From 80cc7bb6c104d733bff60ddda09f19139c61507c Mon Sep 17 00:00:00 2001 From: Kim Phillips Date: Fri, 7 Feb 2020 17:06:11 -0600 Subject: perf stat: Don't report a null stalled cycles per insn metric For data collected on machines with front end stalled cycles supported, such as found on modern AMD CPU families, commit 146540fb545b ("perf stat: Always separate stalled cycles per insn") introduces a new line in CSV output with a leading comma that upsets some automated scripts. Scripts have to use "-e ex_ret_instr" to work around this issue, after upgrading to a version of perf with that commit. We could add "if (have_frontend_stalled && !config->csv_sep)" to the not (total && avg) else clause, to emphasize that CSV users are usually scripts, and are written to do only what is needed, i.e., they wouldn't typically invoke "perf stat" without specifying an explicit event list. But - let alone CSV output - why should users now tolerate a constant 0-reporting extra line in regular terminal output?: BEFORE: $ sudo perf stat --all-cpus -einstructions,cycles -- sleep 1 Performance counter stats for 'system wide': 181,110,981 instructions # 0.58 insn per cycle # 0.00 stalled cycles per insn 309,876,469 cycles 1.002202582 seconds time elapsed The user would not like to see the now permanent: "0.00 stalled cycles per insn" line fixture, as it gives no useful information. So this patch removes the printing of the zeroed stalled cycles line altogether, almost reverting the very original commit fb4605ba47e7 ("perf stat: Check for frontend stalled for metrics"), which seems like it was written to normalize --metric-only column output of common Intel machines at the time: modern Intel machines have ceased to support the genericised frontend stalled metrics AFAICT. AFTER: $ sudo perf stat --all-cpus -einstructions,cycles -- sleep 1 Performance counter stats for 'system wide': 244,071,432 instructions # 0.69 insn per cycle 355,353,490 cycles 1.001862516 seconds time elapsed Output behaviour when stalled cycles is indeed measured is not affected (BEFORE == AFTER): $ sudo perf stat --all-cpus -einstructions,cycles,stalled-cycles-frontend -- sleep 1 Performance counter stats for 'system wide': 247,227,799 instructions # 0.63 insn per cycle # 0.26 stalled cycles per insn 394,745,636 cycles 63,194,485 stalled-cycles-frontend # 16.01% frontend cycles idle 1.002079770 seconds time elapsed Fixes: 146540fb545b ("perf stat: Always separate stalled cycles per insn") Signed-off-by: Kim Phillips Acked-by: Andi Kleen Acked-by: Jiri Olsa Acked-by: Song Liu Cc: Alexander Shishkin Cc: Cong Wang Cc: Davidlohr Bueso Cc: Jin Yao Cc: Kan Liang Cc: Mark Rutland Cc: Namhyung Kim Cc: Peter Zijlstra Link: http://lore.kernel.org/lkml/20200207230613.26709-1-kim.phillips@amd.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/stat-shadow.c | 6 ------ 1 file changed, 6 deletions(-) diff --git a/tools/perf/util/stat-shadow.c b/tools/perf/util/stat-shadow.c index 2c41d47f6f83..90d23cc3c8d4 100644 --- a/tools/perf/util/stat-shadow.c +++ b/tools/perf/util/stat-shadow.c @@ -18,7 +18,6 @@ * AGGR_NONE: Use matching CPU * AGGR_THREAD: Not supported? */ -static bool have_frontend_stalled; struct runtime_stat rt_stat; struct stats walltime_nsecs_stats; @@ -144,7 +143,6 @@ void runtime_stat__exit(struct runtime_stat *st) void perf_stat__init_shadow_stats(void) { - have_frontend_stalled = pmu_have_event("cpu", "stalled-cycles-frontend"); runtime_stat__init(&rt_stat); } @@ -853,10 +851,6 @@ void perf_stat__print_shadow_stats(struct perf_stat_config *config, print_metric(config, ctxp, NULL, "%7.2f ", "stalled cycles per insn", ratio); - } else if (have_frontend_stalled) { - out->new_line(config, ctxp); - print_metric(config, ctxp, NULL, "%7.2f ", - "stalled cycles per insn", 0); } } else if (perf_evsel__match(evsel, HARDWARE, HW_BRANCH_MISSES)) { if (runtime_stat_n(st, STAT_BRANCHES, ctx, cpu) != 0) -- cgit v1.2.3 From 0e71459afcbbf69e92a65085c45515d3f3f02c31 Mon Sep 17 00:00:00 2001 From: Kim Phillips Date: Fri, 7 Feb 2020 17:06:12 -0600 Subject: perf symbols: Update the list of kernel idle symbols The "acpi_idle_do_entry", "acpi_processor_ffh_cstate_enter", and "idle_cpu" symbols appear in 'perf top' output, at least on AMD systems. Add them to perf's idle_symbols list, so they don't dominate 'perf top' output. Signed-off-by: Kim Phillips Acked-by: Jiri Olsa Acked-by: Song Liu Cc: Alexander Shishkin Cc: Andi Kleen Cc: Cong Wang Cc: Davidlohr Bueso Cc: Jin Yao Cc: Kan Liang Cc: Mark Rutland Cc: Namhyung Kim Cc: Peter Zijlstra Link: http://lore.kernel.org/lkml/20200207230613.26709-2-kim.phillips@amd.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/symbol.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c index 3b379b1296f1..f3120c4f47ad 100644 --- a/tools/perf/util/symbol.c +++ b/tools/perf/util/symbol.c @@ -635,9 +635,12 @@ out: static bool symbol__is_idle(const char *name) { const char * const idle_symbols[] = { + "acpi_idle_do_entry", + "acpi_processor_ffh_cstate_enter", "arch_cpu_idle", "cpu_idle", "cpu_startup_entry", + "idle_cpu", "intel_idle", "default_idle", "native_safe_halt", -- cgit v1.2.3 From bc5f15be2c814ca1ff6bb4e62d5b275a8c88cbb1 Mon Sep 17 00:00:00 2001 From: Kim Phillips Date: Mon, 10 Feb 2020 10:31:47 -0600 Subject: perf symbols: Convert symbol__is_idle() to use strlist Use the more optimized strlist implementation to do the idle function lookup. Signed-off-by: Kim Phillips Acked-by: Song Liu Tested-by: Arnaldo Carvalho de Melo Cc: Alexander Shishkin Cc: Andi Kleen Cc: Cong Wang Cc: Davidlohr Bueso Cc: Jin Yao Cc: Jiri Olsa Cc: Kan Liang Cc: Mark Rutland Cc: Namhyung Kim Cc: Peter Zijlstra Link: http://lore.kernel.org/lkml/20200210163147.25358-1-kim.phillips@amd.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/symbol.c | 14 +++++++++----- 1 file changed, 9 insertions(+), 5 deletions(-) diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c index f3120c4f47ad..1077013d8ce2 100644 --- a/tools/perf/util/symbol.c +++ b/tools/perf/util/symbol.c @@ -654,13 +654,17 @@ static bool symbol__is_idle(const char *name) NULL }; int i; + static struct strlist *idle_symbols_list; - for (i = 0; idle_symbols[i]; i++) { - if (!strcmp(idle_symbols[i], name)) - return true; - } + if (idle_symbols_list) + return strlist__has_entry(idle_symbols_list, name); - return false; + idle_symbols_list = strlist__new(NULL, NULL); + + for (i = 0; idle_symbols[i]; i++) + strlist__add(idle_symbols_list, idle_symbols[i]); + + return strlist__has_entry(idle_symbols_list, name); } static int map__process_kallsym_symbol(void *arg, const char *name, -- cgit v1.2.3 From be993e44badc448add6a18d6f12b20615692c4c3 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Fri, 7 Feb 2020 12:57:36 +0100 Subject: arm/ftrace: Fix BE text poking The __patch_text() function already applies __opcode_to_mem_*(), so when __opcode_to_mem_*() is not the identity (BE*), it is applied twice, wrecking the instruction. Fixes: 42e51f187f86 ("arm/ftrace: Use __patch_text()") Reported-by: Dmitry Osipenko Signed-off-by: Peter Zijlstra (Intel) Signed-off-by: Ingo Molnar Tested-by: Dmitry Osipenko --- arch/arm/kernel/ftrace.c | 7 ++----- 1 file changed, 2 