From cde88355f2cd7082f5a208901f99b80f4823872c Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Thu, 24 Mar 2016 11:22:43 -0300 Subject: [PATCH 01/10] perf tests: Fix tarpkg build test error output redirection MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit So we need to trow away just stdout, leaving stderr to be caught by the build tests infrastructure, so that we can see what went wrong when the tarpkg build test fails: $ make -C tools/perf build-test make: Entering directory '/home/acme/git/linux/tools/perf' - tarpkg: ./tests/perf-targz-src-pkg . tests/make:302: recipe for target 'tarpkg' failed make[1]: *** [tarpkg] Error 2 Makefile:102: recipe for target 'build-test' failed make: *** [build-test] Error 2 make: Leaving directory '/home/acme/git/linux/tools/perf' $ cat tools/perf/tarpkg ./tests/perf-targz-src-pkg . PERF_VERSION = 4.5.g05f5ec PERF_VERSION = 4.5.g05f5ec In file included from bench/mem-memcpy-x86-64-asm.S:9:0: bench/../../../arch/x86/lib/memcpy_64.S:5:29: fatal error: asm/cpufeatures.h: No such file or directory compilation terminated. mv: cannot stat ‘bench/.mem-memcpy-x86-64-asm.o.tmp’: No such file or directory make[5]: *** [bench/mem-memcpy-x86-64-asm.o] Error 1 make[5]: *** Waiting for unfinished jobs.... make[4]: *** [bench] Error 2 make[4]: *** Waiting for unfinished jobs.... make[3]: *** [perf-in.o] Error 2 make[3]: *** Waiting for unfinished jobs.... make[2]: *** [all] Error 2 $ So the test flow is: 1. Run: 'make -C tools/perf build-test' 2. One of its tests failed, in this case, the 'tarpkg' one 3. Look at what went wrong, by looking at the output of that test, in tools/perf/tarpkg Admittedly, this should be shortcircuited to showing what went wrong directly from the 'make build-test' step, but lets first fix this tarpkg one and the problem it spotted, which should be fixed by adding some extra file to the tools/perf/MANIFEST so that detached tarballs continue being self contained and build successfully. Cc: Adrian Hunter Cc: Borislav Petkov Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-ynld6egoxolmftcddpnd7oh6@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/tests/perf-targz-src-pkg | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/perf/tests/perf-targz-src-pkg b/tools/perf/tests/perf-targz-src-pkg index 238aa3927c71..f2d9c5fe58e0 100755 --- a/tools/perf/tests/perf-targz-src-pkg +++ b/tools/perf/tests/perf-targz-src-pkg @@ -15,7 +15,7 @@ TMP_DEST=$(mktemp -d) tar xf ${TARBALL} -C $TMP_DEST rm -f ${TARBALL} cd - > /dev/null -make -C $TMP_DEST/perf*/tools/perf > /dev/null 2>&1 +make -C $TMP_DEST/perf*/tools/perf > /dev/null RC=$? rm -rf ${TMP_DEST} exit $RC From 6a1a77baa7d81c8217ed10ba17a84c5f100343d9 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Thu, 24 Mar 2016 11:48:23 -0300 Subject: [PATCH 02/10] perf bench: Fix detached tarball building due to missing 'perf bench memcpy' headers MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit A change on kernel files included by the 'perf bench memcpy' code grew some new include deps, breaking the detached tarball build: $ make -C tools/perf build-test make: Entering directory '/home/acme/git/linux/tools/perf' - tarpkg: ./tests/perf-targz-src-pkg . tests/make:302: recipe for target 'tarpkg' failed make[1]: *** [tarpkg] Error 2 Makefile:102: recipe for target 'build-test' failed make: *** [build-test] Error 2 make: Leaving directory '/home/acme/git/linux/tools/perf' $ cat tools/perf/tarpkg ./tests/perf-targz-src-pkg . PERF_VERSION = 4.5.g05f5ec PERF_VERSION = 4.5.g05f5ec In file included from bench/mem-memcpy-x86-64-asm.S:9:0: bench/../../../arch/x86/lib/memcpy_64.S:5:29: fatal error: asm/cpufeatures.h: No such file or directory compilation terminated. mv: cannot stat ‘bench/.mem-memcpy-x86-64-asm.o.tmp’: No such file or directory make[5]: *** [bench/mem-memcpy-x86-64-asm.o] Error 1 make[5]: *** Waiting for unfinished jobs.... make[4]: *** [bench] Error 2 make[4]: *** Waiting for unfinished jobs.... make[3]: *** [perf-in.o] Error 2 make[3]: *** Waiting for unfinished jobs.... make[2]: *** [all] Error 2 $ Add arch/*/include/asm/*features.h to tools/perf/MANIFEST so that we can continue to use detached tarballs to build perf. Now it builds ok, doing it manually: $ make help | grep perf perf-tar-src-pkg - Build perf-4.5.0.tar source tarball perf-targz-src-pkg - Build perf-4.5.0.tar.gz source tarball perf-tarbz2-src-pkg - Build perf-4.5.0.tar.bz2 source tarball perf-tarxz-src-pkg - Build perf-4.5.0.tar.xz source tarball $ ls -la perf-4.5.0.tar ls: cannot access perf-4.5.0.tar: No such file or directory $ make perf-tar-src-pkg TAR PERF_VERSION = 4.5.g32c25b $ ls -la perf-4.5.0.tar -rw-rw-r--. 1 acme acme 6318080 Mar 24 11:52 perf-4.5.0.tar $ mv perf-4.5.0.tar /tmp $ cd /tmp $ tar xf perf-4.5.0.tar $ cd perf-4.5.0/tools/perf $ make > /dev/null PERF_VERSION = 4.5.g32c25b $ ls -la perf -rwxrwxr-x. 1 acme acme 14046416 Mar 24 11:53 perf $ ./perf --version perf version 4.5.g32c25b $ perf bench Usage: perf bench [] [] # List of all available benchmark collections: sched: Scheduler and IPC benchmarks mem: Memory access benchmarks numa: NUMA scheduling and MM benchmarks futex: Futex stressing benchmarks all: All benchmarks $ perf bench mem # List of available benchmarks for collection 'mem': memcpy: Benchmark for memcpy() functions memset: Benchmark for memset() functions all: Run all memory access benchmarks $ perf bench mem memcpy # Running 'mem/memcpy' benchmark: # function 'default' (Default memcpy() provided by glibc) # Copying 1MB bytes ... 15.024038 GB/sec # function 'x86-64-unrolled' (unrolled memcpy() in arch/x86/lib/memcpy_64.S) # Copying 1MB bytes ... 17.438616 GB/sec # function 'x86-64-movsq' (movsq-based memcpy() in arch/x86/lib/memcpy_64.S) # Copying 1MB bytes ... 25.040064 GB/sec # function 'x86-64-movsb' (movsb-based memcpy() in arch/x86/lib/memcpy_64.S) # Copying 1MB bytes ... 25.040064 GB/sec $ Cc: Adrian Hunter Cc: Borislav Petkov Cc: David Ahern Cc: Hitoshi Mitake Cc: Jiri Olsa Cc: Namhyung Kim Cc: Wang Nan Link: http://lkml.kernel.org/n/tip-2c2sncwffuabw58fj1pw86gu@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/MANIFEST | 1 + 1 file changed, 1 insertion(+) diff --git a/tools/perf/MANIFEST b/tools/perf/MANIFEST index 2e1fa2357528..8c8c6b9ce915 100644 --- a/tools/perf/MANIFEST +++ b/tools/perf/MANIFEST @@ -74,6 +74,7 @@ arch/*/include/uapi/asm/unistd*.h arch/*/include/uapi/asm/perf_regs.h arch/*/lib/memcpy*.S arch/*/lib/memset*.S +arch/*/include/asm/*features.h include/linux/poison.h include/linux/hw_breakpoint.h include/uapi/linux/perf_event.h From a49ac9f83b31e41c8311d64bd2b3f97a23dcb38d Mon Sep 17 00:00:00 2001 From: Huang Rui Date: Fri, 25 Mar 2016 11:18:25 +0800 Subject: [PATCH 03/10] perf/x86: Move events_sysfs_show() outside CPU_SUP_INTEL randconfig builds can sometimes disable CONFIG_CPU_SUP_INTEL while enabling the AMD power reporting PMU driver, resulting in this build failure: arch/x86/kernel/cpu/perf_event.h:663:31: error: 'events_sysfs_show' undeclared here (not in a function) To fix it, move events_sysfs_show() outside of #ifdef CONFIG_CPU_SUP_INTEL. Reported-by: Randy Dunlap Reported-by: build test robot Signed-off-by: Huang Rui Cc: Borislav Petkov Cc: Borislav Petkov Cc: Fengguang