From 2f5d27e6cf14efe652748bad89ee529ed5a5d577 Mon Sep 17 00:00:00 2001 From: Reiji Watanabe Date: Mon, 28 Mar 2022 20:19:24 -0700 Subject: KVM: arm64: selftests: Introduce vcpu_width_config Introduce a test for aarch64 that ensures non-mixed-width vCPUs (all 64bit vCPUs or all 32bit vcPUs) can be configured, and mixed-width vCPUs cannot be configured. Reviewed-by: Andrew Jones Signed-off-by: Reiji Watanabe Reviewed-by: Oliver Upton Signed-off-by: Marc Zyngier Link: https://lore.kernel.org/r/20220329031924.619453-3-reijiw@google.com --- tools/testing/selftests/kvm/.gitignore | 1 + tools/testing/selftests/kvm/Makefile | 1 + .../selftests/kvm/aarch64/vcpu_width_config.c | 122 +++++++++++++++++++++ 3 files changed, 124 insertions(+) create mode 100644 tools/testing/selftests/kvm/aarch64/vcpu_width_config.c (limited to 'tools') diff --git a/tools/testing/selftests/kvm/.gitignore b/tools/testing/selftests/kvm/.gitignore index d1e8f5237469..573d93a1d61f 100644 --- a/tools/testing/selftests/kvm/.gitignore +++ b/tools/testing/selftests/kvm/.gitignore @@ -3,6 +3,7 @@ /aarch64/debug-exceptions /aarch64/get-reg-list /aarch64/psci_cpu_on_test +/aarch64/vcpu_width_config /aarch64/vgic_init /aarch64/vgic_irq /s390x/memop diff --git a/tools/testing/selftests/kvm/Makefile b/tools/testing/selftests/kvm/Makefile index 21c2dbd21a81..681b173aa87c 100644 --- a/tools/testing/selftests/kvm/Makefile +++ b/tools/testing/selftests/kvm/Makefile @@ -106,6 +106,7 @@ TEST_GEN_PROGS_aarch64 += aarch64/arch_timer TEST_GEN_PROGS_aarch64 += aarch64/debug-exceptions TEST_GEN_PROGS_aarch64 += aarch64/get-reg-list TEST_GEN_PROGS_aarch64 += aarch64/psci_cpu_on_test +TEST_GEN_PROGS_aarch64 += aarch64/vcpu_width_config TEST_GEN_PROGS_aarch64 += aarch64/vgic_init TEST_GEN_PROGS_aarch64 += aarch64/vgic_irq TEST_GEN_PROGS_aarch64 += demand_paging_test diff --git a/tools/testing/selftests/kvm/aarch64/vcpu_width_config.c b/tools/testing/selftests/kvm/aarch64/vcpu_width_config.c new file mode 100644 index 000000000000..6e9402679229 --- /dev/null +++ b/tools/testing/selftests/kvm/aarch64/vcpu_width_config.c @@ -0,0 +1,122 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * vcpu_width_config - Test KVM_ARM_VCPU_INIT() with KVM_ARM_VCPU_EL1_32BIT. + * + * Copyright (c) 2022 Google LLC. + * + * This is a test that ensures that non-mixed-width vCPUs (all 64bit vCPUs + * or all 32bit vcPUs) can be configured and mixed-width vCPUs cannot be + * configured. + */ + +#include "kvm_util.h" +#include "processor.h" +#include "test_util.h" + + +/* + * Add a vCPU, run KVM_ARM_VCPU_INIT with @init1, and then + * add another vCPU, and run KVM_ARM_VCPU_INIT with @init2. + */ +static int add_init_2vcpus(struct kvm_vcpu_init *init1, + struct kvm_vcpu_init *init2) +{ + struct kvm_vm *vm; + int ret; + + vm = vm_create(VM_MODE_DEFAULT, DEFAULT_GUEST_PHY_PAGES, O_RDWR); + + vm_vcpu_add(vm, 0); + ret = _vcpu_ioctl(vm, 0, KVM_ARM_VCPU_INIT, init1); + if (ret) + goto free_exit; + + vm_vcpu_add(vm, 1); + ret = _vcpu_ioctl(vm, 1, KVM_ARM_VCPU_INIT, init2); + +free_exit: + kvm_vm_free(vm); + return ret; +} + +/* + * Add two vCPUs, then run KVM_ARM_VCPU_INIT for one vCPU with @init1, + * and run KVM_ARM_VCPU_INIT for another vCPU with @init2. + */ +static int add_2vcpus_init_2vcpus(struct kvm_vcpu_init *init1, + struct kvm_vcpu_init *init2) +{ + struct kvm_vm *vm; + int ret; + + vm = vm_create(VM_MODE_DEFAULT, DEFAULT_GUEST_PHY_PAGES, O_RDWR); + + vm_vcpu_add(vm, 0); + vm_vcpu_add(vm, 1); + + ret = _vcpu_ioctl(vm, 0, KVM_ARM_VCPU_INIT, init1); + if (ret) + goto free_exit; + + ret = _vcpu_ioctl(vm, 1, KVM_ARM_VCPU_INIT, init2); + +free_exit: + kvm_vm_free(vm); + return ret; +} + +/* + * Tests that two 64bit vCPUs can be configured, two 32bit vCPUs can be + * configured, and two mixed-width vCPUs cannot be configured. + * Each of those three cases, configure vCPUs in two different orders. + * The one is running KVM_CREATE_VCPU for 2 vCPUs, and then running + * KVM_ARM_VCPU_INIT for them. + * The other is running KVM_CREATE_VCPU and KVM_ARM_VCPU_INIT for a vCPU, + * and then run those commands for another vCPU. + */ +int main(void) +{ + struct kvm_vcpu_init init1, init2; + struct kvm_vm *vm; + int ret; + + if (!kvm_check_cap(KVM_CAP_ARM_EL1_32BIT)) { + print_skip("KVM_CAP_ARM_EL1_32BIT is not supported"); + exit(KSFT_SKIP); + } + + /* Get the preferred target type and copy that to init2 for later use */ + vm = vm_create(VM_MODE_DEFAULT, DEFAULT_GUEST_PHY_PAGES, O_RDWR); + vm_ioctl(vm, KVM_ARM_PREFERRED_TARGET, &init1); + kvm_vm_free(vm); + init2 = init1; + + /* Test with 64bit vCPUs */ + ret = add_init_2vcpus(&init1, &init1); + TEST_ASSERT(ret == 0, + "Configuring 64bit EL1 vCPUs failed unexpectedly"); + ret = add_2vcpus_init_2vcpus(&init1, &init1); + TEST_ASSERT(ret == 0, + "Configuring 64bit EL1 vCPUs failed unexpectedly"); + + /* Test with 32bit vCPUs */ + init1.features[0] = (1 << KVM_ARM_VCPU_EL1_32BIT); + ret = add_init_2vcpus(&init1, &init1); + TEST_ASSERT(ret == 0, + "Configuring 32bit EL1 vCPUs failed unexpectedly"); + ret = add_2vcpus_init_2vcpus(&init1, &init1); + TEST_ASSERT(ret == 0, + "Configuring 32bit EL1 vCPUs failed unexpectedly"); + + /* Test with mixed-width vCPUs */ + init1.features[0] = 0; + init2.features[0] = (1 << KVM_ARM_VCPU_EL1_32BIT); + ret = add_init_2vcpus(&init1, &init2); + TEST_ASSERT(ret != 0, + "Configuring mixed-width vCPUs worked unexpectedly"); + ret = add_2vcpus_init_2vcpus(&init1, &init2); + TEST_ASSERT(ret != 0, + "Configuring mixed-width vCPUs worked unexpectedly"); + + return 0; +} -- cgit v1.2.3 From 02de9331c4d0c6bddac9c5fa66d91f70adf8612b Mon Sep 17 00:00:00 2001 From: Andrew Jones Date: Wed, 16 Mar 2022 13:51:29 +0100 Subject: KVM: selftests: get-reg-list: Add KVM_REG_ARM_FW_REG(3) When testing a kernel with commit a5905d6af492 ("KVM: arm64: Allow SMCCC_ARCH_WORKAROUND_3 to be discovered and migrated") get-reg-list output vregs: Number blessed registers: 234 vregs: Number registers: 238 vregs: There are 1 new registers. Consider adding them to the blessed reg list with the following lines: KVM_REG_ARM_FW_REG(3), vregs: PASS ... That output inspired two changes: 1) add the new register to the blessed list and 2) explain why "Number registers" is actually four larger than "Number blessed registers" (on the system used for testing), even though only one register is being stated as new. The reason is that some registers are host dependent and they get filtered out when comparing with the blessed list. The system used for the test apparently had three filtered registers. Signed-off-by: Andrew Jones Acked-by: Marc Zyngier Signed-off-by: Marc Zyngier Link: https://lore.kernel.org/r/20220316125129.392128-1-drjones@redhat.com --- tools/testing/selftests/kvm/aarch64/get-reg-list.c | 14 ++++++++++---- 1 file changed, 10 insertions(+), 4 deletions(-) (limited to 'tools') diff --git a/tools/testing/selftests/kvm/aarch64/get-reg-list.c b/tools/testing/selftests/kvm/aarch64/get-reg-list.c index f12147c43464..0b571f3fe64c 100644 --- a/tools/testing/selftests/kvm/aarch64/get-reg-list.c +++ b/tools/testing/selftests/kvm/aarch64/get-reg-list.c @@ -503,8 +503,13 @@ static void run_test(struct vcpu_config *c) ++missing_regs; if (new_regs || missing_regs) { + n = 0; + for_each_reg_filtered(i) + ++n; + printf("%s: Number blessed registers: %5lld\n", config_name(c), blessed_n); - printf("%s: Number registers: %5lld\n", config_name(c), reg_list->n); + printf("%s: Number registers: %5lld (includes %lld filtered registers)\n", + config_name(c), reg_list->n, reg_list->n - n); } if (new_regs) { @@ -683,9 +688,10 @@ static __u64 base_regs[] = { KVM_REG_ARM64 | KVM_REG_SIZE_U64 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(spsr[4]), KVM_REG_ARM64 | KVM_REG_SIZE_U32 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(fp_regs.fpsr), KVM_REG_ARM64 | KVM_REG_SIZE_U32 | KVM_REG_ARM_CORE | KVM_REG_ARM_CORE_REG(fp_regs.fpcr), - KVM_REG_ARM_FW_REG(0), - KVM_REG_ARM_FW_REG(1), - KVM_REG_ARM_FW_REG(2), + KVM_REG_ARM_FW_REG(0), /* KVM_REG_ARM_PSCI_VERSION */ + KVM_REG_ARM_FW_REG(1), /* KVM_REG_ARM_SMCCC_ARCH_WORKAROUND_1 */ + KVM_REG_ARM_FW_REG(2), /* KVM_REG_ARM_SMCCC_ARCH_WORKAROUND_2 */ + KVM_REG_ARM_FW_REG(3), /* KVM_REG_ARM_SMCCC_ARCH_WORKAROUND_3 */ ARM64_SYS_REG(3, 3, 14, 3, 1), /* CNTV_CTL_EL0 */ ARM64_SYS_REG(3, 3, 14, 3, 2), /* CNTV_CVAL_EL0 */ ARM64_SYS_REG(3, 3, 14, 0, 2), -- cgit v1.2.3 From 386ba265a8197716076a88853244f4437b92b167 Mon Sep 17 00:00:00 2001 From: Oliver Upton Date: Wed, 6 Apr 2022 23:56:14 +0000 Subject: selftests: KVM: Don't leak GIC FD across dirty log test iterations dirty_log_perf_test instantiates a VGICv3 for the guest (if supported by hardware) to reduce the overhead of guest exits. However, the test does not actually close the GIC fd when cleaning up the VM between test iterations, meaning that the VM is never actually destroyed in the kernel. While this is generally a bad idea, the bug was detected from the kernel spewing about duplicate debugfs entries as subsequent VMs happen to reuse the same FD even though the debugfs directory is still present. Abstract away the notion of setup/cleanup of the GIC FD from the test by creating arch-specific helpers for test setup/cleanup. Close the GIC FD on VM cleanup and do nothing for the other architectures. Fixes: c340f7899af6 ("KVM: selftests: Add vgic initialization for dirty log perf test for ARM") Reviewed-by: Jing Zhang Signed-off-by: Oliver Upton Signed-off-by: Marc Zyngier Link: https://lore.kernel.org/r/20220406235615.1447180-3-oupton@google.com --- tools/testing/selftests/kvm/dirty_log_perf_test.c | 34 +++++++++++++++++++++-- 1 file changed, 31 insertions(+), 3 deletions(-) (limited to 'tools') diff --git a/tools/testing/selftests/kvm/dirty_log_perf_test.c b/tools/testing/selftests/kvm/dirty_log_perf_test.c index c9d9e513ca04..7b47ae4f952e 100644 --- a/tools/testing/selftests/kvm/dirty_log_perf_test.c +++ b/tools/testing/selftests/kvm/dirty_log_perf_test.c @@ -18,11 +18,40 @@ #include "test_util.h" #include "perf_test_util.h" #include "guest_modes.h" + #ifdef __aarch64__ #include "aarch64/vgic.h" #define GICD_BASE_GPA 0x8000000ULL #define GICR_BASE_GPA 0x80A0000ULL + +static int gic_fd; + +static void arch_setup_vm(struct kvm_vm *vm, unsigned int nr_vcpus) +{ + /* + * The test can still run even if hardware does not support GICv3, as it + * is only an optimization to reduce guest exits. + */ + gic_fd = vgic_v3_setup(vm, nr_vcpus, 64, GICD_BASE_GPA, GICR_BASE_GPA); +} + +static void arch_cleanup_vm(struct kvm_vm *vm) +{ + if (gic_fd > 0) + close(gic_fd); +} + +#else /* __aarch64__ */ + +static void arch_setup_vm(struct kvm_vm *vm, unsigned int nr_vcpus) +{ +} + +static void arch_cleanup_vm(struct kvm_vm *vm) +{ +} + #endif /* How many host loops to run by default (one KVM_GET_DIRTY_LOG for each loop)*/ @@ -206,9 +235,7 @@ static void run_test(enum vm_guest_mode mode, void *arg) vm_enable_cap(vm, &cap); } -#ifdef __aarch64__ - vgic_v3_setup(vm, nr_vcpus, 64, GICD_BASE_GPA, GICR_BASE_GPA); -#endif + arch_setup_vm(vm, nr_vcpus); /* Start the iterations */ iteration = 0; @@ -302,6 +329,7 @@ static void run_test(enum vm_guest_mode mode, void *arg) } free_bitmaps(bitmaps, p->slots); + arch_cleanup_vm(vm); perf_test_destroy_vm(vm); } -- cgit v1.2.3 From 21db83846683d3987666505a3ec38f367708199a Mon Sep 17 00:00:00 2001 From: Oliver Upton Date: Wed, 6 Apr 2022 23:56:15 +0000 Subject: selftests: KVM: Free the GIC FD when cleaning up in arch_timer In order to correctly destroy a VM, all references to the VM must be freed. The arch_timer selftest creates a VGIC for the guest, which itself holds a reference to the VM. Close the GIC FD when cleaning up a VM. Signed-off-by: Oliver Upton Signed-off-by: Marc Zyngier Link: https://lore.kernel.org/r/20220406235615.1447180-4-oupton@google.com --- tools/testing/selftests/kvm/aarch64/arch_timer.c | 15 +++++++++++---- 1 file changed, 11 insertions(+), 4 deletions(-) (limited to 'tools') diff --git a/tools/testing/selftests/kvm/aarch64/arch_timer.c b/tools/testing/selftests/kvm/aarch64/arch_timer.c index b08d30bf71c5..3b940a101bc0 100644 --- a/tools/testing/selftests/kvm/aarch64/arch_timer.c +++ b/tools/testing/selftests/kvm/aarch64/arch_timer.c @@ -362,11 +362,12 @@ static void test_init_timer_irq(struct kvm_vm *vm) pr_debug("ptimer_irq: %d; vtimer_irq: %d\n", ptimer_irq, vtimer_irq); } +static int gic_fd; + static struct kvm_vm *test_vm_create(void) { struct kvm_vm *vm; unsigned int i; - int ret; int nr_vcpus = test_args.nr_vcpus; vm = vm_create_default_with_vcpus(nr_vcpus, 0, 0, guest_code, NULL); @@ -383,8 +384,8 @@ static struct kvm_vm *test_vm_create(void) ucall_init(vm, NULL); test_init_timer_irq(vm); - ret = vgic_v3_setup(vm, nr_vcpus, 64, GICD_BASE_GPA, GICR_BASE_GPA); - if (ret < 0) { + gic_fd = vgic_v3_setup(vm, nr_vcpus, 64, GICD_BASE_GPA, GICR_BASE_GPA); + if (gic_fd < 0) { print_skip("Failed to create vgic-v3"); exit(KSFT_SKIP); } @@ -395,6 +396,12 @@ static struct kvm_vm *test_vm_create(void) return vm; } +static void test_vm_cleanup(struct kvm_vm *vm) +{ + close(gic_fd); + kvm_vm_free(vm); +} + static void test_print_help(char *name) { pr_info("Usage: %s [-h] [-n nr_vcpus] [-i iterations] [-p timer_period_ms]\n", @@ -478,7 +485,7 @@ int main(int argc, char *argv[]) vm = test_vm_create(); test_run(vm); - kvm_vm_free(vm); + test_vm_cleanup(vm); return 0; } -- cgit v1.2.3 From fac3725364397f9a40a101f089b86ea655a58d06 Mon Sep 17 00:00:00 2001 From: Anup Patel Date: Sat, 9 Apr 2022 09:15:44 +0530 Subject: KVM: selftests: riscv: Set PTE A and D bits in VS-stage page table Supporting hardware updates of PTE A and D bits is optional for any RISC-V implementation so current software strategy is to always set these bits in both G-stage (hypervisor) and VS-stage (guest kernel). If PTE A and D bits are not set by software (hypervisor or guest) then RISC-V implementations not supporting hardware updates of these bits will cause traps even for perfectly valid PTEs. Based on above explanation, the VS-stage page table created by various KVM selftest applications is not correct because PTE A and D bits are not set. This patch fixes VS-stage page table programming of PTE A and D bits for KVM selftests. Fixes: 3e06cdf10520 ("KVM: selftests: Add initial support for RISC-V 64-bit") Signed-off-by: Anup Patel Tested-by: Mayuresh Chitale Signed-off-by: Anup Patel --- tools/testing/selftests/kvm/include/riscv/processor.h | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'tools') diff --git a/tools/testing/selftests/kvm/include/riscv/processor.h b/tools/testing/selftests/kvm/include/riscv/processor.h index dc284c6bdbc3..eca5c622efd2 100644 --- a/tools/testing/selftests/kvm/include/riscv/processor.h +++ b/tools/testing/selftests/kvm/include/riscv/processor.h @@ -101,7 +101,9 @@ static inline void set_reg(struct kvm_vm *vm, uint32_t vcpuid, uint64_t id, #define PGTBL_PTE_WRITE_SHIFT 2 #define PGTBL_PTE_READ_MASK 0x0000000000000002ULL #define PGTBL_PTE_READ_SHIFT 1 -#define PGTBL_PTE_PERM_MASK (PGTBL_PTE_EXECUTE_MASK | \ +#define PGTBL_PTE_PERM_MASK (PGTBL_PTE_ACCESSED_MASK | \ + PGTBL_PTE_DIRTY_MASK | \ + PGTBL_PTE_EXECUTE_MASK | \ PGTBL_PTE_WRITE_MASK | \ PGTBL_PTE_READ_MASK) #define PGTBL_PTE_VALID_MASK 0x0000000000000001ULL -- cgit v1.2.3 From ebdef0de2dbc40e697adaa6b3408130f7a7b8351 Mon Sep 17 00:00:00 2001 From: Anup Patel Date: Sat, 9 Apr 2022 09:15:51 +0530 Subject: KVM: selftests: riscv: Fix alignment of the guest_hang() function The guest_hang() function is used as the default exception handler for various KVM selftests applications by setting it's address in the vstvec CSR. The vstvec CSR requires exception handler base address to be at least 4-byte aligned so this patch fixes alignment of the guest_hang() function. Fixes: 3e06cdf10520 ("KVM: selftests: Add initial support for RISC-V 64-bit") Signed-off-by: Anup Patel Tested-by: Mayuresh Chitale Signed-off-by: Anup Patel --- tools/testing/selftests/kvm/lib/riscv/processor.