insertions(+), 5 deletions(-) diff --git a/arch/arm/kernel/ftrace.c b/arch/arm/kernel/ftrace.c index 2a5ff69c28e6..10499d44964a 100644 --- a/arch/arm/kernel/ftrace.c +++ b/arch/arm/kernel/ftrace.c @@ -78,13 +78,10 @@ static int ftrace_modify_code(unsigned long pc, unsigned long old, { unsigned long replaced; - if (IS_ENABLED(CONFIG_THUMB2_KERNEL)) { + if (IS_ENABLED(CONFIG_THUMB2_KERNEL)) old = __opcode_to_mem_thumb32(old); - new = __opcode_to_mem_thumb32(new); - } else { + else old = __opcode_to_mem_arm(old); - new = __opcode_to_mem_arm(new); - } if (validate) { if (probe_kernel_read(&replaced, (void *)pc, MCOUNT_INSN_SIZE)) -- cgit v1.2.3 From 7a7a8f549ddd18126dfa3dedbe42d877614c7995 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Fri, 7 Feb 2020 12:57:37 +0100 Subject: arm/patch: Fix !MMU compile Now that patch.o is unconditionally selected for ftrace, it can also get compiled for !MMU kernels. These (obviously) lack {set,clear}_fixmap() support. Also remove the superfluous __acquire/__release nonsense. Fixes: 42e51f187f86 ("arm/ftrace: Use __patch_text()") Signed-off-by: Peter Zijlstra (Intel) Signed-off-by: Ingo Molnar --- arch/arm/kernel/patch.c | 19 +++++++++---------- 1 file changed, 9 insertions(+), 10 deletions(-) diff --git a/arch/arm/kernel/patch.c b/arch/arm/kernel/patch.c index d0a05a3bdb96..e9e828b6bb30 100644 --- a/arch/arm/kernel/patch.c +++ b/arch/arm/kernel/patch.c @@ -16,10 +16,10 @@ struct patch { unsigned int insn; }; +#ifdef CONFIG_MMU static DEFINE_RAW_SPINLOCK(patch_lock); static void __kprobes *patch_map(void *addr, int fixmap, unsigned long *flags) - __acquires(&patch_lock) { unsigned int uintaddr = (uintptr_t) addr; bool module = !core_kernel_text(uintaddr); @@ -34,8 +34,6 @@ static void __kprobes *patch_map(void *addr, int fixmap, unsigned long *flags) if (flags) raw_spin_lock_irqsave(&patch_lock, *flags); - else - __acquire(&patch_lock); set_fixmap(fixmap, page_to_phys(page)); @@ -43,15 +41,19 @@ static void __kprobes *patch_map(void *addr, int fixmap, unsigned long *flags) } static void __kprobes patch_unmap(int fixmap, unsigned long *flags) - __releases(&patch_lock) { clear_fixmap(fixmap); if (flags) raw_spin_unlock_irqrestore(&patch_lock, *flags); - else - __release(&patch_lock); } +#else +static void __kprobes *patch_map(void *addr, int fixmap, unsigned long *flags) +{ + return addr; +} +static void __kprobes patch_unmap(int fixmap, unsigned long *flags) { } +#endif void __kprobes __patch_text_real(void *addr, unsigned int insn, bool remap) { @@ -64,8 +66,6 @@ void __kprobes __patch_text_real(void *addr, unsigned int insn, bool remap) if (remap) waddr = patch_map(addr, FIX_TEXT_POKE0, &flags); - else - __acquire(&patch_lock); if (thumb2 && __opcode_is_thumb16(insn)) { *(u16 *)waddr = __opcode_to_mem_thumb16(insn); @@ -102,8 +102,7 @@ void __kprobes __patch_text_real(void *addr, unsigned int insn, bool remap) if (waddr != addr) { flush_kernel_vmap_range(waddr, twopage ? size / 2 : size); patch_unmap(FIX_TEXT_POKE0, &flags); - } else - __release(&patch_lock); + } flush_icache_range((uintptr_t)(addr), (uintptr_t)(addr) + size); -- cgit v1.2.3 From eda23b387f6c4bb2971ac7e874a09913f533b22c Mon Sep 17 00:00:00 2001 From: Kan Liang Date: Tue, 28 Jan 2020 10:31:17 -0800 Subject: perf/x86/intel: Add Elkhart Lake support Elkhart Lake also uses Tremont CPU. From the perspective of Intel PMU, there is nothing changed compared with Jacobsville. Share the perf code with Jacobsville. Signed-off-by: Kan Liang Signed-off-by: Peter Zijlstra (Intel) Signed-off-by: Ingo Molnar Reviewed-by: Andi Kleen Link: https://lkml.kernel.org/r/1580236279-35492-1-git-send-email-kan.liang@linux.intel.com --- arch/x86/events/intel/core.c | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/x86/events/intel/core.c b/arch/x86/events/intel/core.c index 3be51aa06e67..dff6623804c2 100644 --- a/arch/x86/events/intel/core.c +++ b/arch/x86/events/intel/core.c @@ -4765,6 +4765,7 @@ __init int intel_pmu_init(void) break; case INTEL_FAM6_ATOM_TREMONT_D: + case INTEL_FAM6_ATOM_TREMONT: x86_pmu.late_ack = true; memcpy(hw_cache_event_ids, glp_hw_cache_event_ids, sizeof(hw_cache_event_ids)); -- cgit v1.2.3 From ecf71fbccb9ac5cb964eb7de59bb9da3755b7885 Mon Sep 17 00:00:00 2001 From: Kan Liang Date: Tue, 28 Jan 2020 10:31:18 -0800 Subject: perf/x86/cstate: Add Tremont support Tremont is Intel's successor to Goldmont Plus. From the perspective of Intel cstate residency counters, there is nothing changed compared with Goldmont Plus and Goldmont. Share glm_cstates with Goldmont Plus and Goldmont. Update the comments for Tremont. Signed-off-by: Kan Liang Signed-off-by: Peter Zijlstra (Intel) Signed-off-by: Ingo Molnar Reviewed-by: Andi Kleen Link: https://lkml.kernel.org/r/1580236279-35492-2-git-send-email-kan.liang@linux.intel.com --- arch/x86/events/intel/cstate.c | 22 +++++++++++++--------- 1 file changed, 13 insertions(+), 9 deletions(-) diff --git a/arch/x86/events/intel/cstate.c b/arch/x86/events/intel/cstate.c index e1daf4151e11..4814c964692c 100644 --- a/arch/x86/events/intel/cstate.c +++ b/arch/x86/events/intel/cstate.c @@ -40,17 +40,18 @@ * Model specific counters: * MSR_CORE_C1_RES: CORE C1 Residency Counter * perf code: 0x00 - * Available model: SLM,AMT,GLM,CNL + * Available model: SLM,AMT,GLM,CNL,TNT * Scope: Core (each processor core has a MSR) * MSR_CORE_C3_RESIDENCY: CORE C3 Residency Counter * perf code: 0x01 * Available model: NHM,WSM,SNB,IVB,HSW,BDW,SKL,GLM, - * CNL,KBL,CML + * CNL,KBL,CML,TNT * Scope: Core * MSR_CORE_C6_RESIDENCY: CORE C6 Residency Counter * perf code: 0x02 * Available model: SLM,AMT,NHM,WSM,SNB,IVB,HSW,BDW, - * SKL,KNL,GLM,CNL,KBL,CML,ICL,TGL + * SKL,KNL,GLM,CNL,KBL,CML,ICL,TGL, + * TNT * Scope: Core * MSR_CORE_C7_RESIDENCY: CORE C7 Residency Counter * perf code: 0x03 @@ -60,17 +61,18 @@ * MSR_PKG_C2_RESIDENCY: Package C2 Residency Counter. * perf code: 0x00 * Available model: SNB,IVB,HSW,BDW,SKL,KNL,GLM,CNL, - * KBL,CML,ICL,TGL + * KBL,CML,ICL,TGL,TNT * Scope: Package (physical package) * MSR_PKG_C3_RESIDENCY: Package C3 Residency Counter. * perf code: 0x01 * Available model: NHM,WSM,SNB,IVB,HSW,BDW,SKL,KNL, - * GLM,CNL,KBL,CML,ICL,TGL + * GLM,CNL,KBL,CML,ICL,TGL,TNT * Scope: Package (physical package) * MSR_PKG_C6_RESIDENCY: Package C6 Residency Counter. * perf code: 0x02 - * Available model: SLM,AMT,NHM,WSM,SNB,IVB,HSW,BDW - * SKL,KNL,GLM,CNL,KBL,CML,ICL,TGL + * Available model: SLM,AMT,NHM,WSM,SNB,IVB,HSW,BDW, + * SKL,KNL,GLM,CNL,KBL,CML,ICL,TGL, + * TNT * Scope: Package (physical package) * MSR_PKG_C7_RESIDENCY: Package C7 Residency Counter. * perf code: 0x03 @@ -87,7 +89,8 @@ * Scope: Package (physical package) * MSR_PKG_C10_RESIDENCY: Package C10 Residency Counter. * perf code: 0x06 - * Available model: HSW ULT,KBL,GLM,CNL,CML,ICL,TGL + * Available model: HSW ULT,KBL,GLM,CNL,CML,ICL,TGL, + * TNT * Scope: Package (physical package) * */ @@ -640,8 +643,9 @@ static const struct x86_cpu_id intel_cstates_match[] __initconst = { X86_CSTATES_MODEL(INTEL_FAM6_ATOM_GOLDMONT, glm_cstates), X86_CSTATES_MODEL(INTEL_FAM6_ATOM_GOLDMONT_D, glm_cstates), - X86_CSTATES_MODEL(INTEL_FAM6_ATOM_GOLDMONT_PLUS, glm_cstates), + X86_CSTATES_MODEL(INTEL_FAM6_ATOM_TREMONT_D, glm_cstates), + X86_CSTATES_MODEL(INTEL_FAM6_ATOM_TREMONT, glm_cstates), X86_CSTATES_MODEL(INTEL_FAM6_ICELAKE_L, icl_cstates), X86_CSTATES_MODEL(INTEL_FAM6_ICELAKE, icl_cstates), -- cgit v1.2.3 From 0aa0e0d6b34b89649e6b5882a7e025a0eb9bd832 Mon Sep 17 00:00:00 2001 From: Kan Liang Date: Tue, 28 Jan 2020 10:31:19 -0800 Subject: perf/x86/msr: Add Tremont support Tremont is Intel's successor to Goldmont Plus. SMI_COUNT MSR is also supported. Signed-off-by: Kan Liang Signed-off-by: Peter Zijlstra (Intel) Signed-off-by: Ingo Molnar Reviewed-by: Andi Kleen Link: https://lkml.kernel.org/r/1580236279-35492-3-git-send-email-kan.liang@linux.intel.com --- arch/x86/events/msr.