Wu Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Sherry Hurwitz Cc: Stephen Rothwell Cc: Thomas Gleixner Cc: acme@kernel.org Cc: kbuild-all@01.org Cc: linux-next@vger.kernel.org Cc: spg_linux_kernel@amd.com Link: http://lkml.kernel.org/r/1458875905-4278-1-git-send-email-ray.huang@amd.com Signed-off-by: Ingo Molnar --- arch/x86/events/perf_event.h | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/arch/x86/events/perf_event.h b/arch/x86/events/perf_event.h index ba6ef18528c9..a6771e2303d2 100644 --- a/arch/x86/events/perf_event.h +++ b/arch/x86/events/perf_event.h @@ -795,6 +795,9 @@ ssize_t intel_event_sysfs_show(char *page, u64 config); struct attribute **merge_attr(struct attribute **a, struct attribute **b); +ssize_t events_sysfs_show(struct device *dev, struct device_attribute *attr, + char *page); + #ifdef CONFIG_CPU_SUP_AMD int amd_pmu_init(void); @@ -925,9 +928,6 @@ int p6_pmu_init(void); int knc_pmu_init(void); -ssize_t events_sysfs_show(struct device *dev, struct device_attribute *attr, - char *page); - static inline int is_ht_workaround_enabled(void) { return !!(x86_pmu.flags & PMU_FL_EXCL_ENABLED); From 379649cfeafef29b77d29860741f4d1c5870b65f Mon Sep 17 00:00:00 2001 From: Sukadev Bhattiprolu Date: Mon, 28 Mar 2016 09:31:41 -0300 Subject: [PATCH 04/10] perf tools: Fix build break on powerpc Commit 531d2410635c ("perf tools: Do not include stringify.h from the kernel sources") seems to have accidentially removed the inclusion of "util/header.h" from "arch/powerpc/util/header.c". "util/header.h" provides the prototype for get_cpuid() and is needed to build perf on Powerpc: arch/powerpc/util/header.c:17:1: error: no previous prototype for 'get_cpuid' [-Werror=missing-prototypes] Reported-by: Michael Ellerman Tested-by: Arnaldo Carvalho de Melo Cc: Stephen Rothwell Fixes: 531d2410635c ("perf tools: Do not include stringify.h from the kernel sources") Signed-off-by: Sukadev Bhattiprolu [ Included "util.h" too, to get the scnprintf() prototype ] Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/arch/powerpc/util/header.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/tools/perf/arch/powerpc/util/header.c b/tools/perf/arch/powerpc/util/header.c index 6138bdef6e63..f8ccee132867 100644 --- a/tools/perf/arch/powerpc/util/header.c +++ b/tools/perf/arch/powerpc/util/header.c @@ -4,6 +4,8 @@ #include #include #include +#include "header.h" +#include "util.h" #define mfspr(rn) ({unsigned long rval; \ asm volatile("mfspr %0," __stringify(rn) \ From 3ea223adcb0c5893a6dc8ed3a84dce264cbb61d6 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Tue, 29 Mar 2016 18:46:04 -0300 Subject: [PATCH 05/10] perf tools: Add missing initialization of perf_sample.cpumode in synthesized samples In 473398a21d28 ("perf tools: Add cpumode to struct perf_sample"), I missed some places where perf_sample fields are directly initialized in addition to what is done in perf_evsel__parse_sample(), namely when synthesizing PERF_RECORD_{MMAP*,COMM,FORK,EXIT} for pre-existing threads and also in intel_pt and intel_bts when synthesizing events from processor trace, the jitdump code also was affected, fix it. The problem was noticed with running: # perf record -e intel_pt//u true # perf script Where the samples wouldn't get resolved because perf_sample.cpumode would be left as zero, i.e. PERF_RECORD_MISC_CPUMODE_UNKNOWN, not resolving as kernel, hypervisor or user cpu modes. Cc: Alexander Shishkin Cc: Andi Kleen Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Peter Zijlstra Cc: Stephane Eranian Cc: Thomas Gleixner Fixes: 473398a21d28 ("perf tools: Add cpumode to struct perf_sample") Link: http://lkml.kernel.org/n/tip-n5sdauxgk24d5nun8kuuu2mh@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/event.c | 23 ++++++++++++++++------- tools/perf/util/intel-bts.c | 1 + tools/perf/util/intel-pt.c | 3 +++ tools/perf/util/jitdump.c | 2 ++ 4 files changed, 22 insertions(+), 7 deletions(-) diff --git a/tools/perf/util/event.c b/tools/perf/util/event.c index 52cf479bc593..dad55d04ffdd 100644 --- a/tools/perf/util/event.c +++ b/tools/perf/util/event.c @@ -56,13 +56,22 @@ const char *perf_event__name(unsigned int id) return perf_event__names[id]; } -static struct perf_sample synth_sample = { +static int perf_tool__process_synth_event(struct perf_tool *tool, + union perf_event *event, + struct machine *machine, + perf_event__handler_t process) +{ + struct perf_sample synth_sample = { .pid = -1, .tid = -1, .time = -1, .stream_id = -1, .cpu = -1, .period = 1, + .cpumode = event->header.misc & PERF_RECORD_MISC_CPUMODE_MASK, + }; + + return process(tool, event, &synth_sample, machine); }; /* @@ -186,7 +195,7 @@ pid_t perf_event__synthesize_comm(struct perf_tool *tool, if (perf_event__prepare_comm(event, pid, machine, &tgid, &ppid) != 0) return -1; - if (process(tool, event, &synth_sample, machine) != 0) + if (perf_tool__process_synth_event(tool, event, machine, process) != 0) return -1; return tgid; @@ -218,7 +227,7 @@ static int perf_event__synthesize_fork(struct perf_tool *tool, event->fork.header.size = (sizeof(event->fork) + machine->id_hdr_size); - if (process(tool, event, &synth_sample, machine) != 0) + if (perf_tool__process_synth_event(tool, event, machine, process) != 0) return -1; return 0; @@ -344,7 +353,7 @@ out: event->mmap2.pid = tgid; event->mmap2.tid = pid; - if (process(tool, event, &synth_sample, machine) != 0) { + if (perf_tool__process_synth_event(tool, event, machine, process) != 0) { rc = -1; break; } @@ -402,7 +411,7 @@ int perf_event__synthesize_modules(struct perf_tool *tool, memcpy(event->mmap.filename, pos->dso->long_name, pos->dso->long_name_len + 1); - if (process(tool, event, &synth_sample, machine) != 0) { + if (perf_tool__process_synth_event(tool, event, machine, process) != 0) { rc = -1; break; } @@ -472,7 +481,7 @@ static int __event__synthesize_thread(union perf_event *comm_event, /* * Send the prepared comm event */ - if (process(tool, comm_event, &synth_sample, machine) != 0) + if (perf_tool__process_synth_event(tool, comm_event, machine, process) != 0) break; rc = 0; @@ -701,7 +710,7 @@ int perf_event__synthesize_kernel_mmap(struct perf_tool *tool, event->mmap.len = map->end - event->mmap.start; event->mmap.pid = machine->pid; - err = process(tool, event, &synth_sample, machine); + err = perf_tool__process_synth_event(tool, event, machine, process); free(event); return err; diff --git a/tools/perf/util/intel-bts.c b/tools/perf/util/intel-bts.c index 6bc3ecd2e7ca..abf1366e2a24 100644 --- a/tools/perf/util/intel-bts.c +++ b/tools/perf/util/intel-bts.c @@ -279,6 +279,7 @@ static int intel_bts_synth_branch_sample(struct intel_bts_queue *btsq, event.sample.header.misc = PERF_RECORD_MISC_USER; event.sample.header.size = sizeof(struct perf_event_header); + sample.cpumode = PERF_RECORD_MISC_USER; sample.ip = le64_to_cpu(branch->from); sample.pid = btsq->pid; sample.tid = btsq->tid; diff --git a/tools/perf/util/intel-pt.c b/tools/perf/util/intel-pt.c index 05d815851be1..407f11b97c8d 100644 --- a/tools/perf/util/intel-pt.c +++ b/tools/perf/util/intel-pt.c @@ -979,6 +979,7 @@ static int intel_pt_synth_branch_sample(struct intel_pt_queue *ptq) if (!pt->timeless_decoding) sample.time = tsc_to_perf_time(ptq->timestamp, &pt->tc); + sample.cpumode = PERF_RECORD_MISC_USER; sample.ip = ptq->state->from_ip; sample.pid = ptq->pid; sample.tid = ptq->tid; @@ -1035,6 +1036,7 @@ static int intel_pt_synth_instruction_sample(struct intel_pt_queue *ptq) if (!pt->timeless_decoding) sample.time = tsc_to_perf_time(ptq->timestamp, &pt->tc); + sample.cpumode = PERF_RECORD_MISC_USER; sample.ip = ptq->state->from_ip; sample.pid = ptq->pid; sample.tid = ptq->tid; @@ -1092,6 +1094,7 @@ static int intel_pt_synth_transaction_sample(struct intel_pt_queue *ptq) if (!pt->timeless_decoding) sample.time = tsc_to_perf_time(ptq->timestamp, &pt->tc); + sample.cpumode = PERF_RECORD_MISC_USER; sample.ip = ptq->state->from_ip; sample.pid = ptq->pid; sample.tid = ptq->tid; diff --git a/tools/perf/util/jitdump.c b/tools/perf/util/jitdump.c index cd272cc21e05..ad0c0bb1fbc7 100644 --- a/tools/perf/util/jitdump.c +++ b/tools/perf/util/jitdump.c @@ -417,6 +417,7 @@ static int jit_repipe_code_load(struct jit_buf_desc *jd, union jr_entry *jr) * use first address as sample address */ memset(&sample, 0, sizeof(sample)); + sample.cpumode = PERF_RECORD_MISC_USER; sample.pid = pid; sample.tid = tid; sample.time = id->time; @@ -505,6 +506,7 @@ static int jit_repipe_code_move(struct jit_buf_desc *jd, union jr_entry *jr) * use first address as sample address */ memset(&sample, 0, sizeof(sample)); + sample.cpumode = PERF_RECORD_MISC_USER; sample.pid = pid; sample.tid = tid; sample.time = id->time; From 909890355507e92bdaf648e73870f6b5df606da8 Mon Sep 17 00:00:00 2001 From: Andres Freund Date: Wed, 30 Mar 2016 21:02:45 +0200 Subject: [PATCH 06/10] perf hists: Fix determination of a callchain node's childlessness The 4b3a3212233a ("perf hists browser: Support flat callchains") commit over-aggressively tried to optimize callchain_node__init_have_children(). That lead to --tui mode not allowing to expand call chain elements if a call chain element had only one parent. That's why --inverted callgraphs looked halfway sane, but plain ones didn't. Revert that individual optimization, it wasn't really related to the rest of the commit. Signed-off-by: Andres Freund Acked-by: Namhyung Kim Tested-by: Arnaldo Carvalho de Melo Cc: Peter Zijlstra Fixes: 4b3a3212233a ("perf hists browser: Support flat callchains") Link: http://lkml.kernel.org/r/20160330190245.GB13305@awork2.anarazel.de Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/ui/browsers/hists.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/perf/ui/browsers/hists.c b/tools/perf/ui/browsers/hists.c index 4b9816555946..2a83414159a6 100644 --- a/tools/perf/ui/browsers/hists.c +++ b/tools/perf/ui/browsers/hists.c @@ -337,7 +337,7 @@ static void callchain_node__init_have_children(struct callchain_node *node, chain = list_entry(node->val.next, struct callchain_list, list); chain->has_children = has_sibling; - if (node->val.next != node->val.prev) { + if (!list_empty(&node->val)) { chain = list_entry(node->val.prev, struct callchain_list, list); chain->has_children = !RB_EMPTY_ROOT(&node->rb_root); } From 9f56c092b99b40ce3cf4c6d0134ff7e513c9f1a6 Mon Sep 17 00:00:00 2001 From: Anton Blanchard Date: Tue, 29 Mar 2016 17:59:44 +1100 Subject: [PATCH 07/10] perf jit: genelf makes assumptions about endian Commit 9b07e27f88b9 ("perf inject: Add jitdump mmap injection support") incorrectly assumed that PowerPC is big endian only. Simplify things by consolidating the define of GEN_ELF_ENDIAN and checking for __BYTE_ORDER == __BIG_ENDIAN. The PowerPC checks were also incorrect, they do not match what gcc emits. We should first look for __powerpc64__, then __powerpc__. Signed-off-by: Anton Blanchard Acked-by: Michael Ellerman Cc: Carl Love Cc: Stephane Eranian Cc: Sukadev Bhattiprolu Cc: linuxppc-dev@lists.ozlabs.org Fixes: 9b07e27f88b9 ("perf inject: Add jitdump mmap injection support") Link: http://lkml.kernel.org/r/20160329175944.33a211cc@kryten Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/genelf.h | 24 ++++++++++-------------- 1 file changed, 10 insertions(+), 14 deletions(-) diff --git a/tools/perf/util/genelf.h b/tools/perf/util/genelf.h index cd67e64a0494..2fbeb59c4bdd 100644 --- a/tools/perf/util/genelf.h +++ b/tools/perf/util/genelf.h @@ -9,36 +9,32 @@ int jit_add_debug_info(Elf *e, uint64_t code_addr, void *debug, int nr_debug_ent #if defined(__arm__) #define GEN_ELF_ARCH EM_ARM -#define GEN_ELF_ENDIAN ELFDATA2LSB #define GEN_ELF_CLASS ELFCLASS32 #elif defined(__aarch64__) #define GEN_ELF_ARCH EM_AARCH64 -#define GEN_ELF_ENDIAN ELFDATA2LSB #define GEN_ELF_CLASS ELFCLASS64 #elif defined(__x86_64__) #define GEN_ELF_ARCH EM_X86_64 -#define GEN_ELF_ENDIAN ELFDATA2LSB #define GEN_ELF_CLASS ELFCLASS64 #elif defined(__i386__) #define GEN_ELF_ARCH EM_386 -#define GEN_ELF_ENDIAN ELFDATA2LSB #define GEN_ELF_CLASS ELFCLASS32 -#elif defined(__ppcle__) -#define GEN_ELF_ARCH EM_PPC -#define GEN_ELF_ENDIAN ELFDATA2LSB +#elif defined(__powerpc64__) +#define GEN_ELF_ARCH EM_PPC64 #define GEN_ELF_CLASS ELFCLASS64 #elif defined(__powerpc__) -#define GEN_ELF_ARCH EM_PPC64 -#define GEN_ELF_ENDIAN ELFDATA2MSB -#define GEN_ELF_CLASS ELFCLASS64 -#elif defined(__powerpcle__) -#define GEN_ELF_ARCH EM_PPC64 -#define GEN_ELF_ENDIAN ELFDATA2LSB -#define GEN_ELF_CLASS ELFCLASS64 +#define GEN_ELF_ARCH EM_PPC +#define GEN_ELF_CLASS ELFCLASS32 #else #error "unsupported architecture" #endif +#if __BYTE_ORDER == __BIG_ENDIAN +#define GEN_ELF_ENDIAN ELFDATA2MSB +#else +#define GEN_ELF_ENDIAN ELFDATA2LSB +#endif + #if GEN_ELF_CLASS == ELFCLASS64 #define elf_newehdr elf64_newehdr #define elf_getshdr elf64_getshdr From 8fdc65391c6ad16461526a685f03262b3b01bfde Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Tue, 29 Mar 2016 09:26:44 +0200 Subject: [PATCH 08/10] perf/core: Fix time tracking bug with multiplexing Stephane reported that commit: 3cbaa5906967 ("perf: Fix ctx time tracking by introducing EVENT_TIME") introduced a regression wrt. time tracking, as easily observed by: > This patch introduce a bug in the time tracking of events when > multiplexing is used. > > The issue is easily reproducible with the following perf run: > > $ perf stat -a -C 0 -e branches,branches,branches,branches,branches,branches -I 1000 > 1.000730239 652,394 branches (66.41%) > 1.000730239 597,809 branches (66.41%) > 1.000730239 593,870 branches (66.63%) > 1.000730239 651,440 branches (67.03%) > 1.000730239 656,725 branches (66.96%) > 1.000730239 branches > > One branches event is shown as not having run. Yet, with > multiplexing, all events should run especially with a 1s (-I 1000) > interval. The delta for time_running comes out to 0. Yet, the event > has run because the kernel is actually multiplexing the events. The > problem is that the time tracking is the kernel and especially in > ctx_sched_out() is wrong now. > > The problem is that in case that the kernel enters ctx_sched_out() with the > following state: > ctx->is_active=0x7 event_type=0x1 > Call Trace: > [] dump_stack+0x63/0x82 > [] ctx_sched_out+0x2bc/0x2d0 > [] perf_mux_hrtimer_handler+0xf6/0x2c0 > [] ? __perf_install_in_context+0x130/0x130 > [] __hrtimer_run_queues+0xf8/0x2f0 > [] hrtimer_interrupt+0xb7/0x1d0 > [] local_apic_timer_interrupt+0x38/0x60 > [] smp_apic_timer_interrupt+0x3d/0x50 > [] apic_timer_interrupt+0x8c/0xa0 > > In that case, the