diff options
Diffstat (limited to 'arch/x86/kernel/apic')
| -rw-r--r-- | arch/x86/kernel/apic/Makefile | 2 | ||||
| -rw-r--r-- | arch/x86/kernel/apic/apic.c | 390 | ||||
| -rw-r--r-- | arch/x86/kernel/apic/apic_common.c | 15 | ||||
| -rw-r--r-- | arch/x86/kernel/apic/apic_flat_64.c | 137 | ||||
| -rw-r--r-- | arch/x86/kernel/apic/apic_noop.c | 6 | ||||
| -rw-r--r-- | arch/x86/kernel/apic/apic_numachip.c | 21 | ||||
| -rw-r--r-- | arch/x86/kernel/apic/bigsmp_32.c | 21 | ||||
| -rw-r--r-- | arch/x86/kernel/apic/io_apic.c | 856 | ||||
| -rw-r--r-- | arch/x86/kernel/apic/local.h | 5 | ||||
| -rw-r--r-- | arch/x86/kernel/apic/msi.c | 8 | ||||
| -rw-r--r-- | arch/x86/kernel/apic/probe_32.c | 10 | ||||
| -rw-r--r-- | arch/x86/kernel/apic/vector.c | 26 | ||||
| -rw-r--r-- | arch/x86/kernel/apic/x2apic_cluster.c | 12 | ||||
| -rw-r--r-- | arch/x86/kernel/apic/x2apic_phys.c | 13 | ||||
| -rw-r--r-- | arch/x86/kernel/apic/x2apic_uv_x.c | 70 |
15 files changed, 479 insertions, 1113 deletions
diff --git a/arch/x86/kernel/apic/Makefile b/arch/x86/kernel/apic/Makefile index 2ee867d796d9..3bf0487cf3b7 100644 --- a/arch/x86/kernel/apic/Makefile +++ b/arch/x86/kernel/apic/Makefile @@ -4,7 +4,7 @@ # # Leads to non-deterministic coverage that is not a function of syscall inputs. -# In particualr, smp_apic_timer_interrupt() is called in random places. +# In particular, smp_apic_timer_interrupt() is called in random places. KCOV_INSTRUMENT := n obj-$(CONFIG_X86_LOCAL_APIC) += apic.o apic_common.o apic_noop.o ipi.o vector.o init.o diff --git a/arch/x86/kernel/apic/apic.c b/arch/x86/kernel/apic/apic.c index 41093cf20acd..e893dc6f11c1 100644 --- a/arch/x86/kernel/apic/apic.c +++ b/arch/x86/kernel/apic/apic.c @@ -19,6 +19,7 @@ #include <linux/kernel_stat.h> #include <linux/mc146818rtc.h> #include <linux/acpi_pmtmr.h> +#include <linux/bitmap.h> #include <linux/clockchips.h> #include <linux/interrupt.h> #include <linux/memblock.h> @@ -67,10 +68,6 @@ #include "local.h" -unsigned int num_processors; - -unsigned disabled_cpus; - /* Processor that is doing the boot up */ u32 boot_cpu_physical_apicid __ro_after_init = BAD_APICID; EXPORT_SYMBOL_GPL(boot_cpu_physical_apicid); @@ -78,18 +75,6 @@ EXPORT_SYMBOL_GPL(boot_cpu_physical_apicid); u8 boot_cpu_apic_version __ro_after_init; /* - * Bitmask of physically existing CPUs: - */ -physid_mask_t phys_cpu_present_map; - -/* - * Processor to be disabled specified by kernel parameter - * disable_cpu_apicid=<int>, mostly used for the kdump 2nd kernel to - * avoid undefined behaviour caused by sending INIT from AP to BSP. - */ -static u32 disabled_cpu_apicid __ro_after_init = BAD_APICID; - -/* * This variable controls which CPUs receive external NMIs. By default, * external NMIs are delivered only to the BSP. */ @@ -108,14 +93,6 @@ static inline bool apic_accessible(void) return x2apic_mode || apic_mmio_base; } -/* - * Map cpu index to physical APIC ID - */ -DEFINE_EARLY_PER_CPU_READ_MOSTLY(u32, x86_cpu_to_apicid, BAD_APICID); -DEFINE_EARLY_PER_CPU_READ_MOSTLY(u32, x86_cpu_to_acpiid, U32_MAX); -EXPORT_EARLY_PER_CPU_SYMBOL(x86_cpu_to_apicid); -EXPORT_EARLY_PER_CPU_SYMBOL(x86_cpu_to_acpiid); - #ifdef CONFIG_X86_32 /* Local APIC was disabled by the BIOS and enabled by the kernel */ static int enabled_via_apicbase __ro_after_init; @@ -261,16 +238,6 @@ u64 native_apic_icr_read(void) return icr1 | ((u64)icr2 << 32); } -#ifdef CONFIG_X86_32 -/** - * get_physical_broadcast - Get number of physical broadcast IDs - */ -int get_physical_broadcast(void) -{ - return modern_apic() ? 0xff : 0xf; -} -#endif - /** * lapic_get_maxlvt - get the maximum number of local vector table entries */ @@ -473,7 +440,19 @@ static int lapic_timer_shutdown(struct clock_event_device *evt) v = apic_read(APIC_LVTT); v |= (APIC_LVT_MASKED | LOCAL_TIMER_VECTOR); apic_write(APIC_LVTT, v); - apic_write(APIC_TMICT, 0); + + /* + * Setting APIC_LVT_MASKED (above) should be enough to tell + * the hardware that this timer will never fire. But AMD + * erratum 411 and some Intel CPU behavior circa 2024 say + * otherwise. Time for belt and suspenders programming: mask + * the timer _and_ zero the counter registers: + */ + if (v & APIC_LVT_TIMER_TSCDEADLINE) + wrmsrl(MSR_IA32_TSC_DEADLINE, 0); + else + apic_write(APIC_TMICT, 0); + return 0; } @@ -530,32 +509,32 @@ static struct clock_event_device lapic_clockevent = { static DEFINE_PER_CPU(struct clock_event_device, lapic_events); static const struct x86_cpu_id deadline_match[] __initconst = { - X86_MATCH_INTEL_FAM6_MODEL_STEPPINGS(HASWELL_X, X86_STEPPINGS(0x2, 0x2), 0x3a), /* EP */ - X86_MATCH_INTEL_FAM6_MODEL_STEPPINGS(HASWELL_X, X86_STEPPINGS(0x4, 0x4), 0x0f), /* EX */ + X86_MATCH_VFM_STEPS(INTEL_HASWELL_X, 0x2, 0x2, 0x3a), /* EP */ + X86_MATCH_VFM_STEPS(INTEL_HASWELL_X, 0x4, 0x4, 0x0f), /* EX */ - X86_MATCH_INTEL_FAM6_MODEL( BROADWELL_X, 0x0b000020), + X86_MATCH_VFM(INTEL_BROADWELL_X, 0x0b000020), - X86_MATCH_INTEL_FAM6_MODEL_STEPPINGS(BROADWELL_D, X86_STEPPINGS(0x2, 0x2), 0x00000011), - X86_MATCH_INTEL_FAM6_MODEL_STEPPINGS(BROADWELL_D, X86_STEPPINGS(0x3, 0x3), 0x0700000e), - X86_MATCH_INTEL_FAM6_MODEL_STEPPINGS(BROADWELL_D, X86_STEPPINGS(0x4, 0x4), 0x0f00000c), - X86_MATCH_INTEL_FAM6_MODEL_STEPPINGS(BROADWELL_D, X86_STEPPINGS(0x5, 0x5), 0x0e000003), + X86_MATCH_VFM_STEPS(INTEL_BROADWELL_D, 0x2, 0x2, 0x00000011), + X86_MATCH_VFM_STEPS(INTEL_BROADWELL_D, 0x3, 0x3, 0x0700000e), + X86_MATCH_VFM_STEPS(INTEL_BROADWELL_D, 0x4, 0x4, 0x0f00000c), + X86_MATCH_VFM_STEPS(INTEL_BROADWELL_D, 0x5, 0x5, 0x0e000003), - X86_MATCH_INTEL_FAM6_MODEL_STEPPINGS(SKYLAKE_X, X86_STEPPINGS(0x3, 0x3), 0x01000136), - X86_MATCH_INTEL_FAM6_MODEL_STEPPINGS(SKYLAKE_X, X86_STEPPINGS(0x4, 0x4), 0x02000014), - X86_MATCH_INTEL_FAM6_MODEL_STEPPINGS(SKYLAKE_X, X86_STEPPINGS(0x5, 0xf), 0), + X86_MATCH_VFM_STEPS(INTEL_SKYLAKE_X, 0x3, 0x3, 0x01000136), + X86_MATCH_VFM_STEPS(INTEL_SKYLAKE_X, 0x4, 0x4, 0x02000014), + X86_MATCH_VFM_STEPS(INTEL_SKYLAKE_X, 0x5, 0xf, 0), - X86_MATCH_INTEL_FAM6_MODEL( HASWELL, 0x22), - X86_MATCH_INTEL_FAM6_MODEL( HASWELL_L, 0x20), - X86_MATCH_INTEL_FAM6_MODEL( HASWELL_G, 0x17), + X86_MATCH_VFM(INTEL_HASWELL, 0x22), + X86_MATCH_VFM(INTEL_HASWELL_L, 0x20), + X86_MATCH_VFM(INTEL_HASWELL_G, 0x17), - X86_MATCH_INTEL_FAM6_MODEL( BROADWELL, 0x25), - X86_MATCH_INTEL_FAM6_MODEL( BROADWELL_G, 0x17), + X86_MATCH_VFM(INTEL_BROADWELL, 0x25), + X86_MATCH_VFM(INTEL_BROADWELL_G, 0x17), - X86_MATCH_INTEL_FAM6_MODEL( SKYLAKE_L, 0xb2), - X86_MATCH_INTEL_FAM6_MODEL( SKYLAKE, 0xb2), + X86_MATCH_VFM(INTEL_SKYLAKE_L, 0xb2), + X86_MATCH_VFM(INTEL_SKYLAKE, 0xb2), - X86_MATCH_INTEL_FAM6_MODEL( KABYLAKE_L, 0x52), - X86_MATCH_INTEL_FAM6_MODEL( KABYLAKE, 0x52), + X86_MATCH_VFM(INTEL_KABYLAKE_L, 0x52), + X86_MATCH_VFM(INTEL_KABYLAKE, 0x52), {}, }; @@ -664,7 +643,7 @@ void lapic_update_tsc_freq(void) static __initdata int lapic_cal_loops = -1; static __initdata long lapic_cal_t1, lapic_cal_t2; static __initdata unsigned long long lapic_cal_tsc1, lapic_cal_tsc2; -static __initdata unsigned long lapic_cal_pm1, lapic_cal_pm2; +static __initdata u32 lapic_cal_pm1, lapic_cal_pm2; static __initdata unsigned long lapic_cal_j1, lapic_cal_j2; /* @@ -674,7 +653,7 @@ static void __init lapic_cal_handler(struct clock_event_device *dev) { unsigned long long tsc = 0; long tapic = apic_read(APIC_TMCCT); - unsigned long pm = acpi_pm_read_early(); + u32 pm = acpi_pm_read_early(); if (boot_cpu_has(X86_FEATURE_TSC)) tsc = rdtsc(); @@ -699,7 +678,7 @@ static void __init lapic_cal_handler(struct clock_event_device *dev) } static int __init -calibrate_by_pmtimer(long deltapm, long *delta, long *deltatsc) +calibrate_by_pmtimer(u32 deltapm, long *delta, long *deltatsc) { const long pm_100ms = PMTMR_TICKS_PER_SEC / 10; const long pm_thresh = pm_100ms / 100; @@ -710,7 +689,7 @@ calibrate_by_pmtimer(long deltapm, long *delta, long *deltatsc) return -1; #endif - apic_printk(APIC_VERBOSE, "... PM-Timer delta = %ld\n", deltapm); + apic_pr_verbose("... PM-Timer delta = %u\n", deltapm); /* Check, if the PM timer is available */ if (!deltapm) @@ -720,14 +699,14 @@ calibrate_by_pmtimer(long deltapm, long *delta, long *deltatsc) if (deltapm > (pm_100ms - pm_thresh) && deltapm < (pm_100ms + pm_thresh)) { - apic_printk(APIC_VERBOSE, "... PM-Timer result ok\n"); + apic_pr_verbose("... PM-Timer result ok\n"); return 0; } res = (((u64)deltapm) * mult) >> 22; do_div(res, 1000000); - pr_warn("APIC calibration not consistent " - "with PM-Timer: %ldms instead of 100ms\n", (long)res); + pr_warn("APIC calibration not consistent with PM-Timer: %ldms instead of 100ms\n", + (long)res); /* Correct the lapic counter value */ res = (((u64)(*delta)) * pm_100ms); @@ -740,9 +719,8 @@ calibrate_by_pmtimer(long deltapm, long *delta, long *deltatsc) if (boot_cpu_has(X86_FEATURE_TSC)) { res = (((u64)(*deltatsc)) * pm_100ms); do_div(res, deltapm); - apic_printk(APIC_VERBOSE, "TSC delta adjusted to " - "PM-Timer: %lu (%ld)\n", - (unsigned long)res, *deltatsc); + apic_pr_verbose("TSC delta adjusted to PM-Timer: %lu (%ld)\n", + (unsigned long)res, *deltatsc); *deltatsc = (long)res; } @@ -782,7 +760,7 @@ bool __init apic_needs_pit(void) /* * If interrupt delivery mode is legacy PIC or virtual wire without - * configuration, the local APIC timer wont be set up. Make sure + * configuration, the local APIC timer won't be set up. Make sure * that the PIT is initialized. */ if (apic_intr_mode == APIC_PIC || @@ -825,8 +803,7 @@ static int __init calibrate_APIC_clock(void) * in the clockevent structure and return. */ if (!lapic_init_clockevent()) { - apic_printk(APIC_VERBOSE, "lapic timer already calibrated %d\n", - lapic_timer_period); + apic_pr_verbose("lapic timer already calibrated %d\n", lapic_timer_period); /* * Direct calibration methods must have an always running * local APIC timer, no need for broadcast timer. @@ -835,8 +812,7 @@ static int __init calibrate_APIC_clock(void) return 0; } - apic_printk(APIC_VERBOSE, "Using local APIC timer interrupts.\n" - "calibrating APIC timer ...\n"); + apic_pr_verbose("Using local APIC timer interrupts. Calibrating APIC timer ...\n"); /* * There are platforms w/o global clockevent devices. Instead of @@ -899,7 +875,7 @@ static int __init calibrate_APIC_clock(void) /* Build delta t1-t2 as apic timer counts down */ delta = lapic_cal_t1 - lapic_cal_t2; - apic_printk(APIC_VERBOSE, "... lapic delta = %ld\n", delta); + apic_pr_verbose("... lapic delta = %ld\n", delta); deltatsc = (long)(lapic_cal_tsc2 - lapic_cal_tsc1); @@ -910,22 +886,19 @@ static int __init calibrate_APIC_clock(void) lapic_timer_period = (delta * APIC_DIVISOR) / LAPIC_CAL_LOOPS; lapic_init_clockevent(); - apic_printk(APIC_VERBOSE, "..... delta %ld\n", delta); - apic_printk(APIC_VERBOSE, "..... mult: %u\n", lapic_clockevent.mult); - apic_printk(APIC_VERBOSE, "..... calibration result: %u\n", - lapic_timer_period); + apic_pr_verbose("..... delta %ld\n", delta); + apic_pr_verbose("..... mult: %u\n", lapic_clockevent.mult); + apic_pr_verbose("..... calibration result: %u\n", lapic_timer_period); if (boot_cpu_has(X86_FEATURE_TSC)) { - apic_printk(APIC_VERBOSE, "..... CPU clock speed is " - "%ld.%04ld MHz.\n", - (deltatsc / LAPIC_CAL_LOOPS) / (1000000 / HZ), - (deltatsc / LAPIC_CAL_LOOPS) % (1000000 / HZ)); + apic_pr_verbose("..... CPU clock speed is %ld.%04ld MHz.\n", + (deltatsc / LAPIC_CAL_LOOPS) / (1000000 / HZ), + (deltatsc / LAPIC_CAL_LOOPS) % (1000000 / HZ)); } - apic_printk(APIC_VERBOSE, "..... host bus clock speed is " - "%u.%04u MHz.\n", - lapic_timer_period / (1000000 / HZ), - lapic_timer_period % (1000000 / HZ)); + apic_pr_verbose("..... host bus clock speed is %u.%04u MHz.\n", + lapic_timer_period / (1000000 / HZ), + lapic_timer_period % (1000000 / HZ)); /* * Do a sanity check on the APIC calibration result @@ -944,7 +917,7 @@ static int __init calibrate_APIC_clock(void) * available. */ if (!pm_referenced && global_clock_event) { - apic_printk(APIC_VERBOSE, "... verify APIC timer\n"); + apic_pr_verbose("... verify APIC timer\n"); /* * Setup the apic timer manually @@ -965,11 +938,11 @@ static int __init calibrate_APIC_clock(void) /* Jiffies delta */ deltaj = lapic_cal_j2 - lapic_cal_j1; - apic_printk(APIC_VERBOSE, "... jiffies delta = %lu\n", deltaj); + apic_pr_verbose("... jiffies delta = %lu\n", deltaj); /* Check, if the jiffies result is consistent */ if (deltaj >= LAPIC_CAL_LOOPS-2 && deltaj <= LAPIC_CAL_LOOPS+2) - apic_printk(APIC_VERBOSE, "... jiffies result ok\n"); + apic_pr_verbose("... jiffies result ok\n"); else levt->features |= CLOCK_EVT_FEAT_DUMMY; } @@ -1254,9 +1227,8 @@ void __init sync_Arb_IDs(void) */ apic_wait_icr_idle(); - apic_printk(APIC_DEBUG, "Synchronizing Arb IDs.\n"); - apic_write(APIC_ICR, APIC_DEST_ALLINC | - APIC_INT_LEVELTRIG | APIC_DM_INIT); + apic_pr_debug("Synchronizing Arb IDs.