From d364847eed890211444ad74496bb549f838c6018 Mon Sep 17 00:00:00 2001 From: Thomas Gleixner Date: Tue, 25 Feb 2020 14:55:15 +0100 Subject: x86/mce/therm_throt: Undo thermal polling properly on CPU offline Chris Wilson reported splats from running the thermal throttling workqueue callback on offlined CPUs. The problem is that that callback should not even run on offlined CPUs but it happens nevertheless because the offlining callback thermal_throttle_offline() does not symmetrically undo the setup work done in its onlining counterpart. IOW, 1. The thermal interrupt vector should be masked out before ... 2. ... cancelling any pending work synchronously so that no new work is enqueued anymore. Do those things and fix the issue properly. [ bp: Write commit message. ] Fixes: f6656208f04e ("x86/mce/therm_throt: Optimize notifications of thermal throttle") Reported-by: Chris Wilson Tested-by: Pandruvada, Srinivas Signed-off-by: Thomas Gleixner Signed-off-by: Borislav Petkov Link: https://lkml.kernel.org/r/158120068234.18291.7938335950259651295@skylake-alporthouse-com --- arch/x86/kernel/cpu/mce/therm_throt.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/cpu/mce/therm_throt.c b/arch/x86/kernel/cpu/mce/therm_throt.c index 58b4ee3cda77..f36dc0742085 100644 --- a/arch/x86/kernel/cpu/mce/therm_throt.c +++ b/arch/x86/kernel/cpu/mce/therm_throt.c @@ -486,9 +486,14 @@ static int thermal_throttle_offline(unsigned int cpu) { struct thermal_state *state = &per_cpu(thermal_state, cpu); struct device *dev = get_cpu_device(cpu); + u32 l; + + /* Mask the thermal vector before draining evtl. pending work */ + l = apic_read(APIC_LVTTHMR); + apic_write(APIC_LVTTHMR, l | APIC_LVT_MASKED); - cancel_delayed_work(&state->package_throttle.therm_work); - cancel_delayed_work(&state->core_throttle.therm_work); + cancel_delayed_work_sync(&state->package_throttle.therm_work); + cancel_delayed_work_sync(&state->core_throttle.therm_work); state->package_throttle.rate_control_active = false; state->core_throttle.rate_control_active = false; -- cgit v1.2.3 From 59b5809655bdafb0767d3fd00a3e41711aab07e6 Mon Sep 17 00:00:00 2001 From: Tony Luck Date: Tue, 25 Feb 2020 17:17:37 -0800 Subject: x86/mce: Fix logic and comments around MSR_PPIN_CTL There are two implemented bits in the PPIN_CTL MSR: Bit 0: LockOut (R/WO) Set 1 to prevent further writes to MSR_PPIN_CTL. Bit 1: Enable_PPIN (R/W) If 1, enables MSR_PPIN to be accessible using RDMSR. If 0, an attempt to read MSR_PPIN will cause #GP. So there are four defined values: 0: PPIN is disabled, PPIN_CTL may be updated 1: PPIN is disabled. PPIN_CTL is locked against updates 2: PPIN is enabled. PPIN_CTL may be updated 3: PPIN is enabled. PPIN_CTL is locked against updates Code would only enable the X86_FEATURE_INTEL_PPIN feature for case "2". When it should have done so for both case "2" and case "3". Fix the final test to just check for the enable bit. Also fix some of the other comments in this function. Fixes: 3f5a7896a509 ("x86/mce: Include the PPIN in MCE records when available") Signed-off-by: Tony Luck Signed-off-by: Borislav Petkov Cc: Link: https://lkml.kernel.org/r/20200226011737.9958-1-tony.luck@intel.com --- arch/x86/kernel/cpu/mce/intel.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/cpu/mce/intel.c b/arch/x86/kernel/cpu/mce/intel.c index 5627b1091b85..f996ffb887bc 100644 --- a/arch/x86/kernel/cpu/mce/intel.c +++ b/arch/x86/kernel/cpu/mce/intel.c @@ -493,17 +493,18 @@ static void intel_ppin_init(struct cpuinfo_x86 *c) return; if ((val & 3UL) == 1UL) { - /* PPIN available but disabled: */ + /* PPIN locked in disabled mode */ return; } - /* If PPIN is disabled, but not locked, try to enable: */ - if (!(val & 3UL)) { + /* If PPIN is disabled, try to enable */ + if (!