test: > if (is_active & EVENT_TIME) > > will be false and the time will not be updated. Time must always be updated on > sched out. Fix this by always updating time if EVENT_TIME was set, as opposed to only updating time when EVENT_TIME changed. Reported-by: Stephane Eranian Tested-by: Stephane Eranian Signed-off-by: Peter Zijlstra (Intel) Cc: Alexander Shishkin Cc: Arnaldo Carvalho de Melo Cc: Jiri Olsa Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: Vince Weaver Cc: kan.liang@intel.com Cc: namhyung@kernel.org Fixes: 3cbaa5906967 ("perf: Fix ctx time tracking by introducing EVENT_TIME") Link: http://lkml.kernel.org/r/20160329072644.GB3408@twins.programming.kicks-ass.net Signed-off-by: Ingo Molnar --- kernel/events/core.c | 14 ++++++++++++-- 1 file changed, 12 insertions(+), 2 deletions(-) diff --git a/kernel/events/core.c b/kernel/events/core.c index de24fbce5277..8c11388e92a5 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -2417,14 +2417,24 @@ static void ctx_sched_out(struct perf_event_context *ctx, cpuctx->task_ctx = NULL; } - is_active ^= ctx->is_active; /* changed bits */ - + /* + * Always update time if it was set; not only when it changes. + * Otherwise we can 'forget' to update time for any but the last + * context we sched out. For example: + * + * ctx_sched_out(.event_type = EVENT_FLEXIBLE) + * ctx_sched_out(.event_type = EVENT_PINNED) + * + * would only update time for the pinned events. + */ if (is_active & EVENT_TIME) { /* update (and stop) ctx time */ update_context_time(ctx); update_cgrp_time_from_cpuctx(cpuctx); } + is_active ^= ctx->is_active; /* changed bits */ + if (!ctx->nr_active || !(is_active & EVENT_ALL)) return; From 201c2f85bd0bc13b712d9c0b3d11251b182e06ae Mon Sep 17 00:00:00 2001 From: Alexander Shishkin Date: Mon, 21 Mar 2016 10:02:42 +0200 Subject: [PATCH 09/10] perf/core: Don't leak event in the syscall error path In the error path, event_file not being NULL is used to determine whether the event itself still needs to be free'd, so fix it up to avoid leaking. Reported-by: Leon Yu Signed-off-by: Alexander Shishkin Signed-off-by: Peter Zijlstra (Intel) Cc: Arnaldo Carvalho de Melo Cc: Arnaldo Carvalho de Melo Cc: Jiri Olsa Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Stephane Eranian Cc: Thomas Gleixner Cc: Vince Weaver Fixes: 130056275ade ("perf: Do not double free") Link: http://lkml.kernel.org/r/87twk06yxp.fsf@ashishki-desk.ger.corp.intel.com Signed-off-by: Ingo Molnar --- kernel/events/core.c | 1 + 1 file changed, 1 insertion(+) diff --git a/kernel/events/core.c b/kernel/events/core.c index 8c11388e92a5..52bedc5a5aaa 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -8542,6 +8542,7 @@ SYSCALL_DEFINE5(perf_event_open, f_flags); if (IS_ERR(event_file)) { err = PTR_ERR(event_file); + event_file = NULL; goto err_context; } From 85dc600263c2291cea33bffa90038808ee64198b Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Mon, 21 Mar 2016 11:47:52 +0100 Subject: [PATCH 10/10] perf/x86/amd/ibs: Fix pmu::stop() nesting Patch 5a50f5291701 ("perf/x86/ibs: Fix race with IBS_STARTING state") closed a big hole while opening another, smaller hole. Signed-off-by: Peter Zijlstra (Intel) Cc: Alexander Shishkin Cc: Arnaldo Carvalho de Melo Cc: Jiri Olsa Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Stephane Eranian Cc: Thomas Gleixner Cc: Vince Weaver Fixes: 5a50f5291701 ("perf/x86/ibs: Fix race with IBS_STARTING state") Signed-off-by: Ingo Molnar --- arch/x86/events/amd/ibs.c | 52 +++++++++++++++++++++++++++++++++------ 1 file changed, 45 insertions(+), 7 deletions(-) diff --git a/arch/x86/events/amd/ibs.c b/arch/x86/events/amd/ibs.c