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'tools') diff --git a/tools/testing/selftests/kvm/lib/riscv/processor.c b/tools/testing/selftests/kvm/lib/riscv/processor.c index d377f2603d98..3961487a4870 100644 --- a/tools/testing/selftests/kvm/lib/riscv/processor.c +++ b/tools/testing/selftests/kvm/lib/riscv/processor.c @@ -268,7 +268,7 @@ void vcpu_dump(FILE *stream, struct kvm_vm *vm, uint32_t vcpuid, uint8_t indent) core.regs.t3, core.regs.t4, core.regs.t5, core.regs.t6); } -static void guest_hang(void) +static void __aligned(16) guest_hang(void) { while (1) ; -- cgit v1.2.3 From 400331f8ffa3bec5c561417e5eec6848464e9160 Mon Sep 17 00:00:00 2001 From: Pawan Gupta Date: Thu, 10 Mar 2022 14:02:09 -0800 Subject: x86/tsx: Disable TSX development mode at boot A microcode update on some Intel processors causes all TSX transactions to always abort by default[*]. Microcode also added functionality to re-enable TSX for development purposes. With this microcode loaded, if tsx=on was passed on the cmdline, and TSX development mode was already enabled before the kernel boot, it may make the system vulnerable to TSX Asynchronous Abort (TAA). To be on safer side, unconditionally disable TSX development mode during boot. If a viable use case appears, this can be revisited later. [*]: Intel TSX Disable Update for Selected Processors, doc ID: 643557 [ bp: Drop unstable web link, massage heavily. ] Suggested-by: Andrew Cooper Suggested-by: Borislav Petkov Signed-off-by: Pawan Gupta Signed-off-by: Borislav Petkov Tested-by: Neelima Krishnan Cc: Link: https://lore.kernel.org/r/347bd844da3a333a9793c6687d4e4eb3b2419a3e.1646943780.git.pawan.kumar.gupta@linux.intel.com --- arch/x86/include/asm/msr-index.h | 4 +-- arch/x86/kernel/cpu/common.c | 2 ++ arch/x86/kernel/cpu/cpu.h | 5 ++-- arch/x86/kernel/cpu/intel.c | 8 ------ arch/x86/kernel/cpu/tsx.c | 50 ++++++++++++++++++++++++++++++++-- tools/arch/x86/include/asm/msr-index.h | 4 +-- 6 files changed, 55 insertions(+), 18 deletions(-) (limited to 'tools') diff --git a/arch/x86/include/asm/msr-index.h b/arch/x86/include/asm/msr-index.h index 0eb90d21049e..ee15311b6be1 100644 --- a/arch/x86/include/asm/msr-index.h +++ b/arch/x86/include/asm/msr-index.h @@ -128,9 +128,9 @@ #define TSX_CTRL_RTM_DISABLE BIT(0) /* Disable RTM feature */ #define TSX_CTRL_CPUID_CLEAR BIT(1) /* Disable TSX enumeration */ -/* SRBDS support */ #define MSR_IA32_MCU_OPT_CTRL 0x00000123 -#define RNGDS_MITG_DIS BIT(0) +#define RNGDS_MITG_DIS BIT(0) /* SRBDS support */ +#define RTM_ALLOW BIT(1) /* TSX development mode */ #define MSR_IA32_SYSENTER_CS 0x00000174 #define MSR_IA32_SYSENTER_ESP 0x00000175 diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c index ed4417500700..e342ae4db3c4 100644 --- a/arch/x86/kernel/cpu/common.c +++ b/arch/x86/kernel/cpu/common.c @@ -1855,6 +1855,8 @@ void identify_secondary_cpu(struct cpuinfo_x86 *c) validate_apic_and_package_id(c); x86_spec_ctrl_setup_ap(); update_srbds_msr(); + + tsx_ap_init(); } static __init int setup_noclflush(char *arg) diff --git a/arch/x86/kernel/cpu/cpu.h b/arch/x86/kernel/cpu/cpu.h index ee6f23f7587d..2a8e584fc991 100644 --- a/arch/x86/kernel/cpu/cpu.h +++ b/arch/x86/kernel/cpu/cpu.h @@ -55,11 +55,10 @@ enum tsx_ctrl_states { extern __ro_after_init enum tsx_ctrl_states tsx_ctrl_state; extern void __init tsx_init(void); -extern void tsx_enable(void); -extern void tsx_disable(void); -extern void tsx_clear_cpuid(void); +void tsx_ap_init(void); #else static inline void tsx_init(void) { } +static inline void tsx_ap_init(void) { } #endif /* CONFIG_CPU_SUP_INTEL */ extern void get_cpu_cap(struct cpuinfo_x86 *c); diff --git a/arch/x86/kernel/cpu/intel.c b/arch/x86/kernel/cpu/intel.c index 8abf995677a4..f7a5370a9b3b 100644 --- a/arch/x86/kernel/cpu/intel.c +++ b/arch/x86/kernel/cpu/intel.c @@ -717,14 +717,6 @@ static void init_intel(struct cpuinfo_x86 *c) init_intel_misc_features(c); - if (tsx_ctrl_state == TSX_CTRL_ENABLE) - tsx_enable(); - else if (tsx_ctrl_state == TSX_CTRL_DISABLE) - tsx_disable(); - else if (tsx_ctrl_state == TSX_CTRL_RTM_ALWAYS_ABORT) - /* See comment over that function for more details. */ - tsx_clear_cpuid(); - split_lock_init(); bus_lock_init(); diff --git a/arch/x86/kernel/cpu/tsx.c b/arch/x86/kernel/cpu/tsx.c index ec6ff8000920..ec7bbac3a9f2 100644 --- a/arch/x86/kernel/cpu/tsx.c +++ b/arch/x86/kernel/cpu/tsx.c @@ -19,7 +19,7 @@ enum tsx_ctrl_states tsx_ctrl_state __ro_after_init = TSX_CTRL_NOT_SUPPORTED; -void tsx_disable(void) +static void tsx_disable(void) { u64 tsx; @@ -39,7 +39,7 @@ void tsx_disable(void) wrmsrl(MSR_IA32_TSX_CTRL, tsx); } -void tsx_enable(void) +static void tsx_enable(void) { u64 tsx; @@ -122,7 +122,7 @@ static enum tsx_ctrl_states x86_get_tsx_auto_mode(void) * That's why, this function's call in init_intel() doesn't clear the * feature flags. */ -void tsx_clear_cpuid(void) +static void tsx_clear_cpuid(void) { u64 msr; @@ -142,11 +142,42 @@ void tsx_clear_cpuid(void) } } +/* + * Disable TSX development mode + * + * When the microcode released in Feb 2022 is applied, TSX will be disabled by + * default on some processors. MSR 0x122 (TSX_CTRL) and MSR 0x123 + * (IA32_MCU_OPT_CTRL) can be used to re-enable TSX for development, doing so is + * not recommended for production deployments. In particular, applying MD_CLEAR + * flows for mitigation of the Intel TSX Asynchronous Abort (TAA) transient + * execution attack may not be effective on these processors when Intel TSX is + * enabled with updated microcode. + */ +static void tsx_dev_mode_disable(void) +{ + u64 mcu_opt_ctrl; + + /* Check if RTM_ALLOW exists */ + if (!boot_cpu_has_bug(X86_BUG_TAA) || !tsx_ctrl_is_supported() || + !cpu_feature_enabled(X86_FEATURE_SRBDS_CTRL)) + return; + + rdmsrl(MSR_IA32_MCU_OPT_CTRL, mcu_opt_ctrl); + + if (mcu_opt_ctrl & RTM_ALLOW) { + mcu_opt_ctrl &= ~RTM_ALLOW; + wrmsrl(MSR_IA32_MCU_OPT_CTRL, mcu_opt_ctrl); + setup_force_cpu_cap(X86_FEATURE_RTM_ALWAYS_ABORT); + } +} + void __init tsx_init(void) { char arg[5] = {}; int ret; + tsx_dev_mode_disable(); + /* * Hardware will always abort a TSX transaction when the CPUID bit * RTM_ALWAYS_ABORT is set. In this case, it is better not to enumerate @@ -215,3 +246,16 @@ void __init tsx_init(void) setup_force_cpu_cap(X86_FEATURE_HLE); } } + +void tsx_ap_init(void) +{ + tsx_dev_mode_disable(); + + if (tsx_ctrl_state == TSX_CTRL_ENABLE) + tsx_enable(); + else if (tsx_ctrl_state == TSX_CTRL_DISABLE) + tsx_disable(); + else if (tsx_ctrl_state == TSX_CTRL_RTM_ALWAYS_ABORT) + /* See comment over that function for more details. */ + tsx_clear_cpuid(); +} diff --git a/tools/arch/x86/include/asm/msr-index.h b/tools/arch/x86/include/asm/msr-index.h index 0eb90d21049e..ee15311b6be1 100644 --- a/tools/arch/x86/include/asm/msr-index.h +++ b/tools/arch/x86/include/asm/msr-index.h @@ -128,9 +128,9 @@ #define TSX_CTRL_RTM_DISABLE BIT(0) /* Disable RTM feature */ #define TSX_CTRL_CPUID_CLEAR BIT(1) /* Disable TSX enumeration */ -/* SRBDS support */ #define MSR_IA32_MCU_OPT_CTRL 0x00000123 -#define RNGDS_MITG_DIS BIT(0) +#define RNGDS_MITG_DIS BIT(0) /* SRBDS support */ +#define RTM_ALLOW BIT(1) /* TSX development mode */ #define MSR_IA32_SYSENTER_CS 0x00000174 #define MSR_IA32_SYSENTER_ESP 0x00000175 -- cgit v1.2.3 From 0c8b6641c8410930e2a2f4a437ac3f987fbf9404 Mon Sep 17 00:00:00 2001 From: Like Xu Date: Wed, 6 Apr 2022 14:37:13 +0800 Subject: selftests: kvm: add tsc_scaling_sync to .gitignore The tsc_scaling_sync's binary should be present in the .gitignore file for the git to ignore it. Signed-off-by: Like Xu Message-Id: <20220406063715.55625-3-likexu@tencent.com> Signed-off-by: Paolo Bonzini --- tools/testing/selftests/kvm/.gitignore | 1 + 1 file changed, 1 insertion(+) (limited to 'tools') diff --git a/tools/testing/selftests/kvm/.gitignore b/tools/testing/selftests/kvm/.gitignore index 573d93a1d61f..0b0e4402bba6 100644 --- a/tools/testing/selftests/kvm/.gitignore +++ b/tools/testing/selftests/kvm/.gitignore @@ -34,6 +34,7 @@ /x86_64/state_test /x86_64/svm_vmcall_test /x86_64/svm_int_ctl_test +/x86_64/tsc_scaling_sync /x86_64/sync_regs_test /x86_64/tsc_msrs_test /x86_64/userspace_io_test -- cgit v1.2.3 From ce64763c63854b4079f2e036638aa881a1fb3fbc Mon Sep 17 00:00:00 2001 From: Athira Rajeev Date: Fri, 8 Apr 2022 12:54:31 +0530 Subject: testing/selftests/mqueue: Fix mq_perf_tests to free the allocated cpu set The selftest "mqueue/mq_perf_tests.c" use CPU_ALLOC to allocate CPU set. This cpu set is used further in pthread_attr_setaffinity_np and by pthread_create in the code. But in current code, allocated cpu set is not freed. Fix this issue by adding CPU_FREE in the "shutdown" function which is called in most of the error/exit path for the cleanup. There are few error paths which exit without using shutdown. Add a common goto error path with CPU_FREE for these cases. Fixes: 7820b0715b6f ("tools/selftests: add mq_perf_tests") Signed-off-by: Athira Rajeev Signed-off-by: Shuah Khan --- tools/testing/selftests/mqueue/mq_perf_tests.c | 25 +++++++++++++++++-------- 1 file changed, 17 insertions(+), 8 deletions(-) (limited to 'tools') diff --git a/tools/testing/selftests/mqueue/mq_perf_tests.c b/tools/testing/selftests/mqueue/mq_perf_tests.c index b019e0b8221c..84fda3b49073 100644 --- a/tools/testing/selftests/mqueue/mq_perf_tests.c +++ b/tools/testing/selftests/mqueue/mq_perf_tests.c @@ -180,6 +180,9 @@ void shutdown(int exit_val, char *err_cause, int line_no) if (in_shutdown++) return; + /* Free the cpu_set allocated using CPU_ALLOC in main function */ + CPU_FREE(cpu_set); + for (i = 0; i < num_cpus_to_pin; i++) if (cpu_threads[i]) { pthread_kill(cpu_threads[i], SIGUSR1); @@ -551,6 +554,12 @@ int main(int argc, char *argv[]) perror("sysconf(_SC_NPROCESSORS_ONLN)"); exit(1); } + + if (getuid() != 0) + ksft_exit_skip("Not running as root, but almost all tests " + "require root in order to modify\nsystem settings. " + "Exiting.\n"); + cpus_online = min(MAX_CPUS, sysconf(_SC_NPROCESSORS_ONLN)); cpu_set = CPU_ALLOC(cpus_online); if (cpu_set == NULL) { @@ -589,7 +598,7 @@ int main(int argc, char *argv[]) cpu_set)) { fprintf(stderr, "Any given CPU may " "only be given once.\n"); - exit(1); + goto err_code; } else CPU_SET_S(cpus_to_pin[cpu], cpu_set_size, cpu_set); @@ -607,7 +616,7 @@ int main(int argc, char *argv[]) queue_path = malloc(strlen(option) + 2); if (!queue_path) { perror("malloc()"); - exit(1); + goto err_code; } queue_path[0] = '/'; queue_path[1] = 0; @@ -622,17 +631,12 @@ int main(int argc, char *argv[]) fprintf(stderr, "Must pass at least one CPU to continuous " "mode.\n"); poptPrintUsage(popt_context, stderr, 0); - exit(1); + goto err_code; } else if (!continuous_mode) { num_cpus_to_pin = 1; cpus_to_pin[0] = cpus_online - 1; } - if (getuid() != 0) - ksft_exit_skip("Not running as root, but almost all tests " - "require root in order to modify\nsystem settings. " - "Exiting.\n"); - max_msgs = fopen(MAX_MSGS, "r+"); max_msgsize = fopen(MAX_MSGSIZE, "r+"); if (!max_msgs) @@ -740,4 +744,9 @@ int main(int argc, char *argv[]) sleep(1); } shutdown(0, "", 0); + +err_code: + CPU_FREE(cpu_set); + exit(1); + } -- cgit v1.2.3 From f034fc50d3c7d9385c20d505ab4cf56b8fd18ac7 Mon Sep 17 00:00:00 2001 From: Adrian Hunter Date: Mon, 11 Apr 2022 09:17:58 +0300 Subject: perf tools: Fix misleading add event PMU debug message Fix incorrect debug message: Attempting to add event pmu 'intel_pt' with '' that may result in non-fatal errors which always appears with perf record -vv and intel_pt e.g. perf record -vv -e intel_pt//u uname The message is incorrect because there will never be non-fatal errors. Suppress the message if the PMU is 'selectable' i.e. meant to be selected directly as an event. Fixes: 4ac22b484d4c79e8 ("perf parse-events: Make add PMU verbose output clearer") Signed-off-by: Adrian Hunter Cc: Ian Rogers Cc: Jiri Olsa Link: http://lore.kernel.org/lkml/20220411061758.2458417-1-adrian.hunter@intel.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/parse-events.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) (limited to 'tools') diff --git a/tools/perf/util/parse-events.c b/tools/perf/util/parse-events.c index 24997925ae00..dd84fed698a3 100644 --- a/tools/perf/util/parse-events.c +++ b/tools/perf/util/parse-events.c @@ -1523,7 +1523,9 @@ int parse_events_add_pmu(struct parse_events_state *parse_state, bool use_uncore_alias; LIST_HEAD(config_terms); - if (verbose > 1) { + pmu = parse_state->fake_pmu ?: perf_pmu__find(name); + + if (verbose > 1 && !(pmu && pmu->selectable)) { fprintf(stderr, "Attempting to add event pmu '%s' with '", name); if (head_config) { @@ -1536,7 +1538,6 @@ int parse_events_add_pmu(struct parse_events_state *parse_state, fprintf(stderr, "' that may result in non-fatal errors\n"); } - pmu = parse_state->fake_pmu ?: perf_pmu__find(name); if (!pmu) { char *err_str; -- cgit v1.2.3 From b2dd71f9f728da695a86b8308feb4f39defe9019 Mon Sep 17 00:00:00 2001 From: "Herton R. Krzesinski" Date: Mon, 4 Apr 2022 18:05:25 -0300 Subject: tools/power/x86/intel-speed-select: fix build failure when using -Wl,--as-needed Build of intel-speed-select will fail if you run: $ LDFLAGS="-Wl,--as-needed" /usr/bin/make V=1 ... gcc -O2 -Wall -g -D_GNU_SOURCE -Iinclude -I/usr/include/libnl3 -Wl,--as-needed -lnl-genl-3 -lnl-3 intel-speed-select-in.o -o intel-speed-select /usr/bin/ld: intel-speed-select-in.o: in function `handle_event': (...)/linux/tools/power/x86/intel-speed-select/hfi-events.c:189: undefined reference to `nlmsg_hdr' ... In this case the problem is that order when linking matters when using the flag -Wl,--as-needed, symbols not used at that point are discarded. So since intel-speed-select-in.o comes after, at that point the libraries/symbols are already discarded and then missing/undefined references are reported. To fix this, make sure we specify LDFLAGS after the object file. Acked-by: Srinivas Pandruvada Signed-off-by: Herton R. Krzesinski Link: https://lore.kernel.org/r/20220404210525.725611-1-herton@redhat.com Signed-off-by: Hans de Goede --- tools/power/x86/intel-speed-select/Makefile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'tools') diff --git a/tools/power/x86/intel-speed-select/Makefile b/tools/power/x86/intel-speed-select/Makefile index 846f785e278d..7221f2f55e8b 100644 --- a/tools/power/x86/intel-speed-select/Makefile +++ b/tools/power/x86/intel-speed-select/Makefile @@ -42,7 +42,7 @@ ISST_IN := $(OUTPUT)intel-speed-select-in.o $(ISST_IN): prepare FORCE $(Q)$(MAKE) $(build)=intel-speed-select $(OUTPUT)intel-speed-select: $(ISST_IN) - $(QUIET_LINK)$(CC) $(CFLAGS) $(LDFLAGS) $< -o $@ + $(QUIET_LINK)$(CC) $(CFLAGS) $< $(LDFLAGS) -o $@ clean: rm -f $(ALL_PROGRAMS) -- cgit v1.2.3 From d73f5d14e0cdd1f39764379250f26163913d7155 Mon Sep 17 00:00:00 2001 From: Lv Ruyi Date: Wed, 13 Apr 2022 09:33:02 +0000 Subject: perf stat: Fix error check return value of hashmap__new(), must use IS_ERR() hashmap__new() returns ERR_PTR(-ENOMEM) when it fails, so we should use IS_ERR() to check it in error handling path. Reported-by: Zeal Robot Signed-off-by: Lv Ruyi Cc: Alexander Shishkin Cc: Ian Rogers Cc: Ingo Molnar Cc: James Clark Cc: Jiri Olsa Cc: Mark Rutland Cc: Namhyung Kim Cc: Peter Zijlstra Link: https://lore.kernel.org/r/20220413093302.2538128-1-lv.ruyi@zte.com.cn Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/stat.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'tools') diff --git a/tools/perf/util/stat.c b/tools/perf/util/stat.c index ee6f03481215..817a2de264b4 100644 --- a/tools/perf/util/stat.c +++ b/tools/perf/util/stat.c @@ -1,5 +1,6 @@ // SPDX-License-Identifier: GPL-2.0 #include +#include #include #include #include @@ -311,7 +312,7 @@ static int check_per_pkg(struct evsel *counter, struct perf_counts_values *vals, if (!mask) { mask = hashmap__new(pkg_id_hash, pkg_id_equal, NULL); - if (!mask) + if (IS_ERR(mask)) return -ENOMEM; counter->per_pkg_mask = mask; -- cgit v1.2.3 From a668cc07f990d2ed19424d5c1a529521a9d1cee1 Mon Sep 17 00:00:00 2001 From: Adrian Hunter Date: Wed, 13 Apr 2022 14:42:32 +0300 Subject: perf tools: Fix segfault accessing sample_id xyarray perf_evsel::sample_id is an xyarray which can cause a segfault when accessed beyond its size. e.g. # perf record -e intel_pt// -C 1 sleep 1 Segmentation fault (core dumped) # That is happening because a dummy event is opened to capture text poke events accross all CPUs, however the mmap logic is allocating according to the number of user_requested_cpus. In general, perf sometimes uses the evsel cpus to open events, and sometimes the evlist user_requested_cpus. However, it is not necessary to determine which case is which because the opened event file descriptors are also in an xyarray, the size of whch can be used to correctly allocate the size of the sample_id xyarray, because there is one ID per file descriptor. Note, in the affected code path, perf_evsel fd array is subsequently used to get the file descriptor for the mmap, so it makes sense for the xyarrays to be the same size there. Fixes: d1a177595b3a824c ("libperf: Adopt perf_evlist__mmap()/munmap() from tools/perf") Fixes: 246eba8e9041c477 ("perf tools: Add support for PERF_RECORD_TEXT_POKE") Signed-off-by: Adrian Hunter Acked-by: Ian Rogers Cc: Adrian Hunter Cc: Jiri Olsa Cc: stable@vger.kernel.org # 5.5+ Link: https://lore.kernel.org/r/20220413114232.26914-1-adrian.hunter@intel.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/lib/perf/evlist.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) (limited to 'tools') diff --git a/tools/lib/perf/evlist.c b/tools/lib/perf/evlist.c index 1b15ba13c477..a09315538a30 100644 --- a/tools/lib/perf/evlist.c +++ b/tools/lib/perf/evlist.c @@ -577,7 +577,6 @@ int perf_evlist__mmap_ops(struct perf_evlist *evlist, { struct perf_evsel *evsel; const struct perf_cpu_map *cpus = evlist->user_requested_cpus; - const struct perf_thread_map *threads = evlist->threads; if (!ops || !ops->get || !ops->mmap) return -EINVAL; @@ -589,7 +588,7 @@ int perf_evlist__mmap_ops(struct perf_evlist *evlist, perf_evlist__for_each_entry(evlist, evsel) { if ((evsel->attr.read_format & PERF_FORMAT_ID) && evsel->sample_id == NULL && - perf_evsel__alloc_id(evsel, perf_cpu_map__nr(cpus), threads->nr) < 0) + perf_evsel__alloc_id(evsel, evsel->fd->max_x, evsel->fd->max_y) < 0) return -ENOMEM; } -- cgit v1.2.3 From 23380e4d53305765789fd1a2cf6bddb07239cd3b Mon Sep 17 00:00:00 2001 From: Alexey Bayduraev Date: Wed, 13 Apr 2022 18:46:40 -0700 Subject: perf record: Fix per-thread option Per-thread mode doesn't have specific CPUs for events, add checks for this case. Minor fix to a pr_debug by Ian Rogers to avoid an out of bound array access. Fixes: 7954f71689f90cb2 ("perf record: Introduce thread affinity and mmap masks") Reported-by: Ian Rogers Signed-off-by: Alexey Bayduraev Signed-off-by: Ian Rogers Cc: Alexander Shishkin Cc: Alexey Bayduraev Cc: Andi Kleen Cc: Ingo Molnar Cc: Jiri Olsa Cc: Mark Rutland Cc: Namhyung Kim Cc: Peter Zijlstra Cc: Riccardo Mancini Cc: Stephane Eranian Link: https://lore.kernel.org/r/20220414014642.3308206-1-irogers@google.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-record.c | 22 +++++++++++++++++----- 1 file changed, 17 insertions(+), 5 deletions(-) (limited to 'tools') diff --git a/tools/perf/builtin-record.c b/tools/perf/builtin-record.c index ba74fab02e62..069825c48d40 100644 --- a/tools/perf/builtin-record.c +++ b/tools/perf/builtin-record.c @@ -989,8 +989,11 @@ static int record__thread_data_init_maps(struct record_thread *thread_data, stru struct mmap *overwrite_mmap = evlist->overwrite_mmap; struct perf_cpu_map *cpus = evlist->core.user_requested_cpus; - thread_data->nr_mmaps = bitmap_weight(thread_data->mask->maps.bits, - thread_data->mask->maps.nbits); + if (cpu_map__is_dummy(cpus)) + thread_data->nr_mmaps = nr_mmaps; + else + thread_data->nr_mmaps = bitmap_weight(thread_data->mask->maps.bits, + thread_data->mask->maps.nbits); if (mmap) { thread_data->maps = zalloc(thread_data->nr_mmaps * sizeof(struct mmap *)); if (!thread_data->maps) @@ -1007,16 +1010,17 @@ static int record__thread_data_init_maps(struct record_thread *thread_data, stru thread_data->nr_mmaps, thread_data->maps, thread_data->overwrite_maps); for (m = 0, tm = 0; m < nr_mmaps && tm < thread_data->nr_mmaps; m++) { - if (test_bit(cpus->map[m].cpu, thread_data->mask->maps.bits)) { + if (cpu_map__is_dummy(cpus) || + test_bit(cpus->map[m].cpu, thread_data->mask->maps.bits)) { if (thread_data->maps) { thread_data->maps[tm] = &mmap[m]; pr_debug2("thread_data[%p]: cpu%d: maps[%d] -> mmap[%d]\n", - thread_data, cpus->map[m].cpu, tm, m); + thread_data, perf_cpu_map__cpu(cpus, m).cpu, tm, m); } if (thread_data->overwrite_maps) { thread_data->overwrite_maps[tm] = &overwrite_mmap[m]; pr_debug2("thread_data[%p]: cpu%d: ow_maps[%d] -> ow_mmap[%d]\n", - thread_data, cpus->map[m].cpu, tm, m); + thread_data, perf_cpu_map__cpu(cpus, m).cpu, tm, m); } tm++; } @@ -3329,6 +3333,9 @@ static void record__mmap_cpu_mask_init(struct mmap_cpu_mask *mask, struct perf_c { int c; + if (cpu_map__is_dummy(cpus)) + return; + for (c = 0; c < cpus->nr; c++) set_bit(cpus->map[c].cpu, mask->bits); } @@ -3680,6 +3687,11 @@ static int record__init_thread_masks(struct record *rec) if (!record__threads_enabled(rec)) return record__init_thread_default_masks(rec, cpus); + if (cpu_map__is_dummy(cpus)) { + pr_err("--per-thread option is mutually exclusive to parallel streaming mode.