diff options
Diffstat (limited to 'arch/s390/kernel')
27 files changed, 495 insertions, 315 deletions
diff --git a/arch/s390/kernel/alternative.c b/arch/s390/kernel/alternative.c index c22ea1c3ef84..cce0ddee2d02 100644 --- a/arch/s390/kernel/alternative.c +++ b/arch/s390/kernel/alternative.c @@ -1,5 +1,8 @@ // SPDX-License-Identifier: GPL-2.0 #include <linux/module.h> +#include <linux/cpu.h> +#include <linux/smp.h> +#include <asm/text-patching.h> #include <asm/alternative.h> #include <asm/facility.h> #include <asm/nospec-branch.h> @@ -110,3 +113,20 @@ void __init apply_alternative_instructions(void) { apply_alternatives(__alt_instructions, __alt_instructions_end); } + +static void do_sync_core(void *info) +{ + sync_core(); +} + +void text_poke_sync(void) +{ + on_each_cpu(do_sync_core, NULL, 1); +} + +void text_poke_sync_lock(void) +{ + cpus_read_lock(); + text_poke_sync(); + cpus_read_unlock(); +} diff --git a/arch/s390/kernel/asm-offsets.c b/arch/s390/kernel/asm-offsets.c index b57da9338588..8e00bb228662 100644 --- a/arch/s390/kernel/asm-offsets.c +++ b/arch/s390/kernel/asm-offsets.c @@ -35,6 +35,7 @@ int main(void) OFFSET(__PT_ORIG_GPR2, pt_regs, orig_gpr2); OFFSET(__PT_FLAGS, pt_regs, flags); OFFSET(__PT_CR1, pt_regs, cr1); + OFFSET(__PT_LAST_BREAK, pt_regs, last_break); DEFINE(__PT_SIZE, sizeof(struct pt_regs)); BLANK(); /* stack_frame offsets */ @@ -45,6 +46,7 @@ int main(void) OFFSET(__SF_SIE_SAVEAREA, stack_frame, empty1[2]); OFFSET(__SF_SIE_REASON, stack_frame, empty1[3]); OFFSET(__SF_SIE_FLAGS, stack_frame, empty1[4]); + DEFINE(STACK_FRAME_OVERHEAD, sizeof(struct stack_frame)); BLANK(); /* idle data offsets */ OFFSET(__CLOCK_IDLE_ENTER, s390_idle_data, clock_idle_enter); @@ -77,7 +79,7 @@ int main(void) OFFSET(__LC_MCCK_CODE, lowcore, mcck_interruption_code); OFFSET(__LC_EXT_DAMAGE_CODE, lowcore, external_damage_code); OFFSET(__LC_MCCK_FAIL_STOR_ADDR, lowcore, failing_storage_address); - OFFSET(__LC_LAST_BREAK, lowcore, breaking_event_addr); + OFFSET(__LC_PGM_LAST_BREAK, lowcore, pgm_last_break); OFFSET(__LC_RETURN_LPSWE, lowcore, return_lpswe); OFFSET(__LC_RETURN_MCCK_LPSWE, lowcore, return_mcck_lpswe); OFFSET(__LC_RST_OLD_PSW, lowcore, restart_old_psw); @@ -126,6 +128,7 @@ int main(void) OFFSET(__LC_PREEMPT_COUNT, lowcore, preempt_count); OFFSET(__LC_GMAP, lowcore, gmap); OFFSET(__LC_BR_R1, lowcore, br_r1_trampoline); + OFFSET(__LC_LAST_BREAK, lowcore, last_break); /* software defined ABI-relevant lowcore locations 0xe00 - 0xe20 */ OFFSET(__LC_DUMP_REIPL, lowcore, ipib); /* hardware defined lowcore locations 0x1000 - 0x18ff */ @@ -139,6 +142,7 @@ int main(void) OFFSET(__LC_TOD_PROGREG_SAVE_AREA, lowcore, tod_progreg_save_area); OFFSET(__LC_CPU_TIMER_SAVE_AREA, lowcore, cpu_timer_save_area); OFFSET(__LC_CLOCK_COMP_SAVE_AREA, lowcore, clock_comp_save_area); + OFFSET(__LC_LAST_BREAK_SAVE_AREA, lowcore, last_break_save_area); OFFSET(__LC_AREGS_SAVE_AREA, lowcore, access_regs_save_area); OFFSET(__LC_CREGS_SAVE_AREA, lowcore, cregs_save_area); OFFSET(__LC_PGM_TDB, lowcore, pgm_tdb); @@ -160,5 +164,6 @@ int main(void) DEFINE(OLDMEM_BASE, PARMAREA + offsetof(struct parmarea, oldmem_base)); DEFINE(OLDMEM_SIZE, PARMAREA + offsetof(struct parmarea, oldmem_size)); DEFINE(COMMAND_LINE, PARMAREA + offsetof(struct parmarea, command_line)); + DEFINE(MAX_COMMAND_LINE_SIZE, PARMAREA + offsetof(struct parmarea, max_command_line_size)); return 0; } diff --git a/arch/s390/kernel/audit.c b/arch/s390/kernel/audit.c index d395c6c9944c..02051a596b87 100644 --- a/arch/s390/kernel/audit.c +++ b/arch/s390/kernel/audit.c @@ -47,15 +47,17 @@ int audit_classify_syscall(int abi, unsigned syscall) #endif switch(syscall) { case __NR_open: - return 2; + return AUDITSC_OPEN; case __NR_openat: - return 3; + return AUDITSC_OPENAT; case __NR_socketcall: - return 4; + return AUDITSC_SOCKETCALL; case __NR_execve: - return 5; + return AUDITSC_EXECVE; + case __NR_openat2: + return AUDITSC_OPENAT2; default: - return 0; + return AUDITSC_NATIVE; } } diff --git a/arch/s390/kernel/compat_audit.c b/arch/s390/kernel/compat_audit.c index 444fb1f66944..a7c46e8310f0 100644 --- a/arch/s390/kernel/compat_audit.c +++ b/arch/s390/kernel/compat_audit.c @@ -1,5 +1,6 @@ // SPDX-License-Identifier: GPL-2.0 #undef __s390x__ +#include <linux/audit_arch.h> #include <asm/unistd.h> #include "audit.h" @@ -32,14 +33,16 @@ int s390_classify_syscall(unsigned syscall) { switch(syscall) { case __NR_open: - return 2; + return AUDITSC_OPEN; case __NR_openat: - return 3; + return AUDITSC_OPENAT; case __NR_socketcall: - return 4; + return AUDITSC_SOCKETCALL; case __NR_execve: - return 5; + return AUDITSC_EXECVE; + case __NR_openat2: + return AUDITSC_OPENAT2; default: - return 1; + return AUDITSC_COMPAT; } } diff --git a/arch/s390/kernel/cpcmd.c b/arch/s390/kernel/cpcmd.c index 54efc279f54e..72e106cfd8c7 100644 --- a/arch/s390/kernel/cpcmd.c +++ b/arch/s390/kernel/cpcmd.c @@ -29,7 +29,7 @@ static int diag8_noresponse(int cmdlen) asm volatile( " diag %[rx],%[ry],0x8\n" : [ry] "+&d" (cmdlen) - : [rx] "d" ((addr_t) cpcmd_buf) + : [rx] "d" (__pa(cpcmd_buf)) : "cc"); return cmdlen; } @@ -39,8 +39,8 @@ static int diag8_response(int cmdlen, char *response, int *rlen) union register_pair rx, ry; int cc; - rx.even = (addr_t) cpcmd_buf; - rx.odd = (addr_t) response; + rx.even = __pa(cpcmd_buf); + rx.odd = __pa(response); ry.even = cmdlen | 0x40000000L; ry.odd = *rlen; asm volatile( diff --git a/arch/s390/kernel/dumpstack.c b/arch/s390/kernel/dumpstack.c index db1bc00229ca..85f326e258df 100644 --- a/arch/s390/kernel/dumpstack.c +++ b/arch/s390/kernel/dumpstack.c @@ -152,7 +152,7 @@ void show_stack(struct task_struct *task, unsigned long *stack, static void show_last_breaking_event(struct pt_regs *regs) { printk("Last Breaking-Event-Address:\n"); - printk(" [<%016lx>] %pSR\n", regs->args[0], (void *)regs->args[0]); + printk(" [<%016lx>] %pSR\n", regs->last_break, (void *)regs->last_break); } void show_registers(struct pt_regs *regs) diff --git a/arch/s390/kernel/early.c b/arch/s390/kernel/early.c index 9857cb046726..3cdf68c53614 100644 --- a/arch/s390/kernel/early.c +++ b/arch/s390/kernel/early.c @@ -280,7 +280,7 @@ char __bootdata(early_command_line)[COMMAND_LINE_SIZE]; static void __init setup_boot_command_line(void) { /* copy arch command line */ - strlcpy(boot_command_line, early_command_line, ARCH_COMMAND_LINE_SIZE); + strlcpy(boot_command_line, early_command_line, COMMAND_LINE_SIZE); } static void __init check_image_bootable(void) @@ -296,6 +296,7 @@ static void __init check_image_bootable(void) void __init startup_init(void) { + sclp_early_adjust_va(); reset_tod_clock(); check_image_bootable(); time_early_init(); diff --git a/arch/s390/kernel/entry.S b/arch/s390/kernel/entry.S index 4c9b967290ae..01bae1d51113 100644 --- a/arch/s390/kernel/entry.S +++ b/arch/s390/kernel/entry.S @@ -52,6 +52,22 @@ STACK_INIT = STACK_SIZE - STACK_FRAME_OVERHEAD - __PT_SIZE _LPP_OFFSET = __LC_LPP + .macro STBEAR address + ALTERNATIVE "", ".insn s,0xb2010000,\address", 193 + .endm + + .macro LBEAR address + ALTERNATIVE "", ".insn s,0xb2000000,\address", 193 + .endm + + .macro LPSWEY address,lpswe + ALTERNATIVE "b \lpswe", ".insn siy,0xeb0000000071,\address,0", 193 + .endm + + .macro MBEAR reg + ALTERNATIVE "", __stringify(mvc __PT_LAST_BREAK(8,\reg),__LC_LAST_BREAK), 193 + .endm + .macro CHECK_STACK savearea #ifdef CONFIG_CHECK_STACK tml %r15,STACK_SIZE - CONFIG_STACK_GUARD @@ -302,6 +318,7 @@ ENTRY(system_call) BPOFF lghi %r14,0 .Lsysc_per: + STBEAR __LC_LAST_BREAK lctlg %c1,%c1,__LC_KERNEL_ASCE lg %r12,__LC_CURRENT lg %r15,__LC_KERNEL_STACK @@ -321,14 +338,16 @@ ENTRY(system_call) xgr %r11,%r11 la %r2,STACK_FRAME_OVERHEAD(%r15) # pointer to pt_regs mvc __PT_R8(64,%r2),__LC_SAVE_AREA_SYNC + MBEAR %r2 lgr %r3,%r14 brasl %r14,__do_syscall lctlg %c1,%c1,__LC_USER_ASCE mvc __LC_RETURN_PSW(16),STACK_FRAME_OVERHEAD+__PT_PSW(%r15) BPEXIT __TI_flags(%r12),_TIF_ISOLATE_BP + LBEAR STACK_FRAME_OVERHEAD+__PT_LAST_BREAK(%r15) lmg %r0,%r15,STACK_FRAME_OVERHEAD+__PT_R0(%r15) stpt __LC_EXIT_TIMER - b __LC_RETURN_LPSWE + LPSWEY __LC_RETURN_PSW,__LC_RETURN_LPSWE ENDPROC(system_call) # @@ -340,9 +359,10 @@ ENTRY(ret_from_fork) lctlg %c1,%c1,__LC_USER_ASCE mvc __LC_RETURN_PSW(16),STACK_FRAME_OVERHEAD+__PT_PSW(%r15) BPEXIT __TI_flags(%r12),_TIF_ISOLATE_BP + LBEAR STACK_FRAME_OVERHEAD+__PT_LAST_BREAK(%r15) lmg %r0,%r15,STACK_FRAME_OVERHEAD+__PT_R0(%r15) stpt __LC_EXIT_TIMER - b __LC_RETURN_LPSWE + LPSWEY __LC_RETURN_PSW,__LC_RETURN_LPSWE ENDPROC(ret_from_fork) /* @@ -382,6 +402,7 @@ ENTRY(pgm_check_handler) xc __SF_BACKCHAIN(8,%r15),__SF_BACKCHAIN(%r15) stmg %r0,%r7,__PT_R0(%r11) mvc __PT_R8(64,%r11),__LC_SAVE_AREA_SYNC + mvc __PT_LAST_BREAK(8,%r11),__LC_PGM_LAST_BREAK stmg %r8,%r9,__PT_PSW(%r11) # clear user controlled registers to prevent speculative use @@ -401,8 +422,9 @@ ENTRY(pgm_check_handler) stpt __LC_EXIT_TIMER .Lpgm_exit_kernel: mvc __LC_RETURN_PSW(16),STACK_FRAME_OVERHEAD+__PT_PSW(%r15) + LBEAR STACK_FRAME_OVERHEAD+__PT_LAST_BREAK(%r15) lmg %r0,%r15,STACK_FRAME_OVERHEAD+__PT_R0(%r15) - b __LC_RETURN_LPSWE + LPSWEY __LC_RETURN_PSW,__LC_RETURN_LPSWE # # single stepped system call @@ -412,7 +434,8 @@ ENTRY(pgm_check_handler) larl %r14,.Lsysc_per stg %r14,__LC_RETURN_PSW+8 lghi %r14,1 - lpswe __LC_RETURN_PSW # branch to .Lsysc_per + LBEAR __LC_PGM_LAST_BREAK + LPSWEY __LC_RETURN_PSW,__LC_RETURN_LPSWE # branch to .Lsysc_per ENDPROC(pgm_check_handler) /* @@ -422,6 +445,7 @@ ENDPROC(pgm_check_handler) ENTRY(\name) STCK __LC_INT_CLOCK stpt __LC_SYS_ENTER_TIMER + STBEAR __LC_LAST_BREAK BPOFF stmg %r8,%r15,__LC_SAVE_AREA_ASYNC lg %r12,__LC_CURRENT @@ -453,6 +477,7 @@ ENTRY(\name) xgr %r10,%r10 xc __PT_FLAGS(8,%r11),__PT_FLAGS(%r11) mvc __PT_R8(64,%r11),__LC_SAVE_AREA_ASYNC + MBEAR %r11 stmg %r8,%r9,__PT_PSW(%r11) tm %r8,0x0001 # coming from user space? jno 1f @@ -465,8 +490,9 @@ ENTRY(\name) lctlg %c1,%c1,__LC_USER_ASCE BPEXIT __TI_flags(%r12),_TIF_ISOLATE_BP stpt __LC_EXIT_TIMER -2: lmg %r0,%r15,__PT_R0(%r11) - b __LC_RETURN_LPSWE +2: LBEAR __PT_LAST_BREAK(%r11) + lmg %r0,%r15,__PT_R0(%r11) + LPSWEY __LC_RETURN_PSW,__LC_RETURN_LPSWE ENDPROC(\name) .endm @@ -505,6 +531,7 @@ ENTRY(mcck_int_handler) BPOFF la %r1,4095 # validate r1 spt __LC_CPU_TIMER_SAVE_AREA-4095(%r1) # validate cpu timer + LBEAR __LC_LAST_BREAK_SAVE_AREA-4095(%r1) # validate bear lmg %r0,%r15,__LC_GPREGS_SAVE_AREA-4095(%r1)# validate gprs lg %r12,__LC_CURRENT lmg %r8,%r9,__LC_MCK_OLD_PSW @@ -591,8 +618,10 @@ ENTRY(mcck_int_handler) jno 0f BPEXIT __TI_flags(%r12),_TIF_ISOLATE_BP stpt __LC_EXIT_TIMER -0: lmg %r11,%r15,__PT_R11(%r11) - b __LC_RETURN_MCCK_LPSWE +0: ALTERNATIVE "", __stringify(lghi %r12,__LC_LAST_BREAK_SAVE_AREA),193 + LBEAR 0(%r12) + lmg %r11,%r15,__PT_R11(%r11) + LPSWEY __LC_RETURN_MCCK_PSW,__LC_RETURN_MCCK_LPSWE .Lmcck_panic: /* diff --git a/arch/s390/kernel/entry.h b/arch/s390/kernel/entry.h index 7f2696e8d511..6083090be1f4 100644 --- a/arch/s390/kernel/entry.h +++ b/arch/s390/kernel/entry.h @@ -70,5 +70,6 @@ extern struct exception_table_entry _stop_amode31_ex_table[]; #define __amode31_data __section(".amode31.data") #define __amode31_ref __section(".amode31.refs") extern long _start_amode31_refs[], _end_amode31_refs[]; +extern unsigned long __amode31_base; #endif /* _ENTRY_H */ diff --git a/arch/s390/kernel/ftrace.c b/arch/s390/kernel/ftrace.c index 1d94ffdf347b..5510c7d10ddc 100644 --- a/arch/s390/kernel/ftrace.c +++ b/arch/s390/kernel/ftrace.c @@ -17,6 +17,7 @@ #include <linux/kprobes.h> #include <trace/syscall.h> #include <asm/asm-offsets.h> +#include <asm/text-patching.h> #include <asm/cacheflush.h> #include <asm/ftrace.lds.h> #include <asm/nospec-branch.h> @@ -80,17 +81,6 @@ asm( #ifdef CONFIG_MODULES static char *ftrace_plt; - -asm( - " .data\n" - "ftrace_plt_template:\n" - " basr %r1,%r0\n" - " lg %r1,0f-.(%r1)\n" - " br %r1\n" - "0: .quad ftrace_caller\n" - "ftrace_plt_template_end:\n" - " .previous\n" -); #endif /* CONFIG_MODULES */ static const char *ftrace_shared_hotpatch_trampoline(const char **end) @@ -116,7 +106,7 @@ static const char *ftrace_shared_hotpatch_trampoline(const char **end) bool ftrace_need_init_nop(void) { - return ftrace_shared_hotpatch_trampoline(NULL); + return true; } int ftrace_init_nop(struct module *mod, struct dyn_ftrace *rec) @@ -175,28 +165,6 @@ int ftrace_modify_call(struct dyn_ftrace *rec, unsigned long old_addr, return 0; } -static void ftrace_generate_nop_insn(struct ftrace_insn *insn) -{ - /* brcl 0,0 */ - insn->opc = 0xc004; - insn->disp = 0; -} - -static void ftrace_generate_call_insn(struct ftrace_insn *insn, - unsigned long ip) -{ - unsigned long target; - - /* brasl r0,ftrace_caller */ - target = FTRACE_ADDR; -#ifdef CONFIG_MODULES - if (is_module_addr((void *)ip)) - target = (unsigned long)ftrace_plt; -#endif /* CONFIG_MODULES */ - insn->opc = 0xc005; - insn->disp = (target - ip) / 2; -} - static void brcl_disable(void *brcl) { u8 op = 0x04; /* set mask field to zero */ @@ -207,23 +175,7 @@ static void brcl_disable(void *brcl) int ftrace_make_nop(struct module *mod, struct dyn_ftrace *rec, unsigned long addr) { - struct ftrace_insn orig, new, old; - - if (ftrace_shared_hotpatch_trampoline(NULL)) { - brcl_disable((void *)rec->ip); - return 0; - } - - if (copy_from_kernel_nofault(&old, (void *) rec->ip, sizeof(old))) - return -EFAULT; - /* Replace ftrace call with a nop. */ - ftrace_generate_call_insn(&orig, rec->ip); - ftrace_generate_nop_insn(&new); - - /* Verify that the to be replaced code matches what we expect. */ - if (memcmp(&orig, &old, sizeof(old))) - return -EINVAL; - s390_kernel_write((void *) rec->ip, &new, sizeof(new)); + brcl_disable((void *)rec->ip); return 0; } @@ -236,23 +188,7 @@ static void brcl_enable(void *brcl) int ftrace_make_call(struct dyn_ftrace *rec, unsigned long addr) { - struct ftrace_insn orig, new, old; - - if (ftrace_shared_hotpatch_trampoline(NULL)) { - brcl_enable((void *)rec->ip); - return 0; - } - - if (copy_from_kernel_nofault(&old, (void *) rec->ip, sizeof(old))) - return -EFAULT; - /* Replace nop with an ftrace call. */ - ftrace_generate_nop_insn(&orig); - ftrace_generate_call_insn(&new, rec->ip); - - /* Verify that the to be replaced code matches what we expect. */ - if (memcmp(&orig, &old, sizeof(old))) - return -EINVAL; - s390_kernel_write((void *) rec->ip, &new, sizeof(new)); + brcl_enable((void *)rec->ip); return 0; } @@ -262,29 +198,18 @@ int ftrace_update_ftrace_func(ftrace_func_t func) return 0; } -int __init ftrace_dyn_arch_init(void) -{ - return 0; -} - void arch_ftrace_update_code(int command) { - if (ftrace_shared_hotpatch_trampoline(NULL)) - ftrace_modify_all_code(command); - else - ftrace_run_stop_machine(command); -} - -static void __ftrace_sync(void *dummy) -{ + ftrace_modify_all_code(command); } int ftrace_arch_code_modify_post_process(void) { - if (ftrace_shared_hotpatch_trampoline(NULL)) { - /* Send SIGP to the other CPUs, so they see the new code. */ - smp_call_function(__ftrace_sync, NULL, 1); - } + /* + * Flush any pre-fetched instructions on all + * CPUs to make the new code visible. + */ + text_poke_sync_lock(); return 0; } @@ -299,10 +224,6 @@ static int __init ftrace_plt_init(void) panic("cannot allocate ftrace plt\n"); start = ftrace_shared_hotpatch_trampoline(&end); - if (!start) { - start = ftrace_plt_template; - end = ftrace_plt_template_end; - } memcpy(ftrace_plt, start, end - start); set_memory_ro((unsigned long)ftrace_plt, 1); return 0; @@ -342,12 +263,14 @@ NOKPROBE_SYMBOL(prepare_ftrace_return); int ftrace_enable_ftrace_graph_caller(void) { brcl_disable(ftrace_graph_caller); + text_poke_sync_lock(); return 0; } int ftrace_disable_ftrace_graph_caller(void) { brcl_enable(ftrace_graph_caller); + text_poke_sync_lock(); return 0; } diff --git a/arch/s390/kernel/head64.S b/arch/s390/kernel/head64.S index 114b5490ad8e..42f9a325a257 100644 --- a/arch/s390/kernel/head64.S +++ b/arch/s390/kernel/head64.S @@ -20,8 +20,6 @@ __HEAD ENTRY(startup_continue) larl %r1,tod_clock_base mvc 0(16,%r1),__LC_BOOT_CLOCK - larl %r13,.LPG1 # get base - lctlg %c0,%c15,.Lctl-.LPG1(%r13) # load control registers # # Setup stack # @@ -42,19 +40,3 @@ ENTRY(startup_continue) .align 16 .LPG1: .Ldw: .quad 0x0002000180000000,0x0000000000000000 -.Lctl: .quad 0x04040000 # cr0: AFP registers & secondary space - .quad 0 # cr1: primary space segment table - .quad 0 # cr2: dispatchable unit control table - .quad 0 # cr3: instruction authorization - .quad 0xffff # cr4: instruction authorization - .quad 0 # cr5: primary-aste origin - .quad 0 # cr6: I/O interrupts - .quad 0 # cr7: secondary space segment table - .quad 0x0000000000008000 # cr8: access registers translation - .quad 0 # cr9: tracing off - .quad 0 # cr10: tracing off - .quad 0 # cr11: tracing off - .quad 0 # cr12: tracing off - .quad 0 # cr13: home space segment table - .quad 0xc0000000 # cr14: machine check handling off - .quad 0 # cr15: linkage stack operations diff --git a/arch/s390/kernel/irq.c b/arch/s390/kernel/irq.c index 3a3145c4a3ba..0df83ecaa2e0 100644 --- a/arch/s390/kernel/irq.c +++ b/arch/s390/kernel/irq.c @@ -140,8 +140,11 @@ void noinstr do_io_irq(struct pt_regs *regs) irq_enter(); - if (user_mode(regs)) + if (user_mode(regs)) { update_timer_sys(); + if (static_branch_likely(&cpu_has_bear)) + current->thread.last_break = regs->last_break; + } from_idle = !user_mode(regs) && regs->psw.addr == (unsigned long)psw_idle_exit; if (from_idle) @@ -171,8 +174,11 @@ void noinstr do_ext_irq(struct pt_regs *regs) irq_enter(); - if (user_mode(regs)) + if (user_mode(regs)) { update_timer_sys(); + if (static_branch_likely(&cpu_has_bear)) + current->thread.last_break = regs->last_break; + } regs->int_code = S390_lowcore.ext_int_code_addr; regs->int_parm = S390_lowcore.ext_params; diff --git a/arch/s390/kernel/jump_label.c b/arch/s390/kernel/jump_label.c index 9156653b56f6..6bec000c6c1c 100644 --- a/arch/s390/kernel/jump_label.c +++ b/arch/s390/kernel/jump_label.c @@ -6,8 +6,9 @@ * Author(s): Jan Glauber <jang@linux.vnet.ibm.com> */ #include <linux/uaccess.h> -#include <linux/stop_machine.h> #include <linux/jump_label.h> +#include <linux/module.h> +#include <asm/text-patching.h> #include <asm/ipl.h> struct insn { @@ -48,9 +49,9 @@ static struct insn orignop = { .offset = JUMP_LABEL_NOP_OFFSET >> 1, }; -static void __jump_label_transform(struct jump_entry *entry, - enum jump_label_type type, - int init) +static void jump_label_transform(struct jump_entry *entry, + enum jump_label_type type, + int init) { void *code = (void *)jump_entry_code(entry); struct insn old, new; @@ -72,19 +73,28 @@ static void __jump_label_transform(struct jump_entry *entry, s390_kernel_write(code, &new, sizeof(new)); } -static void __jump_label_sync(void *dummy) +void arch_jump_label_transform(struct jump_entry *entry, + enum jump_label_type type) { + jump_label_transform(entry, type, 0); + text_poke_sync(); } -void arch_jump_label_transform(struct jump_entry *entry, - enum jump_label_type type) +bool arch_jump_label_transform_queue(struct jump_entry *entry, + enum jump_label_type type) +{ + jump_label_transform(entry, type, 0); + return true; +} + +void arch_jump_label_transform_apply(void) { - __jump_label_transform(entry, type, 0); - smp_call_function(__jump_label_sync, NULL, 1); + text_poke_sync(); } -void arch_jump_label_transform_static(struct jump_entry *entry, - enum jump_label_type type) +void __init_or_module arch_jump_label_transform_static(struct jump_entry *entry, + enum jump_label_type type) { - __jump_label_transform(entry, type, 1); + jump_label_transform(entry, type, 1); + text_poke_sync(); } diff --git a/arch/s390/kernel/kprobes.c b/arch/s390/kernel/kprobes.c index 52d056a5f89f..e27a7d3b0364 100644 --- a/arch/s390/kernel/kprobes.c +++ b/arch/s390/kernel/kprobes.c @@ -7,6 +7,8 @@ * s390 port, used ppc64 as template. Mike Grundy <grundym@us.ibm.com> */ +#define pr_fmt(fmt) "kprobes: " fmt + #include <linux/moduleloader.h> #include <linux/kprobes.h> #include <linux/ptrace.h> @@ -120,9 +122,55 @@ static void s390_free_insn_slot(struct kprobe *p) } NOKPROBE_SYMBOL(s390_free_insn_slot); +/* Check if paddr is at an instruction boundary */ +static bool can_probe(unsigned long paddr) +{ + unsigned long addr, offset = 0; + kprobe_opcode_t insn; + struct kprobe *kp; + + if (paddr & 0x01) + return false; + + if (!kallsyms_lookup_size_offset(paddr, NULL, &offset)) + return false; + + /* Decode instructions */ + addr = paddr - offset; + while (addr < paddr) { + if (copy_from_kernel_nofault(&insn, (void *)addr, sizeof(insn))) + return