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/arch/x86/events/msr.c b/arch/x86/events/msr.c index 6f86650b3f77..a949f6f55991 100644 --- a/arch/x86/events/msr.c +++ b/arch/x86/events/msr.c @@ -75,8 +75,9 @@ static bool test_intel(int idx, void *data) case INTEL_FAM6_ATOM_GOLDMONT: case INTEL_FAM6_ATOM_GOLDMONT_D: - case INTEL_FAM6_ATOM_GOLDMONT_PLUS: + case INTEL_FAM6_ATOM_TREMONT_D: + case INTEL_FAM6_ATOM_TREMONT: case INTEL_FAM6_XEON_PHI_KNL: case INTEL_FAM6_XEON_PHI_KNM: -- cgit v1.2.3 From 25d387287cf0330abf2aad761ce6eee67326a355 Mon Sep 17 00:00:00 2001 From: Kim Phillips Date: Tue, 21 Jan 2020 11:12:31 -0600 Subject: perf/x86/amd: Add missing L2 misses event spec to AMD Family 17h's event map Commit 3fe3331bb285 ("perf/x86/amd: Add event map for AMD Family 17h"), claimed L2 misses were unsupported, due to them not being found in its referenced documentation, whose link has now moved [1]. That old documentation listed PMCx064 unit mask bit 3 as: "LsRdBlkC: LS Read Block C S L X Change to X Miss." and bit 0 as: "IcFillMiss: IC Fill Miss" We now have new public documentation [2] with improved descriptions, that clearly indicate what events those unit mask bits represent: Bit 3 now clearly states: "LsRdBlkC: Data Cache Req Miss in L2 (all types)" and bit 0 is: "IcFillMiss: Instruction Cache Req Miss in L2." So we can now add support for L2 misses in perf's genericised events as PMCx064 with both the above unit masks. [1] The commit's original documentation reference, "Processor Programming Reference (PPR) for AMD Family 17h Model 01h, Revision B1 Processors", originally available here: https://www.amd.com/system/files/TechDocs/54945_PPR_Family_17h_Models_00h-0Fh.pdf is now available here: https://developer.amd.com/wordpress/media/2017/11/54945_PPR_Family_17h_Models_00h-0Fh.pdf [2] "Processor Programming Reference (PPR) for Family 17h Model 31h, Revision B0 Processors", available here: https://developer.amd.com/wp-content/resources/55803_0.54-PUB.pdf Fixes: 3fe3331bb285 ("perf/x86/amd: Add event map for AMD Family 17h") Reported-by: Babu Moger Signed-off-by: Kim Phillips Signed-off-by: Peter Zijlstra (Intel) Signed-off-by: Ingo Molnar Tested-by: Babu Moger Cc: stable@vger.kernel.org Link: https://lkml.kernel.org/r/20200121171232.28839-1-kim.phillips@amd.com --- arch/x86/events/amd/core.c | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/x86/events/amd/core.c b/arch/x86/events/amd/core.c index 1f22b6bbda68..39eb276d0277 100644 --- a/arch/x86/events/amd/core.c +++ b/arch/x86/events/amd/core.c @@ -250,6 +250,7 @@ static const u64 amd_f17h_perfmon_event_map[PERF_COUNT_HW_MAX] = [PERF_COUNT_HW_CPU_CYCLES] = 0x0076, [PERF_COUNT_HW_INSTRUCTIONS] = 0x00c0, [PERF_COUNT_HW_CACHE_REFERENCES] = 0xff60, + [PERF_COUNT_HW_CACHE_MISSES] = 0x0964, [PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = 0x00c2, [PERF_COUNT_HW_BRANCH_MISSES] = 0x00c3, [PERF_COUNT_HW_STALLED_CYCLES_FRONTEND] = 0x0287, -- cgit v1.2.3 From f861854e1b435b27197417f6f90d87188003cb24 Mon Sep 17 00:00:00 2001 From: Kan Liang Date: Tue, 21 Jan 2020 11:01:25 -0800 Subject: perf/x86/intel: Fix inaccurate period in context switch for auto-reload Perf doesn't take the left period into account when auto-reload is enabled with fixed period sampling mode in context switch. Here is the MSR trace of the perf command as below. (The MSR trace is simplified from a ftrace log.) #perf record -e cycles:p -c 2000000 -- ./triad_loop //The MSR trace of task schedule out //perf disable all counters, disable PEBS, disable GP counter 0, //read GP counter 0, and re-enable all counters. //The counter 0 stops at 0xfffffff82840 write_msr: MSR_CORE_PERF_GLOBAL_CTRL(38f), value 0 write_msr: MSR_IA32_PEBS_ENABLE(3f1), value 0 write_msr: MSR_P6_EVNTSEL0(186), value 40003003c rdpmc: 0, value fffffff82840 write_msr: MSR_CORE_PERF_GLOBAL_CTRL(38f), value f000000ff //The MSR trace of the same task schedule in again //perf disable all counters, enable and set GP counter 0, //enable PEBS, and re-enable all counters. //0xffffffe17b80 (-2000000) is written to GP counter 0. write_msr: MSR_CORE_PERF_GLOBAL_CTRL(38f), value 0 write_msr: MSR_IA32_PMC0(4c1), value ffffffe17b80 write_msr: MSR_P6_EVNTSEL0(186), value 40043003c write_msr: MSR_IA32_PEBS_ENABLE(3f1), value 1 write_msr: MSR_CORE_PERF_GLOBAL_CTRL(38f), value f000000ff When the same task schedule in again, the counter should starts from previous left. However, it starts from the fixed period -2000000 again. A special variant of intel_pmu_save_and_restart() is used for auto-reload, which doesn't update the hwc->period_left. When the monitored task schedules in again, perf doesn't know the left period. The fixed period is used, which is inaccurate. With auto-reload, the counter always has a negative counter value. So the left period is -value. Update the period_left in intel_pmu_save_and_restart_reload(). With the patch: //The MSR trace of task schedule out write_msr: MSR_CORE_PERF_GLOBAL_CTRL(38f), value 0 write_msr: MSR_IA32_PEBS_ENABLE(3f1), value 0 write_msr: MSR_P6_EVNTSEL0(186), value 40003003c rdpmc: 0, value ffffffe25cbc write_msr: MSR_CORE_PERF_GLOBAL_CTRL(38f), value f000000ff //The MSR trace of the same task schedule in again write_msr: MSR_CORE_PERF_GLOBAL_CTRL(38f), value 0 write_msr: MSR_IA32_PMC0(4c1), value ffffffe25cbc write_msr: MSR_P6_EVNTSEL0(186), value 40043003c write_msr: MSR_IA32_PEBS_ENABLE(3f1), value 1 write_msr: MSR_CORE_PERF_GLOBAL_CTRL(38f), value f000000ff Fixes: d31fc13fdcb2 ("perf/x86/intel: Fix event update for auto-reload") Signed-off-by: Kan Liang Signed-off-by: Peter Zijlstra (Intel) Signed-off-by: Ingo Molnar Link: https://lkml.kernel.org/r/20200121190125.3389-1-kan.liang@linux.intel.com --- arch/x86/events/intel/ds.