\n"); + return -EINVAL; + } + switch (rec->opts.threads_spec) { case THREAD_SPEC__CPU: ret = record__init_thread_cpu_masks(rec, cpus); -- cgit v1.2.3 From 8cb7a188ac33b71ec24a9081d1521f46dcab53c4 Mon Sep 17 00:00:00 2001 From: Athira Rajeev Date: Tue, 12 Apr 2022 22:10:58 +0530 Subject: perf bench: Fix numa testcase to check if CPU used to bind task is online Perf numa bench test fails with error: Testcase: ./perf bench numa mem -p 2 -t 1 -P 1024 -C 0,8 -M 1,0 -s 20 -zZq --thp 1 --no-data_rand_walk Failure snippet: <<>> Running 'numa/mem' benchmark: # Running main, "perf bench numa numa-mem -p 2 -t 1 -P 1024 -C 0,8 -M 1,0 -s 20 -zZq --thp 1 --no-data_rand_walk" perf: bench/numa.c:333: bind_to_cpumask: Assertion `!(ret)' failed. <<>> The Testcases uses CPU's 0 and 8. In function "parse_setup_cpu_list", There is check to see if cpu number is greater than max cpu's possible in the system ie via "if (bind_cpu_0 >= g->p.nr_cpus || bind_cpu_1 >= g->p.nr_cpus) {". But it could happen that system has say 48 CPU's, but only number of online CPU's is 0-7. Other CPU's are offlined. Since "g->p.nr_cpus" is 48, so function will go ahead and set bit for CPU 8 also in cpumask ( td->bind_cpumask). bind_to_cpumask function is called to set affinity using sched_setaffinity and the cpumask. Since the CPU8 is not present, set affinity will fail here with EINVAL. Fix this issue by adding a check to make sure that, CPU's provided in the input argument values are online before proceeding further and skip the test. For this, include new helper function "is_cpu_online" in "tools/perf/util/header.c". Since "BIT(x)" definition will get included from header.h, remove that from bench/numa.c Reported-by: Disha Goel Signed-off-by: Athira Jajeev Tested-by: Disha Goel Cc: Ian Rogers Cc: Jiri Olsa Cc: Kajol Jain Cc: Madhavan Srinivasan Cc: Michael Ellerman Cc: Nageswara R Sastry Cc: Srikar Dronamraju Cc: linuxppc-dev@lists.ozlabs.org Link: https://lore.kernel.org/r/20220412164059.42654-2-atrajeev@linux.vnet.ibm.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/bench/numa.c | 8 ++++++-- tools/perf/util/header.c | 51 ++++++++++++++++++++++++++++++++++++++++++++++++ tools/perf/util/header.h | 1 + 3 files changed, 58 insertions(+), 2 deletions(-) (limited to 'tools') diff --git a/tools/perf/bench/numa.c b/tools/perf/bench/numa.c index f2640179ada9..c838c248aa2c 100644 --- a/tools/perf/bench/numa.c +++ b/tools/perf/bench/numa.c @@ -34,6 +34,7 @@ #include #include +#include "../util/header.h" #include #include @@ -585,6 +586,11 @@ static int parse_setup_cpu_list(void) return -1; } + if (is_cpu_online(bind_cpu_0) != 1 || is_cpu_online(bind_cpu_1) != 1) { + printf("\nTest not applicable, bind_cpu_0 or bind_cpu_1 is offline\n"); + return -1; + } + BUG_ON(bind_cpu_0 < 0 || bind_cpu_1 < 0); BUG_ON(bind_cpu_0 > bind_cpu_1); @@ -752,8 +758,6 @@ static int parse_nodes_opt(const struct option *opt __maybe_unused, return parse_node_list(arg); } -#define BIT(x) (1ul << x) - static inline uint32_t lfsr_32(uint32_t lfsr) { const uint32_t taps = BIT(1) | BIT(5) | BIT(6) | BIT(31); diff --git a/tools/perf/util/header.c b/tools/perf/util/header.c index d546ff724dbe..a27132e5a5ef 100644 --- a/tools/perf/util/header.c +++ b/tools/perf/util/header.c @@ -983,6 +983,57 @@ static int write_dir_format(struct feat_fd *ff, return do_write(ff, &data->dir.version, sizeof(data->dir.version)); } +/* + * Check whether a CPU is online + * + * Returns: + * 1 -> if CPU is online + * 0 -> if CPU is offline + * -1 -> error case + */ +int is_cpu_online(unsigned int cpu) +{ + char *str; + size_t strlen; + char buf[256]; + int status = -1; + struct stat statbuf; + + snprintf(buf, sizeof(buf), + "/sys/devices/system/cpu/cpu%d", cpu); + if (stat(buf, &statbuf) != 0) + return 0; + + /* + * Check if /sys/devices/system/cpu/cpux/online file + * exists. Some cases cpu0 won't have online file since + * it is not expected to be turned off generally. + * In kernels without CONFIG_HOTPLUG_CPU, this + * file won't exist + */ + snprintf(buf, sizeof(buf), + "/sys/devices/system/cpu/cpu%d/online", cpu); + if (stat(buf, &statbuf) != 0) + return 1; + + /* + * Read online file using sysfs__read_str. + * If read or open fails, return -1. + * If read succeeds, return value from file + * which gets stored in "str" + */ + snprintf(buf, sizeof(buf), + "devices/system/cpu/cpu%d/online", cpu); + + if (sysfs__read_str(buf, &str, &strlen) < 0) + return status; + + status = atoi(str); + + free(str); + return status; +} + #ifdef HAVE_LIBBPF_SUPPORT static int write_bpf_prog_info(struct feat_fd *ff, struct evlist *evlist __maybe_unused) diff --git a/tools/perf/util/header.h b/tools/perf/util/header.h index c9e3265832d9..0eb4bc29a5a4 100644 --- a/tools/perf/util/header.h +++ b/tools/perf/util/header.h @@ -158,6 +158,7 @@ int do_write(struct feat_fd *fd, const void *buf, size_t size); int write_padded(struct feat_fd *fd, const void *bf, size_t count, size_t count_aligned); +int is_cpu_online(unsigned int cpu); /* * arch specific callback */ -- cgit v1.2.3 From f58faed7fb3f784efdbf3b401368dcf51a6e28fa Mon Sep 17 00:00:00 2001 From: Athira Rajeev Date: Tue, 12 Apr 2022 22:10:59 +0530 Subject: perf bench: Fix numa bench to fix usage of affinity for machines with #CPUs > 1K The 'perf bench numa' testcase fails on systems with more than 1K CPUs. Testcase: perf bench numa mem -p 1 -t 3 -P 512 -s 100 -zZ0qcm --thp 1 Snippet of code: <<>> perf: bench/numa.c:302: bind_to_node: Assertion `!(ret)' failed. Aborted (core dumped) <<>> bind_to_node() uses "sched_getaffinity" to save the original cpumask and this call is returning EINVAL ((invalid argument). This happens because the default mask size in glibc is 1024. To overcome this 1024 CPUs mask size limitation of cpu_set_t, change the mask size using the CPU_*_S macros ie, use CPU_ALLOC to allocate cpumask, CPU_ALLOC_SIZE for size. Apart from fixing this for "orig_mask", apply same logic to "mask" as well which is used to setaffinity so that mask size is large enough to represent number of possible CPU's in the system. sched_getaffinity is used in one more place in perf numa bench. It is in "bind_to_cpu" function. Apply the same logic there also. Though currently no failure is reported from there, it is ideal to change getaffinity to work with such system configurations having CPU's more than default mask size supported by glibc. Also fix "sched_setaffinity" to use mask size which is large enough to represent number of possible CPU's in the system. Fixed all places where "bind_cpumask" which is part of "struct thread_data" is used such that bind_cpumask works in all configuration. Reported-by: Disha Goel Signed-off-by: Athira Jajeev Cc: Ian Rogers Cc: Jiri Olsa Cc: Kajol Jain Cc: Madhavan Srinivasan Cc: Michael Ellerman Cc: Nageswara R Sastry Cc: Srikar Dronamraju Cc: linuxppc-dev@lists.ozlabs.org Link: https://lore.kernel.org/r/20220412164059.42654-3-atrajeev@linux.vnet.ibm.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/bench/numa.c | 128 +++++++++++++++++++++++++++++++++++------------- 1 file changed, 95 insertions(+), 33 deletions(-) (limited to 'tools') diff --git a/tools/perf/bench/numa.c b/tools/perf/bench/numa.c index c838c248aa2c..44e1f8a44087 100644 --- a/tools/perf/bench/numa.c +++ b/tools/perf/bench/numa.c @@ -55,7 +55,7 @@ struct thread_data { int curr_cpu; - cpu_set_t bind_cpumask; + cpu_set_t *bind_cpumask; int bind_node; u8 *process_data; int process_nr; @@ -267,71 +267,115 @@ static bool node_has_cpus(int node) return ret; } -static cpu_set_t bind_to_cpu(int target_cpu) +static cpu_set_t *bind_to_cpu(int target_cpu) { - cpu_set_t orig_mask, mask; - int ret; + int nrcpus = numa_num_possible_cpus(); + cpu_set_t *orig_mask, *mask; + size_t size; - ret = sched_getaffinity(0, sizeof(orig_mask), &orig_mask); - BUG_ON(ret); + orig_mask = CPU_ALLOC(nrcpus); + BUG_ON(!orig_mask); + size = CPU_ALLOC_SIZE(nrcpus); + CPU_ZERO_S(size, orig_mask); + + if (sched_getaffinity(0, size, orig_mask)) + goto err_out; + + mask = CPU_ALLOC(nrcpus); + if (!mask) + goto err_out; - CPU_ZERO(&mask); + CPU_ZERO_S(size, mask); if (target_cpu == -1) { int cpu; for (cpu = 0; cpu < g->p.nr_cpus; cpu++) - CPU_SET(cpu, &mask); + CPU_SET_S(cpu, size, mask); } else { - BUG_ON(target_cpu < 0 || target_cpu >= g->p.nr_cpus); - CPU_SET(target_cpu, &mask); + if (target_cpu < 0 || target_cpu >= g->p.nr_cpus) + goto err; + + CPU_SET_S(target_cpu, size, mask); } - ret = sched_setaffinity(0, sizeof(mask), &mask); - BUG_ON(ret); + if (sched_setaffinity(0, size, mask)) + goto err; return orig_mask; + +err: + CPU_FREE(mask); +err_out: + CPU_FREE(orig_mask); + + /* BUG_ON due to failure in allocation of orig_mask/mask */ + BUG_ON(-1); } -static cpu_set_t bind_to_node(int target_node) +static cpu_set_t *bind_to_node(int target_node) { - cpu_set_t orig_mask, mask; + int nrcpus = numa_num_possible_cpus(); + size_t size; + cpu_set_t *orig_mask, *mask; int cpu; - int ret; - ret = sched_getaffinity(0, sizeof(orig_mask), &orig_mask); - BUG_ON(ret); + orig_mask = CPU_ALLOC(nrcpus); + BUG_ON(!orig_mask); + size = CPU_ALLOC_SIZE(nrcpus); + CPU_ZERO_S(size, orig_mask); + + if (sched_getaffinity(0, size, orig_mask)) + goto err_out; + + mask = CPU_ALLOC(nrcpus); + if (!mask) + goto err_out; - CPU_ZERO(&mask); + CPU_ZERO_S(size, mask); if (target_node == NUMA_NO_NODE) { for (cpu = 0; cpu < g->p.nr_cpus; cpu++) - CPU_SET(cpu, &mask); + CPU_SET_S(cpu, size, mask); } else { struct bitmask *cpumask = numa_allocate_cpumask(); - BUG_ON(!cpumask); + if (!cpumask) + goto err; + if (!numa_node_to_cpus(target_node, cpumask)) { for (cpu = 0; cpu < (int)cpumask->size; cpu++) { if (numa_bitmask_isbitset(cpumask, cpu)) - CPU_SET(cpu, &mask); + CPU_SET_S(cpu, size, mask); } } numa_free_cpumask(cpumask); } - ret = sched_setaffinity(0, sizeof(mask), &mask); - BUG_ON(ret); + if (sched_setaffinity(0, size, mask)) + goto err; return orig_mask; + +err: + CPU_FREE(mask); +err_out: + CPU_FREE(orig_mask); + + /* BUG_ON due to failure in allocation of orig_mask/mask */ + BUG_ON(-1); } -static void bind_to_cpumask(cpu_set_t mask) +static void bind_to_cpumask(cpu_set_t *mask) { int ret; + size_t size = CPU_ALLOC_SIZE(numa_num_possible_cpus()); - ret = sched_setaffinity(0, sizeof(mask), &mask); - BUG_ON(ret); + ret = sched_setaffinity(0, size, mask); + if (ret) { + CPU_FREE(mask); + BUG_ON(ret); + } } static void mempol_restore(void) @@ -377,7 +421,7 @@ do { \ static u8 *alloc_data(ssize_t bytes0, int map_flags, int init_zero, int init_cpu0, int thp, int init_random) { - cpu_set_t orig_mask; + cpu_set_t *orig_mask = NULL; ssize_t bytes; u8 *buf; int ret; @@ -435,6 +479,7 @@ static u8 *alloc_data(ssize_t bytes0, int map_flags, /* Restore affinity: */ if (init_cpu0) { bind_to_cpumask(orig_mask); + CPU_FREE(orig_mask); mempol_restore(); } @@ -595,6 +640,7 @@ static int parse_setup_cpu_list(void) BUG_ON(bind_cpu_0 > bind_cpu_1); for (bind_cpu = bind_cpu_0; bind_cpu <= bind_cpu_1; bind_cpu += step) { + size_t size = CPU_ALLOC_SIZE(g->p.nr_cpus); int i; for (i = 0; i < mul; i++) { @@ -614,10 +660,15 @@ static int parse_setup_cpu_list(void) tprintf("%2d", bind_cpu); } - CPU_ZERO(&td->bind_cpumask); + td->bind_cpumask = CPU_ALLOC(g->p.nr_cpus); + BUG_ON(!td->bind_cpumask); + CPU_ZERO_S(size, td->bind_cpumask); for (cpu = bind_cpu; cpu < bind_cpu+bind_len; cpu++) { - BUG_ON(cpu < 0 || cpu >= g->p.nr_cpus); - CPU_SET(cpu, &td->bind_cpumask); + if (cpu < 0 || cpu >= g->p.nr_cpus) { + CPU_FREE(td->bind_cpumask); + BUG_ON(-1); + } + CPU_SET_S(cpu, size, td->bind_cpumask); } t++; } @@ -1245,7 +1296,7 @@ static void *worker_thread(void *__tdata) * by migrating to CPU#0: */ if (first_task && g->p.perturb_secs && (int)(stop.tv_sec - last_perturbance) >= g->p.perturb_secs) { - cpu_set_t orig_mask; + cpu_set_t *orig_mask; int target_cpu; int this_cpu; @@ -1269,6 +1320,7 @@ static void *worker_thread(void *__tdata) printf(" (injecting perturbalance, moved to CPU#%d)\n", target_cpu); bind_to_cpumask(orig_mask); + CPU_FREE(orig_mask); } if (details >= 3) { @@ -1402,21 +1454,31 @@ static void init_thread_data(void) for (t = 0; t < g->p.nr_tasks; t++) { struct thread_data *td = g->threads + t; + size_t cpuset_size = CPU_ALLOC_SIZE(g->p.nr_cpus); int cpu; /* Allow all nodes by default: */ td->bind_node = NUMA_NO_NODE; /* Allow all CPUs by default: */ - CPU_ZERO(&td->bind_cpumask); + td->bind_cpumask = CPU_ALLOC(g->p.nr_cpus); + BUG_ON(!td->bind_cpumask); + CPU_ZERO_S(cpuset_size, td->bind_cpumask); for (cpu = 0; cpu < g->p.nr_cpus; cpu++) - CPU_SET(cpu, &td->bind_cpumask); + CPU_SET_S(cpu, cpuset_size, td->bind_cpumask); } } static void deinit_thread_data(void) { ssize_t size = sizeof(*g->threads)*g->p.nr_tasks; + int t; + + /* Free the bind_cpumask allocated for thread_data */ + for (t = 0; t < g->p.nr_tasks; t++) { + struct thread_data *td = g->threads + t; + CPU_FREE(td->bind_cpumask); + } free_data(g->threads, size); } -- cgit v1.2.3 From d4e5268a08b211b536fed29beb24271ecd85187e Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Fri, 8 Apr 2022 11:45:55 +0200 Subject: x86,objtool: Mark cpu_startup_entry() __noreturn GCC-8 isn't clever enough to figure out that cpu_start_entry() is a noreturn while objtool is. This results in code after the call in start_secondary(). Give GCC a hand so that they all agree on things. vmlinux.o: warning: objtool: start_secondary()+0x10e: unreachable Reported-by: Rick Edgecombe Signed-off-by: Peter Zijlstra (Intel) Acked-by: Josh Poimboeuf Link: https://lore.kernel.org/r/20220408094718.383658532@infradead.org --- include/linux/cpu.h | 2 +- tools/objtool/check.c | 1 + 2 files changed, 2 insertions(+), 1 deletion(-) (limited to 'tools') diff --git a/include/linux/cpu.h b/include/linux/cpu.h index 9cf51e41e697..54dc2f9a2d56 100644 --- a/include/linux/cpu.h +++ b/include/linux/cpu.h @@ -167,7 +167,7 @@ static inline int suspend_disable_secondary_cpus(void) { return 0; } static inline void suspend_enable_secondary_cpus(void) { } #endif /* !CONFIG_PM_SLEEP_SMP */ -void cpu_startup_entry(enum cpuhp_state state); +void __noreturn cpu_startup_entry(enum cpuhp_state state); void cpu_idle_poll_ctrl(bool enable); diff --git a/tools/objtool/check.c b/tools/objtool/check.c index bd0c2c828940..e3a675d6a704 100644 --- a/tools/objtool/check.c +++ b/tools/objtool/check.c @@ -184,6 +184,7 @@ static bool __dead_end_function(struct objtool_file *file, struct symbol *func, "do_group_exit", "stop_this_cpu", "__invalid_creds", + "cpu_startup_entry", }; if (!func) -- cgit v1.2.3 From 4baae989e638e9bf4b7d29bc5e36b581fddcca52 Mon Sep 17 00:00:00 2001 From: Josh Poimboeuf Date: Mon, 18 Apr 2022 09:50:29 -0700 Subject: objtool: Print data address for "!ENDBR" data warnings When a "!ENDBR" warning is reported for a data section, objtool just prints the text address of the relocation target twice, without giving any clues about the location of the original data reference: vmlinux.o: warning: objtool: dcbnl_netdevice_event()+0x0: .text+0xb64680: data relocation to !ENDBR: dcbnl_netdevice_event+0x0 Instead, print the address of the data reference, in addition to the address of the relocation target. vmlinux.o: warning: objtool: dcbnl_nb+0x0: .data..read_mostly+0xe260: data relocation to !ENDBR: dcbnl_netdevice_event+0x0 Fixes: 89bc853eae4a ("objtool: Find unused ENDBR instructions") Signed-off-by: Josh Poimboeuf Signed-off-by: Peter Zijlstra (Intel) Link: https://lkml.kernel.org/r/762e88d51300e8eaf0f933a5b0feae20ac033bea.1650300597.git.jpoimboe@redhat.com --- tools/objtool/check.c | 7 ++----- 1 file changed, 2 insertions(+), 5 deletions(-) (limited to 'tools') diff --git a/tools/objtool/check.c b/tools/objtool/check.c index e3a675d6a704..b822a6d5a172 100644 --- a/tools/objtool/check.c +++ b/tools/objtool/check.c @@ -3817,11 +3817,8 @@ static int validate_ibt(struct objtool_file *file) struct instruction *dest; dest = validate_ibt_reloc(file, reloc); - if (is_data && dest && !dest->noendbr) { - warn_noendbr("data ", reloc->sym->sec, - reloc->sym->offset + reloc->addend, - dest); - } + if (is_data && dest && !dest->noendbr) + warn_noendbr("data ", sec, reloc->offset, dest); } } -- cgit v1.2.3 From 1d08b92fa2c41c43e4efe9787413e9ac9a434f83 Mon Sep 17 00:00:00 2001 From: Josh Poimboeuf Date: Mon, 18 Apr 2022 09:50:30 -0700 Subject: objtool: Use offstr() to print address of missing ENDBR Fixes: 89bc853eae4a ("objtool: Find unused ENDBR instructions") Signed-off-by: Josh Poimboeuf Signed-off-by: Peter Zijlstra (Intel) Link: https://lkml.kernel.org/r/95d12e800c736a3f7d08d61dabb760b2d5251a8e.1650300597.git.jpoimboe@redhat.com --- tools/objtool/check.