\n"); + apic_write(APIC_ICR, APIC_DEST_ALLINC | APIC_INT_LEVELTRIG | APIC_DM_INIT); } enum apic_intr_mode_id apic_intr_mode __ro_after_init; @@ -1442,10 +1414,10 @@ static void lapic_setup_esr(void) if (maxlvt > 3) apic_write(APIC_ESR, 0); value = apic_read(APIC_ESR); - if (value != oldvalue) - apic_printk(APIC_VERBOSE, "ESR value before enabling " - "vector: 0x%08x after: 0x%08x\n", - oldvalue, value); + if (value != oldvalue) { + apic_pr_verbose("ESR value before enabling vector: 0x%08x after: 0x%08x\n", + oldvalue, value); + } } #define APIC_IR_REGS APIC_ISR_NR @@ -1549,9 +1521,6 @@ static void setup_local_APIC(void) apic_write(APIC_ESR, 0); } #endif - /* Validate that the APIC is registered if required */ - BUG_ON(apic->apic_id_registered && !apic->apic_id_registered()); - /* * Intel recommends to set DFR, LDR and TPR before enabling * an APIC. See e.g. "AP-388 82489DX User's Manual" (Intel @@ -1635,10 +1604,10 @@ static void setup_local_APIC(void) value = apic_read(APIC_LVT0) & APIC_LVT_MASKED; if (!cpu && (pic_mode || !value || ioapic_is_disabled)) { value = APIC_DM_EXTINT; - apic_printk(APIC_VERBOSE, "enabled ExtINT on CPU#%d\n", cpu); + apic_pr_verbose("Enabled ExtINT on CPU#%d\n", cpu); } else { value = APIC_DM_EXTINT | APIC_LVT_MASKED; - apic_printk(APIC_VERBOSE, "masked ExtINT on CPU#%d\n", cpu); + apic_pr_verbose("Masked ExtINT on CPU#%d\n", cpu); } apic_write(APIC_LVT0, value); @@ -1690,8 +1659,6 @@ void apic_ap_setup(void) end_local_APIC_setup(); } -static __init void cpu_set_boot_apic(void); - static __init void apic_read_boot_cpu_id(bool x2apic) { /* @@ -1706,7 +1673,8 @@ static __init void apic_read_boot_cpu_id(bool x2apic) boot_cpu_physical_apicid = read_apic_id(); boot_cpu_apic_version = GET_APIC_VERSION(apic_read(APIC_LVR)); } - cpu_set_boot_apic(); + topology_register_boot_apic(boot_cpu_physical_apicid); + x86_32_probe_bigsmp_early(); } #ifdef CONFIG_X86_X2APIC @@ -1724,11 +1692,11 @@ static int x2apic_state; static bool x2apic_hw_locked(void) { - u64 ia32_cap; + u64 x86_arch_cap_msr; u64 msr; - ia32_cap = x86_read_arch_cap_msr(); - if (ia32_cap & ARCH_CAP_XAPIC_DISABLE) { + x86_arch_cap_msr = x86_read_arch_cap_msr(); + if (x86_arch_cap_msr & ARCH_CAP_XAPIC_DISABLE) { rdmsrl(MSR_IA32_XAPIC_DISABLE_STATUS, msr); return (msr & LEGACY_XAPIC_DISABLED); } @@ -1808,16 +1776,13 @@ void x2apic_setup(void) __x2apic_enable(); } -static __init void apic_set_fixmap(void); +static __init void apic_set_fixmap(bool read_apic); static __init void x2apic_disable(void) { - u32 x2apic_id, state = x2apic_state; - - x2apic_mode = 0; - x2apic_state = X2APIC_DISABLED; + u32 x2apic_id; - if (state != X2APIC_ON) + if (x2apic_state < X2APIC_ON) return; x2apic_id = read_apic_id(); @@ -1830,7 +1795,16 @@ static __init void x2apic_disable(void) } __x2apic_disable(); - apic_set_fixmap(); + + x2apic_mode = 0; + x2apic_state = X2APIC_DISABLED; + + /* + * Don't reread the APIC ID as it was already done from + * check_x2apic() and the APIC driver still is a x2APIC variant, + * which fails to do the read after x2APIC was disabled. + */ + apic_set_fixmap(false); } static __init void x2apic_enable(void) @@ -2091,17 +2065,16 @@ void __init init_apic_mappings(void) pr_info("APIC: disable apic facility\n"); apic_disable(); } - num_processors = 1; } } -static __init void apic_set_fixmap(void) +static __init void apic_set_fixmap(bool read_apic) { set_fixmap_nocache(FIX_APIC_BASE, mp_lapic_addr); apic_mmio_base = APIC_BASE; - apic_printk(APIC_VERBOSE, "mapped APIC to %16lx (%16lx)\n", - apic_mmio_base, mp_lapic_addr); - apic_read_boot_cpu_id(false); + apic_pr_verbose("Mapped APIC to %16lx (%16lx)\n", apic_mmio_base, mp_lapic_addr); + if (read_apic) + apic_read_boot_cpu_id(false); } void __init register_lapic_address(unsigned long address) @@ -2111,7 +2084,7 @@ void __init register_lapic_address(unsigned long address) mp_lapic_addr = address; if (!x2apic_mode) - apic_set_fixmap(); + apic_set_fixmap(true); } /* @@ -2202,18 +2175,17 @@ DEFINE_IDTENTRY_SYSVEC(sysvec_error_interrupt) apic_eoi(); atomic_inc(&irq_err_count); - apic_printk(APIC_DEBUG, KERN_DEBUG "APIC error on CPU%d: %02x", - smp_processor_id(), v); + apic_pr_debug("APIC error on CPU%d: %02x", smp_processor_id(), v); v &= 0xff; while (v) { if (v & 0x1) - apic_printk(APIC_DEBUG, KERN_CONT " : %s", error_interrupt_reason[i]); + apic_pr_debug_cont(" : %s", error_interrupt_reason[i]); i++; v >>= 1; } - apic_printk(APIC_DEBUG, KERN_CONT "\n"); + apic_pr_debug_cont("\n"); trace_error_apic_exit(ERROR_APIC_VECTOR); } @@ -2233,8 +2205,7 @@ static void __init connect_bsp_APIC(void) * PIC mode, enable APIC mode in the IMCR, i.e. connect BSP's * local APIC to INT and NMI lines. */ - apic_printk(APIC_VERBOSE, "leaving PIC mode, " - "enabling APIC mode.\n"); + apic_pr_verbose("Leaving PIC mode, enabling APIC mode.\n"); imcr_pic_to_apic(); } #endif @@ -2259,8 +2230,7 @@ void disconnect_bsp_APIC(int virt_wire_setup) * IPIs, won't work beyond this point! The only exception are * INIT IPIs. */ - apic_printk(APIC_VERBOSE, "disabling APIC mode, " - "entering PIC mode.\n"); + apic_pr_verbose("Disabling APIC mode, entering PIC mode.\n"); imcr_apic_to_pic(); return; } @@ -2305,155 +2275,6 @@ void disconnect_bsp_APIC(int virt_wire_setup) apic_write(APIC_LVT1, value); } -/* - * The number of allocated logical CPU IDs. Since logical CPU IDs are allocated - * contiguously, it equals to current allocated max logical CPU ID plus 1. - * All allocated CPU IDs should be in the [0, nr_logical_cpuids) range, - * so the maximum of nr_logical_cpuids is nr_cpu_ids. - * - * NOTE: Reserve 0 for BSP. - */ -static int nr_logical_cpuids = 1; - -/* - * Used to store mapping between logical CPU IDs and APIC IDs. - */ -u32 cpuid_to_apicid[] = { [0 ... NR_CPUS - 1] = BAD_APICID, }; - -bool arch_match_cpu_phys_id(int cpu, u64 phys_id) -{ - return phys_id == (u64)cpuid_to_apicid[cpu]; -} - -#ifdef CONFIG_SMP -static void cpu_mark_primary_thread(unsigned int cpu, unsigned int apicid) -{ - /* Isolate the SMT bit(s) in the APICID and check for 0 */ - u32 mask = (1U << (fls(smp_num_siblings) - 1)) - 1; - - if (smp_num_siblings == 1 || !(apicid & mask)) - cpumask_set_cpu(cpu, &__cpu_primary_thread_mask); -} - -/* - * Due to the utter mess of CPUID evaluation smp_num_siblings is not valid - * during early boot. Initialize the primary thread mask before SMP - * bringup. - */ -static int __init smp_init_primary_thread_mask(void) -{ - unsigned int cpu; - - /* - * XEN/PV provides either none or useless topology information. - * Pretend that all vCPUs are primary threads. - */ - if (xen_pv_domain()) { - cpumask_copy(&__cpu_primary_thread_mask, cpu_possible_mask); - return 0; - } - - for (cpu = 0; cpu < nr_logical_cpuids; cpu++) - cpu_mark_primary_thread(cpu, cpuid_to_apicid[cpu]); - return 0; -} -early_initcall(smp_init_primary_thread_mask); -#else -static inline void cpu_mark_primary_thread(unsigned int cpu, unsigned int apicid) { } -#endif - -/* - * Should use this API to allocate logical CPU IDs to keep nr_logical_cpuids - * and cpuid_to_apicid[] synchronized. - */ -static int allocate_logical_cpuid(int apicid) -{ - int i; - - /* - * cpuid <-> apicid mapping is persistent, so when a cpu is up, - * check if the kernel has allocated a cpuid for it. - */ - for (i = 0; i < nr_logical_cpuids; i++) { - if (cpuid_to_apicid[i] == apicid) - return i; - } - - /* Allocate a new cpuid. */ - if (nr_logical_cpuids >= nr_cpu_ids) { - WARN_ONCE(1, "APIC: NR_CPUS/possible_cpus limit of %u reached. " - "Processor %d/0x%x and the rest are ignored.\n", - nr_cpu_ids, nr_logical_cpuids, apicid); - return -EINVAL; - } - - cpuid_to_apicid[nr_logical_cpuids] = apicid; - return nr_logical_cpuids++; -} - -static void cpu_update_apic(int cpu, u32 apicid) -{ -#if defined(CONFIG_SMP) || defined(CONFIG_X86_64) - early_per_cpu(x86_cpu_to_apicid, cpu) = apicid; -#endif - set_cpu_possible(cpu, true); - physid_set(apicid, phys_cpu_present_map); - set_cpu_present(cpu, true); - num_processors++; - - if (system_state != SYSTEM_BOOTING) - cpu_mark_primary_thread(cpu, apicid); -} - -static __init void cpu_set_boot_apic(void) -{ - cpuid_to_apicid[0] = boot_cpu_physical_apicid; - cpu_update_apic(0, boot_cpu_physical_apicid); - x86_32_probe_bigsmp_early(); -} - -int generic_processor_info(int apicid) -{ - int cpu, max = nr_cpu_ids; - - /* The boot CPU must be set before MADT/MPTABLE parsing happens */ - if (cpuid_to_apicid[0] == BAD_APICID) - panic("Boot CPU APIC not registered yet\n"); - - if (apicid == boot_cpu_physical_apicid) - return 0; - - if (disabled_cpu_apicid == apicid) { - int thiscpu = num_processors + disabled_cpus; - - pr_warn("APIC: Disabling requested cpu. Processor %d/0x%x ignored.\n", - thiscpu, apicid); - - disabled_cpus++; - return -ENODEV; - } - - if (num_processors >= nr_cpu_ids) { - int thiscpu = max + disabled_cpus; - - pr_warn("APIC: NR_CPUS/possible_cpus limit of %i reached. " - "Processor %d/0x%x ignored.\n", max, thiscpu, apicid); - - disabled_cpus++; - return -EINVAL; - } - - cpu = allocate_logical_cpuid(apicid); - if (cpu < 0) { - disabled_cpus++; - return -EINVAL; - } - - cpu_update_apic(cpu, apicid); - return cpu; -} - - void __irq_msi_compose_msg(struct irq_cfg *cfg, struct msi_msg *msg, bool dmar) { @@ -2496,10 +2317,7 @@ EXPORT_SYMBOL_GPL(x86_msi_msg_get_destid); static void __init apic_bsp_up_setup(void) { -#ifdef CONFIG_X86_64 - apic_write(APIC_ID, apic->set_apic_id(boot_cpu_physical_apicid)); -#endif - physid_set_mask_of_physid(boot_cpu_physical_apicid, &phys_cpu_present_map); + reset_phys_cpu_present_map(boot_cpu_physical_apicid); } /** @@ -2764,19 +2582,12 @@ int apic_is_clustered_box(void) /* * APIC command line parameters */ -static int __init setup_disableapic(char *arg) +static int __init setup_nolapic(char *arg) { apic_is_disabled = true; setup_clear_cpu_cap(X86_FEATURE_APIC); return 0; } -early_param("disableapic", setup_disableapic); - -/* same as disableapic, for compatibility */ -static int __init setup_nolapic(char *arg) -{ - return setup_disableapic(arg); -} early_param("nolapic", setup_nolapic); static int __init parse_lapic_timer_c2_ok(char *arg) @@ -2845,15 +2656,6 @@ static int __init lapic_insert_resource(void) */ late_initcall(lapic_insert_resource); -static int __init apic_set_disabled_cpu_apicid(char *arg) -{ - if (!arg || !get_option(&arg, &disabled_cpu_apicid)) - return -EINVAL; - - return 0; -} -early_param("disable_cpu_apicid", apic_set_disabled_cpu_apicid); - static int __init apic_set_extnmi(char *arg) { if (!arg) diff --git a/arch/x86/kernel/apic/apic_common.c b/arch/x86/kernel/apic/apic_common.c index 8a00141073ea..9ef3be866832 100644 --- a/arch/x86/kernel/apic/apic_common.c +++ b/arch/x86/kernel/apic/apic_common.c @@ -18,16 +18,6 @@ u32 apic_flat_calc_apicid(unsigned int cpu) return 1U << cpu; } -bool default_check_apicid_used(physid_mask_t *map, u32 apicid) -{ - return physid_isset(apicid, *map); -} - -void default_ioapic_phys_id_map(physid_mask_t *phys_map, physid_mask_t *retmap) -{ - *retmap = *phys_map; -} - u32 default_cpu_present_to_apicid(int mps_cpu) { if (mps_cpu < nr_cpu_ids && cpu_present(mps_cpu)) @@ -37,11 +27,6 @@ u32 default_cpu_present_to_apicid(int mps_cpu) } EXPORT_SYMBOL_GPL(default_cpu_present_to_apicid); -bool default_apic_id_registered(void) -{ - return physid_isset(read_apic_id(), phys_cpu_present_map); -} - /* * Set up the logical destination ID when the APIC operates in logical * destination mode. diff --git a/arch/x86/kernel/apic/apic_flat_64.c b/arch/x86/kernel/apic/apic_flat_64.c index 7139867d69cd..e0308d8c4e6c 100644 --- a/arch/x86/kernel/apic/apic_flat_64.c +++ b/arch/x86/kernel/apic/apic_flat_64.c @@ -8,143 +8,25 @@ * Martin Bligh, Andi Kleen, James Bottomley, John Stultz, and * James Cleverdon. */ -#include <linux/cpumask.h> #include <linux/export.h> -#include <linux/acpi.h> -#include <asm/jailhouse_para.h> #include <asm/apic.h> #include "local.h" -static struct apic apic_physflat; -static struct apic apic_flat; - -struct apic *apic __ro_after_init = &apic_flat; -EXPORT_SYMBOL_GPL(apic); - -static int flat_acpi_madt_oem_check(char *oem_id, char *oem_table_id) -{ - return 1; -} - -static void _flat_send_IPI_mask(unsigned long mask, int vector) -{ - unsigned long flags; - - local_irq_save(flags); - __default_send_IPI_dest_field(mask, vector, APIC_DEST_LOGICAL); - local_irq_restore(flags); -} - -static void flat_send_IPI_mask(const struct cpumask *cpumask, int vector) -{ - unsigned long mask = cpumask_bits(cpumask)[0]; - - _flat_send_IPI_mask(mask, vector); -} - -static void -flat_send_IPI_mask_allbutself(const struct cpumask *cpumask, int vector) -{ - unsigned long mask = cpumask_bits(cpumask)[0]; - int cpu = smp_processor_id(); - - if (cpu < BITS_PER_LONG) - __clear_bit(cpu, &mask); - - _flat_send_IPI_mask(mask, vector); -} - -static u32 flat_get_apic_id(u32 x) +static u32 physflat_get_apic_id(u32 x) { return (x >> 24) & 0xFF; } -static u32 set_apic_id(u32 id) -{ - return (id & 0xFF) << 24; -} - -static u32 flat_phys_pkg_id(u32 initial_apic_id, int index_msb) -{ - return initial_apic_id >> index_msb; -} - -static int