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/arch/x86/events/intel/ds.c b/arch/x86/events/intel/ds.c index 4b94ae4ae369..dc43cc124e09 100644 --- a/arch/x86/events/intel/ds.c +++ b/arch/x86/events/intel/ds.c @@ -1714,6 +1714,8 @@ intel_pmu_save_and_restart_reload(struct perf_event *event, int count) old = ((s64)(prev_raw_count << shift) >> shift); local64_add(new - old + count * period, &event->count); + local64_set(&hwc->period_left, -new); + perf_event_update_userpage(event); return 0; -- cgit v1.2.3 From c452833387624d1990c9bbb0ee1e98c10c147478 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Tue, 11 Feb 2020 11:58:56 -0300 Subject: tools include UAPI: Sync x86's syscalls_64.tbl, generic unistd.h and fcntl.h to pick up openat2 and pidfd_getfd fddb5d430ad9 ("open: introduce openat2(2) syscall") 9a2cef09c801 ("arch: wire up pidfd_getfd syscall") We also need to grab a copy of uapi/linux/openat2.h since it is now needed by fcntl.h, add it to tools/perf/check_headers.h. $ diff -u tools/perf/arch/x86/entry/syscalls/syscall_64.tbl arch/x86/entry/syscalls/syscall_64.tbl --- tools/perf/arch/x86/entry/syscalls/syscall_64.tbl 2019-12-20 16:43:57.662429958 -0300 +++ arch/x86/entry/syscalls/syscall_64.tbl 2020-02-10 16:36:22.070012468 -0300 @@ -357,6 +357,8 @@ 433 common fspick __x64_sys_fspick 434 common pidfd_open __x64_sys_pidfd_open 435 common clone3 __x64_sys_clone3/ptregs +437 common openat2 __x64_sys_openat2 +438 common pidfd_getfd __x64_sys_pidfd_getfd # # x32-specific system call numbers start at 512 to avoid cache impact $ Update tools/'s copy of that file: $ cp arch/x86/entry/syscalls/syscall_64.tbl tools/perf/arch/x86/entry/syscalls/syscall_64.tbl See the result: $ diff -u /tmp/build/perf/arch/x86/include/generated/asm/syscalls_64.c.before /tmp/build/perf/arch/x86/include/generated/asm/syscalls_64.c --- /tmp/build/perf/arch/x86/include/generated/asm/syscalls_64.c.before 2020-02-10 16:42:59.010636041 -0300 +++ /tmp/build/perf/arch/x86/include/generated/asm/syscalls_64.c 2020-02-10 16:43:24.149958337 -0300 @@ -346,5 +346,7 @@ [433] = "fspick", [434] = "pidfd_open", [435] = "clone3", + [437] = "openat2", + [438] = "pidfd_getfd", }; -#define SYSCALLTBL_x86_64_MAX_ID 435 +#define SYSCALLTBL_x86_64_MAX_ID 438 $ Now one can use: perf trace -e openat2,pidfd_getfd To get just those syscalls or use in things like: perf trace -e open* To get all the open variant (open, openat, openat2, etc) or: perf trace pidfd* To get the pidfd syscalls. Cc: Adrian Hunter Cc: Aleksa Sarai Cc: Al Viro Cc: Christian Brauner Cc: Jiri Olsa Cc: Namhyung Kim Cc: Sargun Dhillon Signed-off-by: Arnaldo Carvalho de Melo --- tools/include/uapi/asm-generic/unistd.h | 7 +++- tools/include/uapi/linux/fcntl.h | 2 +- tools/include/uapi/linux/openat2.h | 39 +++++++++++++++++++++++ tools/perf/arch/x86/entry/syscalls/syscall_64.tbl | 2 ++ tools/perf/check-headers.sh | 1 + 5 files changed, 49 insertions(+), 2 deletions(-) create mode 100644 tools/include/uapi/linux/openat2.h diff --git a/tools/include/uapi/asm-generic/unistd.h b/tools/include/uapi/asm-generic/unistd.h index 1fc8faa6e973..3a3201e4618e 100644 --- a/tools/include/uapi/asm-generic/unistd.h +++ b/tools/include/uapi/asm-generic/unistd.h @@ -851,8 +851,13 @@ __SYSCALL(__NR_pidfd_open, sys_pidfd_open) __SYSCALL(__NR_clone3, sys_clone3) #endif +#define __NR_openat2 437 +__SYSCALL(__NR_openat2, sys_openat2) +#define __NR_pidfd_getfd 438 +__SYSCALL(__NR_pidfd_getfd, sys_pidfd_getfd) + #undef __NR_syscalls -#define __NR_syscalls 436 +#define __NR_syscalls 439 /* * 32 bit systems traditionally used different diff --git a/tools/include/uapi/linux/fcntl.h b/tools/include/uapi/linux/fcntl.h index 1f97b33c840e..ca88b7bce553 100644 --- a/tools/include/uapi/linux/fcntl.h +++ b/tools/include/uapi/linux/fcntl.h @@ -3,6 +3,7 @@ #define _UAPI_LINUX_FCNTL_H #include +#include #define F_SETLEASE (F_LINUX_SPECIFIC_BASE + 0) #define F_GETLEASE (F_LINUX_SPECIFIC_BASE + 1) @@ -100,5 +101,4 @@ #define AT_RECURSIVE 0x8000 /* Apply to the entire subtree */ - #endif /* _UAPI_LINUX_FCNTL_H */ diff --git a/tools/include/uapi/linux/openat2.h b/tools/include/uapi/linux/openat2.h new file mode 100644 index 000000000000..58b1eb711360 --- /dev/null +++ b/tools/include/uapi/linux/openat2.h @@ -0,0 +1,39 @@ +/* SPDX-License-Identifier: GPL-2.0 WITH Linux-syscall-note */ +#ifndef _UAPI_LINUX_OPENAT2_H +#define _UAPI_LINUX_OPENAT2_H + +#include + +/* + * Arguments for how openat2(2) should open the target path. If only @flags and + * @mode are non-zero, then openat2(2) operates very similarly to openat(2). + * + * However, unlike openat(2), unknown or invalid bits in @flags result in + * -EINVAL rather than being silently ignored. @mode must be zero unless one of + * {O_CREAT, O_TMPFILE} are set. + * + * @flags: O_* flags. + * @mode: O_CREAT/O_TMPFILE file mode. + * @resolve: RESOLVE_* flags. + */ +struct open_how { + __u64 flags; + __u64 mode; + __u64 resolve; +}; + +/* how->resolve flags for openat2(2). */ +#define RESOLVE_NO_XDEV 0x01 /* Block mount-point crossings + (includes bind-mounts). */ +#define RESOLVE_NO_MAGICLINKS 0x02 /* Block traversal through procfs-style + "magic-links". */ +#define RESOLVE_NO_SYMLINKS 0x04 /* Block traversal through all symlinks + (implies OEXT_NO_MAGICLINKS) */ +#define RESOLVE_BENEATH 0x08 /* Block "lexical" trickery like + "..", symlinks, and absolute + paths which escape the dirfd. */ +#define RESOLVE_IN_ROOT 0x10 /* Make all jumps to "/" and ".." + be scoped inside the dirfd + (similar to chroot(2)). */ + +#endif /* _UAPI_LINUX_OPENAT2_H */ diff --git a/tools/perf/arch/x86/entry/syscalls/syscall_64.tbl b/tools/perf/arch/x86/entry/syscalls/syscall_64.tbl index c29976eca4a8..44d510bc9b78 100644 --- a/tools/perf/arch/x86/entry/syscalls/syscall_64.tbl +++ b/tools/perf/arch/x86/entry/syscalls/syscall_64.tbl @@ -357,6 +357,8 @@ 433 common fspick __x64_sys_fspick 434 common pidfd_open __x64_sys_pidfd_open 435 common clone3 __x64_sys_clone3/ptregs +437 common openat2 __x64_sys_openat2 +438 common pidfd_getfd __x64_sys_pidfd_getfd # # x32-specific system call numbers start at 512 to avoid cache impact diff --git a/tools/perf/check-headers.sh b/tools/perf/check-headers.sh index 68039a96c1dc..bfb21d049e6c 100755 --- a/tools/perf/check-headers.sh +++ b/tools/perf/check-headers.sh @@ -13,6 +13,7 @@ include/uapi/linux/kcmp.h include/uapi/linux/kvm.h include/uapi/linux/in.h include/uapi/linux/mount.h +include/uapi/linux/openat2.h include/uapi/linux/perf_event.h include/uapi/linux/prctl.h include/uapi/linux/sched.h -- cgit v1.2.3 From 02213cec64bbef66d7ad9ddc3b7c47236da64343 Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Mon, 10 Feb 2020 15:32:15 +0100 Subject: perf maps: Mark module DSOs with kernel type We add kernel module map into machine->kmaps, so it needs to be created as 'struct kmap', which is dependent on its dso having kernel type. Reported-by: Ravi Bangoria Signed-off-by: Jiri Olsa Tested-by: Kim Phillips Tested-by: Ravi Bangoria Cc: Alexander Shishkin Cc: Michael Petlan Cc: Namhyung Kim Cc: Peter Zijlstra Link: http://lore.kernel.org/lkml/20200210143218.24948-2-jolsa@kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/machine.c | 1 + 1 file changed, 1 insertion(+) diff --git a/tools/perf/util/machine.c b/tools/perf/util/machine.c index c8c5410315e8..e3e5490f6de5 100644 --- a/tools/perf/util/machine.c +++ b/tools/perf/util/machine.c @@ -686,6 +686,7 @@ static struct dso *machine__findnew_module_dso(struct machine *machine, dso__set_module_info(dso, m, machine); dso__set_long_name(dso, strdup(filename), true); + dso->kernel = DSO_TYPE_KERNEL; } dso__get(dso); -- cgit v1.2.3 From 4a4eb6154d67f7766cc7eb74e9f1db424073e832 Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Mon, 10 Feb 2020 21:08:47 +0100 Subject: perf maps: Mark ksymbol DSOs with kernel type We add ksymbol map into machine->kmaps, so it needs to be created as 'struct kmap', which is dependent on its dso having kernel