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) (limited to 'tools') diff --git a/tools/objtool/check.c b/tools/objtool/check.c index b822a6d5a172..5285edd41da8 100644 --- a/tools/objtool/check.c +++ b/tools/objtool/check.c @@ -3211,9 +3211,8 @@ validate_ibt_reloc(struct objtool_file *file, struct reloc *reloc) static void warn_noendbr(const char *msg, struct section *sec, unsigned long offset, struct instruction *dest) { - WARN_FUNC("%srelocation to !ENDBR: %s+0x%lx", sec, offset, msg, - dest->func ? dest->func->name : dest->sec->name, - dest->func ? dest->offset - dest->func->offset : dest->offset); + WARN_FUNC("%srelocation to !ENDBR: %s", sec, offset, msg, + offstr(dest->sec, dest->offset)); } static void validate_ibt_dest(struct objtool_file *file, struct instruction *insn, -- cgit v1.2.3 From 26ff604102c98df79c3fe2614d1b9bb068d4c28c Mon Sep 17 00:00:00 2001 From: Josh Poimboeuf Date: Mon, 11 Apr 2022 16:10:30 -0700 Subject: objtool: Don't set 'jump_dest' for sibling calls For most sibling calls, 'jump_dest' is NULL because objtool treats the jump like a call and sets 'call_dest'. But there are a few edge cases where that's not true. Make it consistent to avoid unexpected behavior. Signed-off-by: Josh Poimboeuf Signed-off-by: Peter Zijlstra (Intel) Link: https://lkml.kernel.org/r/8737d6b9d1691831aed73375f444f0f42da3e2c9.1649718562.git.jpoimboe@redhat.com --- tools/objtool/check.c | 35 ++++++++++++++++++++++------------- 1 file changed, 22 insertions(+), 13 deletions(-) (limited to 'tools') diff --git a/tools/objtool/check.c b/tools/objtool/check.c index bd0c2c828940..6f492789c8c0 100644 --- a/tools/objtool/check.c +++ b/tools/objtool/check.c @@ -1271,7 +1271,7 @@ static bool is_first_func_insn(struct objtool_file *file, struct instruction *in */ static int add_jump_destinations(struct objtool_file *file) { - struct instruction *insn; + struct instruction *insn, *jump_dest; struct reloc *reloc; struct section *dest_sec; unsigned long dest_off; @@ -1291,7 +1291,10 @@ static int add_jump_destinations(struct objtool_file *file) add_retpoline_call(file, insn); continue; } else if (insn->func) { - /* internal or external sibling call (with reloc) */ + /* + * External sibling call or internal sibling call with + * STT_FUNC reloc. + */ add_call_dest(file, insn, reloc->sym, true); continue; } else if (reloc->sym->sec->idx) { @@ -1303,8 +1306,8 @@ static int add_jump_destinations(struct objtool_file *file) continue; } - insn->jump_dest = find_insn(file, dest_sec, dest_off); - if (!insn->jump_dest) { + jump_dest = find_insn(file, dest_sec, dest_off); + if (!jump_dest) { /* * This is a special case where an alt instruction @@ -1323,8 +1326,8 @@ static int add_jump_destinations(struct objtool_file *file) /* * Cross-function jump. */ - if (insn->func && insn->jump_dest->func && - insn->func != insn->jump_dest->func) { + if (insn->func && jump_dest->func && + insn->func != jump_dest->func) { /* * For GCC 8+, create parent/child links for any cold @@ -1342,16 +1345,22 @@ static int add_jump_destinations(struct objtool_file *file) * subfunction is through a jump table. */ if (!strstr(insn->func->name, ".cold") && - strstr(insn->jump_dest->func->name, ".cold")) { - insn->func->cfunc = insn->jump_dest->func; - insn->jump_dest->func->pfunc = insn->func; + strstr(jump_dest->func->name, ".cold")) { + insn->func->cfunc = jump_dest->func; + jump_dest->func->pfunc = insn->func; - } else if (!same_function(insn, insn->jump_dest) && - is_first_func_insn(file, insn->jump_dest)) { - /* internal sibling call (without reloc) */ - add_call_dest(file, insn, insn->jump_dest->func, true); + } else if (!same_function(insn, jump_dest) && + is_first_func_insn(file, jump_dest)) { + /* + * Internal sibling call without reloc or with + * STT_SECTION reloc. + */ + add_call_dest(file, insn, jump_dest->func, true); + continue; } } + + insn->jump_dest = jump_dest; } return 0; -- cgit v1.2.3 From 34c861e806478ac2ea4032721defbf1d6967df08 Mon Sep 17 00:00:00 2001 From: Josh Poimboeuf Date: Mon, 11 Apr 2022 16:10:31 -0700 Subject: objtool: Fix sibling call detection in alternatives In add_jump_destinations(), sibling call detection requires 'insn->func' to be valid. But alternative instructions get their 'func' set in handle_group_alt(), which runs *after* add_jump_destinations(). So sibling calls in alternatives code don't get properly detected. Fix that by changing the initialization order: call add_special_section_alts() *before* add_jump_destinations(). This also means the special case for a missing 'jump_dest' in add_jump_destinations() can be removed, as it has already been dealt with. Signed-off-by: Josh Poimboeuf Signed-off-by: Peter Zijlstra (Intel) Link: https://lkml.kernel.org/r/c02e0a0a2a4286b5f848d17c77fdcb7e0caf709c.1649718562.git.jpoimboe@redhat.com --- tools/objtool/check.c | 36 +++++++++++++++++------------------- 1 file changed, 17 insertions(+), 19 deletions(-) (limited to 'tools') diff --git a/tools/objtool/check.c b/tools/objtool/check.c index 6f492789c8c0..0f5d3de30e0d 100644 --- a/tools/objtool/check.c +++ b/tools/objtool/check.c @@ -1277,6 +1277,13 @@ static int add_jump_destinations(struct objtool_file *file) unsigned long dest_off; for_each_insn(file, insn) { + if (insn->jump_dest) { + /* + * handle_group_alt() may have previously set + * 'jump_dest' for some alternatives. + */ + continue; + } if (!is_static_jump(insn)) continue; @@ -1308,15 +1315,6 @@ static int add_jump_destinations(struct objtool_file *file) jump_dest = find_insn(file, dest_sec, dest_off); if (!jump_dest) { - - /* - * This is a special case where an alt instruction - * jumps past the end of the section. These are - * handled later in handle_group_alt(). - */ - if (!strcmp(insn->sec->name, ".altinstr_replacement")) - continue; - WARN_FUNC("can't find jump dest instruction at %s+0x%lx", insn->sec, insn->offset, dest_sec->name, dest_off); @@ -1549,13 +1547,13 @@ static int handle_group_alt(struct objtool_file *file, continue; dest_off = arch_jump_destination(insn); - if (dest_off == special_alt->new_off + special_alt->new_len) + if (dest_off == special_alt->new_off + special_alt->new_len) { insn->jump_dest = next_insn_same_sec(file, last_orig_insn); - - if (!insn->jump_dest) { - WARN_FUNC("can't find alternative jump destination", - insn->sec, insn->offset); - return -1; + if (!insn->jump_dest) { + WARN_FUNC("can't find alternative jump destination", + insn->sec, insn->offset); + return -1; + } } } @@ -2254,14 +2252,14 @@ static int decode_sections(struct objtool_file *file) return ret; /* - * Must be before add_special_section_alts() as that depends on - * jump_dest being set. + * Must be before add_jump_destinations(), which depends on 'func' + * being set for alternatives, to enable proper sibling call detection. */ - ret = add_jump_destinations(file); + ret = add_special_section_alts(file); if (ret) return ret; - ret = add_special_section_alts(file); + ret = add_jump_destinations(file); if (ret) return ret; -- cgit v1.2.3 From 08feafe8d1958febf3a9733a3d1564d8fc23340e Mon Sep 17 00:00:00 2001 From: Josh Poimboeuf Date: Mon, 11 Apr 2022 16:10:32 -0700 Subject: objtool: Fix function fallthrough detection for vmlinux Objtool's function fallthrough detection only works on C objects. The distinction between C and assembly objects no longer makes sense with objtool running on vmlinux.o. Now that copy_user_64.S has been fixed up, and an objtool sibling call detection bug has been fixed, the asm code is in "compliance" and this hack is no longer needed. Remove it. Fixes: ed53a0d97192 ("x86/alternative: Use .ibt_endbr_seal to seal indirect calls") Signed-off-by: Josh Poimboeuf Signed-off-by: Peter Zijlstra (Intel) Link: https://lkml.kernel.org/r/b434cff98eca3a60dcc64c620d7d5d405a0f441c.1649718562.git.jpoimboe@redhat.com --- tools/objtool/check.c | 2 +- tools/objtool/include/objtool/objtool.h | 2 +- tools/objtool/objtool.c | 1 - 3 files changed, 2 insertions(+), 3 deletions(-) (limited to 'tools') diff --git a/tools/objtool/check.c b/tools/objtool/check.c index 0f5d3de30e0d..5f10653eb5c2 100644 --- a/tools/objtool/check.c +++ b/tools/objtool/check.c @@ -3310,7 +3310,7 @@ static int validate_branch(struct objtool_file *file, struct symbol *func, while (1) { next_insn = next_insn_to_validate(file, insn); - if (file->c_file && func && insn->func && func != insn->func->pfunc) { + if (func && insn->func && func != insn->func->pfunc) { WARN("%s() falls through to next function %s()", func->name, insn->func->name); return 1; diff --git a/tools/objtool/include/objtool/objtool.h b/tools/objtool/include/objtool/objtool.h index 7a5c13a78f87..a6e72d916807 100644 --- a/tools/objtool/include/objtool/objtool.h +++ b/tools/objtool/include/objtool/objtool.h @@ -27,7 +27,7 @@ struct objtool_file { struct list_head static_call_list; struct list_head mcount_loc_list; struct list_head endbr_list; - bool ignore_unreachables, c_file, hints, rodata; + bool ignore_unreachables, hints, rodata; unsigned int nr_endbr; unsigned int nr_endbr_int; diff --git a/tools/objtool/objtool.c b/tools/objtool/objtool.c index b09946f4e1d6..843ff3c2f28e 100644 --- a/tools/objtool/objtool.c +++ b/tools/objtool/objtool.c @@ -129,7 +129,6 @@ struct objtool_file *objtool_open_read(const char *_objname) INIT_LIST_HEAD(&file.static_call_list); INIT_LIST_HEAD(&file.mcount_loc_list); INIT_LIST_HEAD(&file.endbr_list); - file.c_file = !vmlinux && find_section_by_name(file.elf, ".comment"); file.ignore_unreachables = no_unreachable; file.hints = false; -- cgit v1.2.3 From 044011fdf162c5dd61c02841930c8f438a9adadb Mon Sep 17 00:00:00 2001 From: Ido Schimmel Date: Tue, 19 Apr 2022 16:51:54 +0300 Subject: selftests: mlxsw: vxlan_flooding: Prevent flooding of unwanted packets The test verifies that packets are correctly flooded by the bridge and the VXLAN device by matching on the encapsulated packets at the other end. However, if packets other than those generated by the test also ingress the bridge (e.g., MLD packets), they will be flooded as well and interfere with the expected count. Make the test more robust by making sure that only the packets generated by the test can ingress the bridge. Drop all the rest using tc filters on the egress of 'br0' and 'h1'. In the software data path, the problem can be solved by matching on the inner destination MAC or dropping unwanted packets at the egress of the VXLAN device, but this is not currently supported by mlxsw. Fixes: 94d302deae25 ("selftests: mlxsw: Add a test for VxLAN flooding") Signed-off-by: Ido Schimmel Reviewed-by: Amit Cohen Signed-off-by: David S. Miller --- .../selftests/drivers/net/mlxsw/vxlan_flooding.sh | 17 +++++++++++++++++ 1 file changed, 17 insertions(+) (limited to 'tools') diff --git a/tools/testing/selftests/drivers/net/mlxsw/vxlan_flooding.sh b/tools/testing/selftests/drivers/net/mlxsw/vxlan_flooding.sh index fedcb7b35af9..af5ea50ed5c0 100755 --- a/tools/testing/selftests/drivers/net/mlxsw/vxlan_flooding.sh +++ b/tools/testing/selftests/drivers/net/mlxsw/vxlan_flooding.sh @@ -172,6 +172,17 @@ flooding_filters_add() local lsb local i + # Prevent unwanted packets from entering the bridge and interfering + # with the test. + tc qdisc add dev br0 clsact + tc filter add dev br0 egress protocol all pref 1 handle 1 \ + matchall skip_hw action drop + tc qdisc add dev $h1 clsact + tc filter add dev $h1 egress protocol all pref 1 handle 1 \ + flower skip_hw dst_mac de:ad:be:ef:13:37 action pass + tc filter add dev $h1 egress protocol all pref 2 handle 2 \ + matchall skip_hw action drop + tc qdisc add dev $rp2 clsact for i in $(eval echo {1..$num_remotes}); do @@ -194,6 +205,12 @@ flooding_filters_del() done tc qdisc del dev $rp2 clsact + + tc filter del dev $h1 egress protocol all pref 2 handle 2 matchall + tc filter del dev $h1 egress protocol all pref 1 handle 1 flower + tc qdisc del dev $h1 clsact + tc filter del dev br0 egress protocol all pref 1 handle 1 matchall + tc qdisc del dev br0 clsact } flooding_check_packets() -- cgit v1.2.3 From 5e6242151d7f17b056a82ca7b860c4ec8eaa7589 Mon Sep 17 00:00:00 2001 From: Ido Schimmel Date: Tue, 19 Apr 2022 16:51:55 +0300 Subject: selftests: mlxsw: vxlan_flooding_ipv6: Prevent flooding of unwanted packets The test verifies that packets are correctly flooded by the bridge and the VXLAN device by matching on the encapsulated packets at the other end. However, if packets other than those generated by the test also ingress the bridge (e.g., MLD packets), they will be flooded as well and interfere with the expected count. Make the test more robust by making sure that only the packets generated by the test can ingress the bridge. Drop all the rest using tc filters on the egress of 'br0' and 'h1'. In the software data path, the problem can be solved by matching on the inner destination MAC or dropping unwanted packets at the egress of the VXLAN device, but this is not currently supported by mlxsw. Fixes: d01724dd2a66 ("selftests: mlxsw: spectrum-2: Add a test for VxLAN flooding with IPv6") Signed-off-by: Ido Schimmel Reviewed-by: Amit Cohen Signed-off-by: David S. Miller --- .../drivers/net/mlxsw/spectrum-2/vxlan_flooding_ipv6.sh | 17 +++++++++++++++++ 1 file changed, 17 insertions(+) (limited to 'tools') diff --git a/tools/testing/selftests/drivers/net/mlxsw/spectrum-2/vxlan_flooding_ipv6.sh b/tools/testing/selftests/drivers/net/mlxsw/spectrum-2/vxlan_flooding_ipv6.sh index 429f7ee735cf..fd23c80eba31 100755 --- a/tools/testing/selftests/drivers/net/mlxsw/spectrum-2/vxlan_flooding_ipv6.sh +++ b/tools/testing/selftests/drivers/net/mlxsw/spectrum-2/vxlan_flooding_ipv6.sh @@ -159,6 +159,17 @@ flooding_remotes_add() local lsb local i + # Prevent unwanted packets from entering the bridge and interfering + # with the test. + tc qdisc add dev br0 clsact + tc filter add dev br0 egress protocol all pref 1 handle 1 \ + matchall skip_hw action drop + tc qdisc add dev $h1 clsact + tc filter add dev $h1 egress protocol all pref 1 handle 1 \ + flower skip_hw dst_mac de:ad:be:ef:13:37 action pass + tc filter add dev $h1 egress protocol all pref 2 handle 2 \ + matchall skip_hw action drop + for i in $(eval echo {1..$num_remotes}); do lsb=$((i + 1)) @@ -195,6 +206,12 @@ flooding_filters_del() done tc qdisc del dev $rp2 clsact + + tc filter del dev $h1 egress protocol all pref 2 handle 2 matchall + tc filter del dev $h1 egress protocol all pref 1 handle 1 flower + tc qdisc del dev $h1 clsact + tc filter del dev br0 egress protocol all pref 1 handle 1 matchall + tc qdisc del dev br0 clsact } flooding_check_packets() -- cgit v1.2.3 From 266a19a0bc4fbfab4d981a47640ca98972a01865 Mon Sep 17 00:00:00 2001 From: Thomas Huth Date: Thu, 14 Apr 2022 12:30:31 +0200 Subject: KVM: selftests: Silence compiler warning in the kvm_page_table_test When compiling kvm_page_table_test.c, I get this compiler warning with gcc 11.2: kvm_page_table_test.c: In function 'pre_init_before_test': ../../../../tools/include/linux/kernel.h:44:24: warning: comparison of distinct pointer types lacks a cast 44 | (void) (&_max1 == &_max2); \ | ^~ kvm_page_table_test.c:281:21: note: in expansion of macro 'max' 281 | alignment = max(0x100000, alignment); | ^~~ Fix it by adjusting the type of the absolute value. Signed-off-by: Thomas Huth Reviewed-by: Claudio Imbrenda Message-Id: <20220414103031.565037-1-thuth@redhat.com> Signed-off-by: Paolo Bonzini --- tools/testing/selftests/kvm/kvm_page_table_test.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'tools') diff --git a/tools/testing/selftests/kvm/kvm_page_table_test.c b/tools/testing/selftests/kvm/kvm_page_table_test.c index ba1fdc3dcf4a..2c4a7563a4f8 100644 --- a/tools/testing/selftests/kvm/kvm_page_table_test.c +++ b/tools/testing/selftests/kvm/kvm_page_table_test.c @@ -278,7 +278,7 @@ static struct kvm_vm *pre_init_before_test(enum vm_guest_mode mode, void *arg) else guest_test_phys_mem = p->phys_offset; #ifdef __s390x__ - alignment = max(0x100000, alignment); + alignment = max(0x100000UL, alignment); #endif guest_test_phys_mem = align_down(guest_test_phys_mem, alignment); -- cgit v1.2.3 From f18b4aebe107d092e384b1ae680b1e1de7a0196d Mon Sep 17 00:00:00 2001 From: Paolo Bonzini Date: Wed, 20 Apr 2022 06:27:27 -0400 Subject: kvm: selftests: do not use bitfields larger than 32-bits for PTEs Red Hat's QE team reported test failure on access_tracking_perf_test: Testing guest mode: PA-bits:ANY, VA-bits:48, 4K pages guest physical test memory offset: 0x3fffbffff000 Populating memory : 0.684014577s Writing to populated memory : 0.006230175s Reading from populated memory : 0.004557805s ==== Test Assertion Failure ==== lib/kvm_util.c:1411: false pid=125806 tid=125809 errno=4 - Interrupted system call 1 0x0000000000402f7c: addr_gpa2hva at kvm_util.c:1411 2 (inlined by) addr_gpa2hva at kvm_util.c:1405 3 0x0000000000401f52: lookup_pfn at access_tracking_perf_test.c:98 4 (inlined by) mark_vcpu_memory_idle at access_tracking_perf_test.c:152 5 (inlined by) vcpu_thread_main at access_tracking_perf_test.c:232 6 0x00007fefe9ff81ce: ?? ??:0 7 0x00007fefe9c64d82: ?? ??:0 No vm physical memory at 0xffbffff000 I can easily reproduce it with a Intel(R) Xeon(R) CPU E5-2630 with 46 bits PA. It turns out that the address translation for clearing idle page tracking returned a wrong result; addr_gva2gpa()'s last step, which is based on "pte[index[0]].pfn", did the calculation with 40 bits length and the high 12 bits got truncated. In above case the GPA address to be returned should be 0x3fffbffff000 for GVA 0xc0000000, but it got truncated into 0xffbffff000 and the subsequent gpa2hva lookup failed. The width of operations on bit fields greater than 32-bit is implementation defined, and differs between GCC (which uses the bitfield precision) and clang (which uses 64-bit arithmetic), so this is a potential minefield. Remove the bit fields and using manual masking instead. Bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=2075036 Reported-by: Nana Liu Reviewed-by: Peter Xu Tested-by: Peter Xu Signed-off-by: Paolo Bonzini --- .../selftests/kvm/include/x86_64/processor.h | 15 ++ tools/testing/selftests/kvm/lib/x86_64/processor.c | 192 +++++++++------------ 2 files changed, 92 insertions(+), 115 deletions(-) (limited to 'tools') diff --git a/tools/testing/selftests/kvm/include/x86_64/processor.h b/tools/testing/selftests/kvm/include/x86_64/processor.h index 37db341d4cc5..86e79af64dea 100644 --- a/tools/testing/selftests/kvm/include/x86_64/processor.h +++ b/tools/testing/selftests/kvm/include/x86_64/processor.h @@ -60,6 +60,21 @@ /* CPUID.0x8000_0001.EDX */ #define CPUID_GBPAGES (1ul << 26) +/* Page table bitfield declarations */ +#define PTE_PRESENT_MASK BIT_ULL(0) +#define PTE_WRITABLE_MASK BIT_ULL(1) +#define PTE_USER_MASK BIT_ULL(2) +#define PTE_ACCESSED_MASK BIT_ULL(5) +#define PTE_DIRTY_MASK BIT_ULL(6) +#define PTE_LARGE_MASK BIT_ULL(7) +#define PTE_GLOBAL_MASK BIT_ULL(8) +#define PTE_NX_MASK BIT_ULL(63) + +#define PAGE_SHIFT 12 + +#define PHYSICAL_PAGE_MASK GENMASK_ULL(51, 12) +#define PTE_GET_PFN(pte) (((pte) & PHYSICAL_PAGE_MASK) >> PAGE_SHIFT) + /* General Registers in 64-Bit Mode */ struct gpr64_regs { u64 rax; diff --git a/tools/testing/selftests/kvm/lib/x86_64/processor.c b/tools/testing/selftests/kvm/lib/x86_64/processor.c index 9f000dfb5594..0dd442c26015 100644 --- a/tools/testing/selftests/kvm/lib/x86_64/processor.c +++ b/tools/testing/selftests/kvm/lib/x86_64/processor.c @@ -19,38 +19,6 @@ vm_vaddr_t exception_handlers; -/* Virtual translation table structure declarations */ -struct pageUpperEntry { - uint64_t present:1; - uint64_t writable:1; - uint64_t user:1; - uint64_t write_through:1; - uint64_t cache_disable:1; - uint64_t accessed:1; - uint64_t ignored_06:1; - uint64_t page_size:1; - uint64_t ignored_11_08:4; - uint64_t pfn:40; - uint64_t ignored_62_52:11; - uint64_t execute_disable:1; -}; - -struct pageTableEntry { - uint64_t present:1; - uint64_t writable:1; - uint64_t user:1; - uint64_t write_through:1; - uint64_t cache_disable:1; - uint64_t accessed:1; - uint64_t dirty:1; - uint64_t reserved_07:1; - uint64_t global:1; - uint64_t ignored_11_09:3; - uint64_t pfn:40; - uint64_t ignored_62_52:11; - uint64_t execute_disable:1; -}; - void regs_dump(FILE *stream, struct kvm_regs *regs, uint8_t indent) { @@ -195,23 +163,21 @@ static void *virt_get_pte(struct kvm_vm *vm, uint64_t pt_pfn, uint64_t vaddr, return &page_table[index]; } -static struct pageUpperEntry *virt_create_upper_pte(struct kvm_vm *vm, - uint64_t pt_pfn, - uint64_t vaddr, - uint64_t paddr, - int level, - enum x86_page_size page_size) +static uint64_t *virt_create_upper_pte(struct kvm_vm *vm, + uint64_t pt_pfn, + uint64_t vaddr, + uint64_t paddr, + int level, + enum x86_page_size page_size) { - struct pageUpperEntry *pte = virt_get_pte(vm, pt_pfn, vaddr, level); - - if (!pte->present) { - pte->writable = true; - pte->present = true; - pte->page_size = (level == page_size); - if (pte->page_size) - pte->pfn = paddr >> vm->page_shift; + uint64_t *pte = virt_get_pte(vm, pt_pfn, vaddr, level); + + if (!