- 03 Mar, 2016 5 commits
-
-
Wang Nan authored
Commits in a BPF patchkit will extract kernel and module synthesizing code into a separated function and call it multiple times. This patch replace 'if (err < 0)' using WARN_ONCE, makes sure the error message show one time. Signed-off-by:
Wang Nan <wangnan0@huawei.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: He Kuang <hekuang@huawei.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Li Zefan <lizefan@huawei.com> Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Zefan Li <lizefan@huawei.com> Cc: pi3orama@163.com Link: http://lkml.kernel.org/r/1456479154-136027-19-git-send-email-wangnan0@huawei.comSigned-off-by:
Arnaldo Carvalho de Melo <acme@redhat.com>
-
Wang Nan authored
After babeltrace commit 5cec03e402aa ("ir: copy variants and sequences when setting a field path"), 'perf data convert' gets incorrect result if there's bpf output data. For example: # perf data convert --to-ctf ./out.ctf # babeltrace ./out.ctf [10:44:31.186045346] (+?.?????????) evt: { cpu_id = 0 }, { perf_ip = 0xFFFFFFFF810E7DD1, perf_tid = 23819, perf_pid = 23819, perf_id = 518, raw_len = 3, raw_data = [ [0] = 0xC028E32F, [1] = 0x815D0100, [2] = 0x1000000 ] } [10:44:31.286101003] (+0.100055657) evt: { cpu_id = 0 }, { perf_ip = 0xFFFFFFFF8105B609, perf_tid = 23819, perf_pid = 23819, perf_id = 518, raw_len = 3, raw_data = [ [0] = 0x35D9F1EB, [1] = 0x15D81, [2] = 0x2 ] } The expected result of the first sample should be: raw_data = [ [0] = 0x2FE328C0, [1] = 0x15D81, [2] = 0x1 ] } however, 'perf data convert' output big endian value to resuling CTF file. The reason is a internal change (or a bug?) of babeltrace. Before this patch, at the first add_bpf_output_values(), byte order of all integer type is uncertain (is 0, neither 1234 (le) nor 4321 (be)). It would be fixed by: perf_evlist__deliver_sample -> process_sample_event -> ctf_stream ... ->bt_ctf_trace_add_stream_class ->bt_ctf_field_type_structure_set_byte_order ->bt_ctf_field_type_integer_set_byte_order during creating the stream. However, the babeltrace commit mentioned above duplicates types in sequence to prevent potential conflict in following call stack and link the newly allocated type into the 'raw_data' sequence: perf_evlist__deliver_sample -> process_sample_event -> ctf_stream ... -> bt_ctf_trace_add_stream_class -> bt_ctf_stream_class_resolve_types ... -> bt_ctf_field_type_sequence_copy ->bt_ctf_field_type_integer_copy This happens before byte order setting, so only the newly allocated type is initialized, the byte order of original type perf choose to create the first raw_data is still uncertain. Byte order in CTF output is not related to byte order in perf.data. Setting it to anything other than BT_CTF_BYTE_ORDER_NATIVE solves this problem (only BT_CTF_BYTE_ORDER_NATIVE needs to be fixed). To reduce behavior changing, set byte order according to compiling options. Signed-off-by: -
Wang Nan authored
bpf_perf_event_output() outputs data through sample->raw_data. This patch adds support to convert those data into CTF. A python script then can be used to process output data from BPF programs. Test result: # cat ./test_bpf_output_2.c /************************ BEGIN **************************/ #include <uapi/linux/bpf.h> struct bpf_map_def { unsigned int type; unsigned int key_size; unsigned int value_size; unsigned int max_entries; }; #define SEC(NAME) __attribute__((section(NAME), used)) static u64 (*ktime_get_ns)(void) = (void *)BPF_FUNC_ktime_get_ns; static int (*trace_printk)(const char *fmt, int fmt_size, ...) = (void *)BPF_FUNC_trace_printk; static int (*get_smp_processor_id)(void) = (void *)BPF_FUNC_get_smp_processor_id; static int (*perf_event_output)(void *, struct bpf_map_def *, int, void *, unsigned long) = (void *)BPF_FUNC_perf_event_output; struct bpf_map_def SEC("maps") channel = { .type = BPF_MAP_TYPE_PERF_EVENT_ARRAY, .key_size = sizeof(int), .value_size = sizeof(u32), .max_entries = __NR_CPUS__, }; static inline int __attribute__((always_inline)) func(void *ctx, int type) { struct { u64 ktime; int type; } __attribute__((packed)) output_data; char error_data[] = "Error: failed to output\n"; int err; output_data.type = type; output_data.ktime = ktime_get_ns(); err = perf_event_output(ctx, &channel, get_smp_processor_id(), &output_data, sizeof(output_data)); if (err) trace_printk(error_data, sizeof(error_data)); return 0; } SEC("func_begin=sys_nanosleep") int func_begin(void *ctx) {return func(ctx, 1);} SEC("func_end=sys_nanosleep%return") int func_end(void *ctx) { return func(ctx, 2);} char _license[] SEC("license") = "GPL"; int _version SEC("version") = LINUX_VERSION_CODE; /************************* END ***************************/ # ./perf record -e bpf-output/no-inherit,name=evt/ \ -e ./test_bpf_output_2.c/map:channel.event=evt/ \ usleep 100000 [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.012 MB perf.data (2 samples) ] # ./perf script usleep 14942 92503.198504: evt: ffffffff810e0ba1 sys_nanosleep (/lib/modules/4.3.0.... usleep 14942 92503.298562: evt: ffffffff810585e9 kretprobe_trampoline_holder (/lib.... # ./perf data convert --to-ctf ./out.ctf [ perf data convert: Converted 'perf.data' into CTF data './out.ctf' ] [ perf data convert: Converted and wrote 0.000 MB (2 samples) ] # babeltrace ./out.ctf [01:41:43.198504134] (+?.?????????) evt: { cpu_id = 0 }, { perf_ip = 0xFFFFFFFF810E0BA1, perf_tid = 14942, perf_pid = 14942, perf_id = 1044, raw_len = 3, raw_data = [ [0] = 0x32C0C07B, [1] = 0x5421, [2] = 0x1 ] } [01:41:43.298562257] (+0.100058123) evt: { cpu_id = 0 }, { perf_ip = 0xFFFFFFFF810585E9, perf_tid = 14942, perf_pid = 14942, perf_id = 1044, raw_len = 3, raw_data = [ [0] = 0x38B77FAA, [1] = 0x5421, [2] = 0x2 ] } # cat ./test_bpf_output_2.py from babeltrace import TraceCollection tc = TraceCollection() tc.add_trace('./out.ctf', 'ctf') d = {1:[], 2:[]} for event in tc.events: if not event.name.startswith('evt'): continue raw_data = event['raw_data'] (time, type) = ((raw_data[0] + (raw_data[1] << 32)), raw_data[2]) d[type].append(time) print(list(map(lambda i: d[2][i] - d[1][i], range(len(d[1]))))); # python3 ./test_bpf_output_2.py [100056879] Committer note: Make sure you have python3-devel installed, not python-devel, which may be for python2, which will lead to some "PyInstance_Type" errors. Also make sure that you use the right libbabeltrace, because it is shipped in Fedora, for instance, but an older version. To build libbabeltrace's python binding one also needs to use: ./configure --enable-python-bindings And then set PYTHONPATH=/usr/local/lib64/python3.4/site-packages/. Signed-off-by:Jiri Olsa <jolsa@kernel.org> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Brendan Gregg <brendan.d.gregg@gmail.com> Cc: Li Zefan <lizefan@huawei.com> Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Zefan Li <lizefan@huawei.com> Cc: pi3orama@163.com Link: http://lkml.kernel.org/r/1456479154-136027-9-git-send-email-wangnan0@huawei.comSigned-off-by:
-
Andi Kleen authored
Only put the frontend/backend stalled cycles into the default perf stat events when the CPU actually supports them. This avoids empty columns with --metric-only on newer Intel CPUs. Committer note: Before: $ perf stat ls Performance counter stats for 'ls': 1.080893 task-clock (msec) # 0.619 CPUs utilized 0 context-switches # 0.000 K/sec 0 cpu-migrations # 0.000 K/sec 97 page-faults # 0.090 M/sec 3,327,741 cycles # 3.079 GHz <not supported> stalled-cycles-frontend <not supported> stalled-cycles-backend 1,609,544 instructions # 0.48 insn per cycle 319,117 branches # 295.235 M/sec 12,246 branch-misses # 3.84% of all branches 0.001746508 seconds time elapsed $ After: $ perf stat ls Performance counter stats for 'ls': 0.693948 task-clock (msec) # 0.662 CPUs utilized 0 context-switches # 0.000 K/sec 0 cpu-migrations # 0.000 K/sec 95 page-faults # 0.137 M/sec 1,792,509 cycles # 2.583 GHz 1,599,047 instructions # 0.89 insn per cycle 316,328 branches # 455.838 M/sec 12,453 branch-misses # 3.94% of all branches 0.001048987 seconds time elapsed $ Signed-off-by:Andi Kleen <ak@linux.intel.com> Acked-by:
-
Jiri Olsa authored
Ingo reported regression on display format of big numbers, which is missing separators (in default perf stat output). triton:~/tip> perf stat -a sleep 1 ... 127008602 cycles # 0.011 GHz 279538533 stalled-cycles-frontend # 220.09% frontend cycles idle 119213269 instructions # 0.94 insn per cycle This is caused by recent change: perf stat: Check existence of frontend/backed stalled cycles that added call to pmu_have_event, that subsequently calls perf_pmu__parse_scale, which has a bug in locale handling. The lc string returned from setlocale, that we use to store old locale value, may be allocated in static storage. Getting a dynamic copy to make it survive another setlocale call. $ perf stat ls ... 2,360,602 cycles # 3.080 GHz 2,703,090 instructions # 1.15 insn per cycle 546,031 branches # 712.511 M/sec Committer note: Since the patch introducing the regression didn't made to perf/core, move it to just before where the regression was introduced, so that we don't break bisection for this feature. Reported-by:Ingo Molnar <mingo@redhat.com> Signed-off-by:
-
- 29 Feb, 2016 32 commits
-
-
Steven Rostedt authored
Currently there's a single function that is used to display a record's data in human readable format. That's pevent_print_event(). Unfortunately, this gives little room for adding other output within the line without updating that function call. I've decided to split that function into 3 parts. pevent_print_event_task() which prints the task comm, pid and the CPU pevent_print_event_time() which outputs the record's timestamp pevent_print_event_data() which outputs the rest of the event data. pevent_print_event() now simply calls these three functions. To save time from doing the search for event from the record's type, I created a new helper function called pevent_find_event_by_record(), which returns the record's event, and this event has to be passed to the above functions. Signed-off-by:
Steven Rostedt <rostedt@goodmis.org> Cc: Namhyung Kim <namhyung@kernel.org> Link: http://lkml.kernel.org/r/20160229090128.43a56704@gandalf.local.homeSigned-off-by:
-
Taeung Song authored
Some tracepoint have multiple fields with the same name, "nr", the first one is a unique syscall ID, the other is a syscall argument: # cat /sys/kernel/debug/tracing/events/syscalls/sys_enter_io_getevents/format name: sys_enter_io_getevents ID: 747 format: field:unsigned short common_type; offset:0; size:2; signed:0; field:unsigned char common_flags; offset:2; size:1; signed:0; field:unsigned char common_preempt_count; offset:3; size:1; signed:0; field:int common_pid; offset:4; size:4; signed:1; field:int nr; offset:8; size:4; signed:1; field:aio_context_t ctx_id; offset:16; size:8; signed:0; field:long min_nr; offset:24; size:8; signed:0; field:long nr; offset:32; size:8; signed:0; field:struct io_event * events; offset:40; size:8; signed:0; field:struct timespec * timeout; offset:48; size:8; signed:0; print fmt: "ctx_id: 0x%08lx, min_nr: 0x%08lx, nr: 0x%08lx, events: 0x%08lx, timeout: 0x%08lx", ((unsigned long)(REC->ctx_id)), ((unsigned long)(REC->min_nr)), ((unsigned long)(REC->nr)), ((unsigned long)(REC->events)), ((unsigned long)(REC->timeout)) # Fix it by renaming the "/format" common tracepoint field "nr" to "__syscall_nr". Signed-off-by:
Taeung Song <treeze.taeung@gmail.com> [ Do not rename the struct member, just the '/format' field name ] Signed-off-by:
Peter Zijlstra <peterz@infradead.org> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Lai Jiangshan <jiangshanlai@gmail.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Thomas Gleixner <tglx@linutronix.de> Link: http://lkml.kernel.org/r/20160226132301.3ae065a4@gandalf.local.homeSigned-off-by:
-
Taeung Song authored
Format fields of a syscall have the first variable '__syscall_nr' or 'nr' that mean the syscall number. But it isn't relevant here so drop it. 'nr' among fields of syscall was renamed '__syscall_nr'. So add exception handling to drop '__syscall_nr' and modify the comment for this excpetion handling. Reported-by:
-
Jiri Olsa authored
The util/python-ext-sources file contains source files required to build the python extension relative to $(srctree)/tools/perf, Such a file path $(FILE).c is handed over to the python extension build system, which builds the final object in the $(PYTHON_EXTBUILD)/tmp/$(FILE).o path. After the build is done all files from $(PYTHON_EXTBUILD)lib/ are carried as the result binaries. Above system fails when we add source file relative to ../lib, which we do for: ../lib/bitmap.c ../lib/find_bit.c ../lib/hweight.c ../lib/rbtree.c All above objects will be built like: $(PYTHON_EXTBUILD)/tmp/../lib/bitmap.c $(PYTHON_EXTBUILD)/tmp/../lib/find_bit.c $(PYTHON_EXTBUILD)/tmp/../lib/hweight.c $(PYTHON_EXTBUILD)/tmp/../lib/rbtree.c which accidentally happens to be final library path: $(PYTHON_EXTBUILD)/lib/ Changing setup.py to pass full paths of source files to Extension build class and thus keep all built objects under $(PYTHON_EXTBUILD)tmp directory. Reported-by:
Jeff Bastian <jbastian@redhat.com> Signed-off-by:
Josh Boyer <jwboyer@fedoraproject.org> Cc: David Ahern <dsahern@gmail.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: stable@vger.kernel.org # v4.2+ Link: http://lkml.kernel.org/r/20160227201350.GB28494@krava.redhat.comSigned-off-by:
-
Thomas Gleixner authored
Required to use it in modular perf drivers. Signed-off-by:
Thomas Gleixner <tglx@linutronix.de> Signed-off-by:
Ingo Molnar <mingo@kernel.org>
-
Thomas Gleixner authored
RAPL is a per package facility and we already have a mechanism for a dedicated per package reader. So there is no point to have multiple CPUs doing the same. The current implementation actually starts two timers on two CPUs if one does: perf stat -C1,2 -e -e power/energy-pkg .... which makes the whole concept of 1 reader per package moot. What's worse is that the above returns the double of the actual energy consumption, but that's a different problem to address and cannot be solved by removing the pointless per cpuness of that mechanism. Signed-off-by:
-
Thomas Gleixner authored
Store the PMU pointer in event->pmu_private and use it instead of the per CPU data. Preparatory step to get rid of the per CPU allocations. The usage sites are the perf fast path, so we keep that even after the conversion to per package storage as a CPU to package lookup involves 3 loads versus 1 with the pmu_private pointer. Signed-off-by:
-
Thomas Gleixner authored
This lock is taken in hard interrupt context even on Preempt-RT. Make it raw so RT does not have to patch it. Signed-off-by:
-
Thomas Gleixner authored
Split out code from init into seperate functions. Tidy up the code and get rid of pointless comments. I wish there would be comments for code which is not obvious.... Signed-off-by:
-
Thomas Gleixner authored
The output is inconsistent. Use a proper pr_fmt prefix and split out the advertisement into a seperate function. Remove the WARN_ON() in the failure case. It's pointless as we already know where it failed. Signed-off-by:
-
Thomas Gleixner authored
No point in doing the same calculation over and over. Do it once in rapl_check_hw_unit(). Signed-off-by:
-
Thomas Gleixner authored
There is no point in having a quirk machinery for a single possible function. Get rid of it and move the quirk to a place where it actually makes sense. Signed-off-by:
-
Thomas Gleixner authored
Like uncore the rapl driver lacks error handling. It leaks memory and leaves the hotplug notifier registered. Add the proper error checks, cleanup the memory and register the hotplug notifier only on success. Signed-off-by:
-
Thomas Gleixner authored
The Knights Landings support added the events and the detection case, but then returns 0 without actually initializing the driver. Signed-off-by:
-
Thomas Gleixner authored
CQM is a strict per package facility. Use the proper cpumasks to lookup the readers. Signed-off-by:
-
Thomas Gleixner authored
Almost every cpumask function is exported, just not the one I need to make the Intel uncore driver modular. Signed-off-by:
-
Thomas Gleixner authored
Andi wanted to do this before, but the patch fell down the cracks. Implement it with the proper error handling. Requested-by:
-
Thomas Gleixner authored
The only missing bit is to completely clear the hardware state on failure exit. This is now a pretty simple exercise. Undo the box->init_box() setup on all packages which have been initialized so far. Signed-off-by:
-
Thomas Gleixner authored
Uncore is a per package facility, but the code tries to mimick a per CPU facility with completely convoluted constructs. Simplify the whole machinery by tracking per package information. While at it, avoid the kfree/alloc dance when a CPU goes offline and online again. There is no point in freeing the box after it was allocated. We just keep proper refcounting and the first CPU which comes online in a package does the initialization/activation of the box. Signed-off-by:
-
Thomas Gleixner authored
