- 03 Jan, 2024 2 commits
-
-
Ian Rogers authored
Reduce from PERF_SAMPLE_MAX_SIZE to "sizeof(*lost) + session->machines.host.id_hdr_size". Suggested-by:
Namhyung Kim <namhyung@kernel.org> Signed-off-by:
Ian Rogers <irogers@google.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Peter Zijlstra <peterz@infradead.org> Link: https://lore.kernel.org/r/20231207021627.1322884-1-irogers@google.comSigned-off-by:
Arnaldo Carvalho de Melo <acme@redhat.com>
-
Ian Rogers authored
Add variants of perf_env__insert_bpf_prog_info(), perf_env__insert_btf() and perf_env__find_btf prefixed with __ to indicate the env->bpf_progs.lock is assumed held. Call these variants when the lock is held to avoid recursively taking it and potentially having a thread deadlock with itself. Fixes: f8dfeae0 ("perf bpf: Show more BPF program info in print_bpf_prog_info()") Signed-off-by:
Jiri Olsa <jolsa@kernel.org> Acked-by:
Song Liu <song@kernel.org> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Huacai Chen <chenhuacai@kernel.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: K Prateek Nayak <kprateek.nayak@amd.com> Cc: Kan Liang <kan.liang@linux.intel.com> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Ming Wang <wangming01@loongson.cn> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Ravi Bangoria <ravi.bangoria@amd.com> Link: https://lore.kernel.org/r/20231207014655.1252484-1-irogers@google.comSigned-off-by:
-
- 24 Dec, 2023 15 commits
-
-
Namhyung Kim authored
This is for a debugging purpose. It'd be useful to see per-instrucion level success/failure stats. $ perf annotate --data-type --insn-stat Annotate Instruction stats total 264, ok 143 (54.2%), bad 121 (45.8%) Name : Good Bad ----------------------------------------------------------- movq : 45 31 movl : 22 11 popq : 0 19 cmpl : 16 3 addq : 8 7 cmpq : 11 3 cmpxchgl : 3 7 cmpxchgq : 8 0 incl : 3 3 movzbl : 4 2 incq : 4 2 decl : 6 0 ... Committer notes: So these are about being able to find the type for accesses from these instructions, we should improve the naming, but it is for debugging, we can improve this later: @@ -3726,6 +3759,10 @@ struct annotated_data_type *hist_entry__get_data_type(struct hist_entry *he) continue; mem_type = find_data_type(ms, ip, op_loc->reg, op_loc->offset); + if (mem_type) + istat->good++; + else + istat->bad++; Signed-off-by: -
Namhyung Kim authored
The --type-stat option is to be used with --data-type and to print detailed failure reasons for the data type annotation. $ perf annotate --data-type --type-stat Annotate data type stats: total 294, ok 116 (39.5%), bad 178 (60.5%) ----------------------------------------------------------- 30 : no_sym 40 : no_insn_ops 33 : no_mem_ops 63 : no_var 4 : no_typeinfo 8 : bad_offset Signed-off-by: -
Namhyung Kim authored
When events are grouped together, it'd be natural to show them at once like in other mode. Handle group leaders with members to collect the number of samples together and display like below: $ perf annotate --data-type --group ... Annotate type: 'struct page' in vmlinux (1 samples): event[0] = cpu/mem-loads,ldlat=30/P event[1] = cpu/mem-stores/P event[2] = dummy:u ============================================================================ samples offset size field 1 0 0 0 64 struct page { 0 0 0 0 8 long unsigned int flags; 0 0 0 8 40 union { 0 0 0 8 40 struct { 0 0 0 8 16 union { 0 0 0 8 16 struct list_head lru { 0 0 0 8 8 struct list_head* next; 0 0 0 16 8 struct list_head* prev; }; 0 0 0 8 16 struct { 0 0 0 8 8 void* __filler; 0 0 0 16 4 unsigned int mlock_count; }; 0 0 0 8 16 struct list_head buddy_list { 0 0 0 8 8 struct list_head* next; 0 0 0 16 8 struct list_head* prev; }; Signed-off-by: -
Namhyung Kim authored