(*pte & PTE_PRESENT_MASK)) { + *pte = PTE_PRESENT_MASK | PTE_WRITABLE_MASK; + if (level == page_size) + *pte |= PTE_LARGE_MASK | (paddr & PHYSICAL_PAGE_MASK); else - pte->pfn = vm_alloc_page_table(vm) >> vm->page_shift; + *pte |= vm_alloc_page_table(vm) & PHYSICAL_PAGE_MASK; } else { /* * Entry already present. Assert that the caller doesn't want @@ -221,7 +187,7 @@ static struct pageUpperEntry *virt_create_upper_pte(struct kvm_vm *vm, TEST_ASSERT(level != page_size, "Cannot create hugepage at level: %u, vaddr: 0x%lx\n", page_size, vaddr); - TEST_ASSERT(!pte->page_size, + TEST_ASSERT(!(*pte & PTE_LARGE_MASK), "Cannot create page table at level: %u, vaddr: 0x%lx\n", level, vaddr); } @@ -232,8 +198,8 @@ void __virt_pg_map(struct kvm_vm *vm, uint64_t vaddr, uint64_t paddr, enum x86_page_size page_size) { const uint64_t pg_size = 1ull << ((page_size * 9) + 12); - struct pageUpperEntry *pml4e, *pdpe, *pde; - struct pageTableEntry *pte; + uint64_t *pml4e, *pdpe, *pde; + uint64_t *pte; TEST_ASSERT(vm->mode == VM_MODE_PXXV48_4K, "Unknown or unsupported guest mode, mode: 0x%x", vm->mode); @@ -257,24 +223,22 @@ void __virt_pg_map(struct kvm_vm *vm, uint64_t vaddr, uint64_t paddr, */ pml4e = virt_create_upper_pte(vm, vm->pgd >> vm->page_shift, vaddr, paddr, 3, page_size); - if (pml4e->page_size) + if (*pml4e & PTE_LARGE_MASK) return; - pdpe = virt_create_upper_pte(vm, pml4e->pfn, vaddr, paddr, 2, page_size); - if (pdpe->page_size) + pdpe = virt_create_upper_pte(vm, PTE_GET_PFN(*pml4e), vaddr, paddr, 2, page_size); + if (*pdpe & PTE_LARGE_MASK) return; - pde = virt_create_upper_pte(vm, pdpe->pfn, vaddr, paddr, 1, page_size); - if (pde->page_size) + pde = virt_create_upper_pte(vm, PTE_GET_PFN(*pdpe), vaddr, paddr, 1, page_size); + if (*pde & PTE_LARGE_MASK) return; /* Fill in page table entry. */ - pte = virt_get_pte(vm, pde->pfn, vaddr, 0); - TEST_ASSERT(!pte->present, + pte = virt_get_pte(vm, PTE_GET_PFN(*pde), vaddr, 0); + TEST_ASSERT(!(*pte & PTE_PRESENT_MASK), "PTE already present for 4k page at vaddr: 0x%lx\n", vaddr); - pte->pfn = paddr >> vm->page_shift; - pte->writable = true; - pte->present = 1; + *pte = PTE_PRESENT_MASK | PTE_WRITABLE_MASK | (paddr & PHYSICAL_PAGE_MASK); } void virt_pg_map(struct kvm_vm *vm, uint64_t vaddr, uint64_t paddr) @@ -282,12 +246,12 @@ void virt_pg_map(struct kvm_vm *vm, uint64_t vaddr, uint64_t paddr) __virt_pg_map(vm, vaddr, paddr, X86_PAGE_SIZE_4K); } -static struct pageTableEntry *_vm_get_page_table_entry(struct kvm_vm *vm, int vcpuid, +static uint64_t *_vm_get_page_table_entry(struct kvm_vm *vm, int vcpuid, uint64_t vaddr) { uint16_t index[4]; - struct pageUpperEntry *pml4e, *pdpe, *pde; - struct pageTableEntry *pte; + uint64_t *pml4e, *pdpe, *pde; + uint64_t *pte; struct kvm_cpuid_entry2 *entry; struct kvm_sregs sregs; int max_phy_addr; @@ -329,30 +293,29 @@ static struct pageTableEntry *_vm_get_page_table_entry(struct kvm_vm *vm, int vc index[3] = (vaddr >> 39) & 0x1ffu; pml4e = addr_gpa2hva(vm, vm->pgd); - TEST_ASSERT(pml4e[index[3]].present, + TEST_ASSERT(pml4e[index[3]] & PTE_PRESENT_MASK, "Expected pml4e to be present for gva: 0x%08lx", vaddr); - TEST_ASSERT((*(uint64_t*)(&pml4e[index[3]]) & - (rsvd_mask | (1ull << 7))) == 0, + TEST_ASSERT((pml4e[index[3]] & (rsvd_mask | PTE_LARGE_MASK)) == 0, "Unexpected reserved bits set."); - pdpe = addr_gpa2hva(vm, pml4e[index[3]].pfn * vm->page_size); - TEST_ASSERT(pdpe[index[2]].present, + pdpe = addr_gpa2hva(vm, PTE_GET_PFN(pml4e[index[3]]) * vm->page_size); + TEST_ASSERT(pdpe[index[2]] & PTE_PRESENT_MASK, "Expected pdpe to be present for gva: 0x%08lx", vaddr); - TEST_ASSERT(pdpe[index[2]].page_size == 0, + TEST_ASSERT(!(pdpe[index[2]] & PTE_LARGE_MASK), "Expected pdpe to map a pde not a 1-GByte page."); - TEST_ASSERT((*(uint64_t*)(&pdpe[index[2]]) & rsvd_mask) == 0, + TEST_ASSERT((pdpe[index[2]] & rsvd_mask) == 0, "Unexpected reserved bits set."); - pde = addr_gpa2hva(vm, pdpe[index[2]].pfn * vm->page_size); - TEST_ASSERT(pde[index[1]].present, + pde = addr_gpa2hva(vm, PTE_GET_PFN(pdpe[index[2]]) * vm->page_size); + TEST_ASSERT(pde[index[1]] & PTE_PRESENT_MASK, "Expected pde to be present for gva: 0x%08lx", vaddr); - TEST_ASSERT(pde[index[1]].page_size == 0, + TEST_ASSERT(!(pde[index[1]] & PTE_LARGE_MASK), "Expected pde to map a pte not a 2-MByte page."); - TEST_ASSERT((*(uint64_t*)(&pde[index[1]]) & rsvd_mask) == 0, + TEST_ASSERT((pde[index[1]] & rsvd_mask) == 0, "Unexpected reserved bits set."); - pte = addr_gpa2hva(vm, pde[index[1]].pfn * vm->page_size); - TEST_ASSERT(pte[index[0]].present, + pte = addr_gpa2hva(vm, PTE_GET_PFN(pde[index[1]]) * vm->page_size); + TEST_ASSERT(pte[index[0]] & PTE_PRESENT_MASK, "Expected pte to be present for gva: 0x%08lx", vaddr); return &pte[index[0]]; @@ -360,7 +323,7 @@ static struct pageTableEntry *_vm_get_page_table_entry(struct kvm_vm *vm, int vc uint64_t vm_get_page_table_entry(struct kvm_vm *vm, int vcpuid, uint64_t vaddr) { - struct pageTableEntry *pte = _vm_get_page_table_entry(vm, vcpuid, vaddr); + uint64_t *pte = _vm_get_page_table_entry(vm, vcpuid, vaddr); return *(uint64_t *)pte; } @@ -368,18 +331,17 @@ uint64_t vm_get_page_table_entry(struct kvm_vm *vm, int vcpuid, uint64_t vaddr) void vm_set_page_table_entry(struct kvm_vm *vm, int vcpuid, uint64_t vaddr, uint64_t pte) { - struct pageTableEntry *new_pte = _vm_get_page_table_entry(vm, vcpuid, - vaddr); + uint64_t *new_pte = _vm_get_page_table_entry(vm, vcpuid, vaddr); *(uint64_t *)new_pte = pte; } void virt_dump(FILE *stream, struct kvm_vm *vm, uint8_t indent) { - struct pageUpperEntry *pml4e, *pml4e_start; - struct pageUpperEntry *pdpe, *pdpe_start; - struct pageUpperEntry *pde, *pde_start; - struct pageTableEntry *pte, *pte_start; + uint64_t *pml4e, *pml4e_start; + uint64_t *pdpe, *pdpe_start; + uint64_t *pde, *pde_start; + uint64_t *pte, *pte_start; if (!vm->pgd_created) return; @@ -389,58 +351,58 @@ void virt_dump(FILE *stream, struct kvm_vm *vm, uint8_t indent) fprintf(stream, "%*s index hvaddr gpaddr " "addr w exec dirty\n", indent, ""); - pml4e_start = (struct pageUpperEntry *) addr_gpa2hva(vm, vm->pgd); + pml4e_start = (uint64_t *) addr_gpa2hva(vm, vm->pgd); for (uint16_t n1 = 0; n1 <= 0x1ffu; n1++) { pml4e = &pml4e_start[n1]; - if (!pml4e->present) + if (!(*pml4e & PTE_PRESENT_MASK)) continue; - fprintf(stream, "%*spml4e 0x%-3zx %p 0x%-12lx 0x%-10lx %u " + fprintf(stream, "%*spml4e 0x%-3zx %p 0x%-12lx 0x%-10llx %u " " %u\n", indent, "", pml4e - pml4e_start, pml4e, - addr_hva2gpa(vm, pml4e), (uint64_t) pml4e->pfn, - pml4e->writable, pml4e->execute_disable); + addr_hva2gpa(vm, pml4e), PTE_GET_PFN(*pml4e), + !!(*pml4e & PTE_WRITABLE_MASK), !!(*pml4e & PTE_NX_MASK)); - pdpe_start = addr_gpa2hva(vm, pml4e->pfn * vm->page_size); + pdpe_start = addr_gpa2hva(vm, *pml4e & PHYSICAL_PAGE_MASK); for (uint16_t n2 = 0; n2 <= 0x1ffu; n2++) { pdpe = &pdpe_start[n2]; - if (!pdpe->present) + if (!(*pdpe & PTE_PRESENT_MASK)) continue; - fprintf(stream, "%*spdpe 0x%-3zx %p 0x%-12lx 0x%-10lx " + fprintf(stream, "%*spdpe 0x%-3zx %p 0x%-12lx 0x%-10llx " "%u %u\n", indent, "", pdpe - pdpe_start, pdpe, addr_hva2gpa(vm, pdpe), - (uint64_t) pdpe->pfn, pdpe->writable, - pdpe->execute_disable); + PTE_GET_PFN(*pdpe), !!(*pdpe & PTE_WRITABLE_MASK), + !!(*pdpe & PTE_NX_MASK)); - pde_start = addr_gpa2hva(vm, pdpe->pfn * vm->page_size); + pde_start = addr_gpa2hva(vm, *pdpe & PHYSICAL_PAGE_MASK); for (uint16_t n3 = 0; n3 <= 0x1ffu; n3++) { pde = &pde_start[n3]; - if (!pde->present) + if (!(*pde & PTE_PRESENT_MASK)) continue; fprintf(stream, "%*spde 0x%-3zx %p " - "0x%-12lx 0x%-10lx %u %u\n", + "0x%-12lx 0x%-10llx %u %u\n", indent, "", pde - pde_start, pde, addr_hva2gpa(vm, pde), - (uint64_t) pde->pfn, pde->writable, - pde->execute_disable); + PTE_GET_PFN(*pde), !!(*pde & PTE_WRITABLE_MASK), + !!(*pde & PTE_NX_MASK)); - pte_start = addr_gpa2hva(vm, pde->pfn * vm->page_size); + pte_start = addr_gpa2hva(vm, *pde & PHYSICAL_PAGE_MASK); for (uint16_t n4 = 0; n4 <= 0x1ffu; n4++) { pte = &pte_start[n4]; - if (!pte->present) + if (!(*pte & PTE_PRESENT_MASK)) continue; fprintf(stream, "%*spte 0x%-3zx %p " - "0x%-12lx 0x%-10lx %u %u " + "0x%-12lx 0x%-10llx %u %u " " %u 0x%-10lx\n", indent, "", pte - pte_start, pte, addr_hva2gpa(vm, pte), - (uint64_t) pte->pfn, - pte->writable, - pte->execute_disable, - pte->dirty, + PTE_GET_PFN(*pte), + !!(*pte & PTE_WRITABLE_MASK), + !!(*pte & PTE_NX_MASK), + !!(*pte & PTE_DIRTY_MASK), ((uint64_t) n1 << 27) | ((uint64_t) n2 << 18) | ((uint64_t) n3 << 9) @@ -558,8 +520,8 @@ static void kvm_seg_set_kernel_data_64bit(struct kvm_vm *vm, uint16_t selector, vm_paddr_t addr_gva2gpa(struct kvm_vm *vm, vm_vaddr_t gva) { uint16_t index[4]; - struct pageUpperEntry *pml4e, *pdpe, *pde; - struct pageTableEntry *pte; + uint64_t *pml4e, *pdpe, *pde; + uint64_t *pte; TEST_ASSERT(vm->mode == VM_MODE_PXXV48_4K, "Attempt to use " "unknown or unsupported guest mode, mode: 0x%x", vm->mode); @@ -572,22 +534,22 @@ vm_paddr_t addr_gva2gpa(struct kvm_vm *vm, vm_vaddr_t gva) if (!vm->pgd_created) goto unmapped_gva; pml4e = addr_gpa2hva(vm, vm->pgd); - if (!pml4e[index[3]].present) + if (!(pml4e[index[3]] & PTE_PRESENT_MASK)) goto unmapped_gva; - pdpe = addr_gpa2hva(vm, pml4e[index[3]].pfn * vm->page_size); - if (!pdpe[index[2]].present) + pdpe = addr_gpa2hva(vm, PTE_GET_PFN(pml4e[index[3]]) * vm->page_size); + if (!(pdpe[index[2]] & PTE_PRESENT_MASK)) goto unmapped_gva; - pde = addr_gpa2hva(vm, pdpe[index[2]].pfn * vm->page_size); - if (!pde[index[1]].present) + pde = addr_gpa2hva(vm, PTE_GET_PFN(pdpe[index[2]]) * vm->page_size); + if (!(pde[index[1]] & PTE_PRESENT_MASK)) goto unmapped_gva; - pte = addr_gpa2hva(vm, pde[index[1]].pfn * vm->page_size); - if (!pte[index[0]].present) + pte = addr_gpa2hva(vm, PTE_GET_PFN(pde[index[1]]) * vm->page_size); + if (!(pte[index[0]] & PTE_PRESENT_MASK)) goto unmapped_gva; - return (pte[index[0]].pfn * vm->page_size) + (gva & 0xfffu); + return (PTE_GET_PFN(pte[index[0]]) * vm->page_size) + (gva & 0xfffu); unmapped_gva: TEST_FAIL("No mapping for vm virtual address, gva: 0x%lx", gva); -- cgit v1.2.3 From e852be8b148e117e25be1c98cf72ee489b05919e Mon Sep 17 00:00:00 2001 From: Paolo Bonzini Date: Wed, 20 Apr 2022 06:27:27 -0400 Subject: kvm: selftests: introduce and use more page size-related constants Clean up code that was hardcoding masks for various fields, now that the masks are included in processor.h. For more cleanup, define PAGE_SIZE and PAGE_MASK just like in Linux. PAGE_SIZE in particular was defined by several tests. Suggested-by: Sean Christopherson Reviewed-by: Peter Xu Signed-off-by: Paolo Bonzini --- tools/testing/selftests/kvm/include/x86_64/processor.h | 2 ++ tools/testing/selftests/kvm/lib/x86_64/processor.c | 12 ++++++------ tools/testing/selftests/kvm/x86_64/amx_test.c | 1 - tools/testing/selftests/kvm/x86_64/emulator_error_test.c | 1 - tools/testing/selftests/kvm/x86_64/smm_test.c | 2 -- tools/testing/selftests/kvm/x86_64/vmx_tsc_adjust_test.c | 1 - tools/testing/selftests/kvm/x86_64/xen_shinfo_test.c | 1 - tools/testing/selftests/kvm/x86_64/xen_vmcall_test.c | 1 - 8 files changed, 8 insertions(+), 13 deletions(-) (limited to 'tools') diff --git a/tools/testing/selftests/kvm/include/x86_64/processor.h b/tools/testing/selftests/kvm/include/x86_64/processor.h index 86e79af64dea..d0d51adec76e 100644 --- a/tools/testing/selftests/kvm/include/x86_64/processor.h +++ b/tools/testing/selftests/kvm/include/x86_64/processor.h @@ -71,6 +71,8 @@ #define PTE_NX_MASK BIT_ULL(63) #define PAGE_SHIFT 12 +#define PAGE_SIZE (1ULL << PAGE_SHIFT) +#define PAGE_MASK (~(PAGE_SIZE-1)) #define PHYSICAL_PAGE_MASK GENMASK_ULL(51, 12) #define PTE_GET_PFN(pte) (((pte) & PHYSICAL_PAGE_MASK) >> PAGE_SHIFT) diff --git a/tools/testing/selftests/kvm/lib/x86_64/processor.c b/tools/testing/selftests/kvm/lib/x86_64/processor.c index 0dd442c26015..33ea5e9955d9 100644 --- a/tools/testing/selftests/kvm/lib/x86_64/processor.c +++ b/tools/testing/selftests/kvm/lib/x86_64/processor.c @@ -255,13 +255,13 @@ static uint64_t *_vm_get_page_table_entry(struct kvm_vm *vm, int vcpuid, struct kvm_cpuid_entry2 *entry; struct kvm_sregs sregs; int max_phy_addr; - /* Set the bottom 52 bits. */ - uint64_t rsvd_mask = 0x000fffffffffffff; + uint64_t rsvd_mask = 0; entry = kvm_get_supported_cpuid_index(0x80000008, 0); max_phy_addr = entry->eax & 0x000000ff; - /* Clear the bottom bits of the reserved mask. */ - rsvd_mask = (rsvd_mask >> max_phy_addr) << max_phy_addr; + /* Set the high bits in the reserved mask. */ + if (max_phy_addr < 52) + rsvd_mask = GENMASK_ULL(51, max_phy_addr); /* * SDM vol 3, fig 4-11 "Formats of CR3 and Paging-Structure Entries @@ -271,7 +271,7 @@ static uint64_t *_vm_get_page_table_entry(struct kvm_vm *vm, int vcpuid, */ vcpu_sregs_get(vm, vcpuid, &sregs); if ((sregs.efer & EFER_NX) == 0) { - rsvd_mask |= (1ull << 63); + rsvd_mask |= PTE_NX_MASK; } TEST_ASSERT(vm->mode == VM_MODE_PXXV48_4K, "Attempt to use " @@ -549,7 +549,7 @@ vm_paddr_t addr_gva2gpa(struct kvm_vm *vm, vm_vaddr_t gva) if (!(pte[index[0]] & PTE_PRESENT_MASK)) goto unmapped_gva; - return (PTE_GET_PFN(pte[index[0]]) * vm->page_size) + (gva & 0xfffu); + return (PTE_GET_PFN(pte[index[0]]) * vm->page_size) + (gva & ~PAGE_MASK); unmapped_gva: TEST_FAIL("No mapping for vm virtual address, gva: 0x%lx", gva); diff --git a/tools/testing/selftests/kvm/x86_64/amx_test.c b/tools/testing/selftests/kvm/x86_64/amx_test.c index 52a3ef6629e8..76f65c22796f 100644 --- a/tools/testing/selftests/kvm/x86_64/amx_test.c +++ b/tools/testing/selftests/kvm/x86_64/amx_test.c @@ -29,7 +29,6 @@ #define X86_FEATURE_XSAVE (1 << 26) #define X86_FEATURE_OSXSAVE (1 << 27) -#define PAGE_SIZE (1 << 12) #define NUM_TILES 8 #define TILE_SIZE 1024 #define XSAVE_SIZE ((NUM_TILES * TILE_SIZE) + PAGE_SIZE) diff --git a/tools/testing/selftests/kvm/x86_64/emulator_error_test.c b/tools/testing/selftests/kvm/x86_64/emulator_error_test.c index f070ff0224fa..aeb3850f81bd 100644 --- a/tools/testing/selftests/kvm/x86_64/emulator_error_test.c +++ b/tools/testing/selftests/kvm/x86_64/emulator_error_test.c @@ -12,7 +12,6 @@ #include "vmx.h" #define VCPU_ID 1 -#define PAGE_SIZE 4096 #define MAXPHYADDR 36 #define MEM_REGION_GVA 0x0000123456789000 diff --git a/tools/testing/selftests/kvm/x86_64/smm_test.c b/tools/testing/selftests/kvm/x86_64/smm_test.c index a626d40fdb48..b4e0c860769e 100644 --- a/tools/testing/selftests/kvm/x86_64/smm_test.c +++ b/tools/testing/selftests/kvm/x86_64/smm_test.c @@ -21,8 +21,6 @@ #define VCPU_ID 1 -#define PAGE_SIZE 4096 - #define SMRAM_SIZE 65536 #define SMRAM_MEMSLOT ((1 << 16) | 1) #define SMRAM_PAGES (SMRAM_SIZE / PAGE_SIZE) diff --git a/tools/testing/selftests/kvm/x86_64/vmx_tsc_adjust_test.c b/tools/testing/selftests/kvm/x86_64/vmx_tsc_adjust_test.c index e683d0ac3e45..19b35c607dc6 100644 --- a/tools/testing/selftests/kvm/x86_64/vmx_tsc_adjust_test.c +++ b/tools/testing/selftests/kvm/x86_64/vmx_tsc_adjust_test.c @@ -32,7 +32,6 @@ #define MSR_IA32_TSC_ADJUST 0x3b #endif -#define PAGE_SIZE 4096 #define VCPU_ID 5 #define TSC_ADJUST_VALUE (1ll << 32) diff --git a/tools/testing/selftests/kvm/x86_64/xen_shinfo_test.c b/tools/testing/selftests/kvm/x86_64/xen_shinfo_test.c index 865e17146815..bcd370827859 100644 --- a/tools/testing/selftests/kvm/x86_64/xen_shinfo_test.c +++ b/tools/testing/selftests/kvm/x86_64/xen_shinfo_test.c @@ -23,7 +23,6 @@ #define SHINFO_REGION_GVA 0xc0000000ULL #define SHINFO_REGION_GPA 0xc0000000ULL #define SHINFO_REGION_SLOT 10 -#define PAGE_SIZE 4096 #define DUMMY_REGION_GPA (SHINFO_REGION_GPA + (2 * PAGE_SIZE)) #define DUMMY_REGION_SLOT 11 diff --git a/tools/testing/selftests/kvm/x86_64/xen_vmcall_test.c b/tools/testing/selftests/kvm/x86_64/xen_vmcall_test.c index adc94452b57c..b30fe9de1d4f 100644 --- a/tools/testing/selftests/kvm/x86_64/xen_vmcall_test.c +++ b/tools/testing/selftests/kvm/x86_64/xen_vmcall_test.c @@ -15,7 +15,6 @@ #define HCALL_REGION_GPA 0xc0000000ULL #define HCALL_REGION_SLOT 10 -#define PAGE_SIZE 4096 static struct kvm_vm *vm; -- cgit v1.2.3 From 9c85a9bae267f6b5e5e374d0d023bbbe9db096d3 Mon Sep 17 00:00:00 2001 From: Sidhartha Kumar Date: Thu, 21 Apr 2022 16:35:49 -0700 Subject: selftest/vm: verify mmap addr in mremap_test Avoid calling mmap with requested addresses that are less than the system's mmap_min_addr. When run as root, mmap returns EACCES when trying to map addresses < mmap_min_addr. This is not one of the error codes for the condition to retry the mmap in the test. Rather than arbitrarily retrying on EACCES, don't attempt an mmap until addr > vm.mmap_min_addr. Add a munmap call after an alignment check as the mappings are retained after the retry and can reach the vm.max_map_count sysctl. Link: https://lkml.kernel.org/r/20220420215721.4868-1-sidhartha.kumar@oracle.com Signed-off-by: Sidhartha Kumar Reviewed-by: Shuah Khan Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- tools/testing/selftests/vm/mremap_test.c | 41 +++++++++++++++++++++++++++++++- 1 file changed, 40 insertions(+), 1 deletion(-) (limited to 'tools') diff --git a/tools/testing/selftests/vm/mremap_test.c b/tools/testing/selftests/vm/mremap_test.c index 7c0b0617b9f8..9a518fee6306 100644 --- a/tools/testing/selftests/vm/mremap_test.c +++ b/tools/testing/selftests/vm/mremap_test.c @@ -6,6 +6,7 @@ #include #include +#include #include #include #include @@ -63,6 +64,35 @@ enum { .expect_failure = should_fail \ } +/* Returns mmap_min_addr sysctl tunable from procfs */ +static unsigned long long get_mmap_min_addr(void) +{ + FILE *fp; + int n_matched; + static unsigned long long addr; + + if (addr) + return addr; + + fp = fopen("/proc/sys/vm/mmap_min_addr", "r"); + if (fp == NULL) { + ksft_print_msg("Failed to open /proc/sys/vm/mmap_min_addr: %s\n", + strerror(errno)); + exit(KSFT_SKIP); + } + + n_matched = fscanf(fp, "%llu", &addr); + if (n_matched != 1) { + ksft_print_msg("Failed to read /proc/sys/vm/mmap_min_addr: %s\n", + strerror(errno)); + fclose(fp); + exit(KSFT_SKIP); + } + + fclose(fp); + return addr; +} + /* * Returns the start address of the mapping on success, else returns * NULL on failure. @@ -71,8 +101,15 @@ static void *get_source_mapping(struct config c) { unsigned long long addr = 0ULL; void *src_addr = NULL; + unsigned long long mmap_min_addr; + + mmap_min_addr = get_mmap_min_addr(); + retry: addr += c.src_alignment; + if (addr < mmap_min_addr) + goto retry; + src_addr = mmap((void *) addr, c.region_size, PROT_READ | PROT_WRITE, MAP_FIXED_NOREPLACE | MAP_ANONYMOUS | MAP_SHARED, -1, 0); @@ -90,8 +127,10 @@ retry: * alignment in the tests. */ if (((unsigned long long) src_addr & (c.src_alignment - 1)) || - !