flat_probe(void) +static int physflat_probe(void) { return 1; } -static struct apic apic_flat __ro_after_init = { - .name = "flat", - .probe = flat_probe, - .acpi_madt_oem_check = flat_acpi_madt_oem_check, - .apic_id_registered = default_apic_id_registered, - - .delivery_mode = APIC_DELIVERY_MODE_FIXED, - .dest_mode_logical = true, - - .disable_esr = 0, - - .init_apic_ldr = default_init_apic_ldr, - .cpu_present_to_apicid = default_cpu_present_to_apicid, - .phys_pkg_id = flat_phys_pkg_id, - - .max_apic_id = 0xFE, - .get_apic_id = flat_get_apic_id, - .set_apic_id = set_apic_id, - - .calc_dest_apicid = apic_flat_calc_apicid, - - .send_IPI = default_send_IPI_single, - .send_IPI_mask = flat_send_IPI_mask, - .send_IPI_mask_allbutself = flat_send_IPI_mask_allbutself, - .send_IPI_allbutself = default_send_IPI_allbutself, - .send_IPI_all = default_send_IPI_all, - .send_IPI_self = default_send_IPI_self, - .nmi_to_offline_cpu = true, - - .read = native_apic_mem_read, - .write = native_apic_mem_write, - .eoi = native_apic_mem_eoi, - .icr_read = native_apic_icr_read, - .icr_write = native_apic_icr_write, - .wait_icr_idle = apic_mem_wait_icr_idle, - .safe_wait_icr_idle = apic_mem_wait_icr_idle_timeout, -}; - -/* - * Physflat mode is used when there are more than 8 CPUs on a system. - * We cannot use logical delivery in this case because the mask - * overflows, so use physical mode. - */ static int physflat_acpi_madt_oem_check(char *oem_id, char *oem_table_id) { -#ifdef CONFIG_ACPI - /* - * Quirk: some x86_64 machines can only use physical APIC mode - * regardless of how many processors are present (x86_64 ES7000 - * is an example). - */ - if (acpi_gbl_FADT.header.revision >= FADT2_REVISION_ID && - (acpi_gbl_FADT.flags & ACPI_FADT_APIC_PHYSICAL)) { - printk(KERN_DEBUG "system APIC only can use physical flat"); - return 1; - } - - if (!strncmp(oem_id, "IBM", 3) && !strncmp(oem_table_id, "EXA", 3)) { - printk(KERN_DEBUG "IBM Summit detected, will use apic physical"); - return 1; - } -#endif - - return 0; -} - -static int physflat_probe(void) -{ - return apic == &apic_physflat || num_possible_cpus() > 8 || jailhouse_paravirt(); + return 1; } static struct apic apic_physflat __ro_after_init = { @@ -152,19 +34,15 @@ static struct apic apic_physflat __ro_after_init = { .name = "physical flat", .probe = physflat_probe, .acpi_madt_oem_check = physflat_acpi_madt_oem_check, - .apic_id_registered = default_apic_id_registered, - .delivery_mode = APIC_DELIVERY_MODE_FIXED, .dest_mode_logical = false, .disable_esr = 0, .cpu_present_to_apicid = default_cpu_present_to_apicid, - .phys_pkg_id = flat_phys_pkg_id, .max_apic_id = 0xFE, - .get_apic_id = flat_get_apic_id, - .set_apic_id = set_apic_id, + .get_apic_id = physflat_get_apic_id, .calc_dest_apicid = apic_default_calc_apicid, @@ -184,8 +62,7 @@ static struct apic apic_physflat __ro_after_init = { .wait_icr_idle = apic_mem_wait_icr_idle, .safe_wait_icr_idle = apic_mem_wait_icr_idle_timeout, }; +apic_driver(apic_physflat); -/* - * We need to check for physflat first, so this order is important. - */ -apic_drivers(apic_physflat, apic_flat); +struct apic *apic __ro_after_init = &apic_physflat; +EXPORT_SYMBOL_GPL(apic); diff --git a/arch/x86/kernel/apic/apic_noop.c b/arch/x86/kernel/apic/apic_noop.c index b00d52ae84fa..b5bb7a2e8340 100644 --- a/arch/x86/kernel/apic/apic_noop.c +++ b/arch/x86/kernel/apic/apic_noop.c @@ -29,7 +29,6 @@ static void noop_send_IPI_self(int vector) { } static void noop_apic_icr_write(u32 low, u32 id) { } static int noop_wakeup_secondary_cpu(u32 apicid, unsigned long start_eip) { return -1; } static u64 noop_apic_icr_read(void) { return 0; } -static u32 noop_phys_pkg_id(u32 cpuid_apic, int index_msb) { return 0; } static u32 noop_get_apic_id(u32 apicid) { return 0; } static void noop_apic_eoi(void) { } @@ -47,17 +46,12 @@ static void noop_apic_write(u32 reg, u32 val) struct apic apic_noop __ro_after_init = { .name = "noop", - .delivery_mode = APIC_DELIVERY_MODE_FIXED, .dest_mode_logical = true, .disable_esr = 0, - .check_apicid_used = default_check_apicid_used, - .ioapic_phys_id_map = default_ioapic_phys_id_map, .cpu_present_to_apicid = default_cpu_present_to_apicid, - .phys_pkg_id = noop_phys_pkg_id, - .max_apic_id = 0xFE, .get_apic_id = noop_get_apic_id, diff --git a/arch/x86/kernel/apic/apic_numachip.c b/arch/x86/kernel/apic/apic_numachip.c index 456a14c44f67..16410f087b7a 100644 --- a/arch/x86/kernel/apic/apic_numachip.c +++ b/arch/x86/kernel/apic/apic_numachip.c @@ -38,11 +38,6 @@ static u32 numachip1_get_apic_id(u32 x) return id; } -static u32 numachip1_set_apic_id(u32 id) -{ - return (id & 0xff) << 24; -} - static u32 numachip2_get_apic_id(u32 x) { u64 mcfg; @@ -51,16 +46,6 @@ static u32 numachip2_get_apic_id(u32 x) return ((mcfg >> (28 - 8)) & 0xfff00) | (x >> 24); } -static u32 numachip2_set_apic_id(u32 id) -{ - return id << 24; -} - -static u32 numachip_phys_pkg_id(u32 initial_apic_id, int index_msb) -{ - return initial_apic_id >> index_msb; -} - static void numachip1_apic_icr_write(int apicid, unsigned int val) { write_lcsr(CSR_G3_EXT_IRQ_GEN, (apicid << 16) | val); @@ -222,17 +207,14 @@ static const struct apic apic_numachip1 __refconst = { .probe = numachip1_probe, .acpi_madt_oem_check = numachip1_acpi_madt_oem_check, - .delivery_mode = APIC_DELIVERY_MODE_FIXED, .dest_mode_logical = false, .disable_esr = 0, .cpu_present_to_apicid = default_cpu_present_to_apicid, - .phys_pkg_id = numachip_phys_pkg_id, .max_apic_id = UINT_MAX, .get_apic_id = numachip1_get_apic_id, - .set_apic_id = numachip1_set_apic_id, .calc_dest_apicid = apic_default_calc_apicid, @@ -259,17 +241,14 @@ static const struct apic apic_numachip2 __refconst = { .probe = numachip2_probe, .acpi_madt_oem_check = numachip2_acpi_madt_oem_check, - .delivery_mode = APIC_DELIVERY_MODE_FIXED, .dest_mode_logical = false, .disable_esr = 0, .cpu_present_to_apicid = default_cpu_present_to_apicid, - .phys_pkg_id = numachip_phys_pkg_id, .max_apic_id = UINT_MAX, .get_apic_id = numachip2_get_apic_id, - .set_apic_id = numachip2_set_apic_id, .calc_dest_apicid = apic_default_calc_apicid, diff --git a/arch/x86/kernel/apic/bigsmp_32.c b/arch/x86/kernel/apic/bigsmp_32.c index 7ee3c486cb33..9285d500d5b4 100644 --- a/arch/x86/kernel/apic/bigsmp_32.c +++ b/arch/x86/kernel/apic/bigsmp_32.c @@ -18,22 +18,6 @@ static u32 bigsmp_get_apic_id(u32 x) return (x >> 24) & 0xFF; } -static bool bigsmp_check_apicid_used(physid_mask_t *map, u32 apicid) -{ - return false; -} - -static void bigsmp_ioapic_phys_id_map(physid_mask_t *phys_map, physid_mask_t *retmap) -{ - /* For clustered we don't have a good way to do this yet - hack */ - physids_promote(0xFFL, retmap); -} - -static u32 bigsmp_phys_pkg_id(u32 cpuid_apic, int index_msb) -{ - return cpuid_apic >> index_msb; -} - static void bigsmp_send_IPI_allbutself(int vector) { default_send_IPI_mask_allbutself_phys(cpu_online_mask, vector); @@ -80,19 +64,14 @@ static struct apic apic_bigsmp __ro_after_init = { .name = "bigsmp", .probe = probe_bigsmp, - .delivery_mode = APIC_DELIVERY_MODE_FIXED, .dest_mode_logical = false, .disable_esr = 1, - .check_apicid_used = bigsmp_check_apicid_used, - .ioapic_phys_id_map = bigsmp_ioapic_phys_id_map, .cpu_present_to_apicid = default_cpu_present_to_apicid, - .phys_pkg_id = bigsmp_phys_pkg_id, .max_apic_id = 0xFE, .get_apic_id = bigsmp_get_apic_id, - .set_apic_id = NULL, .calc_dest_apicid = apic_default_calc_apicid, diff --git a/arch/x86/kernel/apic/io_apic.c b/arch/x86/kernel/apic/io_apic.c index 00da6cf6b07d..eebc360ed1bb 100644 --- a/arch/x86/kernel/apic/io_apic.c +++ b/arch/x86/kernel/apic/io_apic.c @@ -86,8 +86,8 @@ static unsigned int ioapic_dynirq_base; static int ioapic_initialized; struct irq_pin_list { - struct list_head list; - int apic, pin; + struct list_head list; + int apic, pin; }; struct mp_chip_data { @@ -96,7 +96,7 @@ struct mp_chip_data { bool is_level; bool active_low; bool isa_irq; - u32 count; + u32 count; }; struct mp_ioapic_gsi { @@ -105,21 +105,17 @@ struct mp_ioapic_gsi { }; static struct ioapic { - /* - * # of IRQ routing registers - */ - int nr_registers; - /* - * Saved state during suspend/resume, or while enabling intr-remap. - */ - struct IO_APIC_route_entry *saved_registers; + /* # of IRQ routing registers */ + int nr_registers; + /* Saved state during suspend/resume, or while enabling intr-remap. */ + struct IO_APIC_route_entry *saved_registers; /* I/O APIC config */ - struct mpc_ioapic mp_config; + struct mpc_ioapic mp_config; /* IO APIC gsi routing info */ - struct mp_ioapic_gsi gsi_config; - struct ioapic_domain_cfg irqdomain_cfg; - struct irq_domain *irqdomain; - struct resource *iomem_res; + struct mp_ioapic_gsi gsi_config; + struct ioapic_domain_cfg irqdomain_cfg; + struct irq_domain *irqdomain; + struct resource *iomem_res; } ioapics[MAX_IO_APICS]; #define mpc_ioapic_ver(ioapic_idx) ioapics[ioapic_idx].mp_config.apicver @@ -205,10 +201,9 @@ void mp_save_irq(struct mpc_intsrc *m) { int i; - apic_printk(APIC_VERBOSE, "Int: type %d, pol %d, trig %d, bus %02x," - " IRQ %02x, APIC ID %x, APIC INT %02x\n", - m->irqtype, m->irqflag & 3, (m->irqflag >> 2) & 3, m->srcbus, - m->srcbusirq, m->dstapic, m->dstirq); + apic_pr_verbose("Int: type %d, pol %d, trig %d, bus %02x, IRQ %02x, APIC ID %x, APIC INT %02x\n", + m->irqtype, m->irqflag & 3, (m->irqflag >> 2) & 3, m->srcbus, + m->srcbusirq, m->dstapic, m->dstirq); for (i = 0; i < mp_irq_entries; i++) { if (!memcmp(&mp_irqs[i], m, sizeof(*m))) @@ -269,12 +264,14 @@ static __attribute_const__ struct io_apic __iomem *io_apic_base(int idx) static inline void io_apic_eoi(unsigned int apic, unsigned int vector) { struct io_apic __iomem *io_apic = io_apic_base(apic); + writel(vector, &io_apic->eoi); } unsigned int native_io_apic_read(unsigned int apic, unsigned int reg) { struct io_apic __iomem *io_apic = io_apic_base(apic); + writel(reg, &io_apic->index); return readl(&io_apic->data); } @@ -300,14 +297,8 @@ static struct IO_APIC_route_entry __ioapic_read_entry(int apic, int pin) static struct IO_APIC_route_entry ioapic_read_entry(int apic, int pin) { - struct IO_APIC_route_entry entry; - unsigned long flags; - - raw_spin_lock_irqsave(&ioapic_lock, flags); - entry = __ioapic_read_entry(apic, pin); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); - - return entry; + guard(raw_spinlock_irqsave)(&ioapic_lock); + return __ioapic_read_entry(apic, pin); } /* @@ -324,11 +315,8 @@ static void __ioapic_write_entry(int apic, int pin, struct IO_APIC_route_entry e static void ioapic_write_entry(int apic, int pin, struct IO_APIC_route_entry e) { - unsigned long flags; - - raw_spin_lock_irqsave(&ioapic_lock, flags); + guard(raw_spinlock_irqsave)(&ioapic_lock); __ioapic_write_entry(apic, pin, e); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); } /* @@ -339,12 +327,10 @@ static void ioapic_write_entry(int apic, int pin, struct IO_APIC_route_entry e) static void ioapic_mask_entry(int apic, int pin) { struct IO_APIC_route_entry e = { .masked = true }; - unsigned long flags; - raw_spin_lock_irqsave(&ioapic_lock, flags); + guard(raw_spinlock_irqsave)(&ioapic_lock); io_apic_write(apic, 0x10 + 2*pin, e.w1); io_apic_write(apic, 0x11 + 2*pin, e.w2); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); } /* @@ -352,68 +338,39 @@ static void ioapic_mask_entry(int apic, int pin) * shared ISA-space IRQs, so we have to support them. We are super * fast in the common case, and fast for shared ISA-space IRQs. */ -static int __add_pin_to_irq_node(struct mp_chip_data *data, - int node, int apic, int pin) +static bool add_pin_to_irq_node(struct mp_chip_data *data, int node, int apic, int pin) { struct irq_pin_list *entry; - /* don't allow duplicates */ - for_each_irq_pin(entry, data->irq_2_pin) + /* Don't allow duplicates */ + for_each_irq_pin(entry, data->irq_2_pin) { if (entry->apic == apic && entry->pin == pin) - return 0; + return true; + } entry = kzalloc_node(sizeof(struct irq_pin_list), GFP_ATOMIC, node); if (!entry) { - pr_err("can not alloc irq_pin_list (%d,%d,%d)\n", - node, apic, pin); - return -ENOMEM; + pr_err("Cannot allocate irq_pin_list (%d,%d,%d)\n", node, apic, pin); + return false; } + entry->apic = apic; entry->pin = pin; list_add_tail(&entry->list, &data->irq_2_pin); - - return 0; + return true; } static void __remove_pin_from_irq(struct mp_chip_data *data, int apic, int pin) { struct irq_pin_list *tmp, *entry; - list_for_each_entry_safe(entry, tmp, &data->irq_2_pin, list) + list_for_each_entry_safe(entry, tmp, &data->irq_2_pin, list) { if (entry->apic == apic && entry->pin == pin) { list_del(&entry->list); kfree(entry); return; } -} - -static void add_pin_to_irq_node(struct mp_chip_data *data, - int node, int apic, int pin) -{ - if (__add_pin_to_irq_node(data, node, apic, pin)) - panic("IO-APIC: failed to add irq-pin. Can not proceed\n"); -} - -/* - * Reroute an IRQ to a different pin. - */ -static void __init replace_pin_at_irq_node(struct mp_chip_data *data, int