index 3ea25c3917c0..feb90f6730e8 100644 --- a/arch/x86/events/amd/ibs.c +++ b/arch/x86/events/amd/ibs.c @@ -28,10 +28,46 @@ static u32 ibs_caps; #define IBS_FETCH_CONFIG_MASK (IBS_FETCH_RAND_EN | IBS_FETCH_MAX_CNT) #define IBS_OP_CONFIG_MASK IBS_OP_MAX_CNT + +/* + * IBS states: + * + * ENABLED; tracks the pmu::add(), pmu::del() state, when set the counter is taken + * and any further add()s must fail. + * + * STARTED/STOPPING/STOPPED; deal with pmu::start(), pmu::stop() state but are + * complicated by the fact that the IBS hardware can send late NMIs (ie. after + * we've cleared the EN bit). + * + * In order to consume these late NMIs we have the STOPPED state, any NMI that + * happens after we've cleared the EN state will clear this bit and report the + * NMI handled (this is fundamentally racy in the face or multiple NMI sources, + * someone else can consume our BIT and our NMI will go unhandled). + * + * And since we cannot set/clear this separate bit together with the EN bit, + * there are races; if we cleared STARTED early, an NMI could land in + * between clearing STARTED and clearing the EN bit (in fact multiple NMIs + * could happen if the period is small enough), and consume our STOPPED bit + * and trigger streams of unhandled NMIs. + * + * If, however, we clear STARTED late, an NMI can hit between clearing the + * EN bit and clearing STARTED, still see STARTED set and process the event. + * If this event will have the VALID bit clear, we bail properly, but this + * is not a given. With VALID set we can end up calling pmu::stop() again + * (the throttle logic) and trigger the WARNs in there. + * + * So what we do is set STOPPING before clearing EN to avoid the pmu::stop() + * nesting, and clear STARTED late, so that we have a well defined state over + * the clearing of the EN bit. + * + * XXX: we could probably be using !atomic bitops for all this. + */ + enum ibs_states { IBS_ENABLED = 0, IBS_STARTED = 1, IBS_STOPPING = 2, + IBS_STOPPED = 3, IBS_MAX_STATES, }; @@ -377,11 +413,10 @@ static void perf_ibs_start(struct perf_event *event, int flags) perf_ibs_set_period(perf_ibs, hwc, &period); /* - * Set STARTED before enabling the hardware, such that - * a subsequent NMI must observe it. Then clear STOPPING - * such that we don't consume NMIs by accident. + * Set STARTED before enabling the hardware, such that a subsequent NMI + * must observe it. */ - set_bit(IBS_STARTED, pcpu->state); + set_bit(IBS_STARTED, pcpu->state); clear_bit(IBS_STOPPING, pcpu->state); perf_ibs_enable_event(perf_ibs, hwc, period >> 4); @@ -396,6 +431,9 @@ static void perf_ibs_stop(struct perf_event *event, int flags) u64 config; int stopping; + if (test_and_set_bit(IBS_STOPPING, pcpu->state)) + return; + stopping = test_bit(IBS_STARTED, pcpu->state); if (!stopping && (hwc->state & PERF_HES_UPTODATE)) @@ -405,12 +443,12 @@ static void perf_ibs_stop(struct perf_event *event, int flags) if (stopping) { /* - * Set STOPPING before disabling the hardware, such that it + * Set STOPPED before disabling the hardware, such that it * must be visible to NMIs the moment we clear the EN bit, * at which point we can generate an !VALID sample which * we need to consume. */ - set_bit(IBS_STOPPING, pcpu->state); + set_bit(IBS_STOPPED, pcpu->state); perf_ibs_disable_event(perf_ibs, hwc, config); /* * Clear STARTED after disabling the hardware; if it were @@ -556,7 +594,7 @@ fail: * with samples that even have the valid bit cleared. * Mark all this NMIs as handled. */ - if (test_and_clear_bit(IBS_STOPPING, pcpu->state)) + if (test_and_clear_bit(IBS_STOPPED, pcpu->state)) return 1; return 0;