type. Reported-by: Ravi Bangoria Signed-off-by: Jiri Olsa Tested-by: Ravi Bangoria Tested-by: Kim Phillips Cc: Alexander Shishkin Cc: Michael Petlan Cc: Namhyung Kim Cc: Peter Zijlstra Link: http://lore.kernel.org/lkml/20200210200847.GA36715@krava Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/machine.c | 12 ++++++++++-- 1 file changed, 10 insertions(+), 2 deletions(-) diff --git a/tools/perf/util/machine.c b/tools/perf/util/machine.c index e3e5490f6de5..0ad026561c7f 100644 --- a/tools/perf/util/machine.c +++ b/tools/perf/util/machine.c @@ -727,9 +727,17 @@ static int machine__process_ksymbol_register(struct machine *machine, struct map *map = maps__find(&machine->kmaps, event->ksymbol.addr); if (!map) { - map = dso__new_map(event->ksymbol.name); - if (!map) + struct dso *dso = dso__new(event->ksymbol.name); + + if (dso) { + dso->kernel = DSO_TYPE_KERNEL; + map = map__new2(0, dso); + } + + if (!dso || !map) { + dso__put(dso); return -ENOMEM; + } map->start = event->ksymbol.addr; map->end = map->start + event->ksymbol.len; -- cgit v1.2.3 From 7ce66139a99ce57caaf47b64afed5cb6ed02c5ed Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Mon, 10 Feb 2020 15:32:17 +0100 Subject: perf maps: Fix map__clone() for struct kmap The map__clone() function can be called on kernel maps as well, so it needs to duplicate the whole kmap data. Reported-by: Ravi Bangoria Signed-off-by: Jiri Olsa Tested-by: Ravi Bangoria Tested-by: Kim Phillips Cc: Alexander Shishkin Cc: Michael Petlan Cc: Namhyung Kim Cc: Peter Zijlstra Link: http://lore.kernel.org/lkml/20200210143218.24948-4-jolsa@kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/map.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/tools/perf/util/map.c b/tools/perf/util/map.c index f67960bedebb..cea05fc9595c 100644 --- a/tools/perf/util/map.c +++ b/tools/perf/util/map.c @@ -375,8 +375,13 @@ struct symbol *map__find_symbol_by_name(struct map *map, const char *name) struct map *map__clone(struct map *from) { - struct map *map = memdup(from, sizeof(*map)); + size_t size = sizeof(struct map); + struct map *map; + + if (from->dso && from->dso->kernel) + size += sizeof(struct kmap); + map = memdup(from, size); if (map != NULL) { refcount_set(&map->refcnt, 1); RB_CLEAR_NODE(&map->rb_node); -- cgit v1.2.3 From 484214f49bd0948d716832a94e4737ca4dd02c16 Mon Sep 17 00:00:00 2001 From: Jiri Olsa Date: Mon, 10 Feb 2020 15:32:18 +0100 Subject: perf maps: Move kmap::kmaps setup to maps__insert() So the kmaps pointer setup is centralized and we do not need to update it in all those places (2 current places and few more missing) after calling maps__insert(). Reported-by: Ravi Bangoria Signed-off-by: Jiri Olsa Tested-by: Ravi Bangoria Tested-by: Kim Phillips Cc: Alexander Shishkin Cc: Michael Petlan Cc: Namhyung Kim Cc: Peter Zijlstra Link: http://lore.kernel.org/lkml/20200210143218.24948-5-jolsa@kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/machine.c | 13 +------------ tools/perf/util/map.c | 10 ++++++++++ 2 files changed, 11 insertions(+), 12 deletions(-) diff --git a/tools/perf/util/machine.c b/tools/perf/util/machine.c index 0ad026561c7f..fb5c2cd44d30 100644 --- a/tools/perf/util/machine.c +++ b/tools/perf/util/machine.c @@ -981,7 +981,6 @@ int machine__create_extra_kernel_map(struct machine *machine, kmap = map__kmap(map); - kmap->kmaps = &machine->kmaps; strlcpy(kmap->name, xm->name, KMAP_NAME_LEN); maps__insert(&machine->kmaps, map); @@ -1091,9 +1090,6 @@ int __weak machine__create_extra_kernel_maps(struct machine *machine __maybe_unu static int __machine__create_kernel_maps(struct machine *machine, struct dso *kernel) { - struct kmap *kmap; - struct map *map; - /* In case of renewal the kernel map, destroy previous one */ machine__destroy_kernel_maps(machine); @@ -1102,14 +1098,7 @@ __machine__create_kernel_maps(struct machine *machine, struct dso *kernel) return -1; machine->vmlinux_map->map_ip = machine->vmlinux_map->unmap_ip = identity__map_ip; - map = machine__kernel_map(machine); - kmap = map__kmap(map); - if (!kmap) - return -1; - - kmap->kmaps = &machine->kmaps; - maps__insert(&machine->kmaps, map); - + maps__insert(&machine->kmaps, machine->vmlinux_map); return 0; } diff --git a/tools/perf/util/map.c b/tools/perf/util/map.c index cea05fc9595c..a08ca276098e 100644 --- a/tools/perf/util/map.c +++ b/tools/perf/util/map.c @@ -543,6 +543,16 @@ void maps__insert(struct maps *maps, struct map *map) __maps__insert(maps, map); ++maps->nr_maps; + if (map->dso && map->dso->kernel) { + struct kmap *kmap = map__kmap(map); + + if (kmap) + kmap->kmaps = maps; + else + pr_err("Internal error: kernel dso with non kernel map\n"); + } + + /* * If we already performed some search by name, then we need to add the just * inserted map and resort. -- cgit v1.2.3 From c75bec79fc080039e4575a0f239ea7b111aabe88 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Tue, 11 Feb 2020 15:19:42 -0300 Subject: tools headers UAPI: Sync copy of arm64's asm/unistd.h with the kernel sources To get the changes in: 3e3c8ca5a351 ("arm64: Move __ARCH_WANT_SYS_CLONE3 definition to uapi headers") Silencing this tools/perf/ build warning: Warning: Kernel ABI header at 'tools/arch/arm64/include/uapi/asm/unistd.h' differs from latest version at 'arch/arm64/include/uapi/asm/unistd.h' diff -u tools/arch/arm64/include/uapi/asm/unistd.h arch/arm64/include/uapi/asm/unistd.h Which will probably end up enabling the use of "clone3" in 'perf trace -e', haven't checked the build with this change on an arm64 system. Cc: Adrian Hunter Cc: Amanieu d'Antras Cc: Christian Brauner Cc: Jiri Olsa Cc: Namhyung Kim Signed-off-by: Arnaldo Carvalho de Melo --- tools/arch/arm64/include/uapi/asm/unistd.h | 1 + 1 file changed, 1 insertion(+) diff --git a/tools/arch/arm64/include/uapi/asm/unistd.h b/tools/arch/arm64/include/uapi/asm/unistd.h index 4703d218663a..f83a70e07df8 100644 --- a/tools/arch/arm64/include/uapi/asm/unistd.h +++ b/tools/arch/arm64/include/uapi/asm/unistd.h @@ -19,5 +19,6 @@ #define __ARCH_WANT_NEW_STAT #define __ARCH_WANT_SET_GET_RLIMIT #define __ARCH_WANT_TIME32_SYSCALLS +#define __ARCH_WANT_SYS_CLONE3 #include -- cgit v1.2.3 From fc9199d46e644e6a978f69195cb849b21d2f485c Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Tue, 11 Feb 2020 15:25:21 -0300 Subject: tools headers UAPI: Sync prctl.h with the kernel sources To get the changes in: 8d19f1c8e193 ("prctl: PR_{G,S}ET_IO_FLUSHER to support controlling memory reclaim") Which ends up having this effect in tooling, i.e. the addition of the support to those prctl's options: $ tools/perf/trace/beauty/prctl_option.sh > before $ cp include/uapi/linux/prctl.h tools/include/uapi/linux/prctl.h $ git diff diff --git a/tools/include/uapi/linux/prctl.h b/tools/include/uapi/linux/prctl.h index 7da1b37b27aa..07b4f8131e36 100644 --- a/tools/include/uapi/linux/prctl.h +++ b/tools/include/uapi/linux/prctl.h @@ -234,4 +234,8 @@ struct prctl_mm_map { #define PR_GET_TAGGED_ADDR_CTRL 56 # define PR_TAGGED_ADDR_ENABLE (1UL << 0) +/* Control reclaim behavior when allocating memory */ +#define PR_SET_IO_FLUSHER 57 +#define PR_GET_IO_FLUSHER 58 + #endif /* _LINUX_PRCTL_H */ $ tools/perf/trace/beauty/prctl_option.sh > after $ diff -u before