node, - int oldapic, int oldpin, - int newapic, int newpin) -{ - struct irq_pin_list *entry; - - for_each_irq_pin(entry, data->irq_2_pin) { - if (entry->apic == oldapic && entry->pin == oldpin) { - entry->apic = newapic; - entry->pin = newpin; - /* every one is different, right? */ - return; - } } - - /* old apic/pin didn't exist, so just add new ones */ - add_pin_to_irq_node(data, node, newapic, newpin); } static void io_apic_modify_irq(struct mp_chip_data *data, bool masked, @@ -430,12 +387,12 @@ static void io_apic_modify_irq(struct mp_chip_data *data, bool masked, } } +/* + * Synchronize the IO-APIC and the CPU by doing a dummy read from the + * IO-APIC + */ static void io_apic_sync(struct irq_pin_list *entry) { - /* - * Synchronize the IO-APIC and the CPU by doing - * a dummy read from the IO-APIC - */ struct io_apic __iomem *io_apic; io_apic = io_apic_base(entry->apic); @@ -445,11 +402,9 @@ static void io_apic_sync(struct irq_pin_list *entry) static void mask_ioapic_irq(struct irq_data *irq_data) { struct mp_chip_data *data = irq_data->chip_data; - unsigned long flags; - raw_spin_lock_irqsave(&ioapic_lock, flags); + guard(raw_spinlock_irqsave)(&ioapic_lock); io_apic_modify_irq(data, true, &io_apic_sync); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); } static void __unmask_ioapic(struct mp_chip_data *data) @@ -460,11 +415,9 @@ static void __unmask_ioapic(struct mp_chip_data *data) static void unmask_ioapic_irq(struct irq_data *irq_data) { struct mp_chip_data *data = irq_data->chip_data; - unsigned long flags; - raw_spin_lock_irqsave(&ioapic_lock, flags); + guard(raw_spinlock_irqsave)(&ioapic_lock); __unmask_ioapic(data); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); } /* @@ -492,30 +445,24 @@ static void __eoi_ioapic_pin(int apic, int pin, int vector) entry = entry1 = __ioapic_read_entry(apic, pin); - /* - * Mask the entry and change the trigger mode to edge. - */ + /* Mask the entry and change the trigger mode to edge. */ entry1.masked = true; entry1.is_level = false; __ioapic_write_entry(apic, pin, entry1); - /* - * Restore the previous level triggered entry. - */ + /* Restore the previous level triggered entry. */ __ioapic_write_entry(apic, pin, entry); } } static void eoi_ioapic_pin(int vector, struct mp_chip_data *data) { - unsigned long flags; struct irq_pin_list *entry; - raw_spin_lock_irqsave(&ioapic_lock, flags); + guard(raw_spinlock_irqsave)(&ioapic_lock); for_each_irq_pin(entry, data->irq_2_pin) __eoi_ioapic_pin(entry->apic, entry->pin, vector); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); } static void clear_IO_APIC_pin(unsigned int apic, unsigned int pin) @@ -538,8 +485,6 @@ static void clear_IO_APIC_pin(unsigned int apic, unsigned int pin) } if (entry.irr) { - unsigned long flags; - /* * Make sure the trigger mode is set to level. Explicit EOI * doesn't clear the remote-IRR if the trigger mode is not @@ -549,9 +494,8 @@ static void clear_IO_APIC_pin(unsigned int apic, unsigned int pin) entry.is_level = true; ioapic_write_entry(apic, pin, entry); } - raw_spin_lock_irqsave(&ioapic_lock, flags); + guard(raw_spinlock_irqsave)(&ioapic_lock); __eoi_ioapic_pin(apic, pin, entry.vector); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); } /* @@ -586,28 +530,23 @@ static int pirq_entries[MAX_PIRQS] = { static int __init ioapic_pirq_setup(char *str) { - int i, max; - int ints[MAX_PIRQS+1]; + int i, max, ints[MAX_PIRQS+1]; get_options(str, ARRAY_SIZE(ints), ints); - apic_printk(APIC_VERBOSE, KERN_INFO - "PIRQ redirection, working around broken MP-BIOS.\n"); + apic_pr_verbose("PIRQ redirection, working around broken MP-BIOS.\n"); + max = MAX_PIRQS; if (ints[0] < MAX_PIRQS) max = ints[0]; for (i = 0; i < max; i++) { - apic_printk(APIC_VERBOSE, KERN_DEBUG - "... PIRQ%d -> IRQ %d\n", i, ints[i+1]); - /* - * PIRQs are mapped upside down, usually. - */ + apic_pr_verbose("... PIRQ%d -> IRQ %d\n", i, ints[i + 1]); + /* PIRQs are mapped upside down, usually */ pirq_entries[MAX_PIRQS-i-1] = ints[i+1]; } return 1; } - __setup("pirq=", ioapic_pirq_setup); #endif /* CONFIG_X86_32 */ @@ -626,8 +565,7 @@ int save_ioapic_entries(void) } for_each_pin(apic, pin) - ioapics[apic].saved_registers[pin] = - ioapic_read_entry(apic, pin); + ioapics[apic].saved_registers[pin] = ioapic_read_entry(apic, pin); } return err; @@ -668,8 +606,7 @@ int restore_ioapic_entries(void) continue; for_each_pin(apic, pin) - ioapic_write_entry(apic, pin, - ioapics[apic].saved_registers[pin]); + ioapic_write_entry(apic, pin, ioapics[apic].saved_registers[pin]); } return 0; } @@ -681,12 +618,13 @@ static int find_irq_entry(int ioapic_idx, int pin, int type) { int i; - for (i = 0; i < mp_irq_entries; i++) + for (i = 0; i < mp_irq_entries; i++) { if (mp_irqs[i].irqtype == type && (mp_irqs[i].dstapic == mpc_ioapic_id(ioapic_idx) || mp_irqs[i].dstapic == MP_APIC_ALL) && mp_irqs[i].dstirq == pin) return i; + } return -1; } @@ -701,10 +639,8 @@ static int __init find_isa_irq_pin(int irq, int type) for (i = 0; i < mp_irq_entries; i++) { int lbus = mp_irqs[i].srcbus; - if (test_bit(lbus, mp_bus_not_pci) && - (mp_irqs[i].irqtype == type) && + if (test_bit(lbus, mp_bus_not_pci) && (mp_irqs[i].irqtype == type) && (mp_irqs[i].srcbusirq == irq)) - return mp_irqs[i].dstirq; } return -1; @@ -717,8 +653,7 @@ static int __init find_isa_irq_apic(int irq, int type) for (i = 0; i < mp_irq_entries; i++) { int lbus = mp_irqs[i].srcbus; - if (test_bit(lbus, mp_bus_not_pci) && - (mp_irqs[i].irqtype == type) && + if (test_bit(lbus, mp_bus_not_pci) && (mp_irqs[i].irqtype == type) && (mp_irqs[i].srcbusirq == irq)) break; } @@ -726,9 +661,10 @@ static int __init find_isa_irq_apic(int irq, int type) if (i < mp_irq_entries) { int ioapic_idx; - for_each_ioapic(ioapic_idx) + for_each_ioapic(ioapic_idx) { if (mpc_ioapic_id(ioapic_idx) == mp_irqs[i].dstapic) return ioapic_idx; + } } return -1; @@ -769,8 +705,7 @@ static bool EISA_ELCR(unsigned int irq) unsigned int port = PIC_ELCR1 + (irq >> 3); return (inb(port) >> (irq & 7)) & 1; } - apic_printk(APIC_VERBOSE, KERN_INFO - "Broken MPtable reports ISA irq %d\n", irq); + apic_pr_verbose("Broken MPtable reports ISA irq %d\n", irq); return false; } @@ -947,9 +882,9 @@ static bool mp_check_pin_attr(int irq, struct irq_alloc_info *info) static int alloc_irq_from_domain(struct irq_domain *domain, int ioapic, u32 gsi, struct irq_alloc_info *info) { + int type = ioapics[ioapic].irqdomain_cfg.type; bool legacy = false; int irq = -1; - int type = ioapics[ioapic].irqdomain_cfg.type; switch (type) { case IOAPIC_DOMAIN_LEGACY: @@ -971,8 +906,7 @@ static int alloc_irq_from_domain(struct irq_domain *domain, int ioapic, u32 gsi, return -1; } - return __irq_domain_alloc_irqs(domain, irq, 1, - ioapic_alloc_attr_node(info), + return __irq_domain_alloc_irqs(domain, irq, 1, ioapic_alloc_attr_node(info), info, legacy, NULL); } @@ -986,29 +920,26 @@ static int alloc_irq_from_domain(struct irq_domain *domain, int ioapic, u32 gsi, * PIRQs instead of reprogramming the interrupt routing logic. Thus there may be * multiple pins sharing the same legacy IRQ number when ACPI is disabled. */ -static int alloc_isa_irq_from_domain(struct irq_domain *domain, - int irq, int ioapic, int pin, +static int alloc_isa_irq_from_domain(struct irq_domain *domain, int irq, int ioapic, int pin, struct irq_alloc_info *info) { - struct mp_chip_data *data; struct irq_data *irq_data = irq_get_irq_data(irq); int node = ioapic_alloc_attr_node(info); + struct mp_chip_data *data; /* * Legacy ISA IRQ has already been allocated, just add pin to * the pin list associated with this IRQ and program the IOAPIC - * entry. The IOAPIC entry + * entry. */ if (irq_data && irq_data->parent_data) { if (!mp_check_pin_attr(irq, info)) return -EBUSY; - if (__add_pin_to_irq_node(irq_data->chip_data, node, ioapic, - info->ioapic.pin)) + if (!add_pin_to_irq_node(irq_data->chip_data, node, ioapic, info->ioapic.pin)) return -ENOMEM; } else { info->flags |= X86_IRQ_ALLOC_LEGACY; - irq = __irq_domain_alloc_irqs(domain, irq, 1, node, info, true, - NULL); + irq = __irq_domain_alloc_irqs(domain, irq, 1, node, info, true, NULL); if (irq >= 0) { irq_data = irq_domain_get_irq_data(domain, irq); data = irq_data->chip_data; @@ -1022,11 +953,11 @@ static int alloc_isa_irq_from_domain(struct irq_domain *domain, static int mp_map_pin_to_irq(u32 gsi, int idx, int ioapic, int pin, unsigned int flags, struct irq_alloc_info *info) { - int irq; - bool legacy = false; + struct irq_domain *domain = mp_ioapic_irqdomain(ioapic); struct irq_alloc_info tmp; struct mp_chip_data *data; - struct irq_domain *domain = mp_ioapic_irqdomain(ioapic); + bool legacy = false; + int irq; if (!domain) return -ENOSYS; @@ -1046,7 +977,7 @@ static int mp_map_pin_to_irq(u32 gsi, int idx, int ioapic, int pin, return -EINVAL; } - mutex_lock(&ioapic_mutex); + guard(mutex)(&ioapic_mutex); if (!(flags & IOAPIC_MAP_ALLOC)) { if (!legacy) { irq = irq_find_mapping(domain, pin); @@ -1067,8 +998,6 @@ static int mp_map_pin_to_irq(u32 gsi, int idx, int ioapic, int pin, data->count++; } } - mutex_unlock(&ioapic_mutex); - return irq; } @@ -1076,26 +1005,20 @@ static int pin_2_irq(int idx, int ioapic, int pin, unsigned int flags) { u32 gsi = mp_pin_to_gsi(ioapic, pin); - /* - * Debugging check, we are in big trouble if this message pops up! - */ + /* Debugging check, we are in big trouble if this message pops up! */ if (mp_irqs[idx].dstirq != pin) pr_err("broken BIOS or MPTABLE parser, ayiee!!\n"); #ifdef CONFIG_X86_32 - /* - * PCI IRQ command line redirection. Yes, limits are hardcoded. - */ + /* PCI IRQ command line redirection. Yes, limits are hardcoded. */ if ((pin >= 16) && (pin <= 23)) { - if (pirq_entries[pin-16] != -1) { - if (!pirq_entries[pin-16]) { - apic_printk(APIC_VERBOSE, KERN_DEBUG - "disabling PIRQ%d\n", pin-16); + if (pirq_entries[pin - 16] != -1) { + if (!pirq_entries[pin - 16]) { + apic_pr_verbose("Disabling PIRQ%d\n", pin - 16); } else { int irq = pirq_entries[pin-16]; - apic_printk(APIC_VERBOSE, KERN_DEBUG - "using PIRQ%d -> IRQ %d\n", - pin-16, irq); + + apic_pr_verbose("Using PIRQ%d -> IRQ %d\n", pin - 16, irq); return irq; } } @@ -1133,10 +1056,9 @@ void mp_unmap_irq(int irq) if (!data || data->isa_irq) return; - mutex_lock(&ioapic_mutex); + guard(mutex)(&ioapic_mutex); if (--data->count == 0) irq_domain_free_irqs(irq, 1); - mutex_unlock(&ioapic_mutex); } /* @@ -1147,12 +1069,10 @@ int IO_APIC_get_PCI_irq_vector(int bus, int slot, int pin) { int irq, i, best_ioapic = -1, best_idx = -1; - apic_printk(APIC_DEBUG, - "querying PCI -> IRQ mapping bus:%d, slot:%d, pin:%d.\n", - bus, slot, pin); + apic_pr_debug("Querying PCI -> IRQ mapping bus:%d, slot:%d, pin:%d.\n", + bus, slot, pin); if (test_bit(bus, mp_bus_not_pci)) { - apic_printk(APIC_VERBOSE, - "PCI BIOS passed nonexistent PCI bus %d!\n", bus); + apic_pr_verbose("PCI BIOS passed nonexistent PCI bus %d!\n", bus); return -1; } @@ -1197,8 +1117,7 @@ int IO_APIC_get_PCI_irq_vector(int bus, int slot, int pin) return -1; out: - return pin_2_irq(best_idx, best_ioapic, mp_irqs[best_idx].dstirq, - IOAPIC_MAP_ALLOC); + return pin_2_irq(best_idx, best_ioapic, mp_irqs[best_idx].dstirq, IOAPIC_MAP_ALLOC); } EXPORT_SYMBOL(IO_APIC_get_PCI_irq_vector); @@ -1209,17 +1128,16 @@ static void __init setup_IO_APIC_irqs(void) unsigned int ioapic, pin; int idx; - apic_printk(APIC_VERBOSE, KERN_DEBUG "init IO_APIC IRQs\n"); + apic_pr_verbose("Init IO_APIC IRQs\n"); for_each_ioapic_pin(ioapic, pin) { idx = find_irq_entry(ioapic, pin, mp_INT); - if (idx < 0) - apic_printk(APIC_VERBOSE, - KERN_DEBUG " apic %d pin %d not connected\n", - mpc_ioapic_id(ioapic), pin); - else - pin_2_irq(idx, ioapic, pin, - ioapic ? 0 : IOAPIC_MAP_ALLOC); + if (idx < 0) { + apic_pr_verbose("apic %d pin %d not connected\n", + mpc_ioapic_id(ioapic), pin); + } else { + pin_2_irq(idx, ioapic, pin, ioapic ? 0 : IOAPIC_MAP_ALLOC); + } } } @@ -1234,26 +1152,21 @@ static void io_apic_print_entries(unsigned int apic, unsigned int nr_entries) char buf[256]; int i; - printk(KERN_DEBUG "IOAPIC %d:\n", apic); + apic_dbg("IOAPIC %d:\n", apic); for (i = 0; i <= nr_entries; i++) { entry = ioapic_read_entry(apic, i); - snprintf(buf, sizeof(buf), - " pin%02x, %s, %s, %s, V(%02X), IRR(%1d), S(%1d)", - i, - entry.masked ? "disabled" : "enabled ", + snprintf(buf, sizeof(buf), " pin%02x, %s, %s, %s, V(%02X), IRR(%1d), S(%1d)", + i, entry.masked ? "disabled" : "enabled ", entry.is_level ? "level" : "edge ", entry.active_low ? "low " : "high", entry.vector, entry.irr, entry.delivery_status); if (entry.ir_format) { - printk(KERN_DEBUG "%s, remapped, I(%04X), Z(%X)\n", - buf, - (entry.ir_index_15 << 15) | entry.ir_index_0_14, - entry.ir_zero); + apic_dbg("%s, remapped, I(%04X), Z(%X)\n", buf, + (entry.ir_index_15 << 15) | entry.ir_index_0_14, entry.ir_zero); } else { - printk(KERN_DEBUG "%s, %s, D(%02X%02X), M(%1d)\n", buf, - entry.dest_mode_logical ? "logical " : "physical", - entry.virt_destid_8_14, entry.destid_0_7, - entry.delivery_mode); + apic_dbg("%s, %s, D(%02X%02X), M(%1d)\n", buf, + entry.dest_mode_logical ? "logical " : "physical", + entry.virt_destid_8_14, entry.destid_0_7, entry.delivery_mode); } } } @@ -1264,30 +1177,25 @@ static void __init print_IO_APIC(int ioapic_idx) union IO_APIC_reg_01 reg_01; union IO_APIC_reg_02 reg_02; union IO_APIC_reg_03 reg_03; - unsigned long flags; - raw_spin_lock_irqsave(&ioapic_lock, flags); - reg_00.raw = io_apic_read(ioapic_idx, 0); - reg_01.raw = io_apic_read(ioapic_idx, 1); - if (reg_01.bits.version >= 0x10) - reg_02.raw = io_apic_read(ioapic_idx, 2); - if (reg_01.bits.version >= 0x20) - reg_03.raw = io_apic_read(ioapic_idx, 3); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); - - printk(KERN_DEBUG "IO APIC #%d......