023 Helge Deller <deller@gmx.de> + * + * The code is based on the BPF JIT compiler for RV64 by Björn Töpel. + * + * TODO: + * - check if bpf_jit_needs_zext() is needed (currently enabled) + * - implement arch_prepare_bpf_trampoline(), poke(), ... + */ + +#include <linux/bitfield.h> +#include <linux/bpf.h> +#include <linux/filter.h> +#include <linux/libgcc.h> +#include "bpf_jit.h" + +static const int regmap[] = { + [BPF_REG_0] = HPPA_REG_RET0, + [BPF_REG_1] = HPPA_R(5), + [BPF_REG_2] = HPPA_R(6), + [BPF_REG_3] = HPPA_R(7), + [BPF_REG_4] = HPPA_R(8), + [BPF_REG_5] = HPPA_R(9), + [BPF_REG_6] = HPPA_R(10), + [BPF_REG_7] = HPPA_R(11), + [BPF_REG_8] = HPPA_R(12), + [BPF_REG_9] = HPPA_R(13), + [BPF_REG_FP] = HPPA_R(14), + [BPF_REG_AX] = HPPA_R(15), +}; + +/* + * Stack layout during BPF program execution (note: stack grows up): + * + * high + * HPPA64 sp => +----------+ <= HPPA64 fp + * | saved sp | + * | saved rp | + * | ... | HPPA64 callee-saved registers + * | curr args| + * | local var| + * +----------+ <= (BPF FP) + * | | + * | ... | BPF program stack + * | | + * | ... | Function call stack + * | | + * +----------+ + * low + */ + +/* Offset from fp for BPF registers stored on stack. */ +#define STACK_ALIGN FRAME_SIZE + +#define EXIT_PTR_LOAD(reg) hppa64_ldd_im16(-FRAME_SIZE, HPPA_REG_SP, reg) +#define EXIT_PTR_STORE(reg) hppa64_std_im16(reg, -FRAME_SIZE, HPPA_REG_SP) +#define EXIT_PTR_JUMP(reg, nop) hppa_bv(HPPA_REG_ZERO, reg, nop) + +static u8 bpf_to_hppa_reg(int bpf_reg, struct hppa_jit_context *ctx) +{ + u8 reg = regmap[bpf_reg]; + + REG_SET_SEEN(ctx, reg); + return reg; +}; + +static void emit_hppa_copy(const s8 rs, const s8 rd, struct hppa_jit_context *ctx) +{ + REG_SET_SEEN(ctx, rd); + if (OPTIMIZE_HPPA && (rs == rd)) + return; + REG_SET_SEEN(ctx, rs); + emit(hppa_copy(rs, rd), ctx); +} + +static void emit_hppa64_depd(u8 src, u8 pos, u8 len, u8 target, bool no_zero, struct hppa_jit_context *ctx) +{ + int c; + + pos &= (BITS_PER_LONG - 1); + pos = 63 - pos; + len = 64 - len; + c = (len < 32) ? 0x4 : 0; + c |= (pos >= 32) ? 0x2 : 0; + c |= (no_zero) ? 0x1 : 0; + emit(hppa_t10_insn(0x3c, target, src, 0, c, pos & 0x1f, len & 0x1f), ctx); +} + +static void emit_hppa64_shld(u8 src, int num, u8 target, struct hppa_jit_context *ctx) +{ + emit_hppa64_depd(src, 63-num, 64-num, target, 0, ctx); +} + +static void emit_hppa64_extrd(u8 src, u8 pos, u8 len, u8 target, bool signed_op, struct hppa_jit_context *ctx) +{ + int c; + + pos &= (BITS_PER_LONG - 1); + len = 64 - len; + c = (len < 32) ? 0x4 : 0; + c |= (pos >= 32) ? 0x2 : 0; + c |= signed_op ? 0x1 : 0; + emit(hppa_t10_insn(0x36, src, target, 0, c, pos & 0x1f, len & 0x1f), ctx); +} + +static void emit_hppa64_extrw(u8 src, u8 pos, u8 len, u8 target, bool signed_op, struct hppa_jit_context *ctx) +{ + int c; + + pos &= (32 - 1); + len = 32 - len; + c = 0x06 | (signed_op ? 1 : 0); + emit(hppa_t10_insn(0x34, src, target, 0, c, pos, len), ctx); +} + +#define emit_hppa64_zext32(r, target, ctx) \ + emit_hppa64_extrd(r, 63, 32, target, false, ctx) +#define emit_hppa64_sext32(r, target, ctx) \ + emit_hppa64_extrd(r, 63, 32, target, true, ctx) + +static void emit_hppa64_shrd(u8 src, int num, u8 target, bool signed_op, struct hppa_jit_context *ctx) +{ + emit_hppa64_extrd(src, 63-num, 64-num, target, signed_op, ctx); +} + +static void emit_hppa64_shrw(u8 src, int num, u8 target, bool signed_op, struct hppa_jit_context *ctx) +{ + emit_hppa64_extrw(src, 31-num, 32-num, target, signed_op, ctx); +} + +/* Emit variable-length instructions for 32-bit imm */ +static void emit_imm32(u8 rd, s32 imm, struct hppa_jit_context *ctx) +{ + u32 lower = im11(imm); + + REG_SET_SEEN(ctx, rd); + if (OPTIMIZE_HPPA && relative_bits_ok(imm, 14)) { + emit(hppa_ldi(imm, rd), ctx); + return; + } + if (OPTIMIZE_HPPA && lower == imm) { + emit(hppa_ldo(lower, HPPA_REG_ZERO, rd), ctx); + return; + } + emit(hppa_ldil(imm, rd), ctx); + if (OPTIMIZE_HPPA && (lower == 0)) + return; + emit(hppa_ldo(lower, rd, rd), ctx); +} + +static bool is_32b_int(s64 val) +{ + return val == (s32) val; +} + +/* Emit variable-length instructions for 64-bit imm */ +static void emit_imm(u8 rd, s64 imm, u8 tmpreg, struct hppa_jit_context *ctx) +{ + u32 upper32; + + /* get lower 32-bits into rd, sign extended */ + emit_imm32(rd, imm, ctx); + + /* do we have upper 32-bits too ? */ + if (OPTIMIZE_HPPA && is_32b_int(imm)) + return; + + /* load upper 32-bits into lower tmpreg and deposit into rd */ + upper32 = imm >> 32; + if (upper32 || !OPTIMIZE_HPPA) { + emit_imm32(tmpreg, upper32, ctx); + emit_hppa64_depd(tmpreg, 31, 32, rd, 1, ctx); + } else + emit_hppa64_depd(HPPA_REG_ZERO, 31, 32, rd, 1, ctx); + +} + +static int emit_jump(signed long paoff, bool force_far, + struct hppa_jit_context *ctx) +{ + unsigned long pc, addr; + + /* Note: Use 2 instructions for jumps if force_far is set. */ + if (relative_bits_ok(paoff - HPPA_BRANCH_DISPLACEMENT, 22)) { + /* use BL,long branch followed by nop() */ + emit(hppa64_bl_long(paoff - HPPA_BRANCH_DISPLACEMENT), ctx); + if (force_far) + emit(hppa_nop(), ctx); + return 0; + } + + pc = (uintptr_t) &ctx->insns[ctx->ninsns]; + addr = pc + (paoff * HPPA_INSN_SIZE); + /* even the 64-bit kernel runs in memory below 4GB */ + if (WARN_ON_ONCE(addr >> 32)) + return -E2BIG; + emit(hppa_ldil(addr, HPPA_REG_R31), ctx); + emit(hppa_be_l(im11(addr) >> 2, HPPA_REG_R31, NOP_NEXT_INSTR), ctx); + return 0; +} + +static void __build_epilogue(bool is_tail_call, struct hppa_jit_context *ctx) +{ + int i; + + if (is_tail_call) { + /* + * goto *(t0 + 4); + * Skips first instruction of prologue which initializes tail + * call counter. Assumes t0 contains address of target program, + * see emit_bpf_tail_call. + */ + emit(hppa_ldo(1 * HPPA_INSN_SIZE, HPPA_REG_T0, HPPA_REG_T0), ctx); + emit(hppa_bv(HPPA_REG_ZERO, HPPA_REG_T0, EXEC_NEXT_INSTR), ctx); + /* in delay slot: */ + emit(hppa_copy(HPPA_REG_TCC, HPPA_REG_TCC_IN_INIT), ctx); + + return; + } + + /* load epilogue function pointer and jump to it. */ + /* exit point is either at next instruction, or the outest TCC exit function */ + emit(EXIT_PTR_LOAD(HPPA_REG_RP), ctx); + emit(EXIT_PTR_JUMP(HPPA_REG_RP, NOP_NEXT_INSTR), ctx); + + /* NOTE: we are 64-bit and big-endian, so return lower sign-extended 32-bit value */ + emit_hppa64_sext32(regmap[BPF_REG_0], HPPA_REG_RET0, ctx); + + /* Restore callee-saved registers. */ + for (i = 3; i <= 15; i++) { + if (OPTIMIZE_HPPA && !REG_WAS_SEEN(ctx, HPPA_R(i))) + continue; + emit(hppa64_ldd_im16(-REG_SIZE * i, HPPA_REG_SP, HPPA_R(i)), ctx); + } + + /* load original return pointer (stored by outest TCC function) */ + emit(hppa64_ldd_im16(-2*REG_SIZE, HPPA_REG_SP, HPPA_REG_RP), ctx); + emit(hppa_bv(HPPA_REG_ZERO, HPPA_REG_RP, EXEC_NEXT_INSTR), ctx); + /* in delay slot: */ + emit(hppa64_ldd_im5(-REG_SIZE, HPPA_REG_SP, HPPA_REG_SP), ctx); + + emit(hppa_nop(), ctx); // XXX WARUM einer zu wenig ?? +} + +static int emit_branch(u8 op, u8 rd, u8 rs, signed long paoff, + struct hppa_jit_context *ctx) +{ + int e, s; + bool far = false; + int off; + + if (op == BPF_JSET) { + /* + * BPF_JSET is a special case: it has no inverse so translate + * to and() function and compare against zero + */ + emit(hppa_and(rd, rs, HPPA_REG_T0), ctx); + paoff -= 1; /* reduce offset due to hppa_and() above */ + rd = HPPA_REG_T0; + rs = HPPA_REG_ZERO; + op = BPF_JNE; + } + + /* set start after BPF_JSET */ + s = ctx->ninsns; + + if (!relative_branch_ok(paoff - HPPA_BRANCH_DISPLACEMENT + 1, 12)) { + op = invert_bpf_cond(op); + far = true; + } + + /* + * For a far branch, the condition is negated and we jump over the + * branch itself, and the two instructions from emit_jump. + * For a near branch, just use paoff. + */ + off = far ? (2 - HPPA_BRANCH_DISPLACEMENT) : paoff - HPPA_BRANCH_DISPLACEMENT; + + switch (op) { + /* IF (dst COND src) JUMP off */ + case BPF_JEQ: + emit(hppa_beq(rd, rs, off), ctx); + break; + case BPF_JGT: + emit(hppa_bgtu(rd, rs, off), ctx); + break; + case BPF_JLT: + emit(hppa_bltu(rd, rs, off), ctx); + break; + case BPF_JGE: + emit(hppa_bgeu(rd, rs, off), ctx); + break; + case BPF_JLE: + emit(hppa_bleu(rd, rs, off), ctx); + break; + case BPF_JNE: + emit(hppa_bne(rd, rs, off), ctx); + break; + case BPF_JSGT: + emit(hppa_bgt(rd, rs, off), ctx); + break; + case BPF_JSLT: + emit(hppa_blt(rd, rs, off), ctx); + break; + case BPF_JSGE: + emit(hppa_bge(rd, rs, off), ctx); + break; + case BPF_JSLE: + emit(hppa_ble(rd, rs, off), ctx); + break; + default: + WARN_ON(1); + } + + if (far) { + int ret; + e = ctx->ninsns; + /* Adjust for extra insns. */ + paoff -= (e - s); + ret = emit_jump(paoff, true, ctx); + if (ret) + return ret; + } else { + /* + * always allocate 2 nops instead of the far branch to + * reduce translation loops + */ + emit(hppa_nop(), ctx); + emit(hppa_nop(), ctx); + } + return 0; +} + +static void emit_zext_32(u8 reg, struct hppa_jit_context *ctx) +{ + emit_hppa64_zext32(reg, reg, ctx); +} + +static void emit_bpf_tail_call(int insn, struct hppa_jit_context *ctx) +{ + /* + * R1 -> &ctx + * R2 -> &array + * R3 -> index + */ + int off; + const s8 arr_reg = regmap[BPF_REG_2]; + const s8 idx_reg = regmap[BPF_REG_3]; + struct bpf_array bpfa; + struct bpf_prog bpfp; + + /* if there is any tail call, we need to save & restore all registers */ + REG_SET_SEEN_ALL(ctx); + + /* get address of TCC main exit function for error case into rp */ + emit(EXIT_PTR_LOAD(HPPA_REG_RP), ctx); + + /* max_entries = array->map.max_entries; */ + off = offsetof(struct bpf_array, map.max_entries); + BUILD_BUG_ON(sizeof(bpfa.map.max_entries) != 4); + emit(hppa_ldw(off, arr_reg, HPPA_REG_T1), ctx); + + /* + * if (index >= max_entries) + * goto out; + */ + emit(hppa_bltu(idx_reg, HPPA_REG_T1, 2 - HPPA_BRANCH_DISPLACEMENT), ctx); + emit(EXIT_PTR_JUMP(HPPA_REG_RP, NOP_NEXT_INSTR), ctx); + + /* + * if (--tcc < 0) + * goto out; + */ + REG_FORCE_SEEN(ctx, HPPA_REG_TCC); + emit(hppa_ldo(-1, HPPA_REG_TCC, HPPA_REG_TCC), ctx); + emit(hppa_bge(HPPA_REG_TCC, HPPA_REG_ZERO, 