Support data type annotation with new --data-type option. It internally uses type sort key to collect sample histogram for the type and display every members like below. $ perf annotate --data-type ... Annotate type: 'struct cfs_rq' in [kernel.kallsyms] (13 samples): ============================================================================ samples offset size field 13 0 640 struct cfs_rq { 2 0 16 struct load_weight load { 2 0 8 unsigned long weight; 0 8 4 u32 inv_weight; }; 0 16 8 unsigned long runnable_weight; 0 24 4 unsigned int nr_running; 1 28 4 unsigned int h_nr_running; ... For simplicity it prints the number of samples per field for now. But it should be easy to show the overhead percentage instead. The number at the outer struct is a sum of the numbers of the inner members. For example, struct cfs_rq got total 13 samples, and 2 came from the load (struct load_weight) and 1 from h_nr_running. Similarly, the struct load_weight got total 2 samples and they all came from the weight field. I've added two new flags in the symbol_conf for this. The annotate_data_member is to get the members of the type. This is also needed for perf report with typeoff sort key. The annotate_data_sample is to update sample stats for each offset and used only in annotate. Currently it only support stdio output mode, TUI support can be added later. Committer testing: With the perf.data from the previous csets, a very simple, short duration one: # perf annotate --data-type Annotate type: 'struct list_head' in [kernel.kallsyms] (1 samples): ============================================================================ samples offset size field 1 0 16 struct list_head { 0 0 8 struct list_head* next; 1 8 8 struct list_head* prev; }; Annotate type: 'char' in [kernel.kallsyms] (1 samples): ============================================================================ samples offset size field 1 0 1 char ; # Signed-off-by: -
Namhyung Kim authored
The symoff sort key is to print symbol and offset of sample. This is useful for data type profiling to show exact instruction in the function which refers the data. $ perf report -s type,sym,typeoff,symoff --hierarchy ... # Overhead Data Type / Symbol / Data Type Offset / Symbol Offset # .............. ..................................................... # 1.23% struct cfs_rq 0.84% update_blocked_averages 0.19% struct cfs_rq +336 (leaf_cfs_rq_list.next) 0.19% [k] update_blocked_averages+0x96 0.19% struct cfs_rq +0 (load.weight) 0.14% [k] update_blocked_averages+0x104 0.04% [k] update_blocked_averages+0x31c 0.17% struct cfs_rq +404 (throttle_count) 0.12% [k] update_blocked_averages+0x9d 0.05% [k] update_blocked_averages+0x1f9 0.08% struct cfs_rq +272 (propagate) 0.07% [k] update_blocked_averages+0x3d3 0.02% [k] update_blocked_averages+0x45b ... Committer testing: # perf report --stdio -s type,typeoff,symoff # To display the perf.data header info, please use --header/--header-only options. # # # Total Lost Samples: 0 # # Samples: 4 of event 'cpu_atom/mem-loads,ldlat=30/P' # Event count (approx.): 7 # # Overhead Data Type Data Type Offset Symbol Offset # ........ ......... ................ ............. # 42.86% struct list_head struct list_head +8 (prev) [k] __list_del_entry_valid_or_report+0x7 28.57% (unknown) (unknown) +0 (no field) [.] _nl_intern_locale_data+0x25 14.29% char char +0 (no field) [k] strncpy_from_user+0xa5 14.29% (unknown) (unknown) +0 (no field) [.] _dl_lookup_symbol_x+0x50 # # (Tip: To change sampling frequency to 100 Hz: perf record -F 100) # Signed-off-by: -
Namhyung Kim authored