(val & 2UL)) { wrmsrl_safe(MSR_PPIN_CTL, val | 2UL); rdmsrl_safe(MSR_PPIN_CTL, &val); } - if ((val & 3UL) == 2UL) + /* Is the enable bit set? */ + if (val & 2UL) set_cpu_cap(c, X86_FEATURE_INTEL_PPIN); } } -- cgit v1.2.3 From 9a11997e757bcf716c1b199ea7bd1abbadc4b357 Mon Sep 17 00:00:00 2001 From: Wanpeng Li Date: Mon, 2 Mar 2020 15:15:36 +0800 Subject: KVM: X86: Fix dereference null cpufreq policy Naresh Kamboju reported: Linux version 5.6.0-rc4 (oe-user@oe-host) (gcc version (GCC)) #1 SMP Sun Mar 1 22:59:08 UTC 2020 kvm: no hardware support BUG: kernel NULL pointer dereference, address: 000000000000028c #PF: supervisor read access in kernel mode #PF: error_code(0x0000) - not-present page PGD 0 P4D 0 Oops: 0000 [#1] SMP NOPTI CPU: 0 PID: 1 Comm: swapper/0 Not tainted 5.6.0-rc4 #1 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), 04/01/2014 RIP: 0010:kobject_put+0x12/0x1c0 Call Trace: cpufreq_cpu_put+0x15/0x20 kvm_arch_init+0x1f6/0x2b0 kvm_init+0x31/0x290 ? svm_check_processor_compat+0xd/0xd ? svm_check_processor_compat+0xd/0xd svm_init+0x21/0x23 do_one_initcall+0x61/0x2f0 ? rdinit_setup+0x30/0x30 ? rcu_read_lock_sched_held+0x4f/0x80 kernel_init_freeable+0x219/0x279 ? rest_init+0x250/0x250 kernel_init+0xe/0x110 ret_from_fork+0x27/0x50 Modules linked in: CR2: 000000000000028c ---[ end trace 239abf40c55c409b ]--- RIP: 0010:kobject_put+0x12/0x1c0 cpufreq policy which is get by cpufreq_cpu_get() can be NULL if it is failure, this patch takes care of it. Fixes: aaec7c03de (KVM: x86: avoid useless copy of cpufreq policy) Reported-by: Naresh Kamboju Cc: Naresh Kamboju Signed-off-by: Wanpeng Li Signed-off-by: Paolo Bonzini --- arch/x86/kvm/x86.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index 5de200663f51..3156e25b0774 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -7195,10 +7195,12 @@ static void kvm_timer_init(void) cpu = get_cpu(); policy = cpufreq_cpu_get(cpu); - if (policy && policy->cpuinfo.max_freq) - max_tsc_khz = policy->cpuinfo.max_freq; + if (policy) { + if (policy->cpuinfo.max_freq) + max_tsc_khz = policy->cpuinfo.max_freq; + cpufreq_cpu_put(policy); + } put_cpu(); - cpufreq_cpu_put(policy); #endif cpufreq_register_notifier(&kvmclock_cpufreq_notifier_block, CPUFREQ_TRANSITION_NOTIFIER); -- cgit v1.2.3 From aaca21007ba14005feca74a079888c49a9b6f6ac Mon Sep 17 00:00:00 2001 From: Haiwei Li Date: Mon, 2 Mar 2020 20:19:28 +0800 Subject: KVM: SVM: Fix the svm vmexit code for WRMSR In svm, exit_code for MSR writes is not EXIT_REASON_MSR_WRITE which belongs to vmx. According to amd manual, SVM_EXIT_MSR(7ch) is the exit_code of VMEXIT_MSR due to RDMSR or WRMSR access to protected MSR. Additionally, the processor indicates in the VMCB's EXITINFO1 whether a RDMSR(EXITINFO1=0) or WRMSR(EXITINFO1=1) was intercepted. Signed-off-by: Haiwei Li Fixes: 1e9e2622a149 ("KVM: VMX: FIXED+PHYSICAL mode single target IPI fastpath", 2019-11-21) Signed-off-by: Paolo Bonzini --- arch/x86/kvm/svm.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/kvm/svm.c b/arch/x86/kvm/svm.c index 24c0b2ba8fb9..91000501756e 100644 --- a/arch/x86/kvm/svm.c +++ b/arch/x86/kvm/svm.c @@ -6312,7 +6312,8 @@ static void svm_handle_exit_irqoff(struct kvm_vcpu *vcpu, enum exit_fastpath_completion *exit_fastpath) { if (!is_guest_mode(vcpu) && - to_svm(vcpu)->vmcb->control.exit_code == EXIT_REASON_MSR_WRITE) + to_svm(vcpu)->vmcb->control.exit_code == SVM_EXIT_MSR && + to_svm(vcpu)->vmcb->control.exit_info_1) *exit_fastpath = handle_fastpath_set_msr_irqoff(vcpu); } -- cgit v1.2.3 From 342993f96ab24d5864ab1216f46c0b199c2baf8e Mon Sep 17 00:00:00 2001 From: Vitaly Kuznetsov Date: Tue, 3 Mar 2020 15:33:15 +0100 Subject: KVM: x86: clear stale x86_emulate_ctxt->intercept value After commit 07721feee46b ("KVM: nVMX: Don't emulate instructions in guest mode") Hyper-V guests on KVM stopped booting with: kvm_nested_vmexit: rip fffff802987d6169 reason EPT_VIOLATION info1 181 info2 0 int_info 0 int_info_err 0 kvm_page_fault: address febd0000 error_code 181 kvm_emulate_insn: 0:fffff802987d6169: f3 a5 kvm_emulate_insn: 0:fffff802987d6169: f3 a5 FAIL kvm_inj_exception: #UD (0x0) "f3 a5" is a "rep movsw" instruction, which should not be intercepted at all. Commit c44b4c6ab80e ("KVM: emulate: clean up initializations in init_decode_cache") reduced the number of fields cleared by init_decode_cache() claiming that they are being cleared elsewhere, 'intercept', however, is left uncleared if the instruction does not have any of the "slow path" flags (NotImpl, Stack, Op3264, Sse, Mmx, CheckPerm, NearBranch, No16 and of course Intercept itself). Fixes: c44b4c6ab80e ("KVM: emulate: clean up initializations in init_decode_cache") Fixes: 07721feee46b ("KVM: nVMX: Don't emulate instructions in guest mode") Cc: stable@vger.kernel.org Suggested-by: Paolo Bonzini Signed-off-by: Vitaly Kuznetsov Reviewed-by: Sean Christopherson Signed-off-by: Paolo Bonzini --- arch/x86/kvm/emulate.c | 1 + 1 file changed, 1 