For per package oriented services we must be able to rely on the number of CPU packages to be within bounds. Create a tracking facility, which - calculates the number of possible packages depending on nr_cpu_ids after boot - makes sure that the package id is within the number of possible packages. If the apic id is outside we map it to a logical package id if there is enough space available. Provide interfaces for drivers to query the mapping and do translations from physcial to logical ids. Signed-off-by:
-
Thomas Gleixner authored
Store the PMU pointer in event->pmu_private, so we can get rid of the per CPU data storage. We keep it after converting to per package data, because a CPU to package lookup will be 3 loads versus one and these usage sites are in the perf fast path. Signed-off-by:
-
Thomas Gleixner authored
For PMUs which are not per CPU, but e.g. per package/socket, we want to be able to store a reference to the underlying per package/socket facility in the event at init time so we can avoid magic storage constructs in the PMU driver. This allows us to get rid of the per CPU dance in the intel uncore and RAPL drivers and avoids a lookup of the per package data in the perf hotpath. Signed-off-by:
-
Thomas Gleixner authored
No users outside of this file. Signed-off-by:
-
Thomas Gleixner authored
Clean up the code a bit before reworking it completely. Signed-off-by:
-
Thomas Gleixner authored
When tearing down the boxes nothing undoes the hardware state which was setup by box->init_box(). Add a box->exit_box() callback and implement it for the uncores which have an init_box() callback. This misses the cleanup in the error exit pathes, but I cannot be bothered to implement it before cleaning up the rest of the driver, which makes that task way simpler. Signed-off-by:
-
Thomas Gleixner authored
The storage array is size limited, but misses a sanity check Signed-off-by:
-
Thomas Gleixner authored
This driver lacks any form of proper error handling. If initialization fails or hotplug prepare fails, it lets the facility with half initialized stuff around. Fix the state and memory leaks in a first step. As a second step we need to undo the hardware state which is set via uncore_box_init() on some of the uncore implementations. Signed-off-by:
-
Thomas Gleixner authored
No point in doing partial rollbacks. Robustify uncore_exit_type() so it does not dereference type->pmus unconditionally and remove all the partial rollback hackery. Preparatory patch for proper error handling. Signed-off-by:
-
Thomas Gleixner authored
uncore_cpumask_init() is only ever called from intel_uncore_init() where the mask is guaranteed to be empty. Signed-off-by:
-
Peter Zijlstra authored
Alexander volunteered to review perf (kernel) patches. Signed-off-by:
-
Kan Liang authored
BDX-DE and BDX-EP share the same uncore code path. But there is no sbox in BDX-DE. This patch remove SBOX support for BDX-DE. Signed-off-by:
Kan Liang <kan.liang@intel.com> Signed-off-by:
-
Ingo Molnar authored
Signed-off-by:
-
- 28 Feb, 2016 3 commits
-
-
Linus Torvalds authored
-
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tipLinus Torvalds authored
Pull perf fixes from Thomas Gleixner: "A rather largish series of 12 patches addressing a maze of race conditions in the perf core code from Peter Zijlstra" * 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: perf: Robustify task_function_call() perf: Fix scaling vs. perf_install_in_context() perf: Fix scaling vs. perf_event_enable() perf: Fix scaling vs. perf_event_enable_on_exec() perf: Fix ctx time tracking by introducing EVENT_TIME perf: Cure event->pending_disable race perf: Fix race between event install and jump_labels perf: Fix cloning perf: Only update context time when active perf: Allow perf_release() with !event->ctx perf: Do not double free perf: Close install vs. exit race
-
git://git.kernel.org/pub/scm/linux/kernel/git/tip/tipLinus Torvalds authored
Pull x86 fixes from Thomas Gleixner: "This update contains: - Hopefully the last ASM CLAC fixups - A fix for the Quark family related to the IMR lock which makes kexec work again - A off-by-one fix in the MPX code. Ironic, isn't it? - A fix for X86_PAE which addresses once more an unsigned long vs phys_addr_t hickup" * 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: x86/mpx: Fix off-by-one comparison with nr_registers x86/mm: Fix slow_virt_to_phys() for X86_PAE again x86/entry/compat: Add missing CLAC to entry_INT80_32 x86/entry/32: Add an ASM_CLAC to entry_SYSENTER_32 x86/platform/intel/quark: Change the kernel's IMR lock bit to false
-