\n", mpc_ioapic_id(ioapic_idx)); - printk(KERN_DEBUG ".... register #00: %08X\n", reg_00.raw); - printk(KERN_DEBUG "....... : physical APIC id: %02X\n", reg_00.bits.ID); - printk(KERN_DEBUG "....... : Delivery Type: %X\n", reg_00.bits.delivery_type); - printk(KERN_DEBUG "....... : LTS : %X\n", reg_00.bits.LTS); - - printk(KERN_DEBUG ".... register #01: %08X\n", *(int *)®_01); - printk(KERN_DEBUG "....... : max redirection entries: %02X\n", - reg_01.bits.entries); - - printk(KERN_DEBUG "....... : PRQ implemented: %X\n", reg_01.bits.PRQ); - printk(KERN_DEBUG "....... : IO APIC version: %02X\n", - reg_01.bits.version); + scoped_guard (raw_spinlock_irqsave, &ioapic_lock) { + reg_00.raw = io_apic_read(ioapic_idx, 0); + reg_01.raw = io_apic_read(ioapic_idx, 1); + if (reg_01.bits.version >= 0x10) + reg_02.raw = io_apic_read(ioapic_idx, 2); + if (reg_01.bits.version >= 0x20) + reg_03.raw = io_apic_read(ioapic_idx, 3); + } + + apic_dbg("IO APIC #%d......\n", mpc_ioapic_id(ioapic_idx)); + apic_dbg(".... register #00: %08X\n", reg_00.raw); + apic_dbg("....... : physical APIC id: %02X\n", reg_00.bits.ID); + apic_dbg("....... : Delivery Type: %X\n", reg_00.bits.delivery_type); + apic_dbg("....... : LTS : %X\n", reg_00.bits.LTS); + apic_dbg(".... register #01: %08X\n", *(int *)®_01); + apic_dbg("....... : max redirection entries: %02X\n", reg_01.bits.entries); + apic_dbg("....... : PRQ implemented: %X\n", reg_01.bits.PRQ); + apic_dbg("....... : IO APIC version: %02X\n", reg_01.bits.version); /* * Some Intel chipsets with IO APIC VERSION of 0x1? don't have reg_02, @@ -1295,8 +1203,8 @@ static void __init print_IO_APIC(int ioapic_idx) * value, so ignore it if reg_02 == reg_01. */ if (reg_01.bits.version >= 0x10 && reg_02.raw != reg_01.raw) { - printk(KERN_DEBUG ".... register #02: %08X\n", reg_02.raw); - printk(KERN_DEBUG "....... : arbitration: %02X\n", reg_02.bits.arbitration); + apic_dbg(".... register #02: %08X\n", reg_02.raw); + apic_dbg("....... : arbitration: %02X\n", reg_02.bits.arbitration); } /* @@ -1306,11 +1214,11 @@ static void __init print_IO_APIC(int ioapic_idx) */ if (reg_01.bits.version >= 0x20 && reg_03.raw != reg_02.raw && reg_03.raw != reg_01.raw) { - printk(KERN_DEBUG ".... register #03: %08X\n", reg_03.raw); - printk(KERN_DEBUG "....... : Boot DT : %X\n", reg_03.bits.boot_DT); + apic_dbg(".... register #03: %08X\n", reg_03.raw); + apic_dbg("....... : Boot DT : %X\n", reg_03.bits.boot_DT); } - printk(KERN_DEBUG ".... IRQ redirection table:\n"); + apic_dbg(".... IRQ redirection table:\n"); io_apic_print_entries(ioapic_idx, reg_01.bits.entries); } @@ -1319,11 +1227,11 @@ void __init print_IO_APICs(void) int ioapic_idx; unsigned int irq; - printk(KERN_DEBUG "number of MP IRQ sources: %d.\n", mp_irq_entries); - for_each_ioapic(ioapic_idx) - printk(KERN_DEBUG "number of IO-APIC #%d registers: %d.\n", - mpc_ioapic_id(ioapic_idx), - ioapics[ioapic_idx].nr_registers); + apic_dbg("number of MP IRQ sources: %d.\n", mp_irq_entries); + for_each_ioapic(ioapic_idx) { + apic_dbg("number of IO-APIC #%d registers: %d.\n", + mpc_ioapic_id(ioapic_idx), ioapics[ioapic_idx].nr_registers); + } /* * We are a bit conservative about what we expect. We have to @@ -1334,7 +1242,7 @@ void __init print_IO_APICs(void) for_each_ioapic(ioapic_idx) print_IO_APIC(ioapic_idx); - printk(KERN_DEBUG "IRQ to pin mappings:\n"); + apic_dbg("IRQ to pin mappings:\n"); for_each_active_irq(irq) { struct irq_pin_list *entry; struct irq_chip *chip; @@ -1349,7 +1257,7 @@ void __init print_IO_APICs(void) if (list_empty(&data->irq_2_pin)) continue; - printk(KERN_DEBUG "IRQ%d ", irq); + apic_dbg("IRQ%d ", irq); for_each_irq_pin(entry, data->irq_2_pin) pr_cont("-> %d:%d", entry->apic, entry->pin); pr_cont("\n"); @@ -1363,8 +1271,7 @@ static struct { int pin, apic; } ioapic_i8259 = { -1, -1 }; void __init enable_IO_APIC(void) { - int i8259_apic, i8259_pin; - int apic, pin; + int i8259_apic, i8259_pin, apic, pin; if (ioapic_is_disabled) nr_ioapics = 0; @@ -1376,19 +1283,21 @@ void __init enable_IO_APIC(void) /* See if any of the pins is in ExtINT mode */ struct IO_APIC_route_entry entry = ioapic_read_entry(apic, pin); - /* If the interrupt line is enabled and in ExtInt mode - * I have found the pin where the i8259 is connected. + /* + * If the interrupt line is enabled and in ExtInt mode I + * have found the pin where the i8259 is connected. */ - if (!entry.masked && - entry.delivery_mode == APIC_DELIVERY_MODE_EXTINT) { + if (!entry.masked && entry.delivery_mode == APIC_DELIVERY_MODE_EXTINT) { ioapic_i8259.apic = apic; ioapic_i8259.pin = pin; - goto found_i8259; + break; } } - found_i8259: - /* Look to see what if the MP table has reported the ExtINT */ - /* If we could not find the appropriate pin by looking at the ioapic + + /* + * Look to see what if the MP table has reported the ExtINT + * + * If we could not find the appropriate pin by looking at the ioapic * the i8259 probably is not connected the ioapic but give the * mptable a chance anyway. */ @@ -1396,29 +1305,24 @@ void __init enable_IO_APIC(void) i8259_apic = find_isa_irq_apic(0, mp_ExtINT); /* Trust the MP table if nothing is setup in the hardware */ if ((ioapic_i8259.pin == -1) && (i8259_pin >= 0)) { - printk(KERN_WARNING "ExtINT not setup in hardware but reported by MP table\n"); + pr_warn("ExtINT not setup in hardware but reported by MP table\n"); ioapic_i8259.pin = i8259_pin; ioapic_i8259.apic = i8259_apic; } /* Complain if the MP table and the hardware disagree */ if (((ioapic_i8259.apic != i8259_apic) || (ioapic_i8259.pin != i8259_pin)) && - (i8259_pin >= 0) && (ioapic_i8259.pin >= 0)) - { - printk(KERN_WARNING "ExtINT in hardware and MP table differ\n"); - } + (i8259_pin >= 0) && (ioapic_i8259.pin >= 0)) + pr_warn("ExtINT in hardware and MP table differ\n"); - /* - * Do not trust the IO-APIC being empty at bootup - */ + /* Do not trust the IO-APIC being empty at bootup */ clear_IO_APIC(); } void native_restore_boot_irq_mode(void) { /* - * If the i8259 is routed through an IOAPIC - * Put that IOAPIC in virtual wire mode - * so legacy interrupts can be delivered. + * If the i8259 is routed through an IOAPIC Put that IOAPIC in + * virtual wire mode so legacy interrupts can be delivered. */ if (ioapic_i8259.pin != -1) { struct IO_APIC_route_entry entry; @@ -1433,9 +1337,7 @@ void native_restore_boot_irq_mode(void) entry.destid_0_7 = apic_id & 0xFF; entry.virt_destid_8_14 = apic_id >> 8; - /* - * Add it to the IO-APIC irq-routing table: - */ + /* Add it to the IO-APIC irq-routing table */ ioapic_write_entry(ioapic_i8259.apic, ioapic_i8259.pin, entry); } @@ -1458,37 +1360,34 @@ void restore_boot_irq_mode(void) * * by Matt Domsch <Matt_Domsch@dell.com> Tue Dec 21 12:25:05 CST 1999 */ -void __init setup_ioapic_ids_from_mpc_nocheck(void) +static void __init setup_ioapic_ids_from_mpc_nocheck(void) { + DECLARE_BITMAP(phys_id_present_map, MAX_LOCAL_APIC); + const u32 broadcast_id = 0xF; union IO_APIC_reg_00 reg_00; - physid_mask_t phys_id_present_map; - int ioapic_idx; - int i; unsigned char old_id; - unsigned long flags; + int ioapic_idx, i; /* * This is broken; anything with a real cpu count has to * circumvent this idiocy regardless. */ - apic->ioapic_phys_id_map(&phys_cpu_present_map, &phys_id_present_map); + copy_phys_cpu_present_map(phys_id_present_map); /* * Set the IOAPIC ID to the value stored in the MPC table. */ for_each_ioapic(ioapic_idx) { /* Read the register 0 value */ - raw_spin_lock_irqsave(&ioapic_lock, flags); - reg_00.raw = io_apic_read(ioapic_idx, 0); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); + scoped_guard (raw_spinlock_irqsave, &ioapic_lock) + reg_00.raw = io_apic_read(ioapic_idx, 0); old_id = mpc_ioapic_id(ioapic_idx); - if (mpc_ioapic_id(ioapic_idx) >= get_physical_broadcast()) { - printk(KERN_ERR "BIOS bug, IO-APIC#%d ID is %d in the MPC table!...\n", - ioapic_idx, mpc_ioapic_id(ioapic_idx)); - printk(KERN_ERR "... fixing up to %d. (tell your hw vendor)\n", - reg_00.bits.ID); + if (mpc_ioapic_id(ioapic_idx) >= broadcast_id) { + pr_err(FW_BUG "IO-APIC#%d ID is %d in the MPC table!...\n", + ioapic_idx, mpc_ioapic_id(ioapic_idx)); + pr_err("... fixing up to %d. (tell your hw vendor)\n", reg_00.bits.ID); ioapics[ioapic_idx].mp_config.apicid = reg_00.bits.ID; } @@ -1497,61 +1396,54 @@ void __init setup_ioapic_ids_from_mpc_nocheck(void) * system must have a unique ID or we get lots of nice * 'stuck on smp_invalidate_needed IPI wait' messages. */ - if (apic->check_apicid_used(&phys_id_present_map, - mpc_ioapic_id(ioapic_idx))) { - printk(KERN_ERR "BIOS bug, IO-APIC#%d ID %d is already used!...\n", - ioapic_idx, mpc_ioapic_id(ioapic_idx)); - for (i = 0; i < get_physical_broadcast(); i++) - if (!physid_isset(i, phys_id_present_map)) + if (test_bit(mpc_ioapic_id(ioapic_idx), phys_id_present_map)) { + pr_err(FW_BUG "IO-APIC#%d ID %d is already used!...\n", + ioapic_idx, mpc_ioapic_id(ioapic_idx)); + for (i = 0; i < broadcast_id; i++) + if (!test_bit(i, phys_id_present_map)) break; - if (i >= get_physical_broadcast()) + if (i >= broadcast_id) panic("Max APIC ID exceeded!\n"); - printk(KERN_ERR "... fixing up to %d. (tell your hw vendor)\n", - i); - physid_set(i, phys_id_present_map); + pr_err("... fixing up to %d. (tell your hw vendor)\n", i); + set_bit(i, phys_id_present_map); ioapics[ioapic_idx].mp_config.apicid = i; } else { - apic_printk(APIC_VERBOSE, "Setting %d in the phys_id_present_map\n", - mpc_ioapic_id(ioapic_idx)); - physid_set(mpc_ioapic_id(ioapic_idx), phys_id_present_map); + apic_pr_verbose("Setting %d in the phys_id_present_map\n", + mpc_ioapic_id(ioapic_idx)); + set_bit(mpc_ioapic_id(ioapic_idx), phys_id_present_map); } /* - * We need to adjust the IRQ routing table - * if the ID changed. + * We need to adjust the IRQ routing table if the ID + * changed. */ - if (old_id != mpc_ioapic_id(ioapic_idx)) - for (i = 0; i < mp_irq_entries; i++) + if (old_id != mpc_ioapic_id(ioapic_idx)) { + for (i = 0; i < mp_irq_entries; i++) { if (mp_irqs[i].dstapic == old_id) - mp_irqs[i].dstapic - = mpc_ioapic_id(ioapic_idx); + mp_irqs[i].dstapic = mpc_ioapic_id(ioapic_idx); + } + } /* - * Update the ID register according to the right value - * from the MPC table if they are different. + * Update the ID register according to the right value from + * the MPC table if they are different. */ if (mpc_ioapic_id(ioapic_idx) == reg_00.bits.ID) continue; - apic_printk(APIC_VERBOSE, KERN_INFO - "...changing IO-APIC physical APIC ID to %d ...", - mpc_ioapic_id(ioapic_idx)); + apic_pr_verbose("...changing IO-APIC physical APIC ID to %d ...", + mpc_ioapic_id(ioapic_idx)); reg_00.bits.ID = mpc_ioapic_id(ioapic_idx); - raw_spin_lock_irqsave(&ioapic_lock, flags); - io_apic_write(ioapic_idx, 0, reg_00.raw); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); - - /* - * Sanity check - */ - raw_spin_lock_irqsave(&ioapic_lock, flags); - reg_00.raw = io_apic_read(ioapic_idx, 0); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); + scoped_guard (raw_spinlock_irqsave, &ioapic_lock) { + io_apic_write(ioapic_idx, 0, reg_00.raw); + reg_00.raw = io_apic_read(ioapic_idx, 0); + } + /* Sanity check */ if (reg_00.bits.ID != mpc_ioapic_id(ioapic_idx)) pr_cont("could not set ID!\n"); else - apic_printk(APIC_VERBOSE, " ok.\n"); + apic_pr_verbose(" ok.