2 - HPPA_BRANCH_DISPLACEMENT), ctx); + emit(EXIT_PTR_JUMP(HPPA_REG_RP, NOP_NEXT_INSTR), ctx); + + /* + * prog = array->ptrs[index]; + * if (!prog) + * goto out; + */ + BUILD_BUG_ON(sizeof(bpfa.ptrs[0]) != 8); + emit(hppa64_shladd(idx_reg, 3, arr_reg, HPPA_REG_T0), ctx); + off = offsetof(struct bpf_array, ptrs); + BUILD_BUG_ON(off < 16); + emit(hppa64_ldd_im16(off, HPPA_REG_T0, HPPA_REG_T0), ctx); + emit(hppa_bne(HPPA_REG_T0, HPPA_REG_ZERO, 2 - HPPA_BRANCH_DISPLACEMENT), ctx); + emit(EXIT_PTR_JUMP(HPPA_REG_RP, NOP_NEXT_INSTR), ctx); + + /* + * tcc = temp_tcc; + * goto *(prog->bpf_func + 4); + */ + off = offsetof(struct bpf_prog, bpf_func); + BUILD_BUG_ON(off < 16); + BUILD_BUG_ON(sizeof(bpfp.bpf_func) != 8); + emit(hppa64_ldd_im16(off, HPPA_REG_T0, HPPA_REG_T0), ctx); + /* Epilogue jumps to *(t0 + 4). */ + __build_epilogue(true, ctx); +} + +static void init_regs(u8 *rd, u8 *rs, const struct bpf_insn *insn, + struct hppa_jit_context *ctx) +{ + u8 code = insn->code; + + switch (code) { + case BPF_JMP | BPF_JA: + case BPF_JMP | BPF_CALL: + case BPF_JMP | BPF_EXIT: + case BPF_JMP | BPF_TAIL_CALL: + break; + default: + *rd = bpf_to_hppa_reg(insn->dst_reg, ctx); + } + + if (code & (BPF_ALU | BPF_X) || code & (BPF_ALU64 | BPF_X) || + code & (BPF_JMP | BPF_X) || code & (BPF_JMP32 | BPF_X) || + code & BPF_LDX || code & BPF_STX) + *rs = bpf_to_hppa_reg(insn->src_reg, ctx); +} + +static void emit_zext_32_rd_rs(u8 *rd, u8 *rs, struct hppa_jit_context *ctx) +{ + emit_hppa64_zext32(*rd, HPPA_REG_T2, ctx); + *rd = HPPA_REG_T2; + emit_hppa64_zext32(*rs, HPPA_REG_T1, ctx); + *rs = HPPA_REG_T1; +} + +static void emit_sext_32_rd_rs(u8 *rd, u8 *rs, struct hppa_jit_context *ctx) +{ + emit_hppa64_sext32(*rd, HPPA_REG_T2, ctx); + *rd = HPPA_REG_T2; + emit_hppa64_sext32(*rs, HPPA_REG_T1, ctx); + *rs = HPPA_REG_T1; +} + +static void emit_zext_32_rd_t1(u8 *rd, struct hppa_jit_context *ctx) +{ + emit_hppa64_zext32(*rd, HPPA_REG_T2, ctx); + *rd = HPPA_REG_T2; + emit_zext_32(HPPA_REG_T1, ctx); +} + +static void emit_sext_32_rd(u8 *rd, struct hppa_jit_context *ctx) +{ + emit_hppa64_sext32(*rd, HPPA_REG_T2, ctx); + *rd = HPPA_REG_T2; +} + +static bool is_signed_bpf_cond(u8 cond) +{ + return cond == BPF_JSGT || cond == BPF_JSLT || + cond == BPF_JSGE || cond == BPF_JSLE; +} + +static void emit_call(u64 addr, bool fixed, struct hppa_jit_context *ctx) +{ + const int offset_sp = 2*FRAME_SIZE; + + emit(hppa_ldo(offset_sp, HPPA_REG_SP, HPPA_REG_SP), ctx); + + emit_hppa_copy(regmap[BPF_REG_1], HPPA_REG_ARG0, ctx); + emit_hppa_copy(regmap[BPF_REG_2], HPPA_REG_ARG1, ctx); + emit_hppa_copy(regmap[BPF_REG_3], HPPA_REG_ARG2, ctx); + emit_hppa_copy(regmap[BPF_REG_4], HPPA_REG_ARG3, ctx); + emit_hppa_copy(regmap[BPF_REG_5], HPPA_REG_ARG4, ctx); + + /* Backup TCC. */ + REG_FORCE_SEEN(ctx, HPPA_REG_TCC_SAVED); + if (REG_WAS_SEEN(ctx, HPPA_REG_TCC)) + emit(hppa_copy(HPPA_REG_TCC, HPPA_REG_TCC_SAVED), ctx); + + /* + * Use ldil() to load absolute address. Don't use emit_imm as the + * number of emitted instructions should not depend on the value of + * addr. + */ + WARN_ON(addr >> 32); + /* load function address and gp from Elf64_Fdesc descriptor */ + emit(hppa_ldil(addr, HPPA_REG_R31), ctx); + emit(hppa_ldo(im11(addr), HPPA_REG_R31, HPPA_REG_R31), ctx); + emit(hppa64_ldd_im16(offsetof(struct elf64_fdesc, addr), + HPPA_REG_R31, HPPA_REG_RP), ctx); + emit(hppa64_bve_l_rp(HPPA_REG_RP), ctx); + emit(hppa64_ldd_im16(offsetof(struct elf64_fdesc, gp), + HPPA_REG_R31, HPPA_REG_GP), ctx); + + /* Restore TCC. */ + if (REG_WAS_SEEN(ctx, HPPA_REG_TCC)) + emit(hppa_copy(HPPA_REG_TCC_SAVED, HPPA_REG_TCC), ctx); + + emit(hppa_ldo(-offset_sp, HPPA_REG_SP, HPPA_REG_SP), ctx); + + /* Set return value. */ + emit_hppa_copy(HPPA_REG_RET0, regmap[BPF_REG_0], ctx); +} + +static void emit_call_libgcc_ll(void *func, const s8 arg0, + const s8 arg1, u8 opcode, struct hppa_jit_context *ctx) +{ + u64 func_addr; + + if (BPF_CLASS(opcode) == BPF_ALU) { + emit_hppa64_zext32(arg0, HPPA_REG_ARG0, ctx); + emit_hppa64_zext32(arg1, HPPA_REG_ARG1, ctx); + } else { + emit_hppa_copy(arg0, HPPA_REG_ARG0, ctx); + emit_hppa_copy(arg1, HPPA_REG_ARG1, ctx); + } + + /* libcgcc overwrites HPPA_REG_RET0, so keep copy in HPPA_REG_TCC_SAVED */ + if (arg0 != HPPA_REG_RET0) { + REG_SET_SEEN(ctx, HPPA_REG_TCC_SAVED); + emit(hppa_copy(HPPA_REG_RET0, HPPA_REG_TCC_SAVED), ctx); + } + + /* set up stack */ + emit(hppa_ldo(FRAME_SIZE, HPPA_REG_SP, HPPA_REG_SP), ctx); + + func_addr = (uintptr_t) func; + /* load function func_address and gp from Elf64_Fdesc descriptor */ + emit_imm(HPPA_REG_R31, func_addr, arg0, ctx); + emit(hppa64_ldd_im16(offsetof(struct elf64_fdesc, addr), + HPPA_REG_R31, HPPA_REG_RP), ctx); + /* skip the following bve_l instruction if divisor is 0. */ + if (BPF_OP(opcode) == BPF_DIV || BPF_OP(opcode) == BPF_MOD) { + if (BPF_OP(opcode) == BPF_DIV) + emit_hppa_copy(HPPA_REG_ZERO, HPPA_REG_RET0, ctx); + else { + emit_hppa_copy(HPPA_REG_ARG0, HPPA_REG_RET0, ctx); + } + emit(hppa_beq(HPPA_REG_ARG1, HPPA_REG_ZERO, 2 - HPPA_BRANCH_DISPLACEMENT), ctx); + } + emit(hppa64_bve_l_rp(HPPA_REG_RP), ctx); + emit(hppa64_ldd_im16(offsetof(struct elf64_fdesc, gp), + HPPA_REG_R31, HPPA_REG_GP), ctx); + + emit(hppa_ldo(-FRAME_SIZE, HPPA_REG_SP, HPPA_REG_SP), ctx); + + emit_hppa_copy(HPPA_REG_RET0, arg0, ctx); + + /* restore HPPA_REG_RET0 */ + if (arg0 != HPPA_REG_RET0) + emit(hppa_copy(HPPA_REG_TCC_SAVED, HPPA_REG_RET0), ctx); +} + +static void emit_store(const s8 rd, const s8 rs, s16 off, + struct hppa_jit_context *ctx, const u8 size, + const u8 mode) +{ + s8 dstreg; + + /* need to calculate address since offset does not fit in 14 bits? */ + if (relative_bits_ok(off, 14)) + dstreg = rd; + else { + /* need to use R1 here, since addil puts result into R1 */ + dstreg = HPPA_REG_R1; + emit(hppa_addil(off, rd), ctx); + off = im11(off); + } + + switch (size) { + case BPF_B: + emit(hppa_stb(rs, off, dstreg), ctx); + break; + case BPF_H: + emit(hppa_sth(rs, off, dstreg), ctx); + break; + case BPF_W: + emit(hppa_stw(rs, off, dstreg), ctx); + break; + case BPF_DW: + if (off & 7) { + emit(hppa_ldo(off, dstreg, HPPA_REG_R1), ctx); + emit(hppa64_std_im5(rs, 0, HPPA_REG_R1), ctx); + } else if (off >= -16 && off <= 15) + emit(hppa64_std_im5(rs, off, dstreg), ctx); + else + emit(hppa64_std_im16(rs, off, dstreg), ctx); + break; + } +} + +int bpf_jit_emit_insn(const struct bpf_insn *insn, struct hppa_jit_context *ctx, + bool extra_pass) +{ + bool is64 = BPF_CLASS(insn->code) == BPF_ALU64 || + BPF_CLASS(insn->code) == BPF_JMP; + int s, e, ret, i = insn - ctx->prog->insnsi; + s64 paoff; + struct bpf_prog_aux *aux = ctx->prog->aux; + u8 rd = -1, rs = -1, code = insn->code; + s16 off = insn->off; + s32 imm = insn->imm; + + init_regs(&rd, &rs, insn, ctx); + + switch (code) { + /* dst = src */ + case BPF_ALU | BPF_MOV | BPF_X: + case BPF_ALU64 | BPF_MOV | BPF_X: + if (imm == 1) { + /* Special mov32 for zext */ + emit_zext_32(rd, ctx); + break; + } + if (!is64 && !aux->verifier_zext) + emit_hppa64_zext32(rs, rd, ctx); + else + emit_hppa_copy(rs, rd, ctx); + break; + + /* dst = dst OP src */ + case BPF_ALU | BPF_ADD | BPF_X: + case BPF_ALU64 | BPF_ADD | BPF_X: + emit(hppa_add(rd, rs, rd), ctx); + if (!is64 && !aux->verifier_zext) + emit_zext_32(rd, ctx); + break; + case BPF_ALU | BPF_SUB | BPF_X: + case BPF_ALU64 | BPF_SUB | BPF_X: + emit(hppa_sub(rd, rs, rd), ctx); + if (!is64 && !aux->verifier_zext) + emit_zext_32(rd, ctx); + break; + case BPF_ALU | BPF_AND | BPF_X: + case BPF_ALU64 | BPF_AND | BPF_X: + emit(hppa_and(rd, rs, rd), ctx); + if (!is64 && !aux->verifier_zext) + emit_zext_32(rd, ctx); + break; + case BPF_ALU | BPF_OR | BPF_X: + case BPF_ALU64 | BPF_OR | BPF_X: + emit(hppa_or(rd, rs, rd), ctx); + if (!is64 && !aux->verifier_zext) + emit_zext_32(rd, ctx); + break; + case BPF_ALU | BPF_XOR | BPF_X: + case BPF_ALU64 | BPF_XOR | BPF_X: + emit(hppa_xor(rd, rs, rd), ctx); + if (!is64 && !aux->verifier_zext && rs != rd) + emit_zext_32(rd, ctx); + break; + case BPF_ALU | BPF_MUL | BPF_K: + case BPF_ALU64 | BPF_MUL | BPF_K: + emit_imm(HPPA_REG_T1, is64 ? (s64)(s32)imm : (u32)imm, HPPA_REG_T2, ctx); + rs = HPPA_REG_T1; + fallthrough; + case BPF_ALU | BPF_MUL | BPF_X: + case BPF_ALU64 | BPF_MUL | BPF_X: + emit_call_libgcc_ll(__muldi3, rd, rs, code, ctx); + if (!is64 && !aux->verifier_zext) + emit_zext_32(rd, ctx); + break; + case BPF_ALU | BPF_DIV | BPF_K: + case BPF_ALU64 | BPF_DIV | BPF_K: + emit_imm(HPPA_REG_T1, is64 ? (s64)(s32)imm : (u32)imm, HPPA_REG_T2, ctx); + rs = HPPA_REG_T1; + fallthrough; + case BPF_ALU | BPF_DIV | BPF_X: + case BPF_ALU64 | BPF_DIV | BPF_X: + emit_call_libgcc_ll(&hppa_div64, rd, rs, code, ctx); + if (!is64 && !aux->verifier_zext) + emit_zext_32(rd, ctx); + break; + case BPF_ALU | BPF_MOD | BPF_K: + case BPF_ALU64 | BPF_MOD | BPF_K: + emit_imm(HPPA_REG_T1, is64 ? (s64)(s32)imm : (u32)imm, HPPA_REG_T2, ctx); + rs = HPPA_REG_T1; + fallthrough; + case BPF_ALU | BPF_MOD | BPF_X: + case BPF_ALU64 | BPF_MOD | BPF_X: + emit_call_libgcc_ll(&hppa_div64_rem, rd, rs, code, ctx); + if (!is64 && !aux->verifier_zext) + emit_zext_32(rd, ctx); + break; + + case BPF_ALU | BPF_LSH | BPF_X: + case BPF_ALU64 | BPF_LSH | BPF_X: + emit_hppa64_sext32(rs, HPPA_REG_T0, ctx); + emit(hppa64_mtsarcm(HPPA_REG_T0), ctx); + if (is64) + emit(hppa64_depdz_sar(rd, rd), ctx); + else + emit(hppa_depwz_sar(rd, rd), ctx); + if (!is64 && !aux->verifier_zext) + emit_zext_32(rd, ctx); + break; + case BPF_ALU | BPF_RSH | BPF_X: + case BPF_ALU64 | BPF_RSH | BPF_X: + emit(hppa_mtsar(rs), ctx); + if (is64) + emit(hppa64_shrpd_sar(rd, rd), ctx); + else + emit(hppa_shrpw_sar(rd, rd), ctx); + if (!is64 && !aux->verifier_zext) + emit_zext_32(rd, ctx); + break; + case BPF_ALU | BPF_ARSH | BPF_X: + case BPF_ALU64 | BPF_ARSH | BPF_X: + emit_hppa64_sext32(rs, HPPA_REG_T0, ctx); + emit(hppa64_mtsarcm(HPPA_REG_T0), ctx); + if (is64) + emit(hppa_extrd_sar(rd, rd, 1), ctx); + else + emit(hppa_extrws_sar(rd, rd), ctx); + if (!is64 && !aux->verifier_zext) + emit_zext_32(rd, ctx); + break; + + /* dst = -dst */ + case BPF_ALU | BPF_NEG: + case BPF_ALU64 | BPF_NEG: + emit(hppa_sub(HPPA_REG_ZERO, rd, rd), ctx); + if (!is64 && !aux->verifier_zext) + emit_zext_32(rd, ctx); + break; + + /* dst = BSWAP##imm(dst) */ + case BPF_ALU | BPF_END | BPF_FROM_BE: + switch (imm) { + case 16: + /* zero-extend 16 bits into 64 bits */ + emit_hppa64_depd(HPPA_REG_ZERO, 63-16, 64-16, rd, 1, ctx); + break; + case 32: + if (!aux->verifier_zext) + emit_zext_32(rd, ctx); + break; + case 64: + /* Do nothing */ + break; + } + break; + + case BPF_ALU | BPF_END | BPF_FROM_LE: + switch (imm) { + case 16: + emit(hppa_extru(rd, 31 - 8, 8, HPPA_REG_T1), ctx); + emit(hppa_depwz(rd, 23, 8, HPPA_REG_T1), ctx); + emit(hppa_extru(HPPA_REG_T1, 31, 16, rd), ctx); + emit_hppa64_extrd(HPPA_REG_T1, 63, 16, rd, 0, ctx); + break; + case 32: + emit(hppa_shrpw(rd, rd, 16, HPPA_REG_T1), ctx); + emit_hppa64_depd(HPPA_REG_T1, 63-16, 8, HPPA_REG_T1, 1, ctx); + emit(hppa_shrpw(rd, HPPA_REG_T1, 8, HPPA_REG_T1), ctx); + emit_hppa64_extrd(HPPA_REG_T1, 63, 32, rd, 0, ctx); + break; + case 64: + emit(hppa64_permh_3210(rd, HPPA_REG_T1), ctx); + emit(hppa64_hshl(HPPA_REG_T1, 8, HPPA_REG_T2), ctx); + emit(hppa64_hshr_u(HPPA_REG_T1, 8, HPPA_REG_T1), ctx); + emit(hppa_or(HPPA_REG_T2, HPPA_REG_T1, rd), ctx); + break; + default: + pr_err("bpf-jit: BPF_END imm %d invalid\n", imm); + return -1; + } + break; + + /* dst = imm */ + case BPF_ALU | BPF_MOV | BPF_K: + case BPF_ALU64 | BPF_MOV | BPF_K: + emit_imm(rd, imm, HPPA_REG_T2, ctx); + if (!is64 && !aux->verifier_zext) + emit_zext_32(rd, ctx); + break; + + /* dst = dst OP imm */ + case BPF_ALU | BPF_ADD | BPF_K: + case BPF_ALU64 | BPF_ADD | BPF_K: + if (relative_bits_ok(imm, 14)) { + emit(hppa_ldo(imm, rd, rd), ctx); + } else { + emit_imm(HPPA_REG_T1, imm, HPPA_REG_T2, ctx); + emit(hppa_add(rd, HPPA_REG_T1, rd), ctx); + } + if (!is64 && !aux->verifier_zext) + emit_zext_32(rd, ctx); + break; + case BPF_ALU | BPF_SUB | BPF_K: + case BPF_ALU64 | BPF_SUB | BPF_K: + if (relative_bits_ok(-imm, 14)) { + emit(hppa_ldo(-imm, rd, rd), ctx); + } else { + emit_imm(HPPA_REG_T1, imm, HPPA_REG_T2, ctx); + emit(hppa_sub(rd, HPPA_REG_T1, rd), ctx); + } + if (!is64 && !aux->verifier_zext) + emit_zext_32(rd, ctx); + break; + case BPF_ALU | BPF_AND | BPF_K: + case BPF_ALU64 | BPF_AND | BPF_K: + emit_imm(HPPA_REG_T1, imm, HPPA_REG_T2, ctx); + emit(hppa_and(rd, HPPA_REG_T1, rd), ctx); + if (!is64 && !aux->verifier_zext) + emit_zext_32(rd, ctx); + break; + case BPF_ALU | BPF_OR | BPF_K: + case BPF_ALU64 | BPF_OR | BPF_K: + emit_imm(HPPA_REG_T1, imm, HPPA_REG_T2, ctx); + emit(hppa_or(rd, HPPA_REG_T1, rd), ctx); + if (!is64 && !aux->verifier_zext) + emit_zext_32(rd, ctx); + break; + case BPF_ALU | BPF_XOR | BPF_K: + case BPF_ALU64 | BPF_XOR | BPF_K: + emit_imm(HPPA_REG_T1, imm, HPPA_REG_T2, ctx); + emit(hppa_xor(rd, HPPA_REG_T1, rd), ctx); + if (!is64 && !aux->verifier_zext) + emit_zext_32(rd, ctx); + break; + case BPF_ALU | BPF_LSH | BPF_K: + case BPF_ALU64 | BPF_LSH | BPF_K: + if (imm != 0) { + emit_hppa64_shld(rd, imm, rd, ctx); + } + + if (!is64 && !aux->verifier_zext) + emit_zext_32(rd, ctx); + break; + case BPF_ALU | BPF_RSH | BPF_K: + case BPF_ALU64 | BPF_RSH | BPF_K: + if (imm != 0) { + if (is64) + emit_hppa64_shrd(rd, imm, rd, false, ctx); + else + emit_hppa64_shrw(rd, imm, rd, false, ctx); + } + + if (!is64 && !aux->verifier_zext) + emit_zext_32(rd, ctx); + break; + case BPF_ALU | BPF_ARSH | BPF_K: + case BPF_ALU64 | BPF_ARSH | BPF_K: + if (imm != 0) { + if (is64) + emit_hppa64_shrd(rd, imm, rd, true, ctx); + else + emit_hppa64_shrw(rd, imm, rd, true, ctx); + } + + if (!is64 && !aux->verifier_zext) + emit_zext_32(rd, ctx); + break; + + /* JUMP off */ + case BPF_JMP | BPF_JA: + paoff = hppa_offset(i, off, ctx); + ret = emit_jump(paoff, false, ctx); + if (ret) + return ret; + break; + + /* IF (dst COND src) JUMP off */ + case BPF_JMP | BPF_JEQ | BPF_X: + case BPF_JMP32 | BPF_JEQ | BPF_X: + case BPF_JMP | BPF_JGT | BPF_X: + case BPF_JMP32 | BPF_JGT | BPF_X: + case BPF_JMP | BPF_JLT | BPF_X: + case BPF_JMP32 | BPF_JLT | BPF_X: + case BPF_JMP | BPF_JGE | BPF_X: + case BPF_JMP32 | BPF_JGE | BPF_X: + case BPF_JMP | BPF_JLE | BPF_X: + case BPF_JMP32 | BPF_JLE | BPF_X: + case BPF_JMP | BPF_JNE | BPF_X: + case BPF_JMP32 | BPF_JNE | BPF_X: + case BPF_JMP | BPF_JSGT | BPF_X: + case BPF_JMP32 | BPF_JSGT | BPF_X: + case BPF_JMP | BPF_JSLT | BPF_X: + case BPF_JMP32 | BPF_JSLT | BPF_X: + case BPF_JMP | BPF_JSGE | BPF_X: + case BPF_JMP32 | BPF_JSGE | BPF_X: + case BPF_JMP | BPF_JSLE | BPF_X: + case BPF_JMP32 | BPF_JSLE | BPF_X: + case BPF_JMP | BPF_JSET | BPF_X: + case BPF_JMP32 | BPF_JSET | BPF_X: + paoff = hppa_offset(i, off, ctx); + if (!is64) { + s = ctx->ninsns; + if (is_signed_bpf_cond(BPF_OP(code))) + emit_sext_32_rd_rs(&rd, &rs, ctx); + else + emit_zext_32_rd_rs(&rd, &rs, ctx); + e = ctx->ninsns; + + /* Adjust for extra insns */ + paoff -= (e - s); + } + if (BPF_OP(code) == BPF_JSET) { + /* Adjust for and */ + paoff -= 1; + emit(hppa_and(rs, rd, HPPA_REG_T1), ctx); + emit_branch(BPF_JNE, HPPA_REG_T1, HPPA_REG_ZERO, paoff, + ctx); + } else { + emit_branch(BPF_OP(code), rd, rs, paoff, ctx); + } + break; + + /* IF (dst COND imm) JUMP off */ + case BPF_JMP | BPF_JEQ | BPF_K: + case BPF_JMP32 | BPF_JEQ | BPF_K: + case BPF_JMP | BPF_JGT | BPF_K: + case BPF_JMP32 | BPF_JGT | BPF_K: + case BPF_JMP | BPF_JLT | BPF_K: + case BPF_JMP32 | BPF_JLT | BPF_K: + case BPF_JMP | BPF_JGE | BPF_K: + case BPF_JMP32 | BPF_JGE | BPF_K: + case BPF_JMP | BPF_JLE | BPF_K: + case BPF_JMP32 | BPF_JLE | BPF_K: + case BPF_JMP | BPF_JNE | BPF_K: + case BPF_JMP32 | BPF_JNE | BPF_K: + case BPF_JMP | BPF_JSGT | BPF_K: + case BPF_JMP32 | BPF_JSGT | BPF_K: + case BPF_JMP | BPF_JSLT | BPF_K: + case BPF_JMP32 | BPF_JSLT | BPF_K: + case BPF_JMP | BPF_JSGE | BPF_K: + case BPF_JMP32 | BPF_JSGE | BPF_K: + case BPF_JMP | BPF_JSLE | BPF_K: + case BPF_JMP32 | BPF_JSLE | BPF_K: + paoff = hppa_offset(i, off, ctx); + s = ctx->ninsns; + if (imm) { + emit_imm(HPPA_REG_T1, imm, HPPA_REG_T2, ctx); + rs = HPPA_REG_T1; + } else { + rs = HPPA_REG_ZERO; + } + if (!is64) { + if (is_signed_bpf_cond(BPF_OP(code))) + emit_sext_32_rd(&rd, ctx); + else + emit_zext_32_rd_t1(&rd, ctx); + } + e = ctx->ninsns; + + /* Adjust for extra insns */ + paoff -= (e - s); + emit_branch(BPF_OP(code), rd, rs, paoff, ctx); + break; + case BPF_JMP | BPF_JSET | BPF_K: + case BPF_JMP32 | BPF_JSET | BPF_K: + paoff = hppa_offset(i, off, ctx); + s = ctx->ninsns; + emit_imm(HPPA_REG_T1, imm, HPPA_REG_T2, ctx); + emit(hppa_and(HPPA_REG_T1, rd, HPPA_REG_T1), ctx); + /* For jset32, we should clear the upper 32 bits of t1, but + * sign-extension is sufficient here and saves one instruction, + * as t1 is used only in comparison against zero. + */ + if (!is64 && imm < 0) + emit_hppa64_sext32(HPPA_REG_T1, HPPA_REG_T1, ctx); + e = ctx->ninsns; + paoff -= (e - s); + emit_branch(BPF_JNE, HPPA_REG_T1, HPPA_REG_ZERO, paoff, ctx); + break; + /* function call */ + case BPF_JMP | BPF_CALL: + { + bool fixed_addr; + u64 addr; + + ret = bpf_jit_get_func_addr(ctx->prog, insn, extra_pass, + &addr, &fixed_addr); + if (ret < 0) + return ret; + + REG_SET_SEEN_ALL(ctx); + emit_call(addr, fixed_addr, ctx); + break; + } + /* tail call */ + case BPF_JMP | BPF_TAIL_CALL: + emit_bpf_tail_call(i, ctx); + break; + + /* function return */ + case BPF_JMP | BPF_EXIT: + if (i == ctx->prog->len - 1) + break; + + paoff = epilogue_offset(ctx); + ret = emit_jump(paoff, false, ctx); + if (ret) + return ret; + break; + + /* dst = imm64 */ + case BPF_LD | BPF_IMM | BPF_DW: + { + struct bpf_insn insn1 = insn[1]; + u64 imm64 = (u64)insn1.imm << 32 | (u32)imm; + if (bpf_pseudo_func(insn)) + imm64 = (uintptr_t)dereference_function_descriptor((void*)imm64); + emit_imm(rd, imm64, HPPA_REG_T2, ctx); + + return 1; + } + + /* LDX: dst = *(size *)(src + off) */ + case BPF_LDX | BPF_MEM | BPF_B: + case BPF_LDX | BPF_MEM | BPF_H: + case BPF_LDX | BPF_MEM | BPF_W: + case BPF_LDX | BPF_MEM | BPF_DW: + case BPF_LDX | BPF_PROBE_MEM | BPF_B: + case BPF_LDX | BPF_PROBE_MEM | BPF_H: + case BPF_LDX | BPF_PROBE_MEM | BPF_W: + case BPF_LDX | BPF_PROBE_MEM | BPF_DW: + { + u8 srcreg; + + /* need to calculate address since offset does not fit in 14 bits? */ + if (relative_bits_ok(off, 14)) + srcreg = rs; + else { + /* need to use R1 here, since addil puts result into R1 */ + srcreg = HPPA_REG_R1; + BUG_ON(rs == HPPA_REG_R1); + BUG_ON(rd == HPPA_REG_R1); + emit(hppa_addil(off, rs), ctx); + off = im11(off); + } + + switch (BPF_SIZE(code)) { + case BPF_B: + emit(hppa_ldb(off, srcreg, rd), ctx); + if (insn_is_zext(&insn[1])) + return 1; + break; + case BPF_H: + emit(hppa_ldh(off, srcreg, rd), ctx); + if (insn_is_zext(&insn[1])) + return 1; + break; + case BPF_W: + emit(hppa_ldw(off, srcreg, rd), ctx); + if (insn_is_zext(&insn[1])) + return 1; + break; + case BPF_DW: + if (off & 7) { + emit(hppa_ldo(off, srcreg, HPPA_REG_R1), ctx); + emit(hppa64_ldd_reg(HPPA_REG_ZERO, HPPA_REG_R1, rd), ctx); + } else if (off >= -16 && off <= 15) + emit(hppa64_ldd_im5(off, srcreg, rd), ctx); + else + emit(hppa64_ldd_im16(off, srcreg, rd), ctx); + break; + } + break; + } + /* speculation barrier */ + case BPF_ST | BPF_NOSPEC: + break; + + /* ST: *(size *)(dst + off) = imm */ + /* STX: *(size *)(dst + off) = src */ + case BPF_ST | BPF_MEM | BPF_B: + case BPF_ST | BPF_MEM | BPF_H: + case BPF_ST | BPF_MEM | BPF_W: + case BPF_ST | BPF_MEM | BPF_DW: + + case BPF_STX | BPF_MEM | BPF_B: + case BPF_STX | BPF_MEM | BPF_H: + case BPF_STX | BPF_MEM | BPF_W: + case BPF_STX | BPF_MEM | BPF_DW: + if (BPF_CLASS(code) == BPF_ST) { + emit_imm(HPPA_REG_T2, imm, HPPA_REG_T1, ctx); + rs = HPPA_REG_T2; + } + + emit_store(rd, rs, off, ctx, BPF_SIZE(code), BPF_MODE(code)); + break; + + case BPF_STX | BPF_ATOMIC | BPF_W: + case BPF_STX | BPF_ATOMIC | BPF_DW: + pr_info_once( + "bpf-jit: not supported: atomic operation %02x ***\n", + insn->imm); + return -EFAULT; + + default: + pr_err("bpf-jit: unknown opcode %02x\n", code); + return -EINVAL; + } + + return 0; +} + +void bpf_jit_build_prologue(struct hppa_jit_context *ctx) +{ + int bpf_stack_adjust, stack_adjust, i; + unsigned long addr; + s8 reg; + + /* + * stack on hppa grows up, so if tail calls are used we need to + * allocate the maximum stack size + */ + if (REG_ALL_SEEN(ctx)) + bpf_stack_adjust = MAX_BPF_STACK; + else + bpf_stack_adjust = ctx->prog->aux->stack_depth; + bpf_stack_adjust = round_up(bpf_stack_adjust, STACK_ALIGN); + + stack_adjust = FRAME_SIZE + bpf_stack_adjust; + stack_adjust = round_up(stack_adjust, STACK_ALIGN); + + /* + * NOTE: We construct an Elf64_Fdesc descriptor here. + * The first 4 words initialize the TCC and compares them. + * Then follows the virtual address of the eBPF function, + * and the gp for this function. + * + * The first instruction sets the tail-call-counter (TCC) register. + * This instruction is skipped by tail calls. + * Use a temporary register instead of a caller-saved register initially. + */ + REG_FORCE_SEEN(ctx, HPPA_REG_TCC_IN_INIT); + emit(hppa_ldi(MAX_TAIL_CALL_CNT, HPPA_REG_TCC_IN_INIT), ctx); + + /* + * Skip all initializations when called as BPF TAIL call. + */ + emit(hppa_ldi(MAX_TAIL_CALL_CNT, HPPA_REG_R1), ctx); + emit(hppa_beq(HPPA_REG_TCC_IN_INIT, HPPA_REG_R1, 6 - HPPA_BRANCH_DISPLACEMENT), ctx); + emit(hppa64_bl_long(ctx->prologue_len - 3 - HPPA_BRANCH_DISPLACEMENT), ctx); + + /* store entry address of this eBPF function */ + addr = (uintptr_t) &ctx->insns[0]; + emit(addr >> 32, ctx); + emit(addr & 0xffffffff, ctx); + + /* store gp of this eBPF function */ + asm("copy %%r27,%0" : "=r" (addr) ); + emit(addr >> 32, ctx); + emit(addr & 0xffffffff, ctx); + + /* Set up hppa stack frame. */ + emit_hppa_copy(HPPA_REG_SP, HPPA_REG_R1, ctx); + emit(hppa_ldo(stack_adjust, HPPA_REG_SP, HPPA_REG_SP), ctx); + emit(hppa64_std_im5 (HPPA_REG_R1, -REG_SIZE, HPPA_REG_SP), ctx); + emit(hppa64_std_im16(HPPA_REG_RP, -2*REG_SIZE, HPPA_REG_SP), ctx); + + /* Save callee-save registers. */ + for (i = 3; i <= 15; i++) { + if (OPTIMIZE_HPPA && !REG_WAS_SEEN(ctx, HPPA_R(i))) + continue; + emit(hppa64_std_im16(HPPA_R(i), -REG_SIZE * i, HPPA_REG_SP), ctx); + } + + /* load function parameters; load all if we use tail functions */ + #define LOAD_PARAM(arg, dst) \ + if (REG_WAS_SEEN(ctx, regmap[dst]) || \ + REG_WAS_SEEN(ctx, HPPA_REG_TCC)) \ + emit_hppa_copy(arg, regmap[dst], ctx) + LOAD_PARAM(HPPA_REG_ARG0, BPF_REG_1); + LOAD_PARAM(HPPA_REG_ARG1, BPF_REG_2); + LOAD_PARAM(HPPA_REG_ARG2, BPF_REG_3); + LOAD_PARAM(HPPA_REG_ARG3, BPF_REG_4); + LOAD_PARAM(HPPA_REG_ARG4, BPF_REG_5); + #undef LOAD_PARAM + + REG_FORCE_SEEN(ctx, HPPA_REG_T0); + REG_FORCE_SEEN(ctx, HPPA_REG_T1); + REG_FORCE_SEEN(ctx, HPPA_REG_T2); + + /* + * Now really set the tail call counter (TCC) register. + */ + if (REG_WAS_SEEN(ctx, HPPA_REG_TCC)) + emit(hppa_ldi(MAX_TAIL_CALL_CNT, HPPA_REG_TCC), ctx); + + /* + * Save epilogue function pointer for outer TCC call chain. + * The main TCC call stores the final RP on stack. + */ + addr = (uintptr_t) &ctx->insns[ctx->epilogue_offset]; + /* skip first two instructions which jump to exit */ + addr += 2 * HPPA_INSN_SIZE; + emit_imm(HPPA_REG_T2, addr, HPPA_REG_T1, ctx); + emit(EXIT_PTR_STORE(HPPA_REG_T2), ctx); + + /* Set up BPF frame pointer. */ + reg = regmap[BPF_REG_FP]; /* -> HPPA_REG_FP */ + if (REG_WAS_SEEN(ctx, reg)) { + emit(hppa_ldo(-FRAME_SIZE, HPPA_REG_SP, reg), ctx); + } +} + +void bpf_jit_build_epilogue(struct hppa_jit_context *ctx) +{ + __build_epilogue(false, ctx); +} + +bool bpf_jit_supports_kfunc_call(void) +{ + return true; +} diff --git a/arch/parisc/net/bpf_jit_core.c b/arch/parisc/net/bpf_jit_core.c new file mode 100644 index 000000000000..06cbcd6fe87b --- /dev/null +++ b/arch/parisc/net/bpf_jit_core.c @@ -0,0 +1,207 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Common functionality for HPPA32 and HPPA64 BPF JIT compilers + * + * Copyright (c) 2023 Helge Deller <deller@gmx.de> + * + */ + +#include <linux/bpf.h> +#include <linux/filter.h> +#include "bpf_jit.h" + +/* Number of iterations to try until offsets converge. */ +#define NR_JIT_ITERATIONS 35 + +static int build_body(struct hppa_jit_context *ctx, bool extra_pass, int *offset) +{ + const struct bpf_prog *prog = ctx->prog; + int i; + + ctx->reg_seen_collect = true; + for (i = 0; i < prog->len; i++) { + const struct bpf_insn *insn = &prog->insnsi[i]; + int ret; + + ret = bpf_jit_emit_insn(insn, ctx, extra_pass); + /* BPF_LD | BPF_IMM | BPF_DW: skip the next instruction. */ + if (ret > 0) + i++; + if (offset) + offset[i] = ctx->ninsns; + if (ret < 0) + return ret; + } + ctx->reg_seen_collect = false; + return 0; +} + +bool bpf_jit_needs_zext(void) +{ + return true; +} + +struct bpf_prog *bpf_int_jit_compile(struct