false; + + if (insn >> 8 == 0) { + if (insn != BREAKPOINT_INSTRUCTION) { + /* + * Note that QEMU inserts opcode 0x0000 to implement + * software breakpoints for guests. Since the size of + * the original instruction is unknown, stop following + * instructions and prevent setting a kprobe. + */ + return false; + } + /* + * Check if the instruction has been modified by another + * kprobe, in which case the original instruction is + * decoded. + */ + kp = get_kprobe((void *)addr); + if (!kp) { + /* not a kprobe */ + return false; + } + insn = kp->opcode; + } + addr += insn_length(insn >> 8); + } + return addr == paddr; +} + int arch_prepare_kprobe(struct kprobe *p) { - if ((unsigned long) p->addr & 0x01) + if (!can_probe((unsigned long)p->addr)) return -EINVAL; /* Make sure the probe isn't going on a difficult instruction */ if (probe_is_prohibited_opcode(p->addr)) @@ -240,7 +288,7 @@ void arch_prepare_kretprobe(struct kretprobe_instance *ri, struct pt_regs *regs) ri->fp = NULL; /* Replace the return addr with trampoline addr */ - regs->gprs[14] = (unsigned long) &kretprobe_trampoline; + regs->gprs[14] = (unsigned long) &__kretprobe_trampoline; } NOKPROBE_SYMBOL(arch_prepare_kretprobe); @@ -259,7 +307,7 @@ static void kprobe_reenter_check(struct kprobe_ctlblk *kcb, struct kprobe *p) * is a BUG. The code path resides in the .kprobes.text * section and is executed with interrupts disabled. */ - pr_err("Invalid kprobe detected.\n"); + pr_err("Failed to recover from reentered kprobes.\n"); dump_kprobe(p); BUG(); } @@ -332,8 +380,8 @@ NOKPROBE_SYMBOL(kprobe_handler); */ static void __used kretprobe_trampoline_holder(void) { - asm volatile(".global kretprobe_trampoline\n" - "kretprobe_trampoline: bcr 0,0\n"); + asm volatile(".global __kretprobe_trampoline\n" + "__kretprobe_trampoline: bcr 0,0\n"); } /* @@ -341,7 +389,7 @@ static void __used kretprobe_trampoline_holder(void) */ static int trampoline_probe_handler(struct kprobe *p, struct pt_regs *regs) { - regs->psw.addr = __kretprobe_trampoline_handler(regs, &kretprobe_trampoline, NULL); + regs->psw.addr = __kretprobe_trampoline_handler(regs, NULL); /* * By returning a non-zero value, we are telling * kprobe_handler() that we don't want the post_handler @@ -507,7 +555,7 @@ int kprobe_exceptions_notify(struct notifier_block *self, NOKPROBE_SYMBOL(kprobe_exceptions_notify); static struct kprobe trampoline = { - .addr = (kprobe_opcode_t *) &kretprobe_trampoline, + .addr = (kprobe_opcode_t *) &__kretprobe_trampoline, .pre_handler = trampoline_probe_handler }; @@ -518,6 +566,6 @@ int __init arch_init_kprobes(void) int arch_trampoline_kprobe(struct kprobe *p) { - return p->addr == (kprobe_opcode_t *) &kretprobe_trampoline; + return p->addr == (kprobe_opcode_t *) &__kretprobe_trampoline; } NOKPROBE_SYMBOL(arch_trampoline_kprobe); diff --git a/arch/s390/kernel/machine_kexec_file.c b/arch/s390/kernel/machine_kexec_file.c index f9e4baa64b67..528edff085d9 100644 --- a/arch/s390/kernel/machine_kexec_file.c +++ b/arch/s390/kernel/machine_kexec_file.c @@ -216,7 +216,9 @@ void *kexec_file_add_components(struct kimage *image, int (*add_kernel)(struct kimage *image, struct s390_load_data *data)) { + unsigned long max_command_line_size = LEGACY_COMMAND_LINE_SIZE; struct s390_load_data data = {0}; + unsigned long minsize; int ret; data.report = ipl_report_init(&ipl_block); @@ -227,10 +229,23 @@ void *kexec_file_add_components(struct kimage *image, if (ret) goto out; - if (image->cmdline_buf_len >= ARCH_COMMAND_LINE_SIZE) { - ret = -EINVAL; + ret = -EINVAL; + minsize = PARMAREA + offsetof(struct parmarea, command_line); + if (image->kernel_buf_len < minsize) goto out; - } + + if (data.parm->max_command_line_size) + max_command_line_size = data.parm->max_command_line_size; + + if (minsize + max_command_line_size < minsize) + goto out; + + if (image->kernel_buf_len < minsize + max_command_line_size) + goto out; + + if (image->cmdline_buf_len >= max_command_line_size) + goto out; + memcpy(data.parm->command_line, image->cmdline_buf, image->cmdline_buf_len); @@ -307,17 +322,3 @@ int arch_kexec_apply_relocations_add(struct purgatory_info *pi, } return 0; } - -int arch_kexec_kernel_image_probe(struct kimage *image, void *buf, - unsigned long buf_len) -{ - /* A kernel must be at least large enough to contain head.S. During - * load memory in head.S will be accessed, e.g. to register the next - * command line. If the next kernel were smaller the current kernel - * will panic at load. - */ - if (buf_len < HEAD_END) - return -ENOEXEC; - - return kexec_image_probe_default(image, buf, buf_len); -} diff --git a/arch/s390/kernel/mcount.S b/arch/s390/kernel/mcount.S index 6b13797143a7..39bcc0e39a10 100644 --- a/arch/s390/kernel/mcount.S +++ b/arch/s390/kernel/mcount.S @@ -22,10 +22,11 @@ ENTRY(ftrace_stub) BR_EX %r14 ENDPROC(ftrace_stub) -#define STACK_FRAME_SIZE (STACK_FRAME_OVERHEAD + __PT_SIZE) -#define STACK_PTREGS (STACK_FRAME_OVERHEAD) -#define STACK_PTREGS_GPRS (STACK_PTREGS + __PT_GPRS) -#define STACK_PTREGS_PSW (STACK_PTREGS + __PT_PSW) +#define STACK_FRAME_SIZE (STACK_FRAME_OVERHEAD + __PT_SIZE) +#define STACK_PTREGS (STACK_FRAME_OVERHEAD) +#define STACK_PTREGS_GPRS (STACK_PTREGS + __PT_GPRS) +#define STACK_PTREGS_PSW (STACK_PTREGS + __PT_PSW) +#define STACK_PTREGS_ORIG_GPR2 (STACK_PTREGS + __PT_ORIG_GPR2) #ifdef __PACK_STACK /* allocate just enough for r14, r15 and backchain */ #define TRACED_FUNC_FRAME_SIZE 24 @@ -33,13 +34,15 @@ ENDPROC(ftrace_stub) #define TRACED_FUNC_FRAME_SIZE STACK_FRAME_OVERHEAD #endif -ENTRY(ftrace_caller) - .globl ftrace_regs_caller - .set ftrace_regs_caller,ftrace_caller + .macro ftrace_regs_entry, allregs=0 stg %r14,(__SF_GPRS+8*8)(%r15) # save traced function caller + + .if \allregs == 1 lghi %r14,0 # save condition code ipm %r14 # don't put any instructions sllg %r14,%r14,16 # clobbering CC before this point + .endif + lgr %r1,%r15 # allocate stack frame for ftrace_caller to contain traced function aghi %r15,-TRACED_FUNC_FRAME_SIZE @@ -49,13 +52,31 @@ ENTRY(ftrace_caller) # allocate pt_regs and stack frame for ftrace_trace_function aghi %r15,-STACK_FRAME_SIZE stg %r1,(STACK_PTREGS_GPRS+15*8)(%r15) + xc STACK_PTREGS_ORIG_GPR2(8,%r15),STACK_PTREGS_ORIG_GPR2(%r15) + + .if \allregs == 1 stg %r14,(STACK_PTREGS_PSW)(%r15) - lg %r14,(__SF_GPRS+8*8)(%r1) # restore original return address stosm (STACK_PTREGS_PSW)(%r15),0 + .endif + + lg %r14,(__SF_GPRS+8*8)(%r1) # restore original return address aghi %r1,-TRACED_FUNC_FRAME_SIZE stg %r1,__SF_BACKCHAIN(%r15) stg %r0,(STACK_PTREGS_PSW+8)(%r15) stmg %r2,%r14,(STACK_PTREGS_GPRS+2*8)(%r15) + .endm + +SYM_CODE_START(ftrace_regs_caller) + ftrace_regs_entry 1 + j ftrace_common +SYM_CODE_END(ftrace_regs_caller) + +SYM_CODE_START(ftrace_caller) + ftrace_regs_entry 0 + j ftrace_common +SYM_CODE_END(ftrace_caller) + +SYM_CODE_START(ftrace_common) #ifdef