\n"); } } @@ -1596,8 +1488,7 @@ static void __init delay_with_tsc(void) do { rep_nop(); now = rdtsc(); - } while ((now - start) < 40000000000ULL / HZ && - time_before_eq(jiffies, end)); + } while ((now - start) < 40000000000ULL / HZ && time_before_eq(jiffies, end)); } static void __init delay_without_tsc(void) @@ -1658,36 +1549,29 @@ static int __init timer_irq_works(void) * so we 'resend' these IRQs via IPIs, to the same CPU. It's much * better to do it this way as thus we do not have to be aware of * 'pending' interrupts in the IRQ path, except at this point. - */ -/* - * Edge triggered needs to resend any interrupt - * that was delayed but this is now handled in the device - * independent code. - */ - -/* - * Starting up a edge-triggered IO-APIC interrupt is - * nasty - we need to make sure that we get the edge. - * If it is already asserted for some reason, we need - * return 1 to indicate that is was pending. * - * This is not complete - we should be able to fake - * an edge even if it isn't on the 8259A... + * + * Edge triggered needs to resend any interrupt that was delayed but this + * is now handled in the device independent code. + * + * Starting up a edge-triggered IO-APIC interrupt is nasty - we need to + * make sure that we get the edge. If it is already asserted for some + * reason, we need return 1 to indicate that is was pending. + * + * This is not complete - we should be able to fake an edge even if it + * isn't on the 8259A... */ static unsigned int startup_ioapic_irq(struct irq_data *data) { int was_pending = 0, irq = data->irq; - unsigned long flags; - raw_spin_lock_irqsave(&ioapic_lock, flags); + guard(raw_spinlock_irqsave)(&ioapic_lock); if (irq < nr_legacy_irqs()) { legacy_pic->mask(irq); if (legacy_pic->irq_pending(irq)) was_pending = 1; } __unmask_ioapic(data->chip_data); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); - return was_pending; } @@ -1697,9 +1581,8 @@ atomic_t irq_mis_count; static bool io_apic_level_ack_pending(struct mp_chip_data *data) { struct irq_pin_list *entry; - unsigned long flags; - raw_spin_lock_irqsave(&ioapic_lock, flags); + guard(raw_spinlock_irqsave)(&ioapic_lock); for_each_irq_pin(entry, data->irq_2_pin) { struct IO_APIC_route_entry e; int pin; @@ -1707,13 +1590,9 @@ static bool io_apic_level_ack_pending(struct mp_chip_data *data) pin = entry->pin; e.w1 = io_apic_read(entry->apic, 0x10 + pin*2); /* Is the remote IRR bit set? */ - if (e.irr) { - raw_spin_unlock_irqrestore(&ioapic_lock, flags); + if (e.irr) return true; - } } - raw_spin_unlock_irqrestore(&ioapic_lock, flags); - return false; } @@ -1731,7 +1610,8 @@ static inline bool ioapic_prepare_move(struct irq_data *data) static inline void ioapic_finish_move(struct irq_data *data, bool moveit) { if (unlikely(moveit)) { - /* Only migrate the irq if the ack has been received. + /* + * Only migrate the irq if the ack has been received. * * On rare occasions the broadcast level triggered ack gets * delayed going to ioapics, and if we reprogram the @@ -1914,18 +1794,16 @@ static void ioapic_configure_entry(struct irq_data *irqd) __ioapic_write_entry(entry->apic, entry->pin, mpd->entry); } -static int ioapic_set_affinity(struct irq_data *irq_data, - const struct cpumask *mask, bool force) +static int ioapic_set_affinity(struct irq_data *irq_data, const struct cpumask *mask, bool force) { struct irq_data *parent = irq_data->parent_data; - unsigned long flags; int ret; ret = parent->chip->irq_set_affinity(parent, mask, force); - raw_spin_lock_irqsave(&ioapic_lock, flags); + + guard(raw_spinlock_irqsave)(&ioapic_lock); if (ret >= 0 && ret != IRQ_SET_MASK_OK_DONE) ioapic_configure_entry(irq_data); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); return ret; } @@ -1944,9 +1822,8 @@ static int ioapic_set_affinity(struct irq_data *irq_data, * * Verify that the corresponding Remote-IRR bits are clear. */ -static int ioapic_irq_get_chip_state(struct irq_data *irqd, - enum irqchip_irq_state which, - bool *state) +static int ioapic_irq_get_chip_state(struct irq_data *irqd, enum irqchip_irq_state which, + bool *state) { struct mp_chip_data *mcd = irqd->chip_data; struct IO_APIC_route_entry rentry; @@ -1956,7 +1833,8 @@ static int ioapic_irq_get_chip_state(struct irq_data *irqd, return -EINVAL; *state = false; - raw_spin_lock(&ioapic_lock); + + guard(raw_spinlock)(&ioapic_lock); for_each_irq_pin(p, mcd->irq_2_pin) { rentry = __ioapic_read_entry(p->apic, p->pin); /* @@ -1970,7 +1848,6 @@ static int ioapic_irq_get_chip_state(struct irq_data *irqd, break; } } - raw_spin_unlock(&ioapic_lock); return 0; } @@ -1984,7 +1861,7 @@ static struct irq_chip ioapic_chip __read_mostly = { .irq_set_affinity = ioapic_set_affinity, .irq_retrigger = irq_chip_retrigger_hierarchy, .irq_get_irqchip_state = ioapic_irq_get_chip_state, - .flags = IRQCHIP_SKIP_SET_WAKE | + .flags = IRQCHIP_SKIP_SET_WAKE | IRQCHIP_MOVE_DEFERRED | IRQCHIP_AFFINITY_PRE_STARTUP, }; @@ -2011,14 +1888,13 @@ static inline void init_IO_APIC_traps(void) cfg = irq_cfg(irq); if (IO_APIC_IRQ(irq) && cfg && !cfg->vector) { /* - * Hmm.. We don't have an entry for this, - * so default to an old-fashioned 8259 - * interrupt if we can.. + * Hmm.. We don't have an entry for this, so + * default to an old-fashioned 8259 interrupt if we + * can. Otherwise set the dummy interrupt chip. */ if (irq < nr_legacy_irqs()) legacy_pic->make_irq(irq); else - /* Strange. Oh, well.. */ irq_set_chip(irq, &no_irq_chip); } } @@ -2027,20 +1903,17 @@ static inline void init_IO_APIC_traps(void) /* * The local APIC irq-chip implementation: */ - static void mask_lapic_irq(struct irq_data *data) { - unsigned long v; + unsigned long v = apic_read(APIC_LVT0); - v = apic_read(APIC_LVT0); apic_write(APIC_LVT0, v | APIC_LVT_MASKED); } static void unmask_lapic_irq(struct irq_data *data) { - unsigned long v; + unsigned long v = apic_read(APIC_LVT0); - v = apic_read(APIC_LVT0); apic_write(APIC_LVT0, v & ~APIC_LVT_MASKED); } @@ -2059,8 +1932,7 @@ static struct irq_chip lapic_chip __read_mostly = { static void lapic_register_intr(int irq) { irq_clear_status_flags(irq, IRQ_LEVEL); - irq_set_chip_and_handler_name(irq, &lapic_chip, handle_edge_irq, - "edge"); + irq_set_chip_and_handler_name(irq, &lapic_chip, handle_edge_irq, "edge"); } /* @@ -2072,9 +1944,9 @@ static void lapic_register_intr(int irq) */ static inline void __init unlock_ExtINT_logic(void) { - int apic, pin, i; - struct IO_APIC_route_entry entry0, entry1; unsigned char save_control, save_freq_select; + struct IO_APIC_route_entry entry0, entry1; + int apic, pin, i; u32 apic_id; pin = find_isa_irq_pin(8, mp_INT); @@ -2134,10 +2006,10 @@ static int __init disable_timer_pin_setup(char *arg) } early_param("disable_timer_pin_1", disable_timer_pin_setup); -static int mp_alloc_timer_irq(int ioapic, int pin) +static int __init mp_alloc_timer_irq(int ioapic, int pin) { - int irq = -1; struct irq_domain *domain = mp_ioapic_irqdomain(ioapic); + int irq = -1; if (domain) { struct irq_alloc_info info; @@ -2145,21 +2017,36 @@ static int mp_alloc_timer_irq(int ioapic, int pin) ioapic_set_alloc_attr(&info, NUMA_NO_NODE, 0, 0); info.devid = mpc_ioapic_id(ioapic); info.ioapic.pin = pin; - mutex_lock(&ioapic_mutex); + guard(mutex)(&ioapic_mutex); irq = alloc_isa_irq_from_domain(domain, 0, ioapic, pin, &info); - mutex_unlock(&ioapic_mutex); } return irq; } +static void __init replace_pin_at_irq_node(struct mp_chip_data *data, int node, + int oldapic, int oldpin, + int newapic, int newpin) +{ + struct irq_pin_list *entry; + + for_each_irq_pin(entry, data->irq_2_pin) { + if (entry->apic == oldapic && entry->pin == oldpin) { + entry->apic = newapic; + entry->pin = newpin; + return; + } + } + + /* Old apic/pin didn't exist, so just add a new one */ + add_pin_to_irq_node(data, node, newapic, newpin); +} + /* * This code may look a bit paranoid, but it's supposed to cooperate with * a wide range of boards and BIOS bugs. Fortunately only the timer IRQ * is so screwy. Thanks to Brian Perkins for testing/hacking this beast * fanatically on his truly buggy board. - * - * FIXME: really need to revamp this for all platforms. */ static inline void __init check_timer(void) { @@ -2197,9 +2084,8 @@ static inline void __init check_timer(void) pin2 = ioapic_i8259.pin; apic2 = ioapic_i8259.apic; - apic_printk(APIC_QUIET, KERN_INFO "..TIMER: vector=0x%02X " - "apic1=%d pin1=%d apic2=%d pin2=%d\n", - cfg->vector, apic1, pin1, apic2, pin2); + pr_info("..TIMER: vector=0x%02X apic1=%d pin1=%d apic2=%d pin2=%d\n", + cfg->vector, apic1, pin1, apic2, pin2); /* * Some BIOS writers are clueless and report the ExtINTA @@ -2209,7 +2095,7 @@ static inline void __init check_timer(void) * 8259A. */ if (pin1 == -1) { - panic_if_irq_remap("BIOS bug: timer not connected to IO-APIC"); + panic_if_irq_remap(FW_BUG "Timer not connected to IO-APIC"); pin1 = pin2; apic1 = apic2; no_pin1 = 1; @@ -2243,13 +2129,10 @@ static inline void __init check_timer(void) panic_if_irq_remap("timer doesn't work through Interrupt-remapped IO-APIC"); clear_IO_APIC_pin(apic1, pin1); if (!no_pin1) - apic_printk(APIC_QUIET, KERN_ERR "..MP-BIOS bug: " - "8254 timer not connected to IO-APIC\n"); + pr_err("..MP-BIOS bug: 8254 timer not connected to IO-APIC\n"); - apic_printk(APIC_QUIET, KERN_INFO "...trying to set up timer " - "(IRQ0) through the 8259A ...\n"); - apic_printk(APIC_QUIET, KERN_INFO - "..... (found apic %d pin %d) ...\n", apic2, pin2); + pr_info("...trying to set up timer (IRQ0) through the 8259A ...\n"); + pr_info("..... (found apic %d pin %d) ...\n", apic2, pin2); /* * legacy devices should be connected to IO APIC #0 */ @@ -2258,7 +2141,7 @@ static inline void __init check_timer(void) irq_domain_activate_irq(irq_data, false); legacy_pic->unmask(0); if (timer_irq_works()) { - apic_printk(APIC_QUIET, KERN_INFO "....... works.\n"); + pr_info("....... works.\n"); goto out; } /* @@ -2266,26 +2149,24 @@ static inline void __init check_timer(void) */ legacy_pic->mask(0); clear_IO_APIC_pin(apic2, pin2); - apic_printk(APIC_QUIET, KERN_INFO "....... failed.\n"); + pr_info("....... failed.\n"); } - apic_printk(APIC_QUIET, KERN_INFO - "...trying to set up timer as Virtual Wire IRQ...\n"); + pr_info("...trying to set up timer as Virtual Wire IRQ...\n"); lapic_register_intr(0); apic_write(APIC_LVT0, APIC_DM_FIXED | cfg->vector); /* Fixed mode */ legacy_pic->unmask(0); if (timer_irq_works()) { - apic_printk(APIC_QUIET, KERN_INFO "..... works.\n"); + pr_info("..... works.\n"); goto out; } legacy_pic->mask(0); apic_write(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_FIXED | cfg->vector); - apic_printk(APIC_QUIET, KERN_INFO "..... failed.\n"); + pr_info("..... failed.\n"); - apic_printk(APIC_QUIET, KERN_INFO - "...trying to set up timer as ExtINT IRQ...\n"); + pr_info("...trying to set up timer as ExtINT IRQ...\n"); legacy_pic->init(0); legacy_pic->make_irq(0); @@ -2295,14 +2176,15 @@ static inline void __init check_timer(void) unlock_ExtINT_logic(); if (timer_irq_works()) { - apic_printk(APIC_QUIET, KERN_INFO "..... works.\n"); + pr_info("..... works.\n"); goto out; } - apic_printk(APIC_QUIET, KERN_INFO "..... failed :(.\n"); - if (apic_is_x2apic_enabled()) - apic_printk(APIC_QUIET, KERN_INFO - "Perhaps problem with the pre-enabled x2apic mode\n" - "Try booting with x2apic and interrupt-remapping disabled in the bios.\n"); + + pr_info("..... failed :\n"); + if (apic_is_x2apic_enabled()) { + pr_info("Perhaps problem with the pre-enabled x2apic mode\n" + "Try booting with x2apic and interrupt-remapping disabled in the bios.\n"); + } panic("IO-APIC + timer doesn't work! Boot with apic=debug and send a " "report. Then try booting with the 'noapic' option.\n"); out: @@ -2330,11 +2212,11 @@ out: static int mp_irqdomain_create(int ioapic) { - struct irq_domain *parent; + struct mp_ioapic_gsi *gsi_cfg = mp_ioapic_gsi_routing(ioapic); int hwirqs = mp_ioapic_pin_count(ioapic); struct ioapic *ip = &ioapics[ioapic]; struct ioapic_domain_cfg *cfg = &ip->irqdomain_cfg; - struct mp_ioapic_gsi *gsi_cfg = mp_ioapic_gsi_routing(ioapic); + struct irq_domain *parent; struct fwnode_handle *fn; struct irq_fwspec fwspec; @@ -2354,7 +2236,7 @@ static int mp_irqdomain_create(int ioapic) fwspec.param_count = 1; fwspec.param[0] = mpc_ioapic_id(ioapic); - parent = irq_find_matching_fwspec(&fwspec, DOMAIN_BUS_ANY); + parent = irq_find_matching_fwspec(&fwspec, DOMAIN_BUS_GENERIC_MSI); if (!parent) { if (!cfg->dev) irq_domain_free_fwnode(fn); @@ -2370,10 +2252,8 @@ static int mp_irqdomain_create(int ioapic) return -ENOMEM; } - if (cfg->type == IOAPIC_DOMAIN_LEGACY || - cfg->type == IOAPIC_DOMAIN_STRICT) - ioapic_dynirq_base = max(ioapic_dynirq_base, - gsi_cfg->gsi_end + 1); + if (cfg->type == IOAPIC_DOMAIN_LEGACY || cfg->type == IOAPIC_DOMAIN_STRICT) + ioapic_dynirq_base = max(ioapic_dynirq_base, gsi_cfg->gsi_end + 1); return 0; } @@ -2400,13 +2280,11 @@ void __init setup_IO_APIC(void) io_apic_irqs = nr_legacy_irqs() ? ~PIC_IRQS : ~0UL; - apic_printk(APIC_VERBOSE, "ENABLING IO-APIC IRQs\n"); + apic_pr_verbose("ENABLING IO-APIC IRQs\n"); for_each_ioapic(ioapic) BUG_ON(mp_irqdomain_create(ioapic)); - /* - * Set up IO-APIC IRQ routing. - */ + /* Set up IO-APIC IRQ routing. */ x86_init.mpparse.setup_ioapic_ids(); sync_Arb_IDs(); @@ -2420,16 +2298,14 @@ void __init setup_IO_APIC(void) static void resume_ioapic_id(int ioapic_idx) { - unsigned long flags; union IO_APIC_reg_00 reg_00; - raw_spin_lock_irqsave(&ioapic_lock, flags); + guard(raw_spinlock_irqsave)(&ioapic_lock); reg_00.raw = io_apic_read(ioapic_idx, 0); if (reg_00.bits.ID != mpc_ioapic_id(ioapic_idx)) { reg_00.bits.ID = mpc_ioapic_id(ioapic_idx); io_apic_write(ioapic_idx, 0, reg_00.raw); } - raw_spin_unlock_irqrestore(&ioapic_lock, flags); } static void ioapic_resume(void) @@ -2443,8 +2319,8 @@ static void ioapic_resume(void) } static struct syscore_ops