bpf_prog *prog) +{ + unsigned int prog_size = 0, extable_size = 0; + bool tmp_blinded = false, extra_pass = false; + struct bpf_prog *tmp, *orig_prog = prog; + int pass = 0, prev_ninsns = 0, prologue_len, i; + struct hppa_jit_data *jit_data; + struct hppa_jit_context *ctx; + + if (!prog->jit_requested) + return orig_prog; + + tmp = bpf_jit_blind_constants(prog); + if (IS_ERR(tmp)) + return orig_prog; + if (tmp != prog) { + tmp_blinded = true; + prog = tmp; + } + + jit_data = prog->aux->jit_data; + if (!jit_data) { + jit_data = kzalloc(sizeof(*jit_data), GFP_KERNEL); + if (!jit_data) { + prog = orig_prog; + goto out; + } + prog->aux->jit_data = jit_data; + } + + ctx = &jit_data->ctx; + + if (ctx->offset) { + extra_pass = true; + prog_size = sizeof(*ctx->insns) * ctx->ninsns; + goto skip_init_ctx; + } + + ctx->prog = prog; + ctx->offset = kcalloc(prog->len, sizeof(int), GFP_KERNEL); + if (!ctx->offset) { + prog = orig_prog; + goto out_offset; + } + for (i = 0; i < prog->len; i++) { + prev_ninsns += 20; + ctx->offset[i] = prev_ninsns; + } + + for (i = 0; i < NR_JIT_ITERATIONS; i++) { + pass++; + ctx->ninsns = 0; + if (build_body(ctx, extra_pass, ctx->offset)) { + prog = orig_prog; + goto out_offset; + } + ctx->body_len = ctx->ninsns; + bpf_jit_build_prologue(ctx); + ctx->prologue_len = ctx->ninsns - ctx->body_len; + ctx->epilogue_offset = ctx->ninsns; + bpf_jit_build_epilogue(ctx); + + if (ctx->ninsns == prev_ninsns) { + if (jit_data->header) + break; + /* obtain the actual image size */ + extable_size = prog->aux->num_exentries * + sizeof(struct exception_table_entry); + prog_size = sizeof(*ctx->insns) * ctx->ninsns; + + jit_data->header = + bpf_jit_binary_alloc(prog_size + extable_size, + &jit_data->image, + sizeof(long), + bpf_fill_ill_insns); + if (!jit_data->header) { + prog = orig_prog; + goto out_offset; + } + + ctx->insns = (u32 *)jit_data->image; + /* + * Now, when the image is allocated, the image can + * potentially shrink more (auipc/jalr -> jal). + */ + } + prev_ninsns = ctx->ninsns; + } + + if (i == NR_JIT_ITERATIONS) { + pr_err("bpf-jit: image did not converge in <%d passes!\n", i); + if (jit_data->header) + bpf_jit_binary_free(jit_data->header); + prog = orig_prog; + goto out_offset; + } + + if (extable_size) + prog->aux->extable = (void *)ctx->insns + prog_size; + +skip_init_ctx: + pass++; + ctx->ninsns = 0; + + bpf_jit_build_prologue(ctx); + if (build_body(ctx, extra_pass, NULL)) { + bpf_jit_binary_free(jit_data->header); + prog = orig_prog; + goto out_offset; + } + bpf_jit_build_epilogue(ctx); + + if (HPPA_JIT_DEBUG || bpf_jit_enable > 1) { + if (HPPA_JIT_DUMP) + bpf_jit_dump(prog->len, prog_size, pass, ctx->insns); + if (HPPA_JIT_REBOOT) + { extern int machine_restart(char *); machine_restart(""); } + } + + prog->bpf_func = (void *)ctx->insns; + prog->jited = 1; + prog->jited_len = prog_size; + + bpf_flush_icache(jit_data->header, ctx->insns + ctx->ninsns); + + if (!prog->is_func || extra_pass) { + if (bpf_jit_binary_lock_ro(jit_data->header)) { + bpf_jit_binary_free(jit_data->header); + prog->bpf_func = NULL; + prog->jited = 0; + prog->jited_len = 0; + goto out_offset; + } + prologue_len = ctx->epilogue_offset - ctx->body_len; + for (i = 0; i < prog->len; i++) + ctx->offset[i] += prologue_len; + bpf_prog_fill_jited_linfo(prog, ctx->offset); +out_offset: + kfree(ctx->offset); + kfree(jit_data); + prog->aux->jit_data = NULL; + } +out: + if (HPPA_JIT_REBOOT) + { extern int machine_restart(char *); machine_restart(""); } + + if (tmp_blinded) + bpf_jit_prog_release_other(prog, prog == orig_prog ? + tmp : orig_prog); + return prog; +} + +u64 hppa_div64(u64 div, u64 divisor) +{ + div = div64_u64(div, divisor); + return div; +} + +u64 hppa_div64_rem(u64 div, u64 divisor) +{ + u64 rem; + div64_u64_rem(div, divisor, &rem); + return rem; +} diff --git a/arch/parisc/video/Makefile b/arch/parisc/video/Makefile new file mode 100644 index 000000000000..b5db5b42880f --- /dev/null +++ b/arch/parisc/video/Makefile @@ -0,0 +1,3 @@ +# SPDX-License-Identifier: GPL-2.0-only + +obj-$(CONFIG_STI_CORE) += video-sti.o diff --git a/arch/parisc/video/video-sti.c b/arch/parisc/video/video-sti.c new file mode 100644 index 000000000000..564661e87093 --- /dev/null +++ b/arch/parisc/video/video-sti.c @@ -0,0 +1,27 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Copyright (C) 2000 Philipp Rumpf <prumpf@tux.org> + * Copyright (C) 2001-2020 Helge Deller <deller@gmx.de> + * Copyright (C) 2001-2002 Thomas Bogendoerfer <tsbogend@alpha.franken.de> + */ + +#include <linux/module.h> + +#include <video/sticore.h> + +#include <asm/video.h> + +bool video_is_primary_device(struct device *dev) +{ + struct sti_struct *sti; + + sti = sti_get_rom(0); + + /* if no built-in graphics card found, allow any fb driver as default */ + if (!sti) + return true; + + /* return true if it's the default built-in framebuffer driver */ + return (sti->dev == dev); +} +EXPORT_SYMBOL(video_is_primary_device); |