The typeoff sort key shows the data type name, offset and the name of the field. This is useful to see which field in the struct is accessed most frequently. $ perf report -s type,typeoff --hierarchy --stdio ... # Overhead Data Type / Data Type Offset # ............ ............................ # ... 1.23% struct cfs_rq 0.19% struct cfs_rq +404 (throttle_count) 0.19% struct cfs_rq +0 (load.weight) 0.19% struct cfs_rq +336 (leaf_cfs_rq_list.next) 0.09% struct cfs_rq +272 (propagate) 0.09% struct cfs_rq +196 (removed.nr) 0.09% struct cfs_rq +80 (curr) 0.09% struct cfs_rq +544 (lt_b_children_throttled) 0.06% struct cfs_rq +320 (rq) Committer testing: Again with the perf.data from the previous csets: # perf report --stdio -s type,typeoff # To display the perf.data header info, please use --header/--header-only options. # # # Total Lost Samples: 0 # # Samples: 4 of event 'cpu_atom/mem-loads,ldlat=30/P' # Event count (approx.): 7 # # Overhead Data Type Data Type Offset # ........ ......... ................ # 42.86% struct list_head struct list_head +8 (prev) 42.86% (unknown) (unknown) +0 (no field) 14.29% char char +0 (no field) # # (Tip: To see callchains in a more compact form: perf report -g folded) # # perf report --stdio -s dso,type,typeoff # To display the perf.data header info, please use --header/--header-only options. # # # Total Lost Samples: 0 # # Samples: 4 of event 'cpu_atom/mem-loads,ldlat=30/P' # Event count (approx.): 7 # # Overhead Shared Object Data Type Data Type Offset # ........ .................... ......... ................ # 42.86% [kernel.kallsyms] struct list_head struct list_head +8 (prev) 28.57% libc.so.6 (unknown) (unknown) +0 (no field) 14.29% [kernel.kallsyms] char char +0 (no field) 14.29% ld-linux-x86-64.so.2 (unknown) (unknown) +0 (no field) # # (Tip: If you have debuginfo enabled, try: perf report -s sym,srcline) # # Signed-off-by: -
Namhyung Kim authored
The annotated_data_type__update_samples() to get histogram for data type access. It'll be called by perf annotate to show which fields in the data type are accessed frequently. Signed-off-by:
-
Namhyung Kim authored
Add child member field if the current type is a composite type like a struct or union. The member fields are linked in the children list and do the same recursively if the child itself is a composite type. Add 'self' member to the annotated_data_type to handle the members in the same way. Signed-off-by:
-
Namhyung Kim authored
Enable type annotation when the 'type' sort key is used. It shows type of variables the samples access at the moment. Users can see which types are accessed frequently. $ perf report -s dso,type --stdio ... # Overhead Shared Object Data Type # ........ ................. ......... # 35.47% [kernel.kallsyms] (unknown) 1.62% [kernel.kallsyms] struct sched_entry 1.23% [kernel.kallsyms] struct cfs_rq 0.83% [kernel.kallsyms] struct task_struct 0.34% [kernel.kallsyms] struct list_head 0.30% [kernel.kallsyms] struct mem_cgroup ... Committer testing: With the perf.data file collected in the previous cset: # perf report --stdio -s type # To display the perf.data header info, please use --header/--header-only options. # # # Total Lost Samples: 0 # # Samples: 4 of event 'cpu_atom/mem-loads,ldlat=30/P' # Event count (approx.): 7 # # Overhead Data Type # ........ ......... # 42.86% struct list_head 42.86% (unknown) 14.29% char # # (Tip: To record callchains for each sample: perf record -g) # # perf report --stdio -s dso,type # To display the perf.data header info, please use --header/--header-only options. # # # Total Lost Samples: 0 # # Samples: 4 of event 'cpu_atom/mem-loads,ldlat=30/P' # Event count (approx.): 7 # # Overhead Shared Object Data Type # ........ .................... ......... # 42.86% [kernel.kallsyms] struct list_head 28.57% libc.so.6 (unknown) 14.29% [kernel.kallsyms] char 14.29% ld-linux-x86-64.so.2 (unknown) # # (Tip: Save output of perf stat using: perf stat record <target workload>) # # Signed-off-by: -
Namhyung Kim authored