insertion(+) (limited to 'arch/x86') diff --git a/arch/x86/kvm/emulate.c b/arch/x86/kvm/emulate.c index dd19fb3539e0..bc00642e5d3b 100644 --- a/arch/x86/kvm/emulate.c +++ b/arch/x86/kvm/emulate.c @@ -5173,6 +5173,7 @@ int x86_decode_insn(struct x86_emulate_ctxt *ctxt, void *insn, int insn_len) ctxt->fetch.ptr = ctxt->fetch.data; ctxt->fetch.end = ctxt->fetch.data + insn_len; ctxt->opcode_len = 1; + ctxt->intercept = x86_intercept_none; if (insn_len > 0) memcpy(ctxt->fetch.data, insn, insn_len); else { -- cgit v1.2.3 From d718fdc3e752ba51ddb2b5554d3db98a09355cc2 Mon Sep 17 00:00:00 2001 From: Vitaly Kuznetsov Date: Tue, 3 Mar 2020 15:33:16 +0100 Subject: KVM: x86: remove stale comment from struct x86_emulate_ctxt Commit c44b4c6ab80e ("KVM: emulate: clean up initializations in init_decode_cache") did some field shuffling and instead of [opcode_len, _regs) started clearing [has_seg_override, modrm). The comment about clearing fields altogether is not true anymore. Fixes: c44b4c6ab80e ("KVM: emulate: clean up initializations in init_decode_cache") Signed-off-by: Vitaly Kuznetsov Signed-off-by: Paolo Bonzini --- arch/x86/include/asm/kvm_emulate.h | 1 - 1 file changed, 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/include/asm/kvm_emulate.h b/arch/x86/include/asm/kvm_emulate.h index 2a8f2bd2e5cf..c06e8353efd3 100644 --- a/arch/x86/include/asm/kvm_emulate.h +++ b/arch/x86/include/asm/kvm_emulate.h @@ -360,7 +360,6 @@ struct x86_emulate_ctxt { u64 d; unsigned long _eip; struct operand memop; - /* Fields above regs are cleared together. */ unsigned long _regs[NR_VCPU_REGS]; struct operand *memopp; struct fetch_cache fetch; -- cgit v1.2.3 From a754acc3e4bcf0b70f4356fc450fec72875762da Mon Sep 17 00:00:00 2001 From: "Jason A. Donenfeld" Date: Wed, 4 Mar 2020 18:42:21 +0800 Subject: KVM: fix Kconfig menu text for -Werror This was evidently copy and pasted from the i915 driver, but the text wasn't updated. Fixes: 4f337faf1c55 ("KVM: allow disabling -Werror") Signed-off-by: Jason A. Donenfeld Signed-off-by: Paolo Bonzini --- arch/x86/kvm/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'arch/x86') diff --git a/arch/x86/kvm/Kconfig b/arch/x86/kvm/Kconfig index 1bb4927030af..9fea0757db92 100644 --- a/arch/x86/kvm/Kconfig +++ b/arch/x86/kvm/Kconfig @@ -68,7 +68,7 @@ config KVM_WERROR depends on (X86_64 && !KASAN) || !COMPILE_TEST depends on EXPERT help - Add -Werror to the build flags for (and only for) i915.ko. + Add -Werror to the build flags for KVM. If in doubt, say "N". -- cgit v1.2.3 From 80f1f85036355e5581ec0b99913410345ad3491b Mon Sep 17 00:00:00 2001 From: Luke Nelson Date: Thu, 5 Mar 2020 15:44:12 -0800 Subject: bpf, x32: Fix bug with JMP32 JSET BPF_X checking upper bits The current x32 BPF JIT is incorrect for JMP32 JSET BPF_X when the upper 32 bits of operand registers are non-zero in certain situations. The problem is in the following code: case BPF_JMP | BPF_JSET | BPF_X: case BPF_JMP32 | BPF_JSET | BPF_X: ... /* and dreg_lo,sreg_lo */ EMIT2(0x23, add_2reg(0xC0, sreg_lo, dreg_lo)); /* and dreg_hi,sreg_hi */ EMIT2(0x23, add_2reg(0xC0, sreg_hi, dreg_hi)); /* or dreg_lo,dreg_hi */ EMIT2(0x09, add_2reg(0xC0, dreg_lo, dreg_hi)); This code checks the upper bits of the operand registers regardless if the BPF instruction is BPF_JMP32 or BPF_JMP64. Registers dreg_hi and dreg_lo are not loaded from the stack for BPF_JMP32, however, they can still be polluted with values from previous instructions. The following BPF program demonstrates the bug. The jset64 instruction loads the temporary registers and performs the jump, since ((u64)r7 & (u64)r8) is non-zero. The jset32 should _not_ be taken, as the lower 32 bits are all zero, however, the current JIT will take the branch due the pollution of temporary registers from the earlier jset64. mov64 r0, 0 ld64 r7, 0x8000000000000000 ld64 r8, 0x8000000000000000 jset64 r7, r8, 1 exit jset32 r7, r8, 1 mov64 r0, 2 exit The expected return value of this program is 2; under the buggy x32 JIT it returns 0. The fix is to skip using the upper 32 bits for jset32 and compare the upper 32 bits for jset64 only. All tests in test_bpf.ko and selftests/bpf/test_verifier continue to pass with this change. We found this bug using our automated verification tool, Serval. Fixes: 69f827eb6e14 ("x32: bpf: implement jitting of JMP32") Co-developed-by: Xi