((unsigned long long) src_addr & c.src_alignment)) + !((unsigned long long) src_addr & c.src_alignment)) { + munmap(src_addr, c.region_size); goto retry; + } if (!src_addr) goto error; -- cgit v1.2.3 From 18d609daa546c919fd36b62a7b510c18de4b4af8 Mon Sep 17 00:00:00 2001 From: Sidhartha Kumar Date: Thu, 21 Apr 2022 16:35:52 -0700 Subject: selftest/vm: verify remap destination address in mremap_test Because mremap does not have a MAP_FIXED_NOREPLACE flag, it can destroy existing mappings. This causes a segfault when regions such as text are remapped and the permissions are changed. Verify the requested mremap destination address does not overlap any existing mappings by using mmap's MAP_FIXED_NOREPLACE flag. Keep incrementing the destination address until a valid mapping is found or fail the current test once the max address is reached. Link: https://lkml.kernel.org/r/20220420215721.4868-2-sidhartha.kumar@oracle.com Signed-off-by: Sidhartha Kumar Reviewed-by: Shuah Khan Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- tools/testing/selftests/vm/mremap_test.c | 42 +++++++++++++++++++++++++++++--- 1 file changed, 39 insertions(+), 3 deletions(-) (limited to 'tools') diff --git a/tools/testing/selftests/vm/mremap_test.c b/tools/testing/selftests/vm/mremap_test.c index 9a518fee6306..58775dab3cc6 100644 --- a/tools/testing/selftests/vm/mremap_test.c +++ b/tools/testing/selftests/vm/mremap_test.c @@ -10,6 +10,7 @@ #include #include #include +#include #include "../kselftest.h" @@ -64,6 +65,30 @@ enum { .expect_failure = should_fail \ } +/* + * Returns false if the requested remap region overlaps with an + * existing mapping (e.g text, stack) else returns true. + */ +static bool is_remap_region_valid(void *addr, unsigned long long size) +{ + void *remap_addr = NULL; + bool ret = true; + + /* Use MAP_FIXED_NOREPLACE flag to ensure region is not mapped */ + remap_addr = mmap(addr, size, PROT_READ | PROT_WRITE, + MAP_FIXED_NOREPLACE | MAP_ANONYMOUS | MAP_SHARED, + -1, 0); + + if (remap_addr == MAP_FAILED) { + if (errno == EEXIST) + ret = false; + } else { + munmap(remap_addr, size); + } + + return ret; +} + /* Returns mmap_min_addr sysctl tunable from procfs */ static unsigned long long get_mmap_min_addr(void) { @@ -111,8 +136,8 @@ retry: goto retry; src_addr = mmap((void *) addr, c.region_size, PROT_READ | PROT_WRITE, - MAP_FIXED_NOREPLACE | MAP_ANONYMOUS | MAP_SHARED, - -1, 0); + MAP_FIXED_NOREPLACE | MAP_ANONYMOUS | MAP_SHARED, + -1, 0); if (src_addr == MAP_FAILED) { if (errno == EPERM || errno == EEXIST) goto retry; @@ -179,9 +204,20 @@ static long long remap_region(struct config c, unsigned int threshold_mb, if (!((unsigned long long) addr & c.dest_alignment)) addr = (void *) ((unsigned long long) addr | c.dest_alignment); + /* Don't destroy existing mappings unless expected to overlap */ + while (!is_remap_region_valid(addr, c.region_size) && !c.overlapping) { + /* Check for unsigned overflow */ + if (addr + c.dest_alignment < addr) { + ksft_print_msg("Couldn't find a valid region to remap to\n"); + ret = -1; + goto out; + } + addr += c.dest_alignment; + } + clock_gettime(CLOCK_MONOTONIC, &t_start); dest_addr = mremap(src_addr, c.region_size, c.region_size, - MREMAP_MAYMOVE|MREMAP_FIXED, (char *) addr); + MREMAP_MAYMOVE|MREMAP_FIXED, (char *) addr); clock_gettime(CLOCK_MONOTONIC, &t_end); if (dest_addr == MAP_FAILED) { -- cgit v1.2.3 From e5508fc52c76fe42d8bb091fbd7796eeb64b52c4 Mon Sep 17 00:00:00 2001 From: Sidhartha Kumar Date: Thu, 21 Apr 2022 16:35:55 -0700 Subject: selftest/vm: support xfail in mremap_test Use ksft_test_result_xfail for the tests which are expected to fail. Link: https://lkml.kernel.org/r/20220420215721.4868-3-sidhartha.kumar@oracle.com Signed-off-by: Sidhartha Kumar Reviewed-by: Shuah Khan Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- tools/testing/selftests/vm/mremap_test.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'tools') diff --git a/tools/testing/selftests/vm/mremap_test.c b/tools/testing/selftests/vm/mremap_test.c index 58775dab3cc6..db0270127aeb 100644 --- a/tools/testing/selftests/vm/mremap_test.c +++ b/tools/testing/selftests/vm/mremap_test.c @@ -268,7 +268,7 @@ static void run_mremap_test_case(struct test test_case, int *failures, if (remap_time < 0) { if (test_case.expect_failure) - ksft_test_result_pass("%s\n\tExpected mremap failure\n", + ksft_test_result_xfail("%s\n\tExpected mremap failure\n", test_case.name); else { ksft_test_result_fail("%s\n", test_case.name); -- cgit v1.2.3 From 80df2fb95df26c849c6cc137344013cc048a083f Mon Sep 17 00:00:00 2001 From: Sidhartha Kumar Date: Thu, 21 Apr 2022 16:35:58 -0700 Subject: selftest/vm: add skip support to mremap_test Allow the mremap test to be skipped due to errors such as failing to parse the mmap_min_addr sysctl. Link: https://lkml.kernel.org/r/20220420215721.4868-4-sidhartha.kumar@oracle.com Signed-off-by: Sidhartha Kumar Reviewed-by: Shuah Khan Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- tools/testing/selftests/vm/run_vmtests.sh | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-) (limited to 'tools') diff --git a/tools/testing/selftests/vm/run_vmtests.sh b/tools/testing/selftests/vm/run_vmtests.sh index 3b265f140c25..352ba00cf26b 100755 --- a/tools/testing/selftests/vm/run_vmtests.sh +++ b/tools/testing/selftests/vm/run_vmtests.sh @@ -291,11 +291,16 @@ echo "-------------------" echo "running mremap_test" echo "-------------------" ./mremap_test -if [ $? -ne 0 ]; then +ret_val=$? + +if [ $ret_val -eq 0 ]; then + echo "[PASS]" +elif [ $ret_val -eq $ksft_skip ]; then + echo "[SKIP]" + exitcode=$ksft_skip +else echo "[FAIL]" exitcode=1 -else - echo "[PASS]" fi echo "-----------------" -- cgit v1.2.3 From c087c6e7b551b7f208c0b852304f044954cf2bb3 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Sun, 17 Apr 2022 17:03:40 +0200 Subject: objtool: Fix type of reloc::addend Elf{32,64}_Rela::r_addend is of type: Elf{32,64}_Sword, that means that our reloc::addend needs to be long or face tuncation issues when we do elf_rebuild_reloc_section(): - 107: 48 b8 00 00 00 00 00 00 00 00 movabs $0x0,%rax 109: R_X86_64_64 level4_kernel_pgt+0x80000067 + 107: 48 b8 00 00 00 00 00 00 00 00 movabs $0x0,%rax 109: R_X86_64_64 level4_kernel_pgt-0x7fffff99 Fixes: 627fce14809b ("objtool: Add ORC unwind table generation") Signed-off-by: Peter Zijlstra (Intel) Acked-by: Josh Poimboeuf Link: https://lkml.kernel.org/r/20220419203807.596871927@infradead.org --- tools/objtool/check.c | 8 ++++---- tools/objtool/elf.c | 2 +- tools/objtool/include/objtool/elf.h | 4 ++-- 3 files changed, 7 insertions(+), 7 deletions(-) (limited to 'tools') diff --git a/tools/objtool/check.c b/tools/objtool/check.c index 5f10653eb5c2..3f6785415894 100644 --- a/tools/objtool/check.c +++ b/tools/objtool/check.c @@ -559,12 +559,12 @@ static int add_dead_ends(struct objtool_file *file) else if (reloc->addend == reloc->sym->sec->sh.sh_size) { insn = find_last_insn(file, reloc->sym->sec); if (!insn) { - WARN("can't find unreachable insn at %s+0x%x", + WARN("can't find unreachable insn at %s+0x%lx", reloc->sym->sec->name, reloc->addend); return -1; } } else { - WARN("can't find unreachable insn at %s+0x%x", + WARN("can't find unreachable insn at %s+0x%lx", reloc->sym->sec->name, reloc->addend); return -1; } @@ -594,12 +594,12 @@ reachable: else if (reloc->addend == reloc->sym->sec->sh.sh_size) { insn = find_last_insn(file, reloc->sym->sec); if (!insn) { - WARN("can't find reachable insn at %s+0x%x", + WARN("can't find reachable insn at %s+0x%lx", reloc->sym->sec->name, reloc->addend); return -1; } } else { - WARN("can't find reachable insn at %s+0x%x", + WARN("can't find reachable insn at %s+0x%lx", reloc->sym->sec->name, reloc->addend); return -1; } diff --git a/tools/objtool/elf.c b/tools/objtool/elf.c index d7b99a737496..0cfe84ac4cdb 100644 --- a/tools/objtool/elf.c +++ b/tools/objtool/elf.c @@ -546,7 +546,7 @@ static struct section *elf_create_reloc_section(struct elf *elf, int reltype); int elf_add_reloc(struct elf *elf, struct section *sec, unsigned long offset, - unsigned int type, struct symbol *sym, int addend) + unsigned int type, struct symbol *sym, long addend) { struct reloc *reloc; diff --git a/tools/objtool/include/objtool/elf.h b/tools/objtool/include/objtool/elf.h index 22ba7e2b816e..9b36802ed86f 100644 --- a/tools/objtool/include/objtool/elf.h +++ b/tools/objtool/include/objtool/elf.h @@ -73,7 +73,7 @@ struct reloc { struct symbol *sym; unsigned long offset; unsigned int type; - int addend; + long addend; int idx; bool jump_table_start; }; @@ -135,7 +135,7 @@ struct elf *elf_open_read(const char *name, int flags); struct section *elf_create_section(struct elf *elf, const char *name, unsigned int sh_flags, size_t entsize, int nr); int elf_add_reloc(struct elf *elf, struct section *sec, unsigned long offset, - unsigned int type, struct symbol *sym, int addend); + unsigned int type, struct symbol *sym, long addend); int elf_add_reloc_to_insn(struct elf *elf, struct section *sec, unsigned long offset, unsigned int type, struct section *insn_sec, unsigned long insn_off); -- cgit v1.2.3 From 4abff6d48dbcea8200c7ea35ba70c242d128ebf3 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Sun, 17 Apr 2022 17:03:36 +0200 Subject: objtool: Fix code relocs vs weak symbols Occasionally objtool driven code patching (think .static_call_sites .retpoline_sites etc..) goes sideways and it tries to patch an instruction that doesn't match. Much head-scatching and cursing later the problem is as outlined below and affects every section that objtool generates for us, very much including the ORC data. The below uses .static_call_sites because it's convenient for demonstration purposes, but as mentioned the ORC sections, .retpoline_sites and __mount_loc are all similarly affected. Consider: foo-weak.c: extern void __SCT__foo(void); __attribute__((weak)) void foo(void) { return __SCT__foo(); } foo.c: extern void __SCT__foo(void); extern void my_foo(void); void foo(void) { my_foo(); return __SCT__foo(); } These generate the obvious code (gcc -O2 -fcf-protection=none -fno-asynchronous-unwind-tables -c foo*.c): foo-weak.o: 0000000000000000 : 0: e9 00 00 00 00 jmpq 5 1: R_X86_64_PLT32 __SCT__foo-0x4 foo.o: 0000000000000000 : 0: 48 83 ec 08 sub $0x8,%rsp 4: e8 00 00 00 00 callq 9 5: R_X86_64_PLT32 my_foo-0x4 9: 48 83 c4 08 add $0x8,%rsp d: e9 00 00 00 00 jmpq 12 e: R_X86_64_PLT32 __SCT__foo-0x4 Now, when we link these two files together, you get something like (ld -r -o foos.o foo-weak.o foo.o): foos.o: 0000000000000000 : 0: e9 00 00 00 00 jmpq 5 1: R_X86_64_PLT32 __SCT__foo-0x4 5: 66 2e 0f 1f 84 00 00 00 00 00 nopw %cs:0x0(%rax,%rax,1) f: 90 nop 0000000000000010 : 10: 48 83 ec 08 sub $0x8,%rsp 14: e8 00 00 00 00 callq 19 15: R_X86_64_PLT32 my_foo-0x4 19: 48 83 c4 08 add $0x8,%rsp 1d: e9 00 00 00 00 jmpq 22 1e: R_X86_64_PLT32 __SCT__foo-0x4 Noting that ld preserves the weak function text, but strips the symbol off of it (hence objdump doing that funny negative offset thing). This does lead to 'interesting' unused code issues with objtool when ran on linked objects, but that seems to be working (fingers crossed). So far so good.. Now lets consider the objtool static_call output section (readelf output, old binutils): foo-weak.o: Relocation section '.rela.static_call_sites' at offset 0x2c8 contains 1 entry: Offset Info Type Symbol's Value Symbol's Name + Addend 0000000000000000 0000000200000002 R_X86_64_PC32 0000000000000000 .text + 0 0000000000000004 0000000d00000002 R_X86_64_PC32 0000000000000000 __SCT__foo + 1 foo.o: Relocation section '.rela.static_call_sites' at offset 0x310 contains 2 entries: Offset Info Type Symbol's Value Symbol's Name + Addend 0000000000000000 0000000200000002 R_X86_64_PC32 0000000000000000 .text + d 0000000000000004 0000000d00000002 R_X86_64_PC32 0000000000000000 __SCT__foo + 1 foos.o: Relocation section '.rela.static_call_sites' at offset 0x430 contains 4 entries: Offset Info Type Symbol's Value Symbol's Name + Addend 0000000000000000 0000000100000002 R_X86_64_PC32 0000000000000000 .text + 0 0000000000000004 0000000d00000002 R_X86_64_PC32 0000000000000000 __SCT__foo + 1 0000000000000008 0000000100000002 R_X86_64_PC32 0000000000000000 .text + 1d 000000000000000c 0000000d00000002 R_X86_64_PC32 0000000000000000 __SCT__foo + 1 So we have two patch sites, one in the dead code of the weak foo and one in the real foo. All is well. *HOWEVER*, when the toolchain strips unused section symbols it generates things like this (using new enough binutils): foo-weak.o: Relocation section '.rela.static_call_sites' at offset 0x2c8 contains 1 entry: Offset Info Type Symbol's Value Symbol's Name + Addend 0000000000000000 0000000200000002 R_X86_64_PC32 0000000000000000 foo + 0 0000000000000004 0000000d00000002 R_X86_64_PC32 0000000000000000 __SCT__foo + 1 foo.o: Relocation section '.rela.static_call_sites' at offset 0x310 contains 2 entries: Offset Info Type Symbol's Value Symbol's Name + Addend 0000000000000000 0000000200000002 R_X86_64_PC32 0000000000000000 foo + d 0000000000000004 0000000d00000002 R_X86_64_PC32 0000000000000000 __SCT__foo + 1 foos.o: Relocation section '.rela.static_call_sites' at offset 0x430 contains 4 entries: Offset Info Type Symbol's Value Symbol's Name + Addend 0000000000000000 0000000100000002 R_X86_64_PC32 0000000000000000 foo + 0 0000000000000004 0000000d00000002 R_X86_64_PC32 0000000000000000 __SCT__foo + 1 0000000000000008 0000000100000002 R_X86_64_PC32 0000000000000000 foo + d 000000000000000c 0000000d00000002 R_X86_64_PC32 0000000000000000 __SCT__foo + 1 And now we can see how that foos.o .static_call_sites goes side-ways, we now have _two_ patch sites in foo. One for the weak symbol at foo+0 (which is no longer a static_call site!) and one at foo+d which is in fact the right location. This seems to happen when objtool cannot find a section symbol, in which case it falls back to any other symbol to key off of, however in this case that goes terribly wrong! As such, teach objtool to create a section symbol when there isn't one. Fixes: 44f6a7c0755d ("objtool: Fix seg fault with Clang non-section symbols") Signed-off-by: Peter Zijlstra (Intel) Acked-by: Josh Poimboeuf Link: https://lkml.kernel.org/r/20220419203807.655552918@infradead.org --- tools/objtool/elf.c | 187 +++++++++++++++++++++++++++++++++++++++++++++------- 1 file changed, 165 insertions(+), 22 deletions(-) (limited to 'tools') diff --git a/tools/objtool/elf.c b/tools/objtool/elf.c index 0cfe84ac4cdb..ebf2ba5755c1 100644 --- a/tools/objtool/elf.c +++ b/tools/objtool/elf.c @@ -575,37 +575,180 @@ int elf_add_reloc(struct elf *elf, struct section *sec, unsigned long offset, return 0; } -int elf_add_reloc_to_insn(struct elf *elf, struct section *sec, - unsigned long offset, unsigned int type, - struct section *insn_sec, unsigned long insn_off) +/* + * Ensure that any reloc section containing references to @sym is marked + * changed such that it will get re-generated in elf_rebuild_reloc_sections() + * with the new symbol index. + */ +static void elf_dirty_reloc_sym(struct elf *elf, struct symbol *sym) +{ + struct section *sec; + + list_for_each_entry(sec, &elf->sections, list) { + struct reloc *reloc; + + if (sec->changed) + continue; + + list_for_each_entry(reloc, &sec->reloc_list, list) { + if (reloc->sym == sym) { + sec->changed = true; + break; + } + } + } +} + +/* + * Move the first global symbol, as per sh_info, into a new, higher symbol + * index. This fees up the shndx for a new local symbol. + */ +static int elf_move_global_symbol(struct elf *elf, struct section *symtab, + struct section *symtab_shndx) { + Elf_Data *data, *shndx_data = NULL; + Elf32_Word first_non_local; struct symbol *sym; - int addend; + Elf_Scn *s; - if (insn_sec->sym) { - sym = insn_sec->sym; - addend = insn_off; + first_non_local = symtab->sh.sh_info; - } else { - /* - * The Clang assembler strips section symbols, so we have to - * reference the function symbol instead: - */ - sym = find_symbol_containing(insn_sec, insn_off); - if (!sym) { - /* - * Hack alert. This happens when