The 'type' sort key is to aggregate hist entries by data type they access. Add mem_type field to hist_entry struct to save the type. If hist_entry__get_data_type() returns NULL, it'd use the 'unknown_type' instance. Committer testing: Before: # perf mem record sleep 2s [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.037 MB perf.data (4 samples) ] root@number:/home/acme/Downloads# perf report --stdio -s type Error: Unknown --sort key: `type' Usage: perf report [<options>] -s, --sort <key[,key2...]> sort by key(s): overhead overhead_sys overhead_us overhead_guest_sys overhead_guest_us overhead_children sample period pid comm dso symbol parent cpu socket srcline srcfile local_weight weight transaction trace symbol_size dso_size cgroup cgroup_id ipc_null time code_page_size local_ins_lat ins_lat local_p_stage_cyc p_stage_cyc addr local_retire_lat retire_lat simd dso_from dso_to symbol_from symbol_to mispredict abort in_tx cycles srcline_from srcline_to ipc_lbr addr_from addr_to symbol_daddr dso_daddr locked tlb mem snoop dcacheline symbol_iaddr phys_daddr data_page_size blocked # After: # perf report --stdio -s type # To display the perf.data header info, please use --header/--header-only options. # # # Total Lost Samples: 0 # # Samples: 4 of event 'cpu_atom/mem-loads,ldlat=30/P' # Event count (approx.): 7 # # Overhead Data Type # ........ ......... # 100.00% (unknown) # # (Tip: Print event counts in CSV format with: perf stat -x,) # # rpm -q kernel-debuginfo kernel-debuginfo-6.6.4-200.fc39.x86_64 # uname -r 6.6.4-200.fc39.x86_64 # Signed-off-by: -
Namhyung Kim authored
It's the function to find out the type info from the given sample data and will be called from the hist_entry sort logic when 'type' sort key is used. It first calls objdump to disassemble the instructions and figure out information about memory access at the location. Maybe we can do it better by analyzing the instruction directly, but I'll leave it for later work. The memory access is determined by checking instruction operands to have "(" and then extract register name and offset. It'll return NULL if no data type is found. Signed-off-by: -
Namhyung Kim authored
The annotate_get_insn_location() is to get the detailed information of instruction locations like registers and offset. It has source and target operands locations in an array. Each operand can have a register and an offset. The offset is meaningful when mem_ref flag is set. Signed-off-by:
-
Namhyung Kim authored
The evsel__get_arch() is to get architecture info from the environment. It'll be used by other places later so let's factor it out. Also add arch__is() to check the arch info by name. Committer notes: "get" is usually associated with refcounting, so we better rename this at some point to a better name. Signed-off-by:
-
Namhyung Kim authored
To aggregate accesses to the same data type, add 'data_types' tree in DSO to maintain data types and find it by name and size. It might have different data types that happen to have the same name, so it also compares the size of the type. Even if it doesn't 100% guarantee, it reduces the possibility of mis-handling of such conflicts. And I don't think it's common to have different types with the same name. Committer notes: Very few cases on the Linux kernel, but there are some different types with the same name, unsure if there is a debug mode in libbpf dedup that warns about such cases, but there are provisions in pahole for that, see: "emit: Notice type shadowing, i.e. multiple types with the same name (enum, struct, union, etc)" https://git.kernel.org/pub/scm/devel/pahole/pahole.git/commit/?id=4f332dbfd02072e4f410db7bdcda8d6e3422974b $ pahole --compile > vmlinux.h $ rm -f a ; make a cc a.c -o a $ grep __[0-9] vmlinux.h union irte__1 { struct map_info__1; struct map_info__1 { struct map_info__1 * next; /* 0 8 */ $ drivers/iommu/amd/amd_iommu_types.h 'union irte' include/linux/dmar.h 'struct irte' include/linux/device-mapper.h: union map_info { void *ptr; }; include/linux/mtd/map.h: struct map_info { const char *name; unsigned long size; resource_size_t phys; <SNIP> kernel/events/uprobes.c: struct map_info { struct map_info *next; struct mm_struct *mm; unsigned long vaddr; }; Signed-off-by: -