ioapic_syscore_ops = { - .suspend = save_ioapic_entries, - .resume = ioapic_resume, + .suspend = save_ioapic_entries, + .resume = ioapic_resume, }; static int __init ioapic_init_ops(void) @@ -2459,15 +2335,13 @@ device_initcall(ioapic_init_ops); static int io_apic_get_redir_entries(int ioapic) { union IO_APIC_reg_01 reg_01; - unsigned long flags; - raw_spin_lock_irqsave(&ioapic_lock, flags); + guard(raw_spinlock_irqsave)(&ioapic_lock); reg_01.raw = io_apic_read(ioapic, 1); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); - /* The register returns the maximum index redir index - * supported, which is one less than the total number of redir - * entries. + /* + * The register returns the maximum index redir index supported, + * which is one less than the total number of redir entries. */ return reg_01.bits.entries + 1; } @@ -2494,93 +2368,71 @@ unsigned int arch_dynirq_lower_bound(unsigned int from) #ifdef CONFIG_X86_32 static int io_apic_get_unique_id(int ioapic, int apic_id) { + static DECLARE_BITMAP(apic_id_map, MAX_LOCAL_APIC); + const u32 broadcast_id = 0xF; union IO_APIC_reg_00 reg_00; - static physid_mask_t apic_id_map = PHYSID_MASK_NONE; - physid_mask_t tmp; - unsigned long flags; int i = 0; - /* - * The P4 platform supports up to 256 APIC IDs on two separate APIC - * buses (one for LAPICs, one for IOAPICs), where predecessors only - * supports up to 16 on one shared APIC bus. - * - * TBD: Expand LAPIC/IOAPIC support on P4-class systems to take full - * advantage of new APIC bus architecture. - */ + /* Initialize the ID map */ + if (bitmap_empty(apic_id_map, MAX_LOCAL_APIC)) + copy_phys_cpu_present_map(apic_id_map); - if (physids_empty(apic_id_map)) - apic->ioapic_phys_id_map(&phys_cpu_present_map, &apic_id_map); - - raw_spin_lock_irqsave(&ioapic_lock, flags); - reg_00.raw = io_apic_read(ioapic, 0); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); + scoped_guard (raw_spinlock_irqsave, &ioapic_lock) + reg_00.raw = io_apic_read(ioapic, 0); - if (apic_id >= get_physical_broadcast()) { - printk(KERN_WARNING "IOAPIC[%d]: Invalid apic_id %d, trying " - "%d\n", ioapic, apic_id, reg_00.bits.ID); + if (apic_id >= broadcast_id) { + pr_warn("IOAPIC[%d]: Invalid apic_id %d, trying %d\n", + ioapic, apic_id, reg_00.bits.ID); apic_id = reg_00.bits.ID; } - /* - * Every APIC in a system must have a unique ID or we get lots of nice - * 'stuck on smp_invalidate_needed IPI wait' messages. - */ - if (apic->check_apicid_used(&apic_id_map, apic_id)) { - - for (i = 0; i < get_physical_broadcast(); i++) { - if (!apic->check_apicid_used(&apic_id_map, i)) + /* Every APIC in a system must have a unique ID */ + if (test_bit(apic_id, apic_id_map)) { + for (i = 0; i < broadcast_id; i++) { + if (!test_bit(i, apic_id_map)) break; } - if (i == get_physical_broadcast()) + if (i == broadcast_id) panic("Max apic_id exceeded!\n"); - printk(KERN_WARNING "IOAPIC[%d]: apic_id %d already used, " - "trying %d\n", ioapic, apic_id, i); - + pr_warn("IOAPIC[%d]: apic_id %d already used, trying %d\n", ioapic, apic_id, i); apic_id = i; } - physid_set_mask_of_physid(apic_id, &tmp); - physids_or(apic_id_map, apic_id_map, tmp); + set_bit(apic_id, apic_id_map); if (reg_00.bits.ID != apic_id) { reg_00.bits.ID = apic_id; - raw_spin_lock_irqsave(&ioapic_lock, flags); - io_apic_write(ioapic, 0, reg_00.raw); - reg_00.raw = io_apic_read(ioapic, 0); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); + scoped_guard (raw_spinlock_irqsave, &ioapic_lock) { + io_apic_write(ioapic, 0, reg_00.raw); + reg_00.raw = io_apic_read(ioapic, 0); + } /* Sanity check */ if (reg_00.bits.ID != apic_id) { - pr_err("IOAPIC[%d]: Unable to change apic_id!\n", - ioapic); + pr_err("IOAPIC[%d]: Unable to change apic_id!\n", ioapic); return -1; } } - apic_printk(APIC_VERBOSE, KERN_INFO - "IOAPIC[%d]: Assigned apic_id %d\n", ioapic, apic_id); + apic_pr_verbose("IOAPIC[%d]: Assigned apic_id %d\n", ioapic, apic_id); return apic_id; } static u8 io_apic_unique_id(int idx, u8 id) { - if ((boot_cpu_data.x86_vendor == X86_VENDOR_INTEL) && - !APIC_XAPIC(boot_cpu_apic_version)) + if ((boot_cpu_data.x86_vendor == X86_VENDOR_INTEL) && !APIC_XAPIC(boot_cpu_apic_version)) return io_apic_get_unique_id(idx, id); - else - return id; + return id; } #else static u8 io_apic_unique_id(int idx, u8 id) { union IO_APIC_reg_00 reg_00; DECLARE_BITMAP(used, 256); - unsigned long flags; u8 new_id; int i; @@ -2596,26 +2448,23 @@ static u8 io_apic_unique_id(int idx, u8 id) * Read the current id from the ioapic and keep it if * available. */ - raw_spin_lock_irqsave(&ioapic_lock, flags); - reg_00.raw = io_apic_read(idx, 0); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); + scoped_guard (raw_spinlock_irqsave, &ioapic_lock) + reg_00.raw = io_apic_read(idx, 0); + new_id = reg_00.bits.ID; if (!test_bit(new_id, used)) { - apic_printk(APIC_VERBOSE, KERN_INFO - "IOAPIC[%d]: Using reg apic_id %d instead of %d\n", - idx, new_id, id); + apic_pr_verbose("IOAPIC[%d]: Using reg apic_id %d instead of %d\n", + idx, new_id, id); return new_id; } - /* - * Get the next free id and write it to the ioapic. - */ + /* Get the next free id and write it to the ioapic. */ new_id = find_first_zero_bit(used, 256); reg_00.bits.ID = new_id; - raw_spin_lock_irqsave(&ioapic_lock, flags); - io_apic_write(idx, 0, reg_00.raw); - reg_00.raw = io_apic_read(idx, 0); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); + scoped_guard (raw_spinlock_irqsave, &ioapic_lock) { + io_apic_write(idx, 0, reg_00.raw); + reg_00.raw = io_apic_read(idx, 0); + } /* Sanity check */ BUG_ON(reg_00.bits.ID != new_id); @@ -2625,12 +2474,10 @@ static u8 io_apic_unique_id(int idx, u8 id) static int io_apic_get_version(int ioapic) { - union IO_APIC_reg_01 reg_01; - unsigned long flags; + union IO_APIC_reg_01 reg_01; - raw_spin_lock_irqsave(&ioapic_lock, flags); + guard(raw_spinlock_irqsave)(&ioapic_lock); reg_01.raw = io_apic_read(ioapic, 1); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); return reg_01.bits.version; } @@ -2645,8 +2492,8 @@ static struct resource *ioapic_resources; static struct resource * __init ioapic_setup_resources(void) { - unsigned long n; struct resource *res; + unsigned long n; char *mem; int i; @@ -2656,9 +2503,7 @@ static struct resource * __init ioapic_setup_resources(void) n = IOAPIC_RESOURCE_NAME_SIZE + sizeof(struct resource); n *= nr_ioapics; - mem = memblock_alloc(n, SMP_CACHE_BYTES); - if (!mem) - panic("%s: Failed to allocate %lu bytes\n", __func__, n); + mem = memblock_alloc_or_panic(n, SMP_CACHE_BYTES); res = (void *)mem; mem += sizeof(struct resource) * nr_ioapics; @@ -2706,9 +2551,7 @@ void __init io_apic_init_mappings(void) ioapic_phys = mpc_ioapic_addr(i); #ifdef CONFIG_X86_32 if (!ioapic_phys) { - printk(KERN_ERR - "WARNING: bogus zero IO-APIC " - "address found in MPTABLE, " + pr_err("WARNING: bogus zero IO-APIC address found in MPTABLE, " "disabling IO/APIC support!\n"); smp_found_config = 0; ioapic_is_disabled = true; @@ -2719,17 +2562,13 @@ void __init io_apic_init_mappings(void) #ifdef CONFIG_X86_32 fake_ioapic_page: #endif - ioapic_phys = (unsigned long)memblock_alloc(PAGE_SIZE, + ioapic_phys = (unsigned long)memblock_alloc_or_panic(PAGE_SIZE, PAGE_SIZE); - if (!ioapic_phys) - panic("%s: Failed to allocate %lu bytes align=0x%lx\n", - __func__, PAGE_SIZE, PAGE_SIZE); ioapic_phys = __pa(ioapic_phys); } io_apic_set_fixmap(idx, ioapic_phys); - apic_printk(APIC_VERBOSE, "mapped IOAPIC to %08lx (%08lx)\n", - __fix_to_virt(idx) + (ioapic_phys & ~PAGE_MASK), - ioapic_phys); + apic_pr_verbose("mapped IOAPIC to %08lx (%08lx)\n", + __fix_to_virt(idx) + (ioapic_phys & ~PAGE_MASK), ioapic_phys); idx++; ioapic_res->start = ioapic_phys; @@ -2740,13 +2579,12 @@ fake_ioapic_page: void __init ioapic_insert_resources(void) { - int i; struct resource *r = ioapic_resources; + int i; if (!r) { if (nr_ioapics > 0) - printk(KERN_ERR - "IO APIC resources couldn't be allocated.\n"); + pr_err("IO APIC resources couldn't be allocated.\n"); return; } @@ -2766,11 +2604,12 @@ int mp_find_ioapic(u32 gsi) /* Find the IOAPIC that manages this GSI. */ for_each_ioapic(i) { struct mp_ioapic_gsi *gsi_cfg = mp_ioapic_gsi_routing(i); + if (gsi >= gsi_cfg->gsi_base && gsi <= gsi_cfg->gsi_end) return i; } - printk(KERN_ERR "ERROR: Unable to locate IOAPIC for GSI %d\n", gsi); + pr_err("ERROR: Unable to locate IOAPIC for GSI %d\n", gsi); return -1; } @@ -2809,12 +2648,10 @@ static int bad_ioapic_register(int idx) static int find_free_ioapic_entry(void) { - int idx; - - for (idx = 0; idx < MAX_IO_APICS; idx++) + for (int idx = 0; idx < MAX_IO_APICS; idx++) { if (ioapics[idx].nr_registers == 0) return idx; - + } return MAX_IO_APICS; } @@ -2825,8 +2662,7 @@ static int find_free_ioapic_entry(void) * @gsi_base: base of GSI associated with the IOAPIC * @cfg: configuration information for the IOAPIC */ -int mp_register_ioapic(int id, u32 address, u32 gsi_base, - struct ioapic_domain_cfg *cfg) +int mp_register_ioapic(int id, u32 address, u32 gsi_base, struct ioapic_domain_cfg *cfg) { bool hotplug = !!ioapic_initialized; struct mp_ioapic_gsi *gsi_cfg; @@ -2837,12 +2673,13 @@ int mp_register_ioapic(int id, u32 address, u32 gsi_base, pr_warn("Bogus (zero) I/O APIC address found, skipping!\n"); return -EINVAL; } - for_each_ioapic(ioapic) + + for_each_ioapic(ioapic) { if (ioapics[ioapic].mp_config.apicaddr == address) { - pr_warn("address 0x%x conflicts with IOAPIC%d\n", - address, ioapic); + pr_warn("address 0x%x conflicts with IOAPIC%d\n", address, ioapic); return -EEXIST; } + } idx = find_free_ioapic_entry(); if (idx >= MAX_IO_APICS) { @@ -2877,8 +2714,7 @@ int mp_register_ioapic(int id, u32 address, u32 gsi_base, (gsi_end >= gsi_cfg->gsi_base && gsi_end <= gsi_cfg->gsi_end)) { pr_warn("GSI range [%u-%u] for new IOAPIC conflicts with GSI[%u-%u]\n", - gsi_base, gsi_end, - gsi_cfg->gsi_base, gsi_cfg->gsi_end); + gsi_base, gsi_end, gsi_cfg->gsi_base, gsi_cfg->gsi_end); clear_fixmap(FIX_IO_APIC_BASE_0 + idx); return -ENOSPC; } @@ -2912,8 +2748,7 @@ int mp_register_ioapic(int id, u32 address, u32 gsi_base, ioapics[idx].nr_registers = entries; pr_info("IOAPIC[%d]: apic_id %d, version %d, address 0x%x, GSI %d-%d\n", - idx, mpc_ioapic_id(idx), - mpc_ioapic_ver(idx), mpc_ioapic_addr(idx), + idx, mpc_ioapic_id(idx), mpc_ioapic_ver(idx), mpc_ioapic_addr(idx), gsi_cfg->gsi_base, gsi_cfg->gsi_end); return 0; @@ -2924,11 +2759,13 @@ int mp_unregister_ioapic(u32 gsi_base) int ioapic, pin; int found = 0; - for_each_ioapic(ioapic) + for_each_ioapic(ioapic) { if (ioapics[ioapic].gsi_config.gsi_base == gsi_base) { found = 1; break; } + } + if (!found) { pr_warn("can't find IOAPIC for GSI %d\n", gsi_base); return -ENODEV; @@ -2942,8 +2779,7 @@ int mp_unregister_ioapic(u32 gsi_base) if (irq >= 0) { data = irq_get_chip_data(irq); if (data && data->count) { - pr_warn("pin%d on IOAPIC%d is still in use.\n", - pin, ioapic); + pr_warn("pin%d on IOAPIC%d is still in use.\n", pin, ioapic); return -EBUSY; } } @@ -2978,8 +2814,7 @@ static void mp_irqdomain_get_attr(u32 gsi, struct mp_chip_data *data, if (info && info->ioapic.valid) { data->is_level = info->ioapic.is_level; data->active_low = info->ioapic.active_low; - } else if (__acpi_get_override_irq(gsi, &data->is_level, - &data->active_low) < 0) { + } else if (__acpi_get_override_irq(gsi, &data->is_level, &data->active_low) < 0) { /* PCI interrupts are always active low level triggered. */ data->is_level = true; data->active_low = true; @@ -3037,10 +2872,8 @@ int mp_irqdomain_alloc(struct irq_domain *domain, unsigned int virq, return -ENOMEM; ret = irq_domain_alloc_irqs_parent(domain, virq, nr_irqs, info); - if (ret < 0) { - kfree(data); - return ret; - } + if (ret < 0) + goto free_data; INIT_LIST_HEAD(&data->irq_2_pin); irq_data->hwirq = info->ioapic.pin; @@ -3049,7 +2882,10 @@ int mp_irqdomain_alloc(struct irq_domain *domain, unsigned int virq, irq_data->chip_data = data; mp_irqdomain_get_attr(mp_pin_to_gsi(ioapic, pin), data, info); - add_pin_to_irq_node(data, ioapic_alloc_attr_node(info), ioapic, pin); + if (!add_pin_to_irq_node(data, ioapic_alloc_attr_node(info), ioapic, pin)) { + ret = -ENOMEM; + goto free_irqs; + } mp_preconfigure_entry(data); mp_register_handler(virq, data->is_level); @@ -3059,11 +2895,15 @@ int mp_irqdomain_alloc(struct irq_domain *domain, unsigned int virq, legacy_pic->mask(virq); local_irq_restore(flags); - apic_printk(APIC_VERBOSE, KERN_DEBUG - "IOAPIC[%d]: Preconfigured routing entry (%d-%d -> IRQ %d Level:%i ActiveLow:%i)\n", - ioapic, mpc_ioapic_id(ioapic), pin, virq, - data->is_level, data->active_low); + apic_pr_verbose("IOAPIC[%d]: Preconfigured routing entry (%d-%d -> IRQ %d Level:%i ActiveLow:%i)\n", + ioapic, mpc_ioapic_id(ioapic), pin, virq, data->is_level, data->active_low); return 0; + +free_irqs: + irq_domain_free_irqs_parent(domain, virq, nr_irqs); +free_data: + kfree(data); + return ret; } void mp_irqdomain_free(struct irq_domain *domain, unsigned int virq, @@ -3076,22 +2916,17 @@ void mp_irqdomain_free(struct irq_domain *domain, unsigned int virq, irq_data = irq_domain_get_irq_data(domain, virq); if (irq_data && irq_data->chip_data) { data = irq_data->chip_data; - __remove_pin_from_irq(data, mp_irqdomain_ioapic_idx(domain), - (int)irq_data->hwirq); + __remove_pin_from_irq(data, mp_irqdomain_ioapic_idx(domain), (int)irq_data->hwirq); WARN_ON(!list_empty(&data->irq_2_pin)); kfree(irq_data->chip_data); } irq_domain_free_irqs_top(domain, virq, nr_irqs); } -int mp_irqdomain_activate(struct irq_domain *domain, - struct irq_data *irq_data, bool reserve) +int mp_irqdomain_activate(struct irq_domain *domain, struct irq_data *irq_data, bool reserve) { - unsigned long flags; - - raw_spin_lock_irqsave(&ioapic_lock, flags); + guard(raw_spinlock_irqsave)(&ioapic_lock); ioapic_configure_entry(irq_data); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); return 0; } @@ -3099,8 +2934,7 @@ void mp_irqdomain_deactivate(struct irq_domain *domain, struct irq_data *irq_data) { /* It won't be called for IRQ with multiple IOAPIC pins associated */ - ioapic_mask_entry(mp_irqdomain_ioapic_idx(domain), - (int)irq_data->hwirq); + ioapic_mask_entry(mp_irqdomain_ioapic_idx(domain), (int)irq_data->hwirq); } int mp_irqdomain_ioapic_idx(struct irq_domain *domain) diff --git a/arch/x86/kernel/apic/local.h b/arch/x86/kernel/apic/local.h index 9ea6186ea88c..842fe28496be 100644 --- a/arch/x86/kernel/apic/local.h +++ b/arch/x86/kernel/apic/local.h @@ -16,8 +16,6 @@ /* X2APIC */ void __x2apic_send_IPI_dest(unsigned int apicid, int vector, unsigned int dest); u32 x2apic_get_apic_id(u32 id); -u32 x2apic_set_apic_id(u32 id); -u32 x2apic_phys_pkg_id(u32 initial_apicid, int index_msb); void x2apic_send_IPI_all(int vector); void x2apic_send_IPI_allbutself(int vector); @@ -63,9 +61,6 @@ void default_send_IPI_allbutself(int vector); void default_send_IPI_all(int vector); void default_send_IPI_self(int vector); -bool default_apic_id_registered(void); -bool default_check_apicid_used(physid_mask_t *map, u32 apicid); - #ifdef CONFIG_X86_32 void default_send_IPI_mask_sequence_logical(const struct cpumask *mask, int vector); void default_send_IPI_mask_allbutself_logical(const struct cpumask *mask, int vector); diff --git a/arch/x86/kernel/apic/msi.c b/arch/x86/kernel/apic/msi.c index d9651f15ae4f..66bc5d3e79db 100644 --- a/arch/x86/kernel/apic/msi.c +++ b/arch/x86/kernel/apic/msi.c @@ -184,7 +184,6 @@ static int x86_msi_prepare(struct irq_domain *domain, struct device *dev, alloc->type = X86_IRQ_ALLOC_TYPE_PCI_MSI; return 0; case DOMAIN_BUS_PCI_DEVICE_MSIX: - case DOMAIN_BUS_PCI_DEVICE_IMS: alloc->type = X86_IRQ_ALLOC_TYPE_PCI_MSIX; return 0; default: @@ -215,6 +214,7 @@ static bool x86_init_dev_msi_info(struct device *dev, struct irq_domain *domain, if (WARN_ON_ONCE(domain != real_parent)) return false; info->chip->irq_set_affinity = msi_set_affinity; + info->chip->flags |= IRQCHIP_MOVE_DEFERRED; break; case DOMAIN_BUS_DMAR: case DOMAIN_BUS_AMDVI: @@ -229,10 +229,6 @@ static bool x86_init_dev_msi_info(struct device *dev, struct irq_domain *domain, case DOMAIN_BUS_PCI_DEVICE_MSI: case DOMAIN_BUS_PCI_DEVICE_MSIX: break; - case DOMAIN_BUS_PCI_DEVICE_IMS: - if (!(pops->supported_flags & MSI_FLAG_PCI_IMS)) - return false; - break; default: WARN_ON_ONCE(1); return false; @@ -320,7 +316,7 @@ static struct irq_chip dmar_msi_controller = { .irq_retrigger = irq_chip_retrigger_hierarchy, .irq_compose_msi_msg = dmar_msi_compose_msg, .irq_write_msi_msg = dmar_msi_write_msg, - .flags = IRQCHIP_SKIP_SET_WAKE | + .flags = IRQCHIP_SKIP_SET_WAKE | IRQCHIP_MOVE_DEFERRED | IRQCHIP_AFFINITY_PRE_STARTUP, }; diff --git a/arch/x86/kernel/apic/probe_32.c b/arch/x86/kernel/apic/probe_32.c index 5eb3fbe472da..f75ee345c02d 100644 --- a/arch/x86/kernel/apic/probe_32.c +++ b/arch/x86/kernel/apic/probe_32.c @@ -18,11 +18,6 @@ #include "local.h" -static u32 default_phys_pkg_id(u32 cpuid_apic, int index_msb) -{ - return cpuid_apic >> index_msb; -} - static u32 default_get_apic_id(u32 x) { unsigned int ver = GET_APIC_VERSION(apic_read(APIC_LVR)); @@ -43,18 +38,13 @@ static struct apic apic_default __ro_after_init = { .name = "default", .probe = probe_default, - .apic_id_registered = default_apic_id_registered, - .delivery_mode = APIC_DELIVERY_MODE_FIXED, .dest_mode_logical = true, .disable_esr = 0, - .check_apicid_used = default_check_apicid_used, .init_apic_ldr = default_init_apic_ldr, - .ioapic_phys_id_map = default_ioapic_phys_id_map, .cpu_present_to_apicid = default_cpu_present_to_apicid, - .phys_pkg_id = default_phys_pkg_id, .max_apic_id = 0xFE, .get_apic_id = default_get_apic_id, diff --git a/arch/x86/kernel/apic/vector.c b/arch/x86/kernel/apic/vector.c index 319448d87b99..736f62812f5c 100644 --- a/arch/x86/kernel/apic/vector.c +++ b/arch/x86/kernel/apic/vector.c @@ -712,8 +712,8 @@ int __init arch_probe_nr_irqs(void) { int nr; - if (nr_irqs > (NR_VECTORS * nr_cpu_ids)) - nr_irqs = NR_VECTORS * nr_cpu_ids; + if (irq_get_nr_irqs() > NR_VECTORS * nr_cpu_ids) + irq_set_nr_irqs(NR_VECTORS * nr_cpu_ids); nr = (gsi_top + nr_legacy_irqs()) + 8 * nr_cpu_ids; #if defined(CONFIG_PCI_MSI) @@ -725,8 +725,8 @@ int __init arch_probe_nr_irqs(void) else nr += gsi_top * 16; #endif - if (nr < nr_irqs) - nr_irqs = nr; + if (nr < irq_get_nr_irqs()) + irq_set_nr_irqs(nr); /* * We don't know if PIC is present at this point so we need to do @@ -738,8 +738,8 @@ int __init arch_probe_nr_irqs(void) void lapic_assign_legacy_vector(unsigned int irq, bool replace) { /* - * Use assign system here so it wont get accounted as allocated - * and moveable in the cpu hotplug check and it prevents managed + * Use assign system here so it won't get accounted as allocated + * and movable in the cpu hotplug check and it prevents managed * irq reservation from touching it. */ irq_matrix_assign_system(vector_matrix, ISA_IRQ_VECTOR(irq), replace); @@ -965,7 +965,7 @@ static void __vector_cleanup(struct vector_cleanup *cl, bool check_irr) lockdep_assert_held(&vector_lock); hlist_for_each_entry_safe(apicd, tmp, &cl->head, clist) { - unsigned int irr, vector = apicd->prev_vector; + unsigned int vector = apicd->prev_vector; /* * Paranoia: Check if the vector that needs to be cleaned @@ -979,8 +979,7 @@ static void __vector_cleanup(struct vector_cleanup *cl, bool check_irr) * fixup_irqs() was just called to scan IRR for set bits and * forward them to new destination CPUs via IPIs. */ - irr = check_irr ? apic_read(APIC_IRR + (vector / 32 * 0x10)) : 0; - if (irr & (1U << (vector % 32))) { + if (check_irr && is_vector_pending(vector)) { pr_warn_once("Moved interrupt pending in old target APIC %u\n", apicd->irq); rearm = true; continue; @@ -1036,7 +1035,8 @@ static void __vector_schedule_cleanup(struct apic_chip_data *apicd) add_timer_on(&cl->timer, cpu); } } else { - apicd->prev_vector = 0; + pr_warn("IRQ %u schedule cleanup for offline CPU %u\n", apicd->irq, cpu); + free_moved_vector(apicd); } raw_spin_unlock(&vector_lock); } @@ -1073,6 +1073,7 @@ void irq_complete_move(struct irq_cfg *cfg) */ void irq_force_complete_move(struct irq_desc *desc) { + unsigned int cpu = smp_processor_id(); struct apic_chip_data *apicd; struct irq_data *irqd; unsigned int vector; @@ -1097,10 +1098,11 @@ void irq_force_complete_move(struct irq_desc *desc) goto unlock; /* - * If prev_vector is empty, no action required. + * If prev_vector is empty or the descriptor is neither currently + * nor previously on the outgoing CPU no action required. */ vector = apicd->prev_vector; - if (!vector) + if (!vector || (apicd->cpu != cpu && apicd->prev_cpu != cpu)) goto unlock; /* diff --git a/arch/x86/kernel/apic/x2apic_cluster.c b/arch/x86/kernel/apic/x2apic_cluster.c index a8306089c91b..7db83212effb 100644 --- a/arch/x86/kernel/apic/x2apic_cluster.c +++ b/arch/x86/kernel/apic/x2apic_cluster.c @@ -178,13 +178,16 @@ static int x2apic_prepare_cpu(unsigned int cpu) u32 phys_apicid = apic->cpu_present_to_apicid(cpu); u32 cluster = apic_cluster(phys_apicid); u32 logical_apicid = (cluster << 16) | (1 << (phys_apicid & 0xf)); + int node = cpu_to_node(cpu); x86_cpu_to_logical_apicid[cpu] = logical_apicid; - if (alloc_clustermask(cpu, cluster, cpu_to_node(cpu)) < 0) + if (alloc_clustermask(cpu, cluster, node) < 0) return -ENOMEM; - if (!zalloc_cpumask_var(&per_cpu(ipi_mask, cpu), GFP_KERNEL)) + + if (!zalloc_cpumask_var_node(&per_cpu(ipi_mask, cpu), GFP_KERNEL, node)) return -ENOMEM; + return 0; } @@ -227,21 +230,16 @@ static struct apic apic_x2apic_cluster __ro_after_init = { .probe = x2apic_cluster_probe, .acpi_madt_oem_check = x2apic_acpi_madt_oem_check, - .delivery_mode = APIC_DELIVERY_MODE_FIXED, .dest_mode_logical = true, .disable_esr = 0, - .check_apicid_used = NULL, .init_apic_ldr = init_x2apic_ldr, - .ioapic_phys_id_map = NULL, .cpu_present_to_apicid = default_cpu_present_to_apicid, - .phys_pkg_id = x2apic_phys_pkg_id, .max_apic_id = UINT_MAX, .x2apic_set_max_apicid = true, .get_apic_id = x2apic_get_apic_id, - .set_apic_id = x2apic_set_apic_id, .calc_dest_apicid = x2apic_calc_apicid, diff --git a/arch/x86/kernel/apic/x2apic_phys.c b/arch/x86/kernel/apic/x2apic_phys.c index 558a4a8824f4..12d4c35547a6 100644 --- a/arch/x86/kernel/apic/x2apic_phys.c +++ b/arch/x86/kernel/apic/x2apic_phys.c @@ -129,34 +129,21 @@ u32 x2apic_get_apic_id(u32 id) return id; } -u32 x2apic_set_apic_id(u32 id) -{ - return id; -} - -u32 x2apic_phys_pkg_id(u32 initial_apicid, int index_msb) -{ - return initial_apicid >> index_msb; -} - static struct apic apic_x2apic_phys __ro_after_init = { .name = "physical x2apic", .probe = x2apic_phys_probe, .acpi_madt_oem_check = x2apic_acpi_madt_oem_check, - .delivery_mode = APIC_DELIVERY_MODE_FIXED, .dest_mode_logical = false, .disable_esr = 0, .cpu_present_to_apicid = default_cpu_present_to_apicid, - .phys_pkg_id = x2apic_phys_pkg_id, .max_apic_id = UINT_MAX, .x2apic_set_max_apicid = true, .get_apic_id = x2apic_get_apic_id, - .set_apic_id = x2apic_set_apic_id, .calc_dest_apicid = apic_default_calc_apicid, diff --git a/arch/x86/kernel/apic/x2apic_uv_x.c b/arch/x86/kernel/apic/x2apic_uv_x.c index 1b0d7336a28f..7fef504ca508 100644 --- a/arch/x86/kernel/apic/x2apic_uv_x.c +++ b/arch/x86/kernel/apic/x2apic_uv_x.c @@ -241,54 +241,20 @@ static void __init uv_tsc_check_sync(void) is_uv(UV3) ? sname.s3.field : \ undef) -/* [Copied from arch/x86/kernel/cpu/topology.c:detect_extended_topology()] */ - -#define SMT_LEVEL 0 /* Leaf 0xb SMT level */ -#define INVALID_TYPE 0 /* Leaf 0xb sub-leaf types */ -#define SMT_TYPE 1 -#define CORE_TYPE 2 -#define LEAFB_SUBTYPE(ecx) (((ecx) >> 8) & 0xff) -#define BITS_SHIFT_NEXT_LEVEL(eax) ((eax) & 0x1f) - -static void set_x2apic_bits(void) -{ - unsigned int eax, ebx, ecx, edx, sub_index; - unsigned int sid_shift; - - cpuid(0, &eax, &ebx, &ecx, &edx); - if (eax < 0xb) { - pr_info("UV: CPU does not have CPUID.11\n"); - return; - } - - cpuid_count(0xb, SMT_LEVEL, &eax, &ebx, &ecx, &edx); - if (ebx == 0 || (LEAFB_SUBTYPE(ecx) != SMT_TYPE)) { - pr_info("UV: CPUID.11 not implemented\n"); - return; - } - - sid_shift = BITS_SHIFT_NEXT_LEVEL(eax); - sub_index = 1; - do { - cpuid_count(0xb, sub_index, &eax, &ebx, &ecx, &edx); - if (LEAFB_SUBTYPE(ecx) == CORE_TYPE) { - sid_shift = BITS_SHIFT_NEXT_LEVEL(eax); - break; - } - sub_index++; - } while (LEAFB_SUBTYPE(ecx) != INVALID_TYPE); - - uv_cpuid.apicid_shift = 0; - uv_cpuid.apicid_mask = (~(-1 << sid_shift)); - uv_cpuid.socketid_shift = sid_shift; -} - static void __init early_get_apic_socketid_shift(void) { + unsigned int sid_shift = topology_get_domain_shift(TOPO_PKG_DOMAIN); + if (is_uv2_hub() || is_uv3_hub()) uvh_apicid.v = uv_early_read_mmr(UVH_APICID); - set_x2apic_bits(); + if (sid_shift) { + uv_cpuid.apicid_shift = 0; + uv_cpuid.apicid_mask = (~(-1 << sid_shift)); + uv_cpuid.socketid_shift = sid_shift; + } else { + pr_info("UV: CPU does not have valid CPUID.11\n"); + } pr_info("UV: apicid_shift:%d apicid_mask:0x%x\n", uv_cpuid.apicid_shift, uv_cpuid.apicid_mask); pr_info("UV: socketid_shift:%d pnode_mask:0x%x\n", uv_cpuid.socketid_shift, uv_cpuid.pnode_mask); @@ -779,21 +745,6 @@ static void uv_send_IPI_all(int vector) uv_send_IPI_mask(cpu_online_mask, vector); } -static u32 set_apic_id(u32 id) -{ - return id; -} - -static unsigned int uv_read_apic_id(void) -{ - return x2apic_get_apic_id(apic_read(APIC_ID)); -} - -static u32 uv_phys_pkg_id(u32 initial_apicid, int index_msb) -{ - return uv_read_apic_id() >> index_msb; -} - static int uv_probe(void) { return apic == &apic_x2apic_uv_x; @@ -805,17 +756,14 @@ static struct apic apic_x2apic_uv_x __ro_after_init = { .probe = uv_probe, .acpi_madt_oem_check = uv_acpi_madt_oem_check, - .delivery_mode = APIC_DELIVERY_MODE_FIXED, .dest_mode_logical = false, .disable_esr = 0, .cpu_present_to_apicid = default_cpu_present_to_apicid, - .phys_pkg_id = uv_phys_pkg_id, .max_apic_id = UINT_MAX, .get_apic_id = x2apic_get_apic_id, - .set_apic_id = set_apic_id, .calc_dest_apicid = apic_default_calc_apicid, |