CONFIG_HAVE_MARCH_Z196_FEATURES aghik %r2,%r0,-MCOUNT_INSN_SIZE lgrl %r4,function_trace_op @@ -74,24 +95,31 @@ ENTRY(ftrace_caller) #ifdef CONFIG_FUNCTION_GRAPH_TRACER # The j instruction gets runtime patched to a nop instruction. # See ftrace_enable_ftrace_graph_caller. - .globl ftrace_graph_caller -ftrace_graph_caller: - j ftrace_graph_caller_end +SYM_INNER_LABEL(ftrace_graph_caller, SYM_L_GLOBAL) + j .Lftrace_graph_caller_end lmg %r2,%r3,(STACK_PTREGS_GPRS+14*8)(%r15) lg %r4,(STACK_PTREGS_PSW+8)(%r15) brasl %r14,prepare_ftrace_return stg %r2,(STACK_PTREGS_GPRS+14*8)(%r15) -ftrace_graph_caller_end: - .globl ftrace_graph_caller_end +.Lftrace_graph_caller_end: +#endif + lg %r0,(STACK_PTREGS_PSW+8)(%r15) +#ifdef CONFIG_HAVE_MARCH_Z196_FEATURES + ltg %r1,STACK_PTREGS_ORIG_GPR2(%r15) + locgrz %r1,%r0 +#else + lg %r1,STACK_PTREGS_ORIG_GPR2(%r15) + ltgr %r1,%r1 + jnz 0f + lgr %r1,%r0 #endif - lg %r1,(STACK_PTREGS_PSW+8)(%r15) - lmg %r2,%r15,(STACK_PTREGS_GPRS+2*8)(%r15) +0: lmg %r2,%r15,(STACK_PTREGS_GPRS+2*8)(%r15) BR_EX %r1 -ENDPROC(ftrace_caller) +SYM_CODE_END(ftrace_common) #ifdef CONFIG_FUNCTION_GRAPH_TRACER -ENTRY(return_to_handler) +SYM_FUNC_START(return_to_handler) stmg %r2,%r5,32(%r15) lgr %r1,%r15 aghi %r15,-STACK_FRAME_OVERHEAD @@ -101,6 +129,6 @@ ENTRY(return_to_handler) lgr %r14,%r2 lmg %r2,%r5,32(%r15) BR_EX %r14 -ENDPROC(return_to_handler) +SYM_FUNC_END(return_to_handler) #endif diff --git a/arch/s390/kernel/nospec-branch.c b/arch/s390/kernel/nospec-branch.c index 250e4dbf653c..60e6fec27bba 100644 --- a/arch/s390/kernel/nospec-branch.c +++ b/arch/s390/kernel/nospec-branch.c @@ -38,7 +38,7 @@ static int __init nospec_report(void) { if (test_facility(156)) pr_info("Spectre V2 mitigation: etokens\n"); - if (__is_defined(CC_USING_EXPOLINE) && !nospec_disable) + if (nospec_uses_trampoline()) pr_info("Spectre V2 mitigation: execute trampolines\n"); if (__test_facility(82, alt_stfle_fac_list)) pr_info("Spectre V2 mitigation: limited branch prediction\n"); diff --git a/arch/s390/kernel/nospec-sysfs.c b/arch/s390/kernel/nospec-sysfs.c index b4b5c8c21166..52d4353188ad 100644 --- a/arch/s390/kernel/nospec-sysfs.c +++ b/arch/s390/kernel/nospec-sysfs.c @@ -15,7 +15,7 @@ ssize_t cpu_show_spectre_v2(struct device *dev, { if (test_facility(156)) return sprintf(buf, "Mitigation: etokens\n"); - if (__is_defined(CC_USING_EXPOLINE) && !nospec_disable) + if (nospec_uses_trampoline()) return sprintf(buf, "Mitigation: execute trampolines\n"); if (__test_facility(82, alt_stfle_fac_list)) return sprintf(buf, "Mitigation: limited branch prediction\n"); diff --git a/arch/s390/kernel/perf_cpum_cf.c b/arch/s390/kernel/perf_cpum_cf.c index 4a99154fe651..6f431fa9e4d7 100644 --- a/arch/s390/kernel/perf_cpum_cf.c +++ b/arch/s390/kernel/perf_cpum_cf.c @@ -773,22 +773,46 @@ static int __init cpumf_pmu_init(void) * counter set via normal file operations. */ -static atomic_t cfset_opencnt = ATOMIC_INIT(0); /* Excl. access */ +static atomic_t cfset_opencnt = ATOMIC_INIT(0); /* Access count */ static DEFINE_MUTEX(cfset_ctrset_mutex);/* Synchronize access to hardware */ struct cfset_call_on_cpu_parm { /* Parm struct for smp_call_on_cpu */ unsigned int sets; /* Counter set bit mask */ atomic_t cpus_ack; /* # CPUs successfully executed func */ }; -static struct cfset_request { /* CPUs and counter set bit mask */ +static struct cfset_session { /* CPUs and counter set bit mask */ + struct list_head head; /* Head of list of active processes */ +} cfset_session = { + .head = LIST_HEAD_INIT(cfset_session.head) +}; + +struct cfset_request { /* CPUs and counter set bit mask */ unsigned long ctrset; /* Bit mask of counter set to read */ cpumask_t mask; /* CPU mask to read from */ -} cfset_request; + struct list_head node; /* Chain to cfset_session.head */ +}; + +static void cfset_session_init(void) +{ + INIT_LIST_HEAD(&cfset_session.head); +} + +/* Remove current request from global bookkeeping. Maintain a counter set bit + * mask on a per CPU basis. + * Done in process context under mutex protection. + */ +static void cfset_session_del(struct cfset_request *p) +{ + list_del(&p->node); +} -static void cfset_ctrset_clear(void) +/* Add current request to global bookkeeping. Maintain a counter set bit mask + * on a per CPU basis. + * Done in process context under mutex protection. + */ +static void cfset_session_add(struct cfset_request *p) { - cpumask_clear(&cfset_request.mask); - cfset_request.ctrset = 0; + list_add(&p->node, &cfset_session.head); } /* The /dev/hwctr device access uses PMU_F_IN_USE to mark the device access @@ -827,15 +851,23 @@ static void cfset_ioctl_off(void *parm) struct cfset_call_on_cpu_parm *p = parm; int rc; - cpuhw->dev_state = 0; + /* Check if any counter set used by /dev/hwc */ for (rc = CPUMF_CTR_SET_BASIC; rc < CPUMF_CTR_SET_MAX; ++rc) - if ((p->sets & cpumf_ctr_ctl[rc])) - atomic_dec(&cpuhw->ctr_set[rc]); - rc = lcctl(cpuhw->state); /* Keep perf_event_open counter sets */ + if ((p->sets & cpumf_ctr_ctl[rc])) { + if (!atomic_dec_return(&cpuhw->ctr_set[rc])) { + ctr_set_disable(&cpuhw->dev_state, + cpumf_ctr_ctl[rc]); + ctr_set_stop(&cpuhw->dev_state, + cpumf_ctr_ctl[rc]); + } + } + /* Keep perf_event_open counter sets */ + rc = lcctl(cpuhw->dev_state | cpuhw->state); if (rc) pr_err("Counter set stop %#llx of /dev/%s failed rc=%i\n", cpuhw->state, S390_HWCTR_DEVICE, rc); - cpuhw->flags &= ~PMU_F_IN_USE; + if (!cpuhw->dev_state) + cpuhw->flags &= ~PMU_F_IN_USE; debug_sprintf_event(cf_dbg, 4, "%s rc %d state %#llx dev_state %#llx\n", __func__, rc, cpuhw->state, cpuhw->dev_state); } @@ -870,11 +902,26 @@ static void cfset_release_cpu(void *p) debug_sprintf_event(cf_dbg, 4, "%s state %#llx dev_state %#llx\n", __func__, cpuhw->state, cpuhw->dev_state); + cpuhw->dev_state = 0; rc = lcctl(cpuhw->state); /* Keep perf_event_open counter sets */ if (rc) pr_err("Counter set release %#llx of /dev/%s failed rc=%i\n", cpuhw->state, S390_HWCTR_DEVICE, rc); - cpuhw->dev_state = 0; +} + +/* This modifies the process CPU mask to adopt it to the currently online + * CPUs. Offline CPUs can not be addresses. This call terminates the access + * and is usually followed by close() or a new iotcl(..., START, ...) which + * creates a new request structure. + */ +static void cfset_all_stop(struct cfset_request *req) +{ + struct cfset_call_on_cpu_parm p = { + .sets = req->ctrset, + }; + + cpumask_and(&req->mask, &req->mask, cpu_online_mask); + on_each_cpu_mask(&req->mask, cfset_ioctl_off, &p, 1); } /* Release function is also called when application gets terminated without @@ -882,10 +929,19 @@ static void cfset_release_cpu(void *p) */ static int cfset_release(struct inode *inode, struct file *file) { - on_each_cpu(cfset_release_cpu, NULL, 1); + mutex_lock(&cfset_ctrset_mutex); + /* Open followed by close/exit has no private_data */ + if (file->private_data) { + cfset_all_stop(file->private_data); + cfset_session_del(file->private_data); + kfree(file->private_data); + file->private_data = NULL; + } + if (!atomic_dec_return(&cfset_opencnt)) + on_each_cpu(cfset_release_cpu, NULL, 1); + mutex_unlock(&cfset_ctrset_mutex); + hw_perf_event_destroy(NULL); - cfset_ctrset_clear(); - atomic_set(&cfset_opencnt, 0); return 0; } @@ -893,9 +949,10 @@ static int cfset_open(struct inode *inode, struct file *file) { if (!capable(CAP_SYS_ADMIN)) return -EPERM; - /* Only one user space program can open /dev/hwctr */ - if (atomic_xchg(&cfset_opencnt, 1)) - return -EBUSY; + mutex_lock(&cfset_ctrset_mutex); + if (atomic_inc_return(&cfset_opencnt) == 1) + cfset_session_init(); + mutex_unlock(&cfset_ctrset_mutex); cpumf_hw_inuse(); file->private_data = NULL; @@ -903,25 +960,10 @@ static int cfset_open(struct inode *inode, struct file *file) return nonseekable_open(inode, file); } -static int cfset_all_stop(void) +static int cfset_all_start(struct cfset_request *req) { struct cfset_call_on_cpu_parm p = { - .sets = cfset_request.ctrset, - }; - cpumask_var_t mask; - - if (!alloc_cpumask_var(&mask, GFP_KERNEL)) - return -ENOMEM; - cpumask_and(mask, &cfset_request.mask, cpu_online_mask); - on_each_cpu_mask(mask, cfset_ioctl_off, &p, 1); - free_cpumask_var(mask); - return 0; -} - -static int cfset_all_start(void) -{ - struct cfset_call_on_cpu_parm p = { - .sets = cfset_request.ctrset, + .sets = req->ctrset, .cpus_ack = ATOMIC_INIT(0), }; cpumask_var_t mask; @@ -929,7 +971,7 @@ static int cfset_all_start(void) if (!alloc_cpumask_var(&mask, GFP_KERNEL)) return -ENOMEM; - cpumask_and(mask, &cfset_request.mask, cpu_online_mask); + cpumask_and(mask, &req->mask, cpu_online_mask); on_each_cpu_mask(mask, cfset_ioctl_on, &p, 1); if (atomic_read(&p.cpus_ack) != cpumask_weight(mask)) { on_each_cpu_mask(mask, cfset_ioctl_off, &p, 1); @@ -1045,7 +1087,7 @@ static void cfset_cpu_read(void *parm) cpuhw->sets, cpuhw->used); } -static int cfset_all_read(unsigned long arg) +static int cfset_all_read(unsigned long arg, struct cfset_request *req) { struct cfset_call_on_cpu_parm p; cpumask_var_t mask; @@ -1054,46 +1096,53 @@ static int cfset_all_read(unsigned long arg) if (!alloc_cpumask_var(&mask, GFP_KERNEL)) return -ENOMEM; - p.sets = cfset_request.ctrset; - cpumask_and(mask, &cfset_request.mask, cpu_online_mask); + p.sets = req->ctrset; + cpumask_and(mask, &req->mask, cpu_online_mask); on_each_cpu_mask(mask, cfset_cpu_read, &p, 1); rc = cfset_all_copy(arg, mask); free_cpumask_var(mask); return rc; } -static long cfset_ioctl_read(unsigned long arg) +static long cfset_ioctl_read(unsigned long arg, struct cfset_request *req) { struct s390_ctrset_read read; - int ret = 0; + int ret = -ENODATA; - if (copy_from_user(&read, (char __user *)arg, sizeof(read))) - return -EFAULT; - ret = cfset_all_read(arg); + if (req && req->ctrset) { + if (copy_from_user(&read, (char __user *)arg, sizeof(read))) + return -EFAULT; + ret = cfset_all_read(arg, req); + } return ret; } -static long cfset_ioctl_stop(void) +static long cfset_ioctl_stop(struct file *file) { - int ret = ENXIO; - - if (cfset_request.ctrset) { - ret = cfset_all_stop(); - cfset_ctrset_clear(); + struct cfset_request *req = file->private_data; + int ret = -ENXIO; + + if (req) { + cfset_all_stop(req); + cfset_session_del(req); + kfree(req); + file->private_data = NULL; + ret = 0; } return ret; } -static long cfset_ioctl_start(unsigned long arg) +static long cfset_ioctl_start(unsigned long arg, struct file *file) { struct s390_ctrset_start __user *ustart; struct s390_ctrset_start start; + struct cfset_request *preq; void __user *umask; unsigned int len; int ret = 0; size_t need; - if (cfset_request.ctrset) + if (file->private_data) return -EBUSY; ustart = (struct s390_ctrset_start __user *)arg; if (copy_from_user(&start, ustart, sizeof(start))) @@ -1108,25 +1157,36 @@ static long cfset_ioctl_start(unsigned long arg) return -EINVAL; /* Invalid counter set */ if (!start.counter_sets) return -EINVAL; /* No counter set at all? */ - cpumask_clear(&cfset_request.mask); + + preq = kzalloc(sizeof(*preq), GFP_KERNEL); + if (!preq) + return -ENOMEM; + cpumask_clear(&preq->mask); len = min_t(u64, start.cpumask_len, cpumask_size()); umask = (void __user *)start.cpumask; - if (copy_from_user(&cfset_request.mask, umask, len)) + if (copy_from_user(&preq->mask, umask, len)) { + kfree(preq); return -EFAULT; - if (cpumask_empty(&cfset_request.mask)) + } + if (cpumask_empty(&preq->mask)) { + kfree(preq); return -EINVAL; + } need = cfset_needspace(start.counter_sets); - if (put_user(need, &ustart->data_bytes)) - ret = -EFAULT; - if (ret) - goto out; - cfset_request.ctrset = start.counter_sets; - ret = cfset_all_start(); -out: - if (ret) - cfset_ctrset_clear(); - debug_sprintf_event(cf_dbg, 4, "%s sets %#lx need %ld ret %d\n", - __func__, cfset_request.ctrset, need, ret); + if (put_user(need, &ustart->data_bytes)) { + kfree(preq); + return -EFAULT; + } + preq->ctrset = start.counter_sets; + ret = cfset_all_start(preq); + if (!ret) { + cfset_session_add(preq); + file->private_data = preq; + debug_sprintf_event(cf_dbg, 4, "%s set %#lx need %ld ret %d\n", + __func__, preq->ctrset, need, ret); + } else { + kfree(preq); + } return ret; } @@ -1136,7 +1196,7 @@ out: * counter set keeps running until explicitly stopped. Returns the number * of bytes needed to store the counter values. If another S390_HWCTR_START * ioctl subcommand is called without a previous S390_HWCTR_STOP stop - * command, -EBUSY is returned. + * command on the same file descriptor, -EBUSY is returned. * S390_HWCTR_READ: Read the counter set values from specified CPU list given * with the S390_HWCTR_START command. * S390_HWCTR_STOP: Stops the counter sets on the CPU list given with the @@ -1150,13 +1210,13 @@ static long cfset_ioctl(struct file *file, unsigned int cmd, unsigned long arg) mutex_lock(&cfset_ctrset_mutex); switch (cmd) { case S390_HWCTR_START: - ret = cfset_ioctl_start(arg); + ret = cfset_ioctl_start(arg, file); break; case S390_HWCTR_STOP: - ret = cfset_ioctl_stop(); + ret = cfset_ioctl_stop(file); break; case S390_HWCTR_READ: - ret = cfset_ioctl_read(arg); + ret = cfset_ioctl_read(arg, file->private_data); break; default: ret = -ENOTTY; @@ -1182,29 +1242,41 @@ static struct miscdevice cfset_dev = { .fops = &cfset_fops, }; +/* Hotplug add of a CPU. Scan through all active processes and add + * that CPU to the list of CPUs supplied with ioctl(..., START, ...). + */ int cfset_online_cpu(unsigned int cpu) { struct cfset_call_on_cpu_parm p; + struct cfset_request *rp; mutex_lock(&cfset_ctrset_mutex); - if (cfset_request.ctrset) { - p.sets = cfset_request.ctrset; - cfset_ioctl_on(&p); - cpumask_set_cpu(cpu, &cfset_request.mask); + if (!list_empty(&cfset_session.head)) { + list_for_each_entry(rp, &cfset_session.head, node) { + p.sets = rp->ctrset; + cfset_ioctl_on(&p); + cpumask_set_cpu(cpu, &rp->mask); + } } mutex_unlock(&cfset_ctrset_mutex); return 0; } +/* Hotplug remove of a CPU. Scan through all active processes and clear + * that CPU from the list of CPUs supplied with ioctl(..., START, ...). + */ int cfset_offline_cpu(unsigned int cpu) { struct cfset_call_on_cpu_parm p; + struct cfset_request *rp; mutex_lock(&cfset_ctrset_mutex); - if (cfset_request.ctrset) { - p.sets = cfset_request.ctrset; - cfset_ioctl_off(&p); - cpumask_clear_cpu(cpu, &cfset_request.mask); + if (!list_empty(&cfset_session.head)) { + list_for_each_entry(rp, &cfset_session.head, node) { + p.sets = rp->ctrset; + cfset_ioctl_off(&p); + cpumask_clear_cpu(cpu, &rp->mask); + } } mutex_unlock(&cfset_ctrset_mutex); return 0; diff --git a/arch/s390/kernel/process.c b/arch/s390/kernel/process.c index 350e94d0cac2..e8858b2de24b 100644 --- a/arch/s390/kernel/process.c +++ b/arch/s390/kernel/process.c @@ -141,7 +141,7 @@ int copy_thread(unsigned long clone_flags, unsigned long new_stackp, frame->childregs.gprs[10] = arg; frame->childregs.gprs[11] = (unsigned long)do_exit; frame->childregs.orig_gpr2 = -1; - + frame->childregs.last_break = 1; return 0; } frame->childregs = *current_pt_regs(); @@ -181,12 +181,12 @@ void execve_tail(void) asm volatile("sfpc %0" : : "d" (0)); } -unsigned long get_wchan(struct task_struct *p) +unsigned long __get_wchan(struct task_struct *p) { struct unwind_state state; unsigned long ip = 0; - if (!p || p == current || task_is_running(p) || !task_stack_page(p)) + if (!task_stack_page(p)) return 0; if (!try_get_task_stack(p)) diff --git a/arch/s390/kernel/setup.c b/arch/s390/kernel/setup.c index 67e5fff96ee0..40405f2304f1 100644 --- a/arch/s390/kernel/setup.c +++ b/arch/s390/kernel/setup.c @@ -95,10 +95,10 @@ EXPORT_SYMBOL(console_irq); * relocated above 2 GB, because it has to use 31 bit addresses. * Such code and data is part of the .amode31 section. */ -unsigned long __amode31_ref __samode31 = __pa(&_samode31); -unsigned long __amode31_ref __eamode31 = __pa(&_eamode31); -unsigned long __amode31_ref __stext_amode31 = __pa(&_stext_amode31); -unsigned long __amode31_ref __etext_amode31 = __pa(&_etext_amode31); +unsigned long __amode31_ref __samode31 = (unsigned long)&_samode31; +unsigned long __amode31_ref __eamode31 = (unsigned long)&_eamode31; +unsigned long __amode31_ref __stext_amode31 = (unsigned long)&_stext_amode31; +unsigned long __amode31_ref __etext_amode31 = (unsigned long)&_etext_amode31; struct exception_table_entry __amode31_ref *__start_amode31_ex_table = _start_amode31_ex_table; struct exception_table_entry __amode31_ref *__stop_amode31_ex_table = _stop_amode31_ex_table; @@ -149,6 +149,7 @@ struct mem_detect_info __bootdata(mem_detect); struct initrd_data __bootdata(initrd_data); unsigned long __bootdata_preserved(__kaslr_offset); +unsigned long __bootdata(__amode31_base); unsigned int __bootdata_preserved(zlib_dfltcc_support); EXPORT_SYMBOL(zlib_dfltcc_support); u64 __bootdata_preserved(stfle_fac_list[16]); @@ -173,6 +174,8 @@ unsigned long MODULES_END; struct lowcore *lowcore_ptr[NR_CPUS]; EXPORT_SYMBOL(lowcore_ptr); +DEFINE_STATIC_KEY_FALSE(cpu_has_bear); + /* * The Write Back bit position in the physaddr is given by the SLPC PCI. * Leaving the mask zero always uses write through which is safe @@ -593,7 +596,8 @@ static void __init setup_resources(void) * part of the System RAM resource. */ if (crashk_res.end) { - memblock_add_node(crashk_res.start, resource_size(&crashk_res), 0); + memblock_add_node(crashk_res.start, resource_size(&crashk_res), + 0, MEMBLOCK_NONE); memblock_reserve(crashk_res.start, resource_size(&crashk_res)); insert_resource(&iomem_resource, &crashk_res); } @@ -693,7 +697,7 @@ static void __init reserve_crashkernel(void) } if (register_memory_notifier(&kdump_mem_nb)) { - memblock_free(crash_base, crash_size); + memblock_phys_free(crash_base, crash_size); return; } @@ -718,7 +722,7 @@ static void __init reserve_initrd(void) #ifdef CONFIG_BLK_DEV_INITRD if (!initrd_data.start || !initrd_data.size) return; - initrd_start = initrd_data.start; + initrd_start = (unsigned long)__va(initrd_data.start); initrd_end = initrd_start + initrd_data.size; memblock_reserve(initrd_data.start, initrd_data.size); #endif @@ -748,7 +752,7 @@ static void __init free_mem_detect_info(void) get_mem_detect_reserved(&start, &size); if (size) - memblock_free(start, size); + memblock_phys_free(start, size); } static const char * __init get_mem_info_source(void) @@ -793,7 +797,7 @@ static void __init check_initrd(void) if (initrd_data.start && initrd_data.size && !memblock_is_region_memory(initrd_data.start, initrd_data.size)) { pr_err("The initial RAM disk does not fit into the memory\n"); - memblock_free(initrd_data.start, initrd_data.size); + memblock_phys_free(initrd_data.start, initrd_data.size); initrd_start = initrd_end = 0; } #endif @@ -804,12 +808,10 @@ static void __init check_initrd(void) */ static void __init reserve_kernel(void) { - unsigned long start_pfn = PFN_UP(__pa(_end)); - memblock_reserve(0, STARTUP_NORMAL_OFFSET); - memblock_reserve((unsigned long)sclp_early_sccb, EXT_SCCB_READ_SCP); - memblock_reserve((unsigned long)_stext, PFN_PHYS(start_pfn) - - (unsigned long)_stext); + memblock_reserve(__amode31_base, __eamode31 - __samode31); + memblock_reserve(__pa(sclp_early_sccb), EXT_SCCB_READ_SCP); + memblock_reserve(__pa(_stext), _end - _stext); } static void __init setup_memory(void) @@ -831,20 +833,14 @@ static void __init setup_memory(void) static void __init relocate_amode31_section(void) { - unsigned long amode31_addr, amode31_size; - long amode31_offset; + unsigned long amode31_size = __eamode31 - __samode31; + long amode31_offset = __amode31_base - __samode31; long *ptr; - /* Allocate a new AMODE31 capable memory region */ - amode31_size = __eamode31 - __samode31; pr_info("Relocating AMODE31 section of size 0x%08lx\n", amode31_size); - amode31_addr = (unsigned long)memblock_alloc_low(amode31_size, PAGE_SIZE); - if (!amode31_addr) - panic("Failed to allocate memory for AMODE31 section\n"); - amode31_offset = amode31_addr - __samode31; /* Move original AMODE31 section to the new one */ - memmove((void *)amode31_addr, (void *)__samode31, amode31_size); + memmove((void *)__amode31_base, (void *)__samode31, amode31_size); /* Zero out the old AMODE31 section to catch invalid accesses within it */ memset((void *)__samode31, 0, amode31_size); @@ -883,14 +879,12 @@ static void __init setup_randomness(void) { struct sysinfo_3_2_2 *vmms; - vmms = (struct sysinfo_3_2_2 *) memblock_phys_alloc(PAGE_SIZE, - PAGE_SIZE); + vmms = memblock_alloc(PAGE_SIZE, PAGE_SIZE); if (!vmms) panic("Failed to allocate memory for sysinfo structure\n"); - if (stsi(vmms, 3, 2, 2) == 0 && vmms->count) add_device_randomness(&vmms->vm, sizeof(vmms->vm[0]) * vmms->count); - memblock_free((unsigned long) vmms, PAGE_SIZE); + memblock_free(vmms, PAGE_SIZE); } /* @@ -1047,6 +1041,9 @@ void __init setup_arch(char **cmdline_p) smp_detect_cpus(); topology_init_early(); + if (test_facility(193)) + static_branch_enable(&cpu_has_bear); + /* * Create kernel page tables and switch to virtual addressing. */ diff --git a/arch/s390/kernel/smp.c b/arch/s390/kernel/smp.c index 1a04e5bdf655..78a8ea6fd582 100644 --- a/arch/s390/kernel/smp.c +++ b/arch/s390/kernel/smp.c @@ -723,7 +723,7 @@ void __init smp_save_dump_cpus(void) /* Get the CPU registers */ smp_save_cpu_regs(sa, addr, is_boot_cpu, page); } - memblock_free(page, PAGE_SIZE); + memblock_phys_free(page, PAGE_SIZE); diag_amode31_ops.diag308_reset(); pcpu_set_smt(0); } @@ -880,7 +880,7 @@ void __init smp_detect_cpus(void) /* Add CPUs present at boot */ __smp_rescan_cpus(info, true); - memblock_free_early((unsigned long)info, sizeof(*info)); + memblock_phys_free((unsigned long)info, sizeof(*info)); } /* diff --git a/arch/s390/kernel/stacktrace.c b/arch/s390/kernel/stacktrace.c index 101477b3e263..b7bb1981e9ee 100644 --- a/arch/s390/kernel/stacktrace.c +++ b/arch/s390/kernel/stacktrace.c @@ -46,7 +46,7 @@ int arch_stack_walk_reliable(stack_trace_consume_fn consume_entry, * Mark stacktraces with kretprobed functions on them * as unreliable. */ - if (state.ip == (unsigned long)kretprobe_trampoline) + if (state.ip == (unsigned long)__kretprobe_trampoline) return -EINVAL; #endif diff --git a/arch/s390/kernel/syscall.c b/arch/s390/kernel/syscall.c index 8fe2d23b64f4..dc2355c623d6 100644 --- a/arch/s390/kernel/syscall.c +++ b/arch/s390/kernel/syscall.c @@ -154,6 +154,8 @@ void noinstr __do_syscall(struct pt_regs *regs, int per_trap) regs->psw = S390_lowcore.svc_old_psw; regs->int_code = S390_lowcore.svc_int_code; update_timer_sys(); + if (static_branch_likely(&cpu_has_bear)) + current->thread.last_break = regs->last_break; local_irq_enable(); regs->orig_gpr2 = regs->gprs[2]; diff --git a/arch/s390/kernel/traps.c b/arch/s390/kernel/traps.c index bcefc2173de4..6c6f7dcce1a5 100644 --- a/arch/s390/kernel/traps.c +++ b/arch/s390/kernel/traps.c @@ -300,7 +300,6 @@ static void (*pgm_check_table[128])(struct pt_regs *regs); void noinstr __do_pgm_check(struct pt_regs *regs) { - unsigned long last_break = S390_lowcore.breaking_event_addr; unsigned int trapnr; irqentry_state_t state; @@ -311,10 +310,11 @@ void noinstr __do_pgm_check(struct pt_regs *regs) if (user_mode(regs)) { update_timer_sys(); - if (last_break < 4096) - last_break = 1; - current->thread.last_break = last_break; - regs->args[0] = last_break; + if (!static_branch_likely(&cpu_has_bear)) { + if (regs->last_break < 4096) + regs->last_break = 1; + } + current->thread.last_break = regs->last_break; } if (S390_lowcore.pgm_code & 0x0200) { diff --git a/arch/s390/kernel/uv.c b/arch/s390/kernel/uv.c index 5a656c7b7a67..386d4e42b8d3 100644 --- a/arch/s390/kernel/uv.c +++ b/arch/s390/kernel/uv.c @@ -64,7 +64,7 @@ void __init setup_uv(void) } if (uv_init(uv_stor_base, uv_info.uv_base_stor_len)) { - memblock_free(uv_stor_base, uv_info.uv_base_stor_len); + memblock_phys_free(uv_stor_base, uv_info.uv_base_stor_len); goto fail; } @@ -100,7 +100,7 @@ static int uv_pin_shared(unsigned long paddr) * * @paddr: Absolute host address of page to be destroyed */ -int uv_destroy_page(unsigned long paddr) +static int uv_destroy_page(unsigned long paddr) { struct uv_cb_cfs uvcb = { .header.cmd = UVC_CMD_DESTR_SEC_STOR, @@ -121,6 +121,22 @@ int uv_destroy_page(unsigned long paddr) } /* + * The caller must already hold a reference to the page + */ +int uv_destroy_owned_page(unsigned long paddr) +{ + struct page *page = phys_to_page(paddr); + int rc; + + get_page(page); + rc = uv_destroy_page(paddr); + if (!rc) + clear_bit(PG_arch_1, &page->flags); + put_page(page); + return rc; +} + +/* * Requests the Ultravisor to encrypt a guest page and make it * accessible to the host for paging (export). * @@ -140,6 +156,22 @@ int uv_convert_from_secure(unsigned long paddr) } /* + * The caller must already hold a reference to the page + */ +int uv_convert_owned_from_secure(unsigned long paddr) +{ + struct page *page = phys_to_page(paddr); + int rc; + + get_page(page); + rc = uv_convert_from_secure(paddr); + if (!rc) + clear_bit(PG_arch_1, &page->flags); + put_page(page); + return rc; +} + +/* * Calculate the expected ref_count for a page that would otherwise have no * further pins. This was cribbed from similar functions in other places in * the kernel, but with some slight modifications. We know that a secure @@ -165,7 +197,7 @@ static int make_secure_pte(pte_t *ptep, unsigned long addr, { pte_t entry = READ_ONCE(*ptep); struct page *page; - int expected, rc = 0; + int expected, cc = 0; if (!pte_present(entry)) return -ENXIO; @@ -181,12 +213,25 @@ static int make_secure_pte(pte_t *ptep, unsigned long addr, if (!page_ref_freeze(page, expected)) return -EBUSY; set_bit(PG_arch_1, &page->flags); - rc = uv_call(0, (u64)uvcb); + /* + * If the UVC does not succeed or fail immediately, we don't want to + * loop for long, or we might get stall notifications. + * On the other hand, this is a complex scenario and we are holding a lot of + * locks, so we can't easily sleep and reschedule. We try only once, + * and if the UVC returned busy or partial completion, we return + * -EAGAIN and we let the callers deal with it. + */ + cc = __uv_call(0, (u64)uvcb); page_ref_unfreeze(page, expected); - /* Return -ENXIO if the page was not mapped, -EINVAL otherwise */ - if (rc) - rc = uvcb->rc == 0x10a ? -ENXIO : -EINVAL; - return rc; + /* + * Return -ENXIO if the page was not mapped, -EINVAL for other errors. + * If busy or partially completed, return -EAGAIN. + */ + if (cc == UVC_CC_OK) + return 0; + else if (cc == UVC_CC_BUSY || cc == UVC_CC_PARTIAL) + return -EAGAIN; + return uvcb->rc == 0x10a ? -ENXIO : -EINVAL; } /* @@ -212,7 +257,7 @@ again: uaddr = __gmap_translate(gmap, gaddr); if (IS_ERR_VALUE(uaddr)) goto out; - vma = find_vma(gmap->mm, uaddr); + vma = vma_lookup(gmap->mm, uaddr); if (!vma) goto out; /* @@ -239,6 +284,10 @@ out: mmap_read_unlock(gmap->mm); if (rc == -EAGAIN) { + /* + * If we are here because the UVC returned busy or partial + * completion, this is just a useless check, but it is safe. + */ wait_on_page_writeback(page); } else if (rc == -EBUSY) { /* diff --git a/arch/s390/kernel/vmlinux.lds.S b/arch/s390/kernel/vmlinux.lds.S index 63bdb9e1bfc1..42c43521878f 100644 --- a/arch/s390/kernel/vmlinux.lds.S +++ b/arch/s390/kernel/vmlinux.lds.S @@ -212,6 +212,7 @@ SECTIONS QUAD(__dynsym_start) /* dynsym_start */ QUAD(__rela_dyn_start) /* rela_dyn_start */ QUAD(__rela_dyn_end) /* rela_dyn_end */ + QUAD(_eamode31 - _samode31) /* amode31_size */ } :NONE /* Debugging sections. */ |