Wang Signed-off-by: Xi Wang Signed-off-by: Luke Nelson Signed-off-by: Daniel Borkmann Link: https://lore.kernel.org/bpf/20200305234416.31597-1-luke.r.nels@gmail.com --- arch/x86/net/bpf_jit_comp32.c | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/net/bpf_jit_comp32.c b/arch/x86/net/bpf_jit_comp32.c index 393d251798c0..4d2a7a764602 100644 --- a/arch/x86/net/bpf_jit_comp32.c +++ b/arch/x86/net/bpf_jit_comp32.c @@ -2039,10 +2039,12 @@ static int do_jit(struct bpf_prog *bpf_prog, int *addrs, u8 *image, } /* and dreg_lo,sreg_lo */ EMIT2(0x23, add_2reg(0xC0, sreg_lo, dreg_lo)); - /* and dreg_hi,sreg_hi */ - EMIT2(0x23, add_2reg(0xC0, sreg_hi, dreg_hi)); - /* or dreg_lo,dreg_hi */ - EMIT2(0x09, add_2reg(0xC0, dreg_lo, dreg_hi)); + if (is_jmp64) { + /* and dreg_hi,sreg_hi */ + EMIT2(0x23, add_2reg(0xC0, sreg_hi, dreg_hi)); + /* or dreg_lo,dreg_hi */ + EMIT2(0x09, add_2reg(0xC0, dreg_lo, dreg_hi)); + } goto emit_cond_jmp; } case BPF_JMP | BPF_JSET | BPF_K: -- cgit v1.2.3 From 985e537a4082b4635754a57f4f95430790afee6a Mon Sep 17 00:00:00 2001 From: Tom Lendacky Date: Tue, 10 Mar 2020 18:35:57 +0100 Subject: x86/ioremap: Map EFI runtime services data as encrypted for SEV The dmidecode program fails to properly decode the SMBIOS data supplied by OVMF/UEFI when running in an SEV guest. The SMBIOS area, under SEV, is encrypted and resides in reserved memory that is marked as EFI runtime services data. As a result, when memremap() is attempted for the SMBIOS data, it can't be mapped as regular RAM (through try_ram_remap()) and, since the address isn't part of the iomem resources list, it isn't mapped encrypted through the fallback ioremap(). Add a new __ioremap_check_other() to deal with memory types like EFI_RUNTIME_SERVICES_DATA which are not covered by the resource ranges. This allows any runtime services data which has been created encrypted, to be mapped encrypted too. [ bp: Move functionality to a separate function. ] Signed-off-by: Tom Lendacky Signed-off-by: Borislav Petkov Reviewed-by: Joerg Roedel Tested-by: Joerg Roedel Cc: # 5.3 Link: https://lkml.kernel.org/r/2d9e16eb5b53dc82665c95c6764b7407719df7a0.1582645327.git.thomas.lendacky@amd.com --- arch/x86/mm/ioremap.c | 18 ++++++++++++++++++ 1 file changed, 18 insertions(+) (limited to 'arch/x86') diff --git a/arch/x86/mm/ioremap.c b/arch/x86/mm/ioremap.c index 44e4beb4239f..935a91e1fd77 100644 --- a/arch/x86/mm/ioremap.c +++ b/arch/x86/mm/ioremap.c @@ -106,6 +106,19 @@ static unsigned int __ioremap_check_encrypted(struct resource *res) return 0; } +/* + * The EFI runtime services data area is not covered by walk_mem_res(), but must + * be mapped encrypted when SEV is active. + */ +static void __ioremap_check_other(resource_size_t addr, struct ioremap_desc *desc) +{ + if (!sev_active()) + return; + + if (efi_mem_type(addr) == EFI_RUNTIME_SERVICES_DATA) + desc->flags |= IORES_MAP_ENCRYPTED; +} + static int __ioremap_collect_map_flags(struct resource *res, void *arg) { struct ioremap_desc *desc = arg; @@ -124,6 +137,9 @@ static int __ioremap_collect_map_flags(struct resource *res, void *arg) * To avoid multiple resource walks, this function walks resources marked as * IORESOURCE_MEM and IORESOURCE_BUSY and looking for system RAM and/or a * resource described not as IORES_DESC_NONE (e.g. IORES_DESC_ACPI_TABLES). + * + * After that, deal with misc other ranges in __ioremap_check_other() which do + * not fall into the above category. */ static void __ioremap_check_mem(resource_size_t addr, unsigned long size, struct ioremap_desc *desc) @@ -135,6 +151,8 @@ static void __ioremap_check_mem(resource_size_t addr, unsigned long size, memset(desc, 0, sizeof(struct ioremap_desc)); walk_mem_res(start, end, desc, __ioremap_collect_map_flags); + + __ioremap_check_other(addr, desc); } /* -- cgit v1.2.3 From f967140dfb7442e2db0868b03b961f9c59418a1b Mon Sep 17 00:00:00 2001 From: Kim Phillips Date: Wed, 11 Mar 2020 14:13:21 -0500 Subject: perf/amd/uncore: Replace manual sampling check with CAP_NO_INTERRUPT flag Enable the sampling check in kernel/events/core.c::perf_event_open(), which returns the more appropriate -EOPNOTSUPP. BEFORE: $ sudo perf record -a -e instructions,l3_request_g1.caching_l3_cache_accesses true Error: The sys_perf_event_open() syscall returned with 22 (Invalid argument) for event (l3_request_g1.caching_l3_cache_accesses). /bin/dmesg | grep -i perf may provide additional information. With nothing relevant in dmesg. AFTER: $ sudo perf record -a -e instructions,l3_request_g1.caching_l3_cache_accesses true Error: l3_request_g1.caching_l3_cache_accesses: PMU Hardware doesn't support sampling/overflow-interrupts. Try 'perf stat' Fixes: c43ca5091a37 ("perf/x86/amd: Add support for AMD NB and L2I "uncore" counters") Signed-off-by: Kim Phillips Signed-off-by: Borislav Petkov Acked-by: Peter Zijlstra Cc: stable@vger.kernel.org Link: https://lkml.kernel.org/r/20200311191323.13124-1-kim.phillips@amd.com --- arch/x86/events/amd/uncore.c | 17 +++++++---------- 1 file changed, 7 insertions(+), 10 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/events/amd/uncore.c b/arch/x86/events/amd/uncore.c index a6ea07f2aa84..4d867a752f0e 100644 --- a/arch/x86/events/amd/uncore.c +++ b/arch/x86/events/amd/uncore.c @@ -190,15 +190,12 @@ static int amd_uncore_event_init(struct perf_event *event) /* * NB and Last level cache counters (MSRs) are shared across all cores - * that share the same NB / Last level cache. Interrupts can be directed - * to a single target core, however, event counts generated by processes - * running on other cores cannot be masked out. So we do not support - * sampling and per-thread events. + * that share the same NB / Last level cache. On family 16h and below, + * Interrupts can be directed to a single target core, however, event + * counts generated by processes running on other cores cannot be masked + * out. So we do not support sampling and per-thread events via + * CAP_NO_INTERRUPT, and we do not enable counter overflow interrupts: */ - if (is_sampling_event(event) || event->attach_state & PERF_ATTACH_TASK) - return -EINVAL; - - /* and we do not enable counter overflow interrupts */ hwc->config = event->attr.config & AMD64_RAW_EVENT_MASK_NB; hwc->idx = -1; @@ -306,7 +303,7 @@ static struct pmu amd_nb_pmu = { .start = amd_uncore_start, .stop = amd_uncore_stop, .read = amd_uncore_read, - .capabilities = PERF_PMU_CAP_NO_EXCLUDE, + .capabilities = PERF_PMU_CAP_NO_EXCLUDE | PERF_PMU_CAP_NO_INTERRUPT, }; static struct pmu amd_llc_pmu = { @@ -317,7 +314,7 @@ static struct pmu amd_llc_pmu = { .start = amd_uncore_start, .stop = amd_uncore_stop, .read = amd_uncore_read, - .capabilities = PERF_PMU_CAP_NO_EXCLUDE, + .capabilities = PERF_PMU_CAP_NO_EXCLUDE | PERF_PMU_CAP_NO_INTERRUPT, }; static struct amd_uncore *amd_uncore_alloc(unsigned int cpu) -- cgit v1.2.3 From 469ff207b4c4033540b50bc59587dc915faa1367 Mon Sep 17 00:00:00 2001 From: Peter Xu Date: Thu, 12 Mar 2020 16:58:30 -0400 Subject: x86/vector: Remove warning on managed interrupt migration The vector management code assumes that managed interrupts cannot be migrated away from an online CPU. free_moved_vector() has a WARN_ON_ONCE() which triggers when a managed interrupt vector association on a online CPU is cleared. The CPU offline code uses a different mechanism which cannot trigger this. This assumption is not longer correct because the new CPU isolation feature which affects the placement of managed interrupts must be able to move a managed interrupt away from an online CPU. There are two reasons why this can happen: 1) When the interrupt is activated the affinity mask which was established in irq_create_affinity_masks() is handed in to the vector allocation code. This mask contains all CPUs to which the interrupt can be made affine to, but this does not take the CPU isolation 'managed_irq' mask into account. When the interrupt is finally requested by the device driver then the affinity is checked again and the CPU isolation 'managed_irq' mask is taken into account, which moves the interrupt to a non-isolated CPU if possible. 