after --- before 2020-02-11 15:24:35.339289912 -0300 +++ after 2020-02-11 15:24:56.319711315 -0300 @@ -51,6 +51,8 @@ [54] = "PAC_RESET_KEYS", [55] = "SET_TAGGED_ADDR_CTRL", [56] = "GET_TAGGED_ADDR_CTRL", + [57] = "SET_IO_FLUSHER", + [58] = "GET_IO_FLUSHER", }; static const char *prctl_set_mm_options[] = { [1] = "START_CODE", $ Cc: Adrian Hunter Cc: Christian Brauner Cc: Jiri Olsa Cc: Mike Christie Cc: Namhyung Kim Signed-off-by: Arnaldo Carvalho de Melo --- tools/include/uapi/linux/prctl.h | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/tools/include/uapi/linux/prctl.h b/tools/include/uapi/linux/prctl.h index 7da1b37b27aa..07b4f8131e36 100644 --- a/tools/include/uapi/linux/prctl.h +++ b/tools/include/uapi/linux/prctl.h @@ -234,4 +234,8 @@ struct prctl_mm_map { #define PR_GET_TAGGED_ADDR_CTRL 56 # define PR_TAGGED_ADDR_ENABLE (1UL << 0) +/* Control reclaim behavior when allocating memory */ +#define PR_SET_IO_FLUSHER 57 +#define PR_GET_IO_FLUSHER 58 + #endif /* _LINUX_PRCTL_H */ -- cgit v1.2.3 From c0134b3366ba5f0aba41d56006b574d3be7f5ed3 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Tue, 11 Feb 2020 15:46:10 -0300 Subject: perf beauty prctl: Export the 'options' strarray So that we can use it with strtoul, allowing string to number conversions in filter expressions. Cc: Adrian Hunter Cc: Christian Brauner Cc: Jiri Olsa Cc: Mike Christie Cc: Namhyung Kim Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/trace/beauty/beauty.h | 2 ++ tools/perf/trace/beauty/prctl.c | 3 ++- 2 files changed, 4 insertions(+), 1 deletion(-) diff --git a/tools/perf/trace/beauty/beauty.h b/tools/perf/trace/beauty/beauty.h index 5a61043c2ff7..d6dfe68a7612 100644 --- a/tools/perf/trace/beauty/beauty.h +++ b/tools/perf/trace/beauty/beauty.h @@ -213,6 +213,8 @@ size_t syscall_arg__scnprintf_x86_arch_prctl_code(char *bf, size_t size, struct size_t syscall_arg__scnprintf_prctl_option(char *bf, size_t size, struct syscall_arg *arg); #define SCA_PRCTL_OPTION syscall_arg__scnprintf_prctl_option +extern struct strarray strarray__prctl_options; + size_t syscall_arg__scnprintf_prctl_arg2(char *bf, size_t size, struct syscall_arg *arg); #define SCA_PRCTL_ARG2 syscall_arg__scnprintf_prctl_arg2 diff --git a/tools/perf/trace/beauty/prctl.c b/tools/perf/trace/beauty/prctl.c index ba2179abed00..6fe5ad5f5d3a 100644 --- a/tools/perf/trace/beauty/prctl.c +++ b/tools/perf/trace/beauty/prctl.c @@ -11,9 +11,10 @@ #include "trace/beauty/generated/prctl_option_array.c" +DEFINE_STRARRAY(prctl_options, "PR_"); + static size_t prctl__scnprintf_option(int option, char *bf, size_t size, bool show_prefix) { - static DEFINE_STRARRAY(prctl_options, "PR_"); return strarray__scnprintf(&strarray__prctl_options, bf, size, "%d", show_prefix, option); } -- cgit v1.2.3 From d7a07b293216e5561705303751bc0d213e9fb328 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Tue, 11 Feb 2020 15:54:08 -0300 Subject: perf trace: Resolve prctl's 'option' arg strings to numbers # perf trace -e syscalls:sys_enter_prctl --filter="option==SET_NAME" 0.000 Socket Thread/3860 syscalls:sys_enter_prctl(option: SET_NAME, arg2: 0x7fc50b9733e8) 0.053 SSL Cert #78/3860 syscalls:sys_enter_prctl(option: SET_NAME, arg2: 0x7fc50b9733e8) ^C # If one uses '-v' with 'perf trace', we can see the filter it puts in place: New filter for syscalls:sys_enter_prctl: (option==0xf) && (common_pid != 3859 && common_pid != 2757) We still need to allow using plain '-e prctl' and have this turn into creating a 'syscalls:sys_enter_prctl' event so that the filter can be applied only to it as right now '-e prctl' ends up using the 'raw_syscalls:sys_enter/sys_exit'. The end goal is to have something like: # perf trace -e prctl/option==SET_NAME/ And have that use tracepoint filters or eBPF ones. Cc: Adrian Hunter Cc: Christian Brauner Cc: Jiri Olsa Cc: Mike Christie Cc: Namhyung Kim Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/builtin-trace.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/tools/perf/builtin-trace.c b/tools/perf/builtin-trace.c index 46a72ecac427..01d542007c8b 100644 --- a/tools/perf/builtin-trace.c +++ b/tools/perf/builtin-trace.c @@ -1065,7 +1065,9 @@ static struct syscall_fmt syscall_fmts[] = { { .name = "poll", .timeout = true, }, { .name = "ppoll", .timeout = true, }, { .name = "prctl", - .arg = { [0] = { .scnprintf = SCA_PRCTL_OPTION, /* option */ }, + .arg = { [0] = { .scnprintf = SCA_PRCTL_OPTION, /* option */ + .strtoul = STUL_STRARRAY, + .parm = &strarray__prctl_options, }, [1] = { .scnprintf = SCA_PRCTL_ARG2, /* arg2 */ }, [2] = { .scnprintf = SCA_PRCTL_ARG3, /* arg3 */ }, }, }, { .name = "pread", .alias = "pread64", }, -- cgit v1.2.3 From d6d829d92c6e82b2627d3bb0058403ff15ee0592 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Tue, 11 Feb 2020 16:03:47 -0300 Subject: tools headers UAPI: Sync sched.h with the kernel To get the changes in: 769071ac9f20 ("ns: Introduce Time Namespace") Silencing this tools/perf build warning: Warning: Kernel ABI header at 'tools/include/uapi/linux/sched.h' differs from latest version at 'include/uapi/linux/sched.h' diff -u tools/include/uapi/linux/sched.h include/uapi/linux/sched.h Which enables 'perf trace' to decode the CLONE_NEWTIME bit in the 'flags' argument to the clone syscalls. Example of clone flags being decoded: [root@quaco ~]# perf trace -e clone* 0.000 qemu-system-x8/23923 clone(clone_flags: VM|FS|FILES|SIGHAND|THREAD|SYSVSEM|SETTLS|PARENT_SETTID|CHILD_CLEARTID, newsp: 0x7f0dad7f9870, parent_tidptr: 0x7f0dad7fa9d0, child_tidptr: 0x7f0dad7fa9d0, tls: 0x7f0dad7fa700) = 6806 (qemu-system-x86) ? qemu-system-x8/6806 ... [continued]: clone()) = 0 ^C[root@quaco ~]# At some point this should enable things like: # perf trace -e 'clone*/clone_flags&NEWTIME/' Cc: Adrian Hunter Cc: Andrei Vagin Cc: Jiri Olsa Cc: Namhyung Kim Cc: Thomas Gleixner Signed-off-by: Arnaldo Carvalho de Melo --- tools/include/uapi/linux/sched.h | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/tools/include/uapi/linux/sched.h b/tools/include/uapi/linux/sched.h index 4a0217832464..2e3bc22c6f20 100644 --- a/tools/include/uapi/linux/sched.h +++ b/tools/include/uapi/linux/sched.h @@ -36,6 +36,12 @@ /* Flags for the clone3() syscall. */ #define CLONE_CLEAR_SIGHAND 0x100000000ULL /* Clear any signal handler and reset to SIG_DFL. */ +/* + * cloning flags intersect with CSIGNAL so can be used with unshare and clone3 + * syscalls only: + */ +#define CLONE_NEWTIME 0x00000080 /* New time namespace */ + #ifndef __ASSEMBLY__ /** * struct clone_args - arguments for the clone3 syscall -- cgit v1.2.3 From f65b9dba5733de8e285cf9d7e8672b46dd0cb709 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Wed, 12 Feb 2020 10:13:25 -0300 Subject: tools headers uapi: Sync linux/fscrypt.h with the kernel sources To pick the changes from: e933adde6f97 ("fscrypt: include in UAPI header") 93edd392cad7 ("fscrypt: support passing a keyring key to FS_IOC_ADD_ENCRYPTION_KEY") That don't trigger any changes in tooling. This silences this perf build warning: Warning: Kernel ABI header at 'tools/include/uapi/linux/fscrypt.h' differs from latest version at 'include/uapi/linux/fscrypt.h' diff -u tools/include/uapi/linux/fscrypt.h include/uapi/linux/fscrypt.h