we need to reference - * the NOP pad insn immediately after the function. - */ - sym = find_symbol_containing(insn_sec, insn_off - 1); + sym = find_symbol_by_index(elf, first_non_local); + if (!sym) { + WARN("no non-local symbols !?"); + return first_non_local; + } + + s = elf_getscn(elf->elf, symtab->idx); + if (!s) { + WARN_ELF("elf_getscn"); + return -1; + } + + data = elf_newdata(s); + if (!data) { + WARN_ELF("elf_newdata"); + return -1; + } + + data->d_buf = &sym->sym; + data->d_size = sizeof(sym->sym); + data->d_align = 1; + data->d_type = ELF_T_SYM; + + sym->idx = symtab->sh.sh_size / sizeof(sym->sym); + elf_dirty_reloc_sym(elf, sym); + + symtab->sh.sh_info += 1; + symtab->sh.sh_size += data->d_size; + symtab->changed = true; + + if (symtab_shndx) { + s = elf_getscn(elf->elf, symtab_shndx->idx); + if (!s) { + WARN_ELF("elf_getscn"); + return -1; } - if (!sym) { - WARN("can't find symbol containing %s+0x%lx", insn_sec->name, insn_off); + shndx_data = elf_newdata(s); + if (!shndx_data) { + WARN_ELF("elf_newshndx_data"); return -1; } - addend = insn_off - sym->offset; + shndx_data->d_buf = &sym->sec->idx; + shndx_data->d_size = sizeof(Elf32_Word); + shndx_data->d_align = 4; + shndx_data->d_type = ELF_T_WORD; + + symtab_shndx->sh.sh_size += 4; + symtab_shndx->changed = true; + } + + return first_non_local; +} + +static struct symbol * +elf_create_section_symbol(struct elf *elf, struct section *sec) +{ + struct section *symtab, *symtab_shndx; + Elf_Data *shndx_data = NULL; + struct symbol *sym; + Elf32_Word shndx; + + symtab = find_section_by_name(elf, ".symtab"); + if (symtab) { + symtab_shndx = find_section_by_name(elf, ".symtab_shndx"); + if (symtab_shndx) + shndx_data = symtab_shndx->data; + } else { + WARN("no .symtab"); + return NULL; + } + + sym = malloc(sizeof(*sym)); + if (!sym) { + perror("malloc"); + return NULL; + } + memset(sym, 0, sizeof(*sym)); + + sym->idx = elf_move_global_symbol(elf, symtab, symtab_shndx); + if (sym->idx < 0) { + WARN("elf_move_global_symbol"); + return NULL; + } + + sym->name = sec->name; + sym->sec = sec; + + // st_name 0 + sym->sym.st_info = GELF_ST_INFO(STB_LOCAL, STT_SECTION); + // st_other 0 + // st_value 0 + // st_size 0 + shndx = sec->idx; + if (shndx >= SHN_UNDEF && shndx < SHN_LORESERVE) { + sym->sym.st_shndx = shndx; + if (!shndx_data) + shndx = 0; + } else { + sym->sym.st_shndx = SHN_XINDEX; + if (!shndx_data) { + WARN("no .symtab_shndx"); + return NULL; + } + } + + if (!gelf_update_symshndx(symtab->data, shndx_data, sym->idx, &sym->sym, shndx)) { + WARN_ELF("gelf_update_symshndx"); + return NULL; + } + + elf_add_symbol(elf, sym); + + return sym; +} + +int elf_add_reloc_to_insn(struct elf *elf, struct section *sec, + unsigned long offset, unsigned int type, + struct section *insn_sec, unsigned long insn_off) +{ + struct symbol *sym = insn_sec->sym; + int addend = insn_off; + + if (!sym) { + /* + * Due to how weak functions work, we must use section based + * relocations. Symbol based relocations would result in the + * weak and non-weak function annotations being overlaid on the + * non-weak function after linking. + */ + sym = elf_create_section_symbol(elf, insn_sec); + if (!sym) + return -1; + + insn_sec->sym = sym; } return elf_add_reloc(elf, sec, offset, type, sym, addend); -- cgit v1.2.3 From b9663a6ff8289a095d56d9a3a3f9c185a7b7b0d7 Mon Sep 17 00:00:00 2001 From: "Matthew Wilcox (Oracle)" Date: Fri, 22 Apr 2022 13:20:21 -0400 Subject: tools: Add kmem_cache_alloc_lru() Turn kmem_cache_alloc() into a wrapper around kmem_cache_alloc_lru(). Fixes: 9bbdc0f32409 ("xarray: use kmem_cache_alloc_lru to allocate xa_node") Signed-off-by: Matthew Wilcox (Oracle) Reported-by: Liam R. Howlett Reported-by: Li Wang --- tools/include/linux/slab.h | 8 +++++++- tools/testing/radix-tree/linux.c | 3 ++- 2 files changed, 9 insertions(+), 2 deletions(-) (limited to 'tools') diff --git a/tools/include/linux/slab.h b/tools/include/linux/slab.h index f41d8a0eb1a4..0616409513eb 100644 --- a/tools/include/linux/slab.h +++ b/tools/include/linux/slab.h @@ -28,7 +28,13 @@ static inline void *kzalloc(size_t size, gfp_t gfp) return kmalloc(size, gfp | __GFP_ZERO); } -void *kmem_cache_alloc(struct kmem_cache *cachep, int flags); +struct list_lru; + +void *kmem_cache_alloc_lru(struct kmem_cache *cachep, struct list_lru *, int flags); +static inline void *kmem_cache_alloc(struct kmem_cache *cachep, int flags) +{ + return kmem_cache_alloc_lru(cachep, NULL, flags); +} void kmem_cache_free(struct kmem_cache *cachep, void *objp); struct kmem_cache *kmem_cache_create(const char *name, unsigned int size, diff --git a/tools/testing/radix-tree/linux.c b/tools/testing/radix-tree/linux.c index 81539f543954..d5c1bcba86fe 100644 --- a/tools/testing/radix-tree/linux.c +++ b/tools/testing/radix-tree/linux.c @@ -25,7 +25,8 @@ struct kmem_cache { void (*ctor)(void *); }; -void *kmem_cache_alloc(struct kmem_cache *cachep, int gfp) +void *kmem_cache_alloc_lru(struct kmem_cache *cachep, struct list_lru *lru, + int gfp) { void *p; -- cgit v1.2.3 From d22588d73b18fd12fd971e2dab7fa5ddf978e496 Mon Sep 17 00:00:00 2001 From: Guilherme Amadio Date: Sat, 16 Apr 2022 09:45:55 +0200 Subject: perf clang: Fix header include for LLVM >= 14 The header TargetRegistry.h has moved in LLVM/clang 14. Committer notes: The problem as noticed when building in ubuntu:22.04: 90 98.61 ubuntu:22.04 : FAIL gcc version 11.2.0 (Ubuntu 11.2.0-19ubuntu1) util/c++/clang.cpp:23:10: fatal error: llvm/Support/TargetRegistry.h: No such file or directory 23 | #include "llvm/Support/TargetRegistry.h" | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ compilation terminated. Fixed after applying this patch. Reported-by: Arnaldo Carvalho de Melo Signed-off-by: Guilherme Amadio Tested-by: Arnaldo Carvalho de Melo Link: https://twitter.com/GuilhermeAmadio/status/1514970524232921088 Link: http://lore.kernel.org/lkml/Ylp0M/VYgHOxtcnF@gentoo.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/c++/clang.cpp | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'tools') diff --git a/tools/perf/util/c++/clang.cpp b/tools/perf/util/c++/clang.cpp index df7b18fb6b6e..1aad7d6d34aa 100644 --- a/tools/perf/util/c++/clang.cpp +++ b/tools/perf/util/c++/clang.cpp @@ -20,7 +20,11 @@ #include "llvm/Option/Option.h" #include "llvm/Support/FileSystem.h" #include "llvm/Support/ManagedStatic.h" +#if CLANG_VERSION_MAJOR >= 14 +#include "llvm/MC/TargetRegistry.h" +#else #include "llvm/Support/TargetRegistry.h" +#endif #include "llvm/Support/TargetSelect.h" #include "llvm/Target/TargetMachine.h" #include "llvm/Target/TargetOptions.h" -- cgit v1.2.3 From c6d8df01064333dcf140eda996abdb60a60e24b3 Mon Sep 17 00:00:00 2001 From: Leo Yan Date: Sun, 17 Apr 2022 19:48:37 +0800 Subject: perf script: Always allow field 'data_src' for auxtrace If use command 'perf script -F,+data_src' to dump memory samples with Arm SPE trace data, it reports error: # perf script -F,+data_src Samples for 'dummy:u' event do not have DATA_SRC attribute set. Cannot print 'data_src' field. This is because the 'dummy:u' event is absent DATA_SRC bit in its sample type, so if a file contains AUX area tracing data then always allow field 'data_src' to be selected as an option for perf script. Fixes: e55ed3423c1bb29f ("perf arm-spe: Synthesize memory event") Signed-off-by: Leo Yan Cc: Adrian Hunter Cc: Alexander Shishkin Cc: German Gomez Cc: Ingo Molnar Cc: James Clark Cc: Jiri Olsa Cc: Leo Yan Cc: Mark Rutland Cc: Namhyung Kim Cc: Peter Zijlstra Link: https://lore.kernel.org/r/20220417114837.839896-1-leo.yan@linaro.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-script.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'tools') diff --git a/tools/perf/builtin-script.c b/tools/perf/builtin-script.c index a2f117936188..cf5eab5431b4 100644 --- a/tools/perf/builtin-script.c +++ b/tools/perf/builtin-script.c @@ -461,7 +461,7 @@ static int evsel__check_attr(struct evsel *evsel, struct perf_session *session) return -EINVAL; if (PRINT_FIELD(DATA_SRC) && - evsel__check_stype(evsel, PERF_SAMPLE_DATA_SRC, "DATA_SRC", PERF_OUTPUT_DATA_SRC)) + evsel__do_check_stype(evsel, PERF_SAMPLE_DATA_SRC, "DATA_SRC", PERF_OUTPUT_DATA_SRC, allow_user_set)) return -EINVAL; if (PRINT_FIELD(WEIGHT) && -- cgit v1.2.3 From ccb17caecfbd542f49a2a79ae088136ba8bfb794 Mon Sep 17 00:00:00 2001 From: Leo Yan Date: Thu, 14 Apr 2022 20:32:01 +0800 Subject: perf report: Set PERF_SAMPLE_DATA_SRC bit for Arm SPE event Since commit bb30acae4c4dacfa ("perf report: Bail out --mem-mode if mem info is not available") "perf mem report" and "perf report --mem-mode" don't report result if the PERF_SAMPLE_DATA_SRC bit is missed in sample type. The commit ffab487052054162 ("perf: arm-spe: Fix perf report --mem-mode") partially fixes the issue. It adds PERF_SAMPLE_DATA_SRC bit for Arm SPE event, this allows the perf data file generated by kernel v5.18-rc1 or later version can be reported properly. On the other hand, perf tool still fails to be backward compatibility for a data file recorded by an older version's perf which contains Arm SPE trace data. This patch is a workaround in reporting phase, when detects ARM SPE PMU event and without PERF_SAMPLE_DATA_SRC bit, it will force to set the bit in the sample type and give a warning info. Fixes: bb30acae4c4dacfa ("perf report: Bail out --mem-mode if mem info is not available") Reviewed-by: James Clark Signed-off-by: Leo Yan Tested-by: German Gomez Cc: Alexander Shishkin Cc: Ingo Molnar Cc: Jiri Olsa Cc: Mark Rutland Cc: Namhyung Kim Cc: Peter Zijlstra Cc: Ravi Bangoria Link: https://lore.kernel.org/r/20220414123201.842754-1-leo.yan@linaro.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-report.c | 14 ++++++++++++++ 1 file changed, 14 insertions(+) (limited to 'tools') diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c index 1ad75c7ba074..afe4a5539ecc 100644 --- a/tools/perf/builtin-report.c +++ b/tools/perf/builtin-report.c @@ -353,6 +353,7 @@ static int report__setup_sample_type(struct report *rep) struct perf_session *session = rep->session; u64 sample_type = evlist__combined_sample_type(session->evlist); bool is_pipe = perf_data__is_pipe(session->data); + struct evsel *evsel; if (session->itrace_synth_opts->callchain || session->itrace_synth_opts->add_callchain || @@ -407,6 +408,19 @@ static int report__setup_sample_type(struct report *rep) } if (sort__mode == SORT_MODE__MEMORY) { + /* + * FIXUP: prior to kernel 5.18, Arm SPE missed to set + * PERF_SAMPLE_DATA_SRC bit in sample type. For backward + * compatibility, set the bit if it's an old perf data file. + */ + evlist__for_each_entry(session->evlist, evsel) { + if (strstr(evsel->name, "arm_spe") && + !(sample_type & PERF_SAMPLE_DATA_SRC)) { + evsel->core.attr.sample_type |= PERF_SAMPLE_DATA_SRC; + sample_type |= PERF_SAMPLE_DATA_SRC; + } + } + if (!is_pipe && !(sample_type & PERF_SAMPLE_DATA_SRC)) { ui__error("Selected --mem-mode but no mem data. " "Did you call perf record without -d?\n"); -- cgit v1.2.3 From 5bb017d4b97a0f135f43ef77091b7edcce4dcee6 Mon Sep 17 00:00:00 2001 From: Thomas Richter Date: Wed, 20 Apr 2022 08:29:21 +0200 Subject: perf test: Fix error message for test case 71 on s390, where it is not supported Test case 71 'Convert perf time to TSC' is not supported on s390. Subtest 71.1 is skipped with the correct message, but subtest 71.2 is not skipped and fails. The root cause is function evlist__open() called from test__perf_time_to_tsc(). evlist__open() returns -ENOENT because the event cycles:u is not supported by the selected PMU, for example platform s390 on z/VM or an x86_64 virtual machine. The PMU driver returns -ENOENT in this case. This error is leads to the failure. Fix this by returning TEST_SKIP on -ENOENT. Output before: 71: Convert perf time to TSC: 71.1: TSC support: Skip (This architecture does not support) 71.2: Perf time to TSC: FAILED! Output after: 71: Convert perf time to TSC: 71.1: TSC support: Skip (This architecture does not support) 71.2: Perf time to TSC: Skip (perf_read_tsc_conversion is not supported) This also happens on an x86_64 virtual machine: # uname -m x86_64 $ ./perf test -F 71 71: Convert perf time to TSC : 71.1: TSC support : Ok 71.2: Perf time to TSC : FAILED! $ Committer testing: Continues to work on x86_64: $ perf test 71 71: Convert perf time to TSC : 71.1: TSC support : Ok 71.2: Perf time to TSC : Ok $ Fixes: 290fa68bdc458863 ("perf test tsc: Fix error message when not supported") Signed-off-by: Thomas Richter Acked-by: Sumanth Korikkar Tested-by: Arnaldo Carvalho de Melo Cc: Adrian Hunter Cc: Chengdong Li Cc: chengdongli@tencent.com Cc: Heiko Carstens Cc: Sven Schnelle Cc: Vasily Gorbik Link: https://lore.kernel.org/r/20220420062921.1211825-1-tmricht@linux.ibm.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/tests/perf-time-to-tsc.c | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'tools') diff --git a/tools/perf/tests/perf-time-to-tsc.c b/tools/perf/tests/perf-time-to-tsc.c index cc6df49a65a1..4ad0dfbc8b21 100644 --- a/tools/perf/tests/perf-time-to-tsc.c +++ b/tools/perf/tests/perf-time-to-tsc.c @@ -123,6 +123,10 @@ static int test__perf_time_to_tsc(struct test_suite *test __maybe_unused, int su evsel->core.attr.enable_on_exec = 0; } + if (evlist__open(evlist) == -ENOENT) { + err = TEST_SKIP; + goto out_err; + } CHECK__(evlist__open(evlist)); CHECK__(evlist__mmap(evlist, UINT_MAX)); -- cgit v1.2.3 From 00f3d2ed9dac8fc8674a021765a0772f74c6127b Mon Sep 17 00:00:00 2001 From: "Jason A. Donenfeld" Date: Thu, 21 Apr 2022 15:48:04 +0200 Subject: wireguard: selftests: enable ACPI for SMP It turns out that by having CONFIG_ACPI=n, we've been failing to boot additional CPUs, and so these systems were functionally UP. The code bloat is unfortunate for build times, but I don't see an alternative. So this commit sets CONFIG_ACPI=y for x86_64 and i686 configs. Signed-off-by: Jason A. Donenfeld Signed-off-by: Jakub Kicinski --- tools/testing/selftests/wireguard/qemu/arch/i686.config | 1 + tools/testing/selftests/wireguard/qemu/arch/x86_64.config | 1 + 2 files changed, 2 insertions(+) (limited to 'tools') diff --git a/tools/testing/selftests/wireguard/qemu/arch/i686.config b/tools/testing/selftests/wireguard/qemu/arch/i686.config index a85025d7206e..a9b4fe795048 100644 --- a/tools/testing/selftests/wireguard/qemu/arch/i686.config +++ b/tools/testing/selftests/wireguard/qemu/arch/i686.config @@ -1,3 +1,4 @@ +CONFIG_ACPI=y CONFIG_SERIAL_8250=y CONFIG_SERIAL_8250_CONSOLE=y CONFIG_CMDLINE_BOOL=y diff --git a/tools/testing/selftests/wireguard/qemu/arch/x86_64.config b/tools/testing/selftests/wireguard/qemu/arch/x86_64.config index 00a1ef4869d5..45dd53a0d760 100644 --- a/tools/testing/selftests/wireguard/qemu/arch/x86_64.config +++ b/tools/testing/selftests/wireguard/qemu/arch/x86_64.config @@ -1,3 +1,4 @@ +CONFIG_ACPI=y CONFIG_SERIAL_8250=y CONFIG_SERIAL_8250_CONSOLE=y CONFIG_CMDLINE_BOOL=y -- cgit v1.2.3 From de8fd138430ccac4b8f7b812e5c6f8963b5ccf07 Mon Sep 17 00:00:00 2001 From: Adrian Hunter Date: Thu, 28 Apr 2022 12:31:09 +0300 Subject: perf intel-pt: Fix timeless decoding with perf.data directory Intel PT does not capture data in separate directories, so do not use separate directory processing because it doesn't work for timeless decoding. It also looks like it doesn't support one_mmap handling. Example: Before: # perf record --kcore -a -e intel_pt/tsc=0/k sleep 0.1 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 1.799 MB perf.data ] # perf script --itrace=bep | head # After: # perf script --itrace=bep | head perf 21073 [000] psb: psb offs: 0 ffffffffaa68faf4 native_write_msr+0x4 ([kernel.kallsyms]) perf 21073 [000] cbr: cbr: 45 freq: 4505 MHz (161%) ffffffffaa68faf4 native_write_msr+0x4 ([kernel.kallsyms]) perf 21073 [000] 1 branches:k: 0 [unknown] ([unknown]) => ffffffffaa68faf6 native_write_msr+0x6 ([kernel.kallsyms]) perf 21073 [000] 1 branches:k: ffffffffaa68faf8 native_write_msr+0x8 ([kernel.kallsyms]) => ffffffffaa61aab0 pt_config_start+0x60 ([kernel.kallsyms]) perf 21073 [000] 1 branches:k: ffffffffaa61aabd pt_config_start+0x6d ([kernel.kallsyms]) => ffffffffaa61b8ad pt_event_start+0x27d ([kernel.kallsyms]) perf 21073 [000] 1 branches:k: ffffffffaa61b8bb pt_event_start+0x28b ([kernel.kallsyms]) => ffffffffaa61ba60 pt_event_add+0x40 ([kernel.kallsyms]) perf 21073 [000] 1 branches:k: ffffffffaa61ba76 pt_event_add+0x56 ([kernel.kallsyms]) => ffffffffaa880e86 event_sched_in+0xc6 ([kernel.kallsyms]) perf 21073 [000] 1 branches:k: ffffffffaa880e9b event_sched_in+0xdb ([kernel.kallsyms]) => ffffffffaa880ea5 event_sched_in+0xe5 ([kernel.kallsyms]) perf 21073 [000] 1 branches:k: ffffffffaa880eba event_sched_in+0xfa ([kernel.kallsyms]) => ffffffffaa880f96 event_sched_in+0x1d6 ([kernel.kallsyms]) perf 21073 [000] 1 branches:k: ffffffffaa880fc8 event_sched_in+0x208 ([kernel.kallsyms]) => ffffffffaa880ec0 event_sched_in+0x100 ([kernel.kallsyms]) Fixes: bb6be405c4a2a5 ("perf session: Load data directory files for analysis") Cc: stable@vger.kernel.org Signed-off-by: Adrian Hunter Signed-off-by: Arnaldo Carvalho de Melo Link: https://lore.kernel.org/r/20220428093109.274641-1-adrian.hunter@intel.com Cc: Ian Rogers Cc: Arnaldo Carvalho de Melo Cc: Namhyung Kim Cc: Alexey Bayduraev Cc: Jiri Olsa Cc: linux-kernel@vger.kernel.org --- tools/perf/util/session.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'tools') diff --git a/tools/perf/util/session.c b/tools/perf/util/session.c index 45a30040ec8d..f9a320694b85 100644 --- a/tools/perf/util/session.c +++ b/tools/perf/util/session.c @@ -2576,7 +2576,7 @@ int perf_session__process_events(struct perf_session *session) if (perf_data__is_pipe(session->data)) return __perf_session__process_pipe_events(session); - if (perf_data__is_dir(session->data)) + if (perf_data__is_dir(session->data) && session->data->dir.nr) return __perf_session__process_dir_events(session); return __perf_session__process_events(session); -- cgit v1.2.3 From 4e13f6706d5aee1a6b835a44f6cf4971a921dcb8 Mon Sep 17 00:00:00 2001 From: Timothy Hayes Date: Thu, 21 Apr 2022 17:52:03 +0100 Subject: perf arm-spe: Fix addresses of synthesized SPE events This patch corrects a bug whereby synthesized events from SPE samples are missing virtual addresses. Fixes: 54f7815efef7fad9 ("perf arm-spe: Fill address info for samples") Reviewed-by: Leo Yan Signed-off-by: Timothy Hayes Cc: Alexander Shishkin Cc: bpf@vger.kernel.org Cc: Jiri Olsa Cc: John Fastabend Cc: John Garry Cc: KP Singh Cc: Leo Yan Cc: linux-arm-kernel@lists.infradead.org Cc: Mark Rutland Cc: Martin KaFai Lau Cc: Mathieu Poirier Cc: Namhyung Kim Cc: netdev@vger.kernel.org Cc: Song Liu Cc: Will Deacon Cc: Yonghong Song Link: https://lore.kernel.org/r/20220421165205.117662-2-timothy.hayes@arm.