2) The interrupt can be affine to an isolated CPU because the non-isolated CPUs in the calculated affinity mask are not online. Once a non-isolated CPU which is in the mask comes online the interrupt is migrated to this non-isolated CPU In both cases the regular online migration mechanism is used which triggers the WARN_ON_ONCE() in free_moved_vector(). Case #1 could have been addressed by taking the isolation mask into account, but that would require a massive code change in the activation logic and the eventual migration event was accepted as a reasonable tradeoff when the isolation feature was developed. But even if #1 would be addressed, #2 would still trigger it. Of course the warning in free_moved_vector() was overlooked at that time and the above two cases which have been discussed during patch review have obviously never been tested before the final submission. So keep it simple and remove the warning. [ tglx: Rewrote changelog and added a comment to free_moved_vector() ] Fixes: 11ea68f553e2 ("genirq, sched/isolation: Isolate from handling managed interrupts") Signed-off-by: Peter Xu Signed-off-by: Thomas Gleixner Reviewed-by: Ming Lei Link: https://lkml.kernel.org/r/20200312205830.81796-1-peterx@redhat.com --- arch/x86/kernel/apic/vector.c | 14 ++++++++------ 1 file changed, 8 insertions(+), 6 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kernel/apic/vector.c b/arch/x86/kernel/apic/vector.c index 2c5676b0a6e7..48293d15f1e1 100644 --- a/arch/x86/kernel/apic/vector.c +++ b/arch/x86/kernel/apic/vector.c @@ -838,13 +838,15 @@ static void free_moved_vector(struct apic_chip_data *apicd) bool managed = apicd->is_managed; /* - * This should never happen. Managed interrupts are not - * migrated except on CPU down, which does not involve the - * cleanup vector. But try to keep the accounting correct - * nevertheless. + * Managed interrupts are usually not migrated away + * from an online CPU, but CPU isolation 'managed_irq' + * can make that happen. + * 1) Activation does not take the isolation into account + * to keep the code simple + * 2) Migration away from an isolated CPU can happen when + * a non-isolated CPU which is in the calculated + * affinity mask comes online. */ - WARN_ON_ONCE(managed); - trace_vector_free_moved(apicd->irq, cpu, vector, managed); irq_matrix_free(vector_matrix, cpu, vector, managed); per_cpu(vector_irq, cpu)[vector] = VECTOR_UNUSED; -- cgit v1.2.3 From 7a57c09bb1cb89239f38f690b87cdf2c7db76c34 Mon Sep 17 00:00:00 2001 From: Sean Christopherson Date: Thu, 12 Mar 2020 11:04:16 -0700 Subject: KVM: VMX: Condition ENCLS-exiting enabling on CPU support for SGX1 Enable ENCLS-exiting (and thus set vmcs.ENCLS_EXITING_BITMAP) only if the CPU supports SGX1. Per Intel's SDM, all ENCLS leafs #UD if SGX1 is not supported[*], i.e. intercepting ENCLS to inject a #UD is unnecessary. Avoiding ENCLS-exiting even when it is reported as supported by the CPU works around a reported issue where SGX is "hard" disabled after an S3 suspend/resume cycle, i.e. CPUID.0x7.SGX=0 and the VMCS field/control are enumerated as unsupported. While the root cause of the S3 issue is unknown, it's definitely _not_ a KVM (or kernel) bug, i.e. this is a workaround for what is most likely a hardware or firmware issue. As a bonus side effect, KVM saves a VMWRITE when first preparing vmcs01 and vmcs02. Note, SGX must be disabled in BIOS to take advantage of this workaround [*] The additional ENCLS CPUID check on SGX1 exists so that SGX can be globally "soft" disabled post-reset, e.g. if #MC bits in MCi_CTL are cleared. Soft disabled meaning disabling SGX without clearing the primary CPUID bit (in leaf 0x7) and without poking into non-SGX CPU paths, e.g. for the VMCS controls. Fixes: 0b665d304028 ("KVM: vmx: Inject #UD for SGX ENCLS instruction in guest") Reported-by: Toni Spets Signed-off-by: Sean Christopherson Signed-off-by: Paolo Bonzini --- arch/x86/kvm/vmx/vmx.c | 16 ++++++++++++++-- 1 file changed, 14 insertions(+), 2 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kvm/vmx/vmx.c b/arch/x86/kvm/vmx/vmx.c index 40b1e6138cd5..26f8f31563e9 100644 --- a/arch/x86/kvm/vmx/vmx.c +++ b/arch/x86/kvm/vmx/vmx.c @@ -2338,6 +2338,17 @@ static void hardware_disable(void) kvm_cpu_vmxoff(); } +/* + * There is no X86_FEATURE for SGX yet, but anyway we need to query CPUID + * directly instead of going through cpu_has(), to ensure KVM is trapping + * ENCLS whenever it's supported in hardware. It does not matter whether + * the host OS supports or has enabled SGX. + */ +static bool cpu_has_sgx(void) +{ + return cpuid_eax(0) >= 0x12 && (cpuid_eax(0x12) & BIT(0)); +} + static __init int adjust_vmx_controls(u32 ctl_min, u32 ctl_opt, u32 msr, u32 *result) { @@ -2418,8 +2429,9 @@ static __init int setup_vmcs_config(struct vmcs_config *vmcs_conf, SECONDARY_EXEC_ENABLE_USR_WAIT_PAUSE | SECONDARY_EXEC_PT_USE_GPA | SECONDARY_EXEC_PT_CONCEAL_VMX | - SECONDARY_EXEC_ENABLE_VMFUNC | - SECONDARY_EXEC_ENCLS_EXITING; + SECONDARY_EXEC_ENABLE_VMFUNC; + if (cpu_has_sgx()) + opt2 |= SECONDARY_EXEC_ENCLS_EXITING; if (adjust_vmx_controls(min2, opt2, MSR_IA32_VMX_PROCBASED_CTLS2, &_cpu_based_2nd_exec_control) < 0) -- cgit v1.2.3 From 0c22056f8c43265da21a3dfe7b7e1379e5ff7c72 Mon Sep 17 00:00:00 2001 From: Nitesh Narayan Lal Date: Fri, 13 Mar 2020 09:16:24 -0400 Subject: KVM: x86: Initializing all kvm_lapic_irq fields in ioapic_write_indirect Previously all fields of structure kvm_lapic_irq were not initialized before it was passed to kvm_bitmap_or_dest_vcpus(). Which will cause an issue when any of those fields are used for processing a request. For example not initializing the msi_redir_hint field before passing to the kvm_bitmap_or_dest_vcpus(), may lead to a misbehavior of kvm_apic_map_get_dest_lapic(). This will specifically happen when the kvm_lowest_prio_delivery() returns TRUE due to a non-zero garbage value of msi_redir_hint, which should not happen as the request belongs to APIC fixed delivery mode and we do not want to deliver the interrupt only to the lowest priority candidate. This patch initializes all the fields of kvm_lapic_irq based on the values of ioapic redirect_entry object before passing it on to kvm_bitmap_or_dest_vcpus(). Fixes: 7ee30bc132c6 ("KVM: x86: deliver KVM IOAPIC scan request to target vCPUs") Signed-off-by: Nitesh Narayan Lal Reviewed-by: Vitaly Kuznetsov [Set level to false since the value doesn't really matter. Suggested by Vitaly Kuznetsov. - Paolo] Signed-off-by: Paolo Bonzini --- arch/x86/kvm/ioapic.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kvm/ioapic.c b/arch/x86/kvm/ioapic.c index 7668fed1ce65..750ff0b29404 100644 --- a/arch/x86/kvm/ioapic.c +++ b/arch/x86/kvm/ioapic.c @@ -378,12 +378,15 @@ static void ioapic_write_indirect(struct kvm_ioapic *ioapic, u32 val) if (e->fields.delivery_mode == APIC_DM_FIXED) { struct kvm_lapic_irq irq; - irq.shorthand = APIC_DEST_NOSHORT; irq.vector = e->fields.vector; irq.delivery_mode = e->fields.delivery_mode << 8; - irq.dest_id = e->fields.dest_id; irq.dest_mode = kvm_lapic_irq_dest_mode(!!e->fields.dest_mode); + irq.level = false; + irq.trig_mode = e->fields.trig_mode; + irq.shorthand = APIC_DEST_NOSHORT; + irq.dest_id = e->fields.dest_id; + irq.msi_redir_hint = false; bitmap_zero(&vcpu_bitmap, 16); kvm_bitmap_or_dest_vcpus(ioapic->kvm, &irq, &vcpu_bitmap); -- cgit v1.2.3 From 95fa10103dabc38be5de8efdfced5e67576ed896 Mon Sep 17 00:00:00 2001 From: Vitaly Kuznetsov Date: Mon, 9 Mar 2020 16:52:11 +0100 Subject: KVM: nVMX: avoid NULL pointer dereference with incorrect EVMCS GPAs When an EVMCS enabled L1 guest on KVM will tries doing enlightened VMEnter with EVMCS GPA = 0 the host crashes because the evmcs_gpa != vmx->nested.hv_evmcs_vmptr condition in nested_vmx_handle_enlightened_vmptrld() will evaluate to false (as nested.hv_evmcs_vmptr is zeroed after init). The crash will happen on vmx->nested.hv_evmcs pointer dereference. Another problematic EVMCS ptr value is '-1' but it only causes host crash after nested_release_evmcs() invocation. The problem is exactly the same as with '0', we mistakenly think that the EVMCS pointer hasn't changed and thus nested.hv_evmcs_vmptr is valid. Resolve the issue by adding an additional !vmx->nested.hv_evmcs check to nested_vmx_handle_enlightened_vmptrld(), this way we will always be trying kvm_vcpu_map() when nested.hv_evmcs is NULL and this is supposed to catch all invalid EVMCS GPAs. Also, initialize hv_evmcs_vmptr to '0' in nested_release_evmcs() to be consistent with initialization where we don't currently set hv_evmcs_vmptr to '-1'. Cc: stable@vger.kernel.org Signed-off-by: Vitaly Kuznetsov Signed-off-by: Paolo Bonzini --- arch/x86/kvm/vmx/nested.