Cc: Adrian Hunter Cc: Eric Biggers Cc: Jiri Olsa Cc: Namhyung Kim Signed-off-by: Arnaldo Carvalho de Melo --- tools/include/uapi/linux/fscrypt.h | 14 +++++++++++++- 1 file changed, 13 insertions(+), 1 deletion(-) diff --git a/tools/include/uapi/linux/fscrypt.h b/tools/include/uapi/linux/fscrypt.h index 1beb174ad950..0d8a6f47711c 100644 --- a/tools/include/uapi/linux/fscrypt.h +++ b/tools/include/uapi/linux/fscrypt.h @@ -8,6 +8,7 @@ #ifndef _UAPI_LINUX_FSCRYPT_H #define _UAPI_LINUX_FSCRYPT_H +#include #include /* Encryption policy flags */ @@ -109,11 +110,22 @@ struct fscrypt_key_specifier { } u; }; +/* + * Payload of Linux keyring key of type "fscrypt-provisioning", referenced by + * fscrypt_add_key_arg::key_id as an alternative to fscrypt_add_key_arg::raw. + */ +struct fscrypt_provisioning_key_payload { + __u32 type; + __u32 __reserved; + __u8 raw[]; +}; + /* Struct passed to FS_IOC_ADD_ENCRYPTION_KEY */ struct fscrypt_add_key_arg { struct fscrypt_key_specifier key_spec; __u32 raw_size; - __u32 __reserved[9]; + __u32 key_id; + __u32 __reserved[8]; __u8 raw[]; }; -- cgit v1.2.3 From 365f9cc195a7fae8ac541129cd2a31ad87e46221 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Wed, 12 Feb 2020 10:25:27 -0300 Subject: tools headers UAPI: Sync drm/i915_drm.h with the kernel sources To pick the change in: cc662126b413 ("drm/i915: Introduce DRM_I915_GEM_MMAP_OFFSET") That don't result in any changes in tooling, just silences this perf build warning: Warning: Kernel ABI header at 'tools/include/uapi/drm/i915_drm.h' differs from latest version at 'include/uapi/drm/i915_drm.h' diff -u tools/include/uapi/drm/i915_drm.h include/uapi/drm/i915_drm.h Cc: Abdiel Janulgue Cc: Adrian Hunter Cc: Chris Wilson Cc: Jiri Olsa Cc: Namhyung Kim Signed-off-by: Arnaldo Carvalho de Melo --- tools/include/uapi/drm/i915_drm.h | 32 ++++++++++++++++++++++++++++++++ 1 file changed, 32 insertions(+) diff --git a/tools/include/uapi/drm/i915_drm.h b/tools/include/uapi/drm/i915_drm.h index 5400d7e057f1..829c0a48577f 100644 --- a/tools/include/uapi/drm/i915_drm.h +++ b/tools/include/uapi/drm/i915_drm.h @@ -395,6 +395,7 @@ typedef struct _drm_i915_sarea { #define DRM_IOCTL_I915_GEM_PWRITE DRM_IOW (DRM_COMMAND_BASE + DRM_I915_GEM_PWRITE, struct drm_i915_gem_pwrite) #define DRM_IOCTL_I915_GEM_MMAP DRM_IOWR(DRM_COMMAND_BASE + DRM_I915_GEM_MMAP, struct drm_i915_gem_mmap) #define DRM_IOCTL_I915_GEM_MMAP_GTT DRM_IOWR(DRM_COMMAND_BASE + DRM_I915_GEM_MMAP_GTT, struct drm_i915_gem_mmap_gtt) +#define DRM_IOCTL_I915_GEM_MMAP_OFFSET DRM_IOWR(DRM_COMMAND_BASE + DRM_I915_GEM_MMAP_GTT, struct drm_i915_gem_mmap_offset) #define DRM_IOCTL_I915_GEM_SET_DOMAIN DRM_IOW (DRM_COMMAND_BASE + DRM_I915_GEM_SET_DOMAIN, struct drm_i915_gem_set_domain) #define DRM_IOCTL_I915_GEM_SW_FINISH DRM_IOW (DRM_COMMAND_BASE + DRM_I915_GEM_SW_FINISH, struct drm_i915_gem_sw_finish) #define DRM_IOCTL_I915_GEM_SET_TILING DRM_IOWR (DRM_COMMAND_BASE + DRM_I915_GEM_SET_TILING, struct drm_i915_gem_set_tiling) @@ -793,6 +794,37 @@ struct drm_i915_gem_mmap_gtt { __u64 offset; }; +struct drm_i915_gem_mmap_offset { + /** Handle for the object being mapped. */ + __u32 handle; + __u32 pad; + /** + * Fake offset to use for subsequent mmap call + * + * This is a fixed-size type for 32/64 compatibility. + */ + __u64 offset; + + /** + * Flags for extended behaviour. + * + * It is mandatory that one of the MMAP_OFFSET types + * (GTT, WC, WB, UC, etc) should be included. + */ + __u64 flags; +#define I915_MMAP_OFFSET_GTT 0 +#define I915_MMAP_OFFSET_WC 1 +#define I915_MMAP_OFFSET_WB 2 +#define I915_MMAP_OFFSET_UC 3 + + /* + * Zero-terminated chain of extensions. + * + * No current extensions defined; mbz. + */ + __u64 extensions; +}; + struct drm_i915_gem_set_domain { /** Handle for the object */ __u32 handle; -- cgit v1.2.3 From df5a5f3cf24608457bb5e57297dd9f0d528be58f Mon Sep 17 00:00:00 2001 From: John Garry Date: Fri, 13 Dec 2019 21:54:15 +0800 Subject: perf tools: Add arm64 version of get_cpuid() Add an arm64 version of get_cpuid(), which is used for various annotation and headers - for example, I now get the CPUID in "perf report --header", as shown in this snippet: # hostname : ubuntu # os release : 5.5.0-rc1-dirty # perf version : 5.5.rc1.gbf8a13dc9851 # arch : aarch64 # nrcpus online : 96 # nrcpus avail : 96 # cpuid : 0x00000000480fd010 Since much of the code to read the MIDR is already in get_cpuid_str(), factor out this code. Tester notes: I tested this patch on my new ARM64 Kunpeng 920 server. [root@node1 zsk]# ./perf --version perf version 5.6.rc1.g2cdb955b7252 Both perf list and perf stat can work. Signed-off-by: John Garry Tested-by: Shaokun Zhang Cc: Alexander Shishkin Cc: Jiri Olsa Cc: Mark Rutland Cc: Namhyung Kim Cc: Peter Zijlstra Cc: Will Deacon Cc: linux-arm-kernel@lists.infradead.org Cc: linuxarm@huawei.com Link: http://lore.kernel.org/lkml/1576245255-210926-1-git-send-email-john.garry@huawei.com Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/arch/arm64/util/header.c | 63 ++++++++++++++++++++++++++++--------- 1 file changed, 48 insertions(+), 15 deletions(-) diff --git a/tools/perf/arch/arm64/util/header.c b/tools/perf/arch/arm64/util/header.c index a32e4b72a98f..d730666ab95d 100644 --- a/tools/perf/arch/arm64/util/header.c +++ b/tools/perf/arch/arm64/util/header.c @@ -1,8 +1,10 @@ #include #include #include +#include #include #include +#include #include "debug.h" #include "header.h" @@ -12,26 +14,21 @@ #define MIDR_VARIANT_SHIFT 20 #define MIDR_VARIANT_MASK (0xf << MIDR_VARIANT_SHIFT) -char *get_cpuid_str(struct perf_pmu *pmu) +static int _get_cpuid(char *buf, size_t sz, struct perf_cpu_map *cpus) { - char *buf = NULL; - char path[PATH_MAX]; const char *sysfs = sysfs__mountpoint(); - int cpu; u64 midr = 0; - struct perf_cpu_map *cpus; - FILE *file; + int cpu; - if (!sysfs || !pmu || !pmu->cpus) - return NULL; + if (!sysfs || sz < MIDR_SIZE) + return EINVAL; - buf = malloc(MIDR_SIZE); - if (!buf) - return NULL; + cpus = perf_cpu_map__get(cpus); - /* read midr from list of cpus mapped to this pmu */ - cpus = perf_cpu_map__get(pmu->cpus); for (cpu = 0; cpu < perf_cpu_map__nr(cpus); cpu++) { + char path[PATH_MAX]; + FILE *file; + scnprintf(path, PATH_MAX, "%s/devices/system/cpu/cpu%d"MIDR, sysfs, cpus->map[cpu]); @@ -57,12 +54,48 @@ char *get_cpuid_str(struct perf_pmu *pmu) break; } - if (!midr) { + perf_cpu_map__put(cpus); + + if (!midr) + return EINVAL; + + return 0; +} + +int get_cpuid(char *buf, size_t sz) +{ + struct perf_cpu_map *cpus = perf_cpu_map__new(NULL); + int ret; + + if (!cpus) + return EINVAL; + + ret = _get_cpuid(buf, sz, cpus); + + perf_cpu_map__put(cpus); + + return ret; +} + +char *get_cpuid_str(struct perf_pmu *pmu) +{ + char *buf = NULL; + int res; + + if (!pmu || !pmu->cpus) + return NULL; + + buf = malloc(MIDR_SIZE); + if (!buf) + return NULL; + + /* read midr from list of cpus mapped to this pmu */ + res = _get_cpuid(buf, MIDR_SIZE, pmu->cpus); + if (res) { pr_err("failed to get cpuid string for PMU %s\n", pmu->name); free(buf); buf = NULL; } - perf_cpu_map__put(cpus); return buf; } -- cgit v1.2.3 From 47f8d94ac5241e30a5ab0b6b91b963b54765ec7c Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Wed, 12 