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/arm-spe.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'tools') diff --git a/tools/perf/util/arm-spe.c b/tools/perf/util/arm-spe.c index d2b64e3f588b..151cc38a171c 100644 --- a/tools/perf/util/arm-spe.c +++ b/tools/perf/util/arm-spe.c @@ -1036,7 +1036,7 @@ arm_spe_synth_events(struct arm_spe *spe, struct perf_session *session) attr.sample_type = evsel->core.attr.sample_type & PERF_SAMPLE_MASK; attr.sample_type |= PERF_SAMPLE_IP | PERF_SAMPLE_TID | PERF_SAMPLE_PERIOD | PERF_SAMPLE_DATA_SRC | - PERF_SAMPLE_WEIGHT; + PERF_SAMPLE_WEIGHT | PERF_SAMPLE_ADDR; if (spe->timeless_decoding) attr.sample_type &= ~(u64)PERF_SAMPLE_TIME; else -- cgit v1.2.3 From 7599b70a3c85357d3a57319ae90d419adea0544b Mon Sep 17 00:00:00 2001 From: Timothy Hayes Date: Thu, 21 Apr 2022 17:52:04 +0100 Subject: perf arm-spe: Fix SPE events with phys addresses This patch corrects a bug whereby SPE collection is invoked with pa_enable=1 but synthesized events fail to show physical addresses. Reviewed-by: Leo Yan Signed-off-by: Timothy Hayes Cc: Alexander Shishkin Cc: Jiri Olsa Cc: John Fastabend Cc: John Garry Cc: KP Singh Cc: Mark Rutland Cc: Martin KaFai Lau Cc: Mathieu Poirier Cc: Namhyung Kim Cc: Song Liu Cc: Will Deacon Cc: Yonghong Song Cc: bpf@vger.kernel.org Cc: linux-arm-kernel@lists.infradead.org Cc: netdev@vger.kernel.org Link: https://lore.kernel.org/r/20220421165205.117662-3-timothy.hayes@arm.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/arch/arm64/util/arm-spe.c | 10 ++++++++++ tools/perf/util/arm-spe.c | 3 ++- 2 files changed, 12 insertions(+), 1 deletion(-) (limited to 'tools') diff --git a/tools/perf/arch/arm64/util/arm-spe.c b/tools/perf/arch/arm64/util/arm-spe.c index af4d63af8072..e8b577d33e53 100644 --- a/tools/perf/arch/arm64/util/arm-spe.c +++ b/tools/perf/arch/arm64/util/arm-spe.c @@ -148,6 +148,7 @@ static int arm_spe_recording_options(struct auxtrace_record *itr, bool privileged = perf_event_paranoid_check(-1); struct evsel *tracking_evsel; int err; + u64 bit; sper->evlist = evlist; @@ -245,6 +246,15 @@ static int arm_spe_recording_options(struct auxtrace_record *itr, */ evsel__set_sample_bit(arm_spe_evsel, DATA_SRC); + /* + * The PHYS_ADDR flag does not affect the driver behaviour, it is used to + * inform that the resulting output's SPE samples contain physical addresses + * where applicable. + */ + bit = perf_pmu__format_bits(&arm_spe_pmu->format, "pa_enable"); + if (arm_spe_evsel->core.attr.config & bit) + evsel__set_sample_bit(arm_spe_evsel, PHYS_ADDR); + /* Add dummy event to keep tracking */ err = parse_events(evlist, "dummy:u", NULL); if (err) diff --git a/tools/perf/util/arm-spe.c b/tools/perf/util/arm-spe.c index 151cc38a171c..1a80151baed9 100644 --- a/tools/perf/util/arm-spe.c +++ b/tools/perf/util/arm-spe.c @@ -1033,7 +1033,8 @@ arm_spe_synth_events(struct arm_spe *spe, struct perf_session *session) memset(&attr, 0, sizeof(struct perf_event_attr)); attr.size = sizeof(struct perf_event_attr); attr.type = PERF_TYPE_HARDWARE; - attr.sample_type = evsel->core.attr.sample_type & PERF_SAMPLE_MASK; + attr.sample_type = evsel->core.attr.sample_type & + (PERF_SAMPLE_MASK | PERF_SAMPLE_PHYS_ADDR); attr.sample_type |= PERF_SAMPLE_IP | PERF_SAMPLE_TID | PERF_SAMPLE_PERIOD | PERF_SAMPLE_DATA_SRC | PERF_SAMPLE_WEIGHT | PERF_SAMPLE_ADDR; -- cgit v1.2.3 From 3b9a8c8b9ac2a57df8b7f05c6ffa89652ac1ba8c Mon Sep 17 00:00:00 2001 From: Timothy Hayes Date: Thu, 21 Apr 2022 17:52:05 +0100 Subject: perf test: Add perf_event_attr test for Arm SPE Adds a perf_event_attr test for Arm SPE in which the presence of physical addresses are checked when SPE unit is run with pa_enable=1. Reviewed-by: Leo Yan Signed-off-by: Timothy Hayes Tested-by: Leo Yan Cc: Alexander Shishkin Cc: Jiri Olsa Cc: John Fastabend Cc: John Garry Cc: KP Singh Cc: Mark Rutland Cc: Martin KaFai Lau Cc: Mathieu Poirier Cc: Namhyung Kim Cc: Song Liu Cc: Will Deacon Cc: Yonghong Song Cc: bpf@vger.kernel.org Cc: linux-arm-kernel@lists.infradead.org Cc: netdev@vger.kernel.org Link: https://lore.kernel.org/r/20220421165205.117662-4-timothy.hayes@arm.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/tests/attr/README | 1 + tools/perf/tests/attr/test-record-spe-physical-address | 12 ++++++++++++ 2 files changed, 13 insertions(+) create mode 100644 tools/perf/tests/attr/test-record-spe-physical-address (limited to 'tools') diff --git a/tools/perf/tests/attr/README b/tools/perf/tests/attr/README index 454505d343fa..eb3f7d4bb324 100644 --- a/tools/perf/tests/attr/README +++ b/tools/perf/tests/attr/README @@ -60,6 +60,7 @@ Following tests are defined (with perf commands): perf record -R kill (test-record-raw) perf record -c 2 -e arm_spe_0// -- kill (test-record-spe-period) perf record -e arm_spe_0/period=3/ -- kill (test-record-spe-period-term) + perf record -e arm_spe_0/pa_enable=1/ -- kill (test-record-spe-physical-address) perf stat -e cycles kill (test-stat-basic) perf stat kill (test-stat-default) perf stat -d kill (test-stat-detailed-1) diff --git a/tools/perf/tests/attr/test-record-spe-physical-address b/tools/perf/tests/attr/test-record-spe-physical-address new file mode 100644 index 000000000000..7ebcf5012ce3 --- /dev/null +++ b/tools/perf/tests/attr/test-record-spe-physical-address @@ -0,0 +1,12 @@ +[config] +command = record +args = --no-bpf-event -e arm_spe_0/pa_enable=1/ -- kill >/dev/null 2>&1 +ret = 1 +arch = aarch64 + +[event-10:base-record-spe] +# 622727 is the decimal of IP|TID|TIME|CPU|IDENTIFIER|DATA_SRC|PHYS_ADDR +sample_type=622727 + +# dummy event +[event-1:base-record-spe] \ No newline at end of file -- cgit v1.2.3 From 838425f2defe5262906b698752d28fd2fca1aac2 Mon Sep 17 00:00:00 2001 From: Namhyung Kim Date: Fri, 15 Apr 2022 17:40:46 -0700 Subject: perf symbol: Pass is_kallsyms to symbols__fixup_end() The symbol fixup is necessary for symbols in kallsyms since they don't have size info. So we use the next symbol's address to calculate the size. Now it's also used for user binaries because sometimes they miss size for hand-written asm functions. There's a arch-specific function to handle kallsyms differently but currently it cannot distinguish kallsyms from others. Pass this information explicitly to handle it properly. Note that those arch functions will be moved to the generic function so I didn't added it to the arch-functions. Fixes: 3cf6a32f3f2a4594 ("perf symbols: Fix symbol size calculation condition") Signed-off-by: Namhyung Kim Acked-by: Ian Rogers Cc: Heiko Carstens Cc: Ingo Molnar Cc: Jiri Olsa Cc: John Garry Cc: Leo Yan Cc: Mark Rutland Cc: Masami Hiramatsu Cc: Mathieu Poirier Cc: Michael Ellerman Cc: Michael Petlan Cc: Peter Zijlstra Cc: Song Liu Cc: Will Deacon Cc: linux-s390@vger.kernel.org Cc: linuxppc-dev@lists.ozlabs.org Link: https://lore.kernel.org/r/20220416004048.1514900-2-namhyung@kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/symbol-elf.c | 2 +- tools/perf/util/symbol.c | 7 ++++--- tools/perf/util/symbol.h | 2 +- 3 files changed, 6 insertions(+), 5 deletions(-) (limited to 'tools') diff --git a/tools/perf/util/symbol-elf.c b/tools/perf/util/symbol-elf.c index 31cd59a2b66e..ecd377938eea 100644 --- a/tools/perf/util/symbol-elf.c +++ b/tools/perf/util/symbol-elf.c @@ -1290,7 +1290,7 @@ dso__load_sym_internal(struct dso *dso, struct map *map, struct symsrc *syms_ss, * For misannotated, zeroed, ASM function sizes. */ if (nr > 0) { - symbols__fixup_end(&dso->symbols); + symbols__fixup_end(&dso->symbols, false); symbols__fixup_duplicate(&dso->symbols); if (kmap) { /* diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c index dea0fc495185..1b85cc1422a9 100644 --- a/tools/perf/util/symbol.c +++ b/tools/perf/util/symbol.c @@ -217,7 +217,8 @@ again: } } -void symbols__fixup_end(struct rb_root_cached *symbols) +void symbols__fixup_end(struct rb_root_cached *symbols, + bool is_kallsyms __maybe_unused) { struct rb_node *nd, *prevnd = rb_first_cached(symbols); struct symbol *curr, *prev; @@ -1467,7 +1468,7 @@ int __dso__load_kallsyms(struct dso *dso, const char *filename, if (kallsyms__delta(kmap, filename, &delta)) return -1; - symbols__fixup_end(&dso->symbols); + symbols__fixup_end(&dso->symbols, true); symbols__fixup_duplicate(&dso->symbols); if (dso->kernel == DSO_SPACE__KERNEL_GUEST) @@ -1659,7 +1660,7 @@ int dso__load_bfd_symbols(struct dso *dso, const char *debugfile) #undef bfd_asymbol_section #endif - symbols__fixup_end(&dso->symbols); + symbols__fixup_end(&dso->symbols, false); symbols__fixup_duplicate(&dso->symbols); dso->adjust_symbols = 1; diff --git a/tools/perf/util/symbol.h b/tools/perf/util/symbol.h index fbf866d82dcc..5fcdd1f94c56 100644 --- a/tools/perf/util/symbol.h +++ b/tools/perf/util/symbol.h @@ -203,7 +203,7 @@ void __symbols__insert(struct rb_root_cached *symbols, struct symbol *sym, bool kernel); void symbols__insert(struct rb_root_cached *symbols, struct symbol *sym); void symbols__fixup_duplicate(struct rb_root_cached *symbols); -void symbols__fixup_end(struct rb_root_cached *symbols); +void symbols__fixup_end(struct rb_root_cached *symbols, bool is_kallsyms); void maps__fixup_end(struct maps *maps); typedef int (*mapfn_t)(u64 start, u64 len, u64 pgoff, void *data); -- cgit v1.2.3 From 8799ebce84d672aae1dc3170510f6a3e66f96b11 Mon Sep 17 00:00:00 2001 From: Namhyung Kim Date: Fri, 15 Apr 2022 17:40:47 -0700 Subject: perf symbol: Update symbols__fixup_end() Now arch-specific functions all do the same thing. When it fixes the symbol address it needs to check the boundary between the kernel image and modules. For the last symbol in the previous region, it cannot know the exact size as it's discarded already. Thus it just uses a small page size (4096) and rounds it up like the last symbol. Fixes: 3cf6a32f3f2a4594 ("perf symbols: Fix symbol size calculation condition") Signed-off-by: Namhyung Kim Acked-by: Ian Rogers Cc: Heiko Carstens Cc: Ingo Molnar Cc: Jiri Olsa Cc: John Garry Cc: Leo Yan Cc: Mark Rutland Cc: Masami Hiramatsu Cc: Mathieu Poirier Cc: Michael Ellerman Cc: Michael Petlan Cc: Peter Zijlstra Cc: Song Liu Cc: Will Deacon Cc: linux-s390@vger.kernel.org Cc: linuxppc-dev@lists.ozlabs.org Link: https://lore.kernel.org/r/20220416004048.1514900-3-namhyung@kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/symbol.c | 29 +++++++++++++++++++++++++---- 1 file changed, 25 insertions(+), 4 deletions(-) (limited to 'tools') diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c index 1b85cc1422a9..623094e866fd 100644 --- a/tools/perf/util/symbol.c +++ b/tools/perf/util/symbol.c @@ -217,8 +217,8 @@ again: } } -void symbols__fixup_end(struct rb_root_cached *symbols, - bool is_kallsyms __maybe_unused) +/* Update zero-sized symbols using the address of the next symbol */ +void symbols__fixup_end(struct rb_root_cached *symbols, bool is_kallsyms) { struct rb_node *nd, *prevnd = rb_first_cached(symbols); struct symbol *curr, *prev; @@ -232,8 +232,29 @@ void symbols__fixup_end(struct rb_root_cached *symbols, prev = curr; curr = rb_entry(nd, struct symbol, rb_node); - if (prev->end == prev->start || prev->end != curr->start) - arch__symbols__fixup_end(prev, curr); + /* + * On some architecture kernel text segment start is located at + * some low memory address, while modules are located at high + * memory addresses (or vice versa). The gap between end of + * kernel text segment and beginning of first module's text + * segment is very big. Therefore do not fill this gap and do + * not assign it to the kernel dso map (kallsyms). + * + * In kallsyms, it determines module symbols using '[' character + * like in: + * ffffffffc1937000 T hdmi_driver_init [snd_hda_codec_hdmi] + */ + if (prev->end == prev->start) { + /* Last kernel/module symbol mapped to end of page */ + if (is_kallsyms && (!strchr(prev->name, '[') != + !strchr(curr->name, '['))) + prev->end = roundup(prev->end + 4096, 4096); + else + prev->end = curr->start; + + pr_debug4("%s sym:%s end:%#" PRIx64 "\n", + __func__, prev->name, prev->end); + } } /* Last entry */ -- cgit v1.2.3 From a5d20d42a2f2dc2b2f9e9361912062732414090d Mon Sep 17 00:00:00 2001 From: Namhyung Kim Date: Fri, 15 Apr 2022 17:40:48 -0700 Subject: perf symbol: Remove arch__symbols__fixup_end() Now the generic code can handle kallsyms fixup properly so no need to keep the arch-functions anymore. Fixes: 3cf6a32f3f2a4594 ("perf symbols: Fix symbol size calculation condition") Signed-off-by: Namhyung Kim Acked-by: Ian Rogers Cc: Heiko Carstens Cc: Ingo Molnar Cc: Jiri Olsa Cc: John Garry Cc: Leo Yan Cc: Mark Rutland Cc: Masami Hiramatsu Cc: Mathieu Poirier Cc: Michael Ellerman Cc: Michael Petlan Cc: Peter Zijlstra Cc: Song Liu Cc: Will Deacon Cc: linux-s390@vger.kernel.org Cc: linuxppc-dev@lists.ozlabs.org Link: https://lore.kernel.org/r/20220416004048.1514900-4-namhyung@kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/arch/arm64/util/machine.c | 21 --------------------- tools/perf/arch/powerpc/util/Build | 1 - tools/perf/arch/powerpc/util/machine.c | 25 ------------------------- tools/perf/arch/s390/util/machine.c | 16 ---------------- tools/perf/util/symbol.c | 5 ----- tools/perf/util/symbol.h | 1 - 6 files changed, 69 deletions(-) delete mode 100644 tools/perf/arch/powerpc/util/machine.c (limited to 'tools') diff --git a/tools/perf/arch/arm64/util/machine.c b/tools/perf/arch/arm64/util/machine.c index d2ce31e28cd7..41c1596e5207 100644 --- a/tools/perf/arch/arm64/util/machine.c +++ b/tools/perf/arch/arm64/util/machine.c @@ -8,27 +8,6 @@ #include "callchain.h" #include "record.h" -/* On arm64, kernel text segment starts at high memory address, - * for example 0xffff 0000 8xxx xxxx. Modules start at a low memory - * address, like 0xffff 0000 00ax xxxx. When only small amount of - * memory is used by modules, gap between end of module's text segment - * and start of kernel text segment may reach 2G. - * Therefore do not fill this gap and do not assign it to the kernel dso map. - */ - -#define SYMBOL_LIMIT (1 << 12) /* 4K */ - -void arch__symbols__fixup_end(struct symbol *p, struct symbol *c) -{ - if ((strchr(p->name, '[') && strchr(c->name, '[') == NULL) || - (strchr(p->name, '[') == NULL && strchr(c->name, '['))) - /* Limit range of last symbol in module and kernel */ - p->end += SYMBOL_LIMIT; - else - p->end = c->start; - pr_debug4("%s sym:%s end:%#" PRIx64 "\n", __func__, p->name, p->end); -} - void arch__add_leaf_frame_record_opts(struct record_opts *opts) { opts->sample_user_regs |= sample_reg_masks[PERF_REG_ARM64_LR].mask; diff --git a/tools/perf/arch/powerpc/util/Build b/tools/perf/arch/powerpc/util/Build index 8a79c4126e5b..0115f3166568 100644 --- a/tools/perf/arch/powerpc/util/Build +++ b/tools/perf/arch/powerpc/util/Build @@ -1,5 +1,4 @@ perf-y += header.o -perf-y += machine.o perf-y += kvm-stat.o perf-y += perf_regs.o perf-y += mem-events.o diff --git a/tools/perf/arch/powerpc/util/machine.c b/tools/perf/arch/powerpc/util/machine.c deleted file mode 100644 index e652a1aa8132..000000000000 --- a/tools/perf/arch/powerpc/util/machine.c +++ /dev/null @@ -1,25 +0,0 @@ -// SPDX-License-Identifier: GPL-2.0 - -#include -#include -#include -#include // page_size -#include "debug.h" -#include "symbol.h" - -/* On powerpc kernel text segment start at memory addresses, 0xc000000000000000 - * whereas the modules are located at very high memory addresses, - * for example 0xc00800000xxxxxxx. The gap between end of kernel text segment - * and beginning of first module's text segment is very high. - * Therefore do not fill this gap and do not assign it to the kernel dso map. - */ - -void arch__symbols__fixup_end(struct symbol *p, struct symbol *c) -{ - if (strchr(p->name, '[') == NULL && strchr(c->name, '[')) - /* Limit the range of last kernel symbol */ - p->end += page_size; - else - p->end = c->start; - pr_debug4("%s sym:%s end:%#" PRIx64 "\n", __func__, p->name, p->end); -} diff --git a/tools/perf/arch/s390/util/machine.c b/tools/perf/arch/s390/util/machine.c index 7644a4f6d4a4..98bc3f39d5f3 100644 --- a/tools/perf/arch/s390/util/machine.c +++ b/tools/perf/arch/s390/util/machine.c @@ -35,19 +35,3 @@ int arch__fix_module_text_start(u64 *start, u64 *size, const char *name) return 0; } - -/* On s390 kernel text segment start is located at very low memory addresses, - * for example 0x10000. Modules are located at very high memory addresses, - * for example 0x3ff xxxx xxxx. The gap between end of kernel text segment - * and beginning of first module's text segment is very big. - * Therefore do not fill this gap and do not assign it to the kernel dso map. - */ -void arch__symbols__fixup_end(struct symbol *p, struct symbol *c) -{ - if (strchr(p->name, '[') == NULL && strchr(c->name, '[')) - /* Last kernel symbol mapped to end of page */ - p->end = roundup(p->end, page_size); - else - p->end = c->start; - pr_debug4("%s sym:%s end:%#" PRIx64 "\n", __func__, p->name, p->end); -} diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c index 623094e866fd..f72baf636724 100644 --- a/tools/perf/util/symbol.c +++ b/tools/perf/util/symbol.c @@ -101,11 +101,6 @@ static int prefix_underscores_count(const char *str) return tail - str; } -void __weak arch__symbols__fixup_end(struct symbol *p, struct symbol *c) -{ - p->end = c->start; -} - const char * __weak arch__normalize_symbol_name(const char *name) { return name; diff --git a/tools/perf/util/symbol.h b/tools/perf/util/symbol.h index 5fcdd1f94c56..0b893dcc8ea6 100644 --- a/tools/perf/util/symbol.h +++ b/tools/perf/util/symbol.h @@ -241,7 +241,6 @@ const char *arch__normalize_symbol_name(const char *name); #define SYMBOL_A 0 #define SYMBOL_B 1 -void arch__symbols__fixup_end(struct symbol *p, struct symbol *c); int arch__compare_symbol_names(const char *namea, const char *nameb); int arch__compare_symbol_names_n(const char *namea, const char *nameb, unsigned int n); -- cgit v1.2.3