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/kvm/vmx/nested.c b/arch/x86/kvm/vmx/nested.c index e920d7834d73..9750e590c89d 100644 --- a/arch/x86/kvm/vmx/nested.c +++ b/arch/x86/kvm/vmx/nested.c @@ -224,7 +224,7 @@ static inline void nested_release_evmcs(struct kvm_vcpu *vcpu) return; kvm_vcpu_unmap(vcpu, &vmx->nested.hv_evmcs_map, true); - vmx->nested.hv_evmcs_vmptr = -1ull; + vmx->nested.hv_evmcs_vmptr = 0; vmx->nested.hv_evmcs = NULL; } @@ -1923,7 +1923,8 @@ static int nested_vmx_handle_enlightened_vmptrld(struct kvm_vcpu *vcpu, if (!nested_enlightened_vmentry(vcpu, &evmcs_gpa)) return 1; - if (unlikely(evmcs_gpa != vmx->nested.hv_evmcs_vmptr)) { + if (unlikely(!vmx->nested.hv_evmcs || + evmcs_gpa != vmx->nested.hv_evmcs_vmptr)) { if (!vmx->nested.hv_evmcs) vmx->nested.current_vmptr = -1ull; -- cgit v1.2.3 From 870b4333a62e45b0b2000d14b301b7b8b8cad9da Mon Sep 17 00:00:00 2001 From: Borislav Petkov Date: Wed, 18 Mar 2020 19:27:48 +0100 Subject: x86/ioremap: Fix CONFIG_EFI=n build In order to use efi_mem_type(), one needs CONFIG_EFI enabled. Otherwise that function is undefined. Use IS_ENABLED() to check and avoid the ifdeffery as the compiler optimizes away the following unreachable code then. Fixes: 985e537a4082 ("x86/ioremap: Map EFI runtime services data as encrypted for SEV") Reported-by: Randy Dunlap Signed-off-by: Borislav Petkov Tested-by: Randy Dunlap Cc: Tom Lendacky Cc: Link: https://lkml.kernel.org/r/7561e981-0d9b-d62c-0ef2-ce6007aff1ab@infradead.org --- arch/x86/mm/ioremap.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'arch/x86') diff --git a/arch/x86/mm/ioremap.c b/arch/x86/mm/ioremap.c index 935a91e1fd77..18c637c0dc6f 100644 --- a/arch/x86/mm/ioremap.c +++ b/arch/x86/mm/ioremap.c @@ -115,6 +115,9 @@ static void __ioremap_check_other(resource_size_t addr, struct ioremap_desc *des if (!sev_active()) return; + if (!IS_ENABLED(CONFIG_EFI)) + return; + if (efi_mem_type(addr) == EFI_RUNTIME_SERVICES_DATA) desc->flags |= IORES_MAP_ENCRYPTED; } -- cgit v1.2.3 From 763802b53a427ed3cbd419dbba255c414fdd9e7c Mon Sep 17 00:00:00 2001 From: Joerg Roedel Date: Sat, 21 Mar 2020 18:22:41 -0700 Subject: x86/mm: split vmalloc_sync_all() Commit 3f8fd02b1bf1 ("mm/vmalloc: Sync unmappings in __purge_vmap_area_lazy()") introduced a call to vmalloc_sync_all() in the vunmap() code-path. While this change was necessary to maintain correctness on x86-32-pae kernels, it also adds additional cycles for architectures that don't need it. Specifically on x86-64 with CONFIG_VMAP_STACK=y some people reported severe performance regressions in micro-benchmarks because it now also calls the x86-64 implementation of vmalloc_sync_all() on vunmap(). But the vmalloc_sync_all() implementation on x86-64 is only needed for newly created mappings. To avoid the unnecessary work on x86-64 and to gain the performance back, split up vmalloc_sync_all() into two functions: * vmalloc_sync_mappings(), and * vmalloc_sync_unmappings() Most call-sites to vmalloc_sync_all() only care about new mappings being synchronized. The only exception is the new call-site added in the above mentioned commit. Shile Zhang directed us to a report of an 80% regression in reaim throughput. Fixes: 3f8fd02b1bf1 ("mm/vmalloc: Sync unmappings in __purge_vmap_area_lazy()") Reported-by: kernel test robot Reported-by: Shile Zhang Signed-off-by: Joerg Roedel Signed-off-by: Andrew Morton Tested-by: Borislav Petkov Acked-by: Rafael J. Wysocki [GHES] Cc: Dave Hansen Cc: Andy Lutomirski Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: Ingo Molnar Cc: Link: http://lkml.kernel.org/r/20191009124418.8286-1-joro@8bytes.org Link: https://lists.01.org/hyperkitty/list/lkp@lists.01.org/thread/4D3JPPHBNOSPFK2KEPC6KGKS6J25AIDB/ Link: http://lkml.kernel.org/r/20191113095530.228959-1-shile.zhang@linux.alibaba.com Signed-off-by: Linus Torvalds --- arch/x86/mm/fault.c | 26 ++++++++++++++++++++++++-- 1 file changed, 24 insertions(+), 2 deletions(-) (limited to 'arch/x86') diff --git a/arch/x86/mm/fault.c b/arch/x86/mm/fault.c index fa4ea09593ab..629fdf13f846 100644 --- a/arch/x86/mm/fault.c +++ b/arch/x86/mm/fault.c @@ -190,7 +190,7 @@ static inline pmd_t *vmalloc_sync_one(pgd_t *pgd, unsigned long address) return pmd_k; } -void vmalloc_sync_all(void) +static void vmalloc_sync(void) { unsigned long address; @@ -217,6 +217,16 @@ void vmalloc_sync_all(void) } } +void vmalloc_sync_mappings(void) +{ + vmalloc_sync(); +} + +void vmalloc_sync_unmappings(void) +{ + vmalloc_sync(); +} + /* * 32-bit: * @@ -319,11 +329,23 @@ out: #else /* CONFIG_X86_64: */ -void vmalloc_sync_all(void) +void vmalloc_sync_mappings(void) { + /* + * 64-bit mappings might allocate new p4d/pud pages + * that need to be propagated to all tasks' PGDs. + */ sync_global_pgds(VMALLOC_START & PGDIR_MASK, VMALLOC_END); } +void vmalloc_sync_unmappings(void) +{ + /* + * Unmappings never allocate or free p4d/pud pages. + * No work is required here. + */ +} + /* * 64-bit: * -- cgit v1.2.3