Feb 2020 10:53:06 -0300 Subject: tools headers UAPI: Sync asm-generic/mman-common.h with the kernel To pick the changes from: d41938d2cbee ("mm: Reserve asm-generic prot flags 0x10 and 0x20 for arch use") No changes in tooling, just a rebuild as files needed got touched. This addresses the following perf build warning: Warning: Kernel ABI header at 'tools/include/uapi/asm-generic/mman-common.h' differs from latest version at 'include/uapi/asm-generic/mman-common.h' diff -u tools/include/uapi/asm-generic/mman-common.h include/uapi/asm-generic/mman-common.h Cc: Adrian Hunter Cc: Dave Martin Cc: Jiri Olsa Cc: Namhyung Kim Cc: Will Deacon Signed-off-by: Arnaldo Carvalho de Melo --- tools/include/uapi/asm-generic/mman-common.h | 2 ++ 1 file changed, 2 insertions(+) diff --git a/tools/include/uapi/asm-generic/mman-common.h b/tools/include/uapi/asm-generic/mman-common.h index c160a5354eb6..f94f65d429be 100644 --- a/tools/include/uapi/asm-generic/mman-common.h +++ b/tools/include/uapi/asm-generic/mman-common.h @@ -11,6 +11,8 @@ #define PROT_WRITE 0x2 /* page can be written */ #define PROT_EXEC 0x4 /* page can be executed */ #define PROT_SEM 0x8 /* page may be used for atomic ops */ +/* 0x10 reserved for arch-specific use */ +/* 0x20 reserved for arch-specific use */ #define PROT_NONE 0x0 /* page can not be accessed */ #define PROT_GROWSDOWN 0x01000000 /* mprotect flag: extend change to start of growsdown vma */ #define PROT_GROWSUP 0x02000000 /* mprotect flag: extend change to end of growsup vma */ -- cgit v1.2.3 From 8c65582f82ee736b63b3c0cd9c7c5b4572f1f4d6 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Wed, 12 Feb 2020 11:04:23 -0300 Subject: tools include UAPI: Sync sound/asound.h copy Picking the changes from: 46b770f720bd ("ALSA: uapi: Fix sparse warning") a103a3989993 ("ALSA: control: Fix incompatible protocol error") bd3eb4e87eb3 ("ALSA: ctl: bump protocol version up to v2.1.0") ff16351e3f30 ("ALSA: ctl: remove dimen member from elem_info structure") 542283566679 ("ALSA: ctl: remove unused macro for timestamping of elem_value") 7fd7d6c50451 ("ALSA: uapi: Fix typos and header inclusion in asound.h") 1cfaef961703 ("ALSA: bump uapi version numbers") 80fe7430c708 ("ALSA: add new 32-bit layout for snd_pcm_mmap_status/control") 07094ae6f952 ("ALSA: Avoid using timespec for struct snd_timer_tread") d9e5582c4bb2 ("ALSA: Avoid using timespec for struct snd_rawmidi_status") 3ddee7f88aaf ("ALSA: Avoid using timespec for struct snd_pcm_status") a4e7dd35b9da ("ALSA: Avoid using timespec for struct snd_ctl_elem_value") a07804cc7472 ("ALSA: Avoid using timespec for struct snd_timer_status") Which entails no changes in the tooling side. To silence this perf tools build warning: Warning: Kernel ABI header at 'tools/include/uapi/sound/asound.h' differs from latest version at 'include/uapi/sound/asound.h' diff -u tools/include/uapi/sound/asound.h include/uapi/sound/asound.h Cc: Adrian Hunter Cc: Arnd Bergmann Cc: Baolin Wang Cc: Jiri Olsa Cc: Namhyung Kim Cc: Ranjani Sridharan Cc: Takashi Iwai Cc: Takashi Sakamoto Signed-off-by: Arnaldo Carvalho de Melo --- tools/include/uapi/sound/asound.h | 155 ++++++++++++++++++++++++++++++++------ 1 file changed, 132 insertions(+), 23 deletions(-) diff --git a/tools/include/uapi/sound/asound.h b/tools/include/uapi/sound/asound.h index df1153cea0b7..535a7229e1d9 100644 --- a/tools/include/uapi/sound/asound.h +++ b/tools/include/uapi/sound/asound.h @@ -26,7 +26,9 @@ #if defined(__KERNEL__) || defined(__linux__) #include +#include #else +#include #include #endif @@ -154,7 +156,7 @@ struct snd_hwdep_dsp_image { * * *****************************************************************************/ -#define SNDRV_PCM_VERSION SNDRV_PROTOCOL_VERSION(2, 0, 14) +#define SNDRV_PCM_VERSION SNDRV_PROTOCOL_VERSION(2, 0, 15) typedef unsigned long snd_pcm_uframes_t; typedef signed long snd_pcm_sframes_t; @@ -301,7 +303,9 @@ typedef int __bitwise snd_pcm_subformat_t; #define SNDRV_PCM_INFO_DRAIN_TRIGGER 0x40000000 /* internal kernel flag - trigger in drain */ #define SNDRV_PCM_INFO_FIFO_IN_FRAMES 0x80000000 /* internal kernel flag - FIFO size is in frames */ - +#if (__BITS_PER_LONG == 32 && defined(__USE_TIME_BITS64)) || defined __KERNEL__ +#define __SND_STRUCT_TIME64 +#endif typedef int __bitwise snd_pcm_state_t; #define SNDRV_PCM_STATE_OPEN ((__force snd_pcm_state_t) 0) /* stream is open */ @@ -317,8 +321,17 @@ typedef int __bitwise snd_pcm_state_t; enum { SNDRV_PCM_MMAP_OFFSET_DATA = 0x00000000, - SNDRV_PCM_MMAP_OFFSET_STATUS = 0x80000000, - SNDRV_PCM_MMAP_OFFSET_CONTROL = 0x81000000, + SNDRV_PCM_MMAP_OFFSET_STATUS_OLD = 0x80000000, + SNDRV_PCM_MMAP_OFFSET_CONTROL_OLD = 0x81000000, + SNDRV_PCM_MMAP_OFFSET_STATUS_NEW = 0x82000000, + SNDRV_PCM_MMAP_OFFSET_CONTROL_NEW = 0x83000000, +#ifdef __SND_STRUCT_TIME64 + SNDRV_PCM_MMAP_OFFSET_STATUS = SNDRV_PCM_MMAP_OFFSET_STATUS_NEW, + SNDRV_PCM_MMAP_OFFSET_CONTROL = SNDRV_PCM_MMAP_OFFSET_CONTROL_NEW, +#else + SNDRV_PCM_MMAP_OFFSET_STATUS = SNDRV_PCM_MMAP_OFFSET_STATUS_OLD, + SNDRV_PCM_MMAP_OFFSET_CONTROL = SNDRV_PCM_MMAP_OFFSET_CONTROL_OLD, +#endif }; union snd_pcm_sync_id { @@ -456,8 +469,13 @@ enum { SNDRV_PCM_AUDIO_TSTAMP_TYPE_LAST = SNDRV_PCM_AUDIO_TSTAMP_TYPE_LINK_SYNCHRONIZED }; +#ifndef __KERNEL__ +/* explicit padding avoids incompatibility between i386 and x86-64 */ +typedef struct { unsigned char pad[sizeof(time_t) - sizeof(int)]; } __time_pad; + struct snd_pcm_status { snd_pcm_state_t state; /* stream state */ + __time_pad pad1; /* align to timespec */ struct timespec trigger_tstamp; /* time when stream was started/stopped/paused */ struct timespec tstamp; /* reference timestamp */ snd_pcm_uframes_t appl_ptr; /* appl ptr */ @@ -473,17 +491,48 @@ struct snd_pcm_status { __u32 audio_tstamp_accuracy; /* in ns units, only valid if indicated in audio_tstamp_data */ unsigned char reserved[52-2*sizeof(struct timespec)]; /* must be filled with zero */ }; +#endif + +/* + * For mmap operations, we need the 64-bit layout, both for compat mode, + * and for y2038 compatibility. For 64-bit applications, the two definitions + * are identical, so we keep the traditional version. + */ +#ifdef __SND_STRUCT_TIME64 +#define __snd_pcm_mmap_status64 snd_pcm_mmap_status +#define __snd_pcm_mmap_control64 snd_pcm_mmap_control +#define __snd_pcm_sync_ptr64 snd_pcm_sync_ptr +#ifdef __KERNEL__ +#define __snd_timespec64 __kernel_timespec +#else +#define __snd_timespec64 timespec +#endif +struct __snd_timespec { + __s32 tv_sec; + __s32 tv_nsec; +}; +#else +#define __snd_pcm_mmap_status snd_pcm_mmap_status +#define __snd_pcm_mmap_control snd_pcm_mmap_control +#define __snd_pcm_sync_ptr snd_pcm_sync_ptr +#define __snd_timespec timespec +struct __snd_timespec64 { + __s64 tv_sec; + __s64 tv_nsec; +}; -struct snd_pcm_mmap_status { +#endif + +struct __snd_pcm_mmap_status { snd_pcm_state_t state; /* RO: state - SNDRV_PCM_STATE_XXXX */ int pad1; /* Needed for 64 bit alignment */ snd_pcm_uframes_t hw_ptr; /* RO: hw ptr (0...boundary-1) */ - struct timespec tstamp; /* Timestamp */ + struct __snd_timespec tstamp; /* Timestamp */ snd_pcm_state_t suspended_state; /* RO: suspended stream state */ - struct timespec audio_tstamp; /* from sample counter or