Namhyung Kim authored
The find_data_type() is to get a data type from the memory access at the given address (IP) using a register and an offset. It requires DWARF debug info in the DSO and searches the list of variables and function parameters in the scope. In a pseudo code, it does basically the following: find_data_type(dso, ip, reg, offset) { pc = map__rip_2objdump(ip); CU = dwarf_addrdie(dso->dwarf, pc); scopes = die_get_scopes(CU, pc); for_each_scope(S, scopes) { V = die_find_variable_by_reg(S, pc, reg); if (V && V.type == pointer_type) { T = die_get_real_type(V); if (offset < T.size) return T; } } return NULL; } Committer notes: The 'size' variable in check_variable() is 64-bit, so use PRIu64 and inttypes.h to debug it. Ditto at find_data_type_die(). Signed-off-by:
-
- 23 Dec, 2023 2 commits
-
-
Namhyung Kim authored
The get_dwarf_regnum() returns a DWARF register number from a register name string according to the psABI. Also add two pseudo encodings of DWARF_REG_PC which is a register that are used by PC-relative addressing and DWARF_REG_FB which is a frame base register. They need to be handled in a special way. Signed-off-by:
-
Namhyung Kim authored
The die_get_typename_from_type() is to get the name of the given DIE in C-style type name. The difference from die_get_typename() is that it does not retrieve the DW_AT_type and use the given DIE directly. This will be used when users know the type DIE already. Signed-off-by:
-
- 21 Dec, 2023 2 commits
-
-
JiaLong.Yang authored
HX-C2000 is a new CPU made by HEXIN Technologies Co., Ltd. And a new PVN 0x0066 has been applied from the OpenPower Community for this CPU. Here is a patch to make perf tool run in the CPU. Reviewed-by:
Madhavan Srinivasan <maddy@linux.ibm.com> Signed-off-by:
JiaLong.Yang <jialong.yang@shingroup.cn> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Ian Rogers <irogers@google.com> Cc: Ingo Molnar <mingo@redhat.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: shenghui.qu@shingroup.cn Cc: Zhao Ke <ke.zhao@shingroup.cn> Cc: zhijie.ren@shingroup.cn Link: https://lore.kernel.org/r/20231221060242.4532-1-jialong.yang@shingroup.cnSigned-off-by:
-
Jing Zhang authored
cmn.json contains UTF-8 characters in brief description which could break the perf build on some distros. Fix this issue by removing the UTF-8 characters from cmn.json. without this fix: $find tools/perf/pmu-events/ -name "*.json" | xargs file -i | grep -v us-ascii tools/perf/pmu-events/arch/arm64/arm/cmn/sys/cmn.json: application/json; charset=utf-8 with it: $ file -i tools/perf/pmu-events/arch/arm64/arm/cmn/sys/cmn.json tools/perf/pmu-events/arch/arm64/arm/cmn/sys/cmn.json: text/plain; charset=us-ascii Fixes: 0b4de7bd ("perf jevents: Add support for Arm CMN PMU aliasing") Reported-by:
Arnaldo Carvalho de Melo <acme@kernel.com> Signed-off-by:
Jing Zhang <renyu.zj@linux.alibaba.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Heiko Carstens <hca@linux.ibm.com> Cc: Ian Rogers <irogers@google.com> Cc: Jing Zhang <renyu.zj@linux.alibaba.com> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Kajol Jain <kjain@linux.ibm.com> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com> Cc: Thomas Richter <tmricht@linux.ibm.com> Link: https://lore.kernel.org/r/1703138593-50486-1-git-send-email-renyu.zj@linux.alibaba.comSigned-off-by:
-
- 20 Dec, 2023 15 commits
-
-
Ian Rogers authored
Maps are sometimes made overlapping, in particular kernel maps. If the end of a map overlaps the start of the next, shorten the overlapping map. This should remove potential non-determinism in maps__find, ie finding maps by address. Signed-off-by:
-
Ian Rogers authored
Avoid exposing the implementation of maps so that the internals can be refactored. Signed-off-by:
-
Ian Rogers authored
Use to remove map_rb_node use from machine.c. Signed-off-by:
-
Ian Rogers authored
Avoid bpf_lock_contention_read touching the internal maps data structure by adding a helper function. As access is done directly on the map in maps, hold the read lock to stop it being removed. Signed-off-by:
-
Ian Rogers authored
Rename maps__clone() to maps__copy_from() to be more intention revealing of its behavior. Pass the underlying maps rather than the thread. Signed-off-by:
-
Ian Rogers authored
Simplify merge in for the simple case of a non-overlapping map. Signed-off-by:
-
Ian Rogers authored
Rename to maps__fixup_overlap_and_insert() as the given mapping is always inserted. Factor out first_ending_after() as a utility function. Minor variable name changes. Switch to using debug_file() rather than passing a debug FILE*. Signed-off-by:
-
Ian Rogers authored
Some dumping call backs need to be passed a FILE*. Expose debug file via an accessor API for a consistent way to do this. Catch the unlikely failure of it not being set. Switch two cases where stderr was being used instead of debug_file. Signed-off-by:
-
Ian Rogers authored
Removing maps wasn't being done under the write lock. Similar to maps__for_each_map(), iterate the entries but in this case remove the entry based on the result of the callback. If an entry was removed then maps_by_name() also needs updating, so add missed flush. In dso__load_kcore(), the test of map to save would always be false with REFCNT_CHECKING because of a missing RC_CHK_ACCESS/RC_CHK_EQUAL. Signed-off-by:
-
Ian Rogers authored
Reduce scope of maps__for_each_entry() as maps__for_each_map() is a safer alternative holding the maps lock during iteration. Signed-off-by:
-
Ian Rogers authored
Switch machine__thread_dso_type() from loop macro maps__for_each_entry() to maps__for_each_map() function that takes a callback. The function holds the maps lock, which should be held during iteration. Signed-off-by:
-
Ian Rogers authored
Switch read_unwind_spec_eh_frame() from loop macro maps__for_each_entry() to maps__for_each_map() function that takes a callback. The function holds the maps lock, which should be held during iteration. Committer notes: Fixed up conflict with: 4fb54994 ("perf unwind-libunwind: Fix base address for .eh_frame") Signed-off-by:
-
Ruidong Tian authored
arm-cs-trace-disasm ignore disam the first branch sample, For example as follow, the instructions beteween 0x0000ffffae878750 and 0x0000ffffae878754 is lose: ARM CoreSight Trace Data Assembler Dump Event type: branches:uH Sample = { cpu: 0000 addr: 0x0000ffffae878750 phys_addr: 0x0000000000000000 ip: 0x0000000000000000 pid: 4003489 tid: 4003489 period: 1 time: 26765151766034 } Event type: branches:uH Sample = { cpu: 0000 addr: 0x0000000000000000 phys_addr: 0x0000000000000000 ip: 0x0000ffffae878754 pid: 4003489 tid: 4003489 period: 1 time: 26765151766034 } Initialize cpu_data earlier to fix it: ARM CoreSight Trace Data Assembler Dump Event type: branches:uH Sample = { cpu: 0000 addr: 0x0000000000000000 phys_addr: 0x0000000000000000 ip: 0x0000ffffae878754 pid: 4003489 tid: 4003489 period: 1 time: 26765151766034 } 0000000000028740 <ioctl>: (base address is 0x0000ffffae850000) 28750: b13ffc1f cmn x0, #4095 28754: 54000042 b.hs 0x2875c <ioctl+0x1c> test 4003489/4003489 [0000] 26765.151766034 __GI___ioctl+0x14 /usr/lib64/libc-2.32.so Event type: branches:uH Sample = { cpu: 0000 addr: 0x0000ffffa67535ac phys_addr: 0x0000000000000000 ip: 0x0000000000000000 pid: 4003489 tid: 4003489 period: 1 time: 26765151766034 } Reviewed-by:James Clark <james.clark@arm.com> Signed-off-by:
Ruidong Tian <tianruidong@linux.alibaba.com> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Al Grant <al.grant@arm.com> Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com> Cc: Leo Yan <leo.yan@linaro.org> Cc: Mathieu Poirier <mathieu.poirier@linaro.org> Cc: Mike Leach <mike.leach@linaro.org> Cc: Suzuki Poulouse <suzuki.poulose@arm.com> Cc: Tor Jeremiassen <tor@ti.com> Link: https://lore.kernel.org/r/20231214123304.34087-4-tianruidong@linux.alibaba.comSigned-off-by:
-
Ruidong Tian authored
For exectable ELF file, which e_type is ET_EXEC, dso start address is a absolute address other than offset. Just set vm_start to zero when dso start is 0x400000, which means it is a exectable file. Reviewed-by:
-
Veronika Molnarova authored
Archives generated by the command 'perf archive' have to be unpacked manually. Following the addition of option '--all' now there also exist a nested structure of tars, and after further discussion with Red Hat Global Support Services, they found a feature correctly unpacking archives of 'perf archive' convenient. Option '--unpack' of 'perf archive' unpacks archives generated by the command 'perf archive' as well as archives generated when used with option '--all'. The 'perf.data' file is placed in the current directory, while debug symbols are unpacked in '~/.debug' directory. A tar filename can be passed as an argument, and if not provided the command tries to find a viable perf.tar file for unpacking. Signed-off-by:
Veronika Molnarova <vmolnaro@redhat.com> Tested-by:
-
- 19 Dec, 2023 4 commits
-
-
Veronika Molnarova authored
'perf archive' has limited functionality and people from Red Hat Global Support Services sent a request for a new feature that would pack perf.data file together with an archive with debug symbols created by the command 'perf archive' as customers were being confused and often would forget to send perf.data file with the debug symbols. With this patch 'perf archive' now accepts an option '--all' that generates archive 'perf.all-hostname-date-time.tar.bz2' that holds file 'perf.data' and a sub-tar 'perf.symbols.tar.bz2' with debug symbols. The functionality of the command 'perf archive' was not changed. Committer testing: Run 'perf record' on a Intel 14900K machine, hybrid: root@number:~# perf record -a sleep 5s [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 4.006 MB perf.data (15427 samples) ] root@number:~# perf archive --all Now please run: $ tar xvf perf.all-number-20231219-104854.tar.bz2 && tar xvf perf.symbols.tar.bz2 -C ~/.debug wherever you need to run 'perf report' on. root@number:~# root@number:~# perf report --header-only # ======== # captured on : Tue Dec 19 10:48:48 2023 # header version : 1 # data offset : 1008 # data size : 4199936 # feat offset : 4200944 # hostname : number # os release : 6.6.4-200.fc39.x86_64 # perf version : 6.7.rc6.gca90f8e17b84 # arch : x86_64 # nrcpus online : 28 # nrcpus avail : 28 # cpudesc : Intel(R) Core(TM) i7-14700K # cpuid : GenuineIntel,6,183,1 # total memory : 32610508 kB # cmdline : /home/acme/bin/perf (deleted) record -a sleep 5s # event : name = cpu_atom/cycles/P, , id = { 5088024, 5088025, 5088026, 5088027, 5088028, 5088029, 5088030, 5088031, 5088032, 5088033, 5088034, 5088035 }, type = 0 (PERF_TYPE_HARDWARE), size> # event : name = cpu_core/cycles/P, , id = { 5088036, 5088037, 5088038, 5088039, 5088040, 5088041, 5088042, 5088043, 5088044, 5088045, 5088046, 5088047, 5088048, 5088049, 5088050, 5088051 },> # event : name = dummy:u, , id = { 5088052, 5088053, 5088054, 5088055, 5088056, 5088057, 5088058, 5088059, 5088060, 5088061, 5088062, 5088063, 5088064, 5088065, 5088066, 5088067, 5088068, 50> # CPU_TOPOLOGY info available, use -I to display # NUMA_TOPOLOGY info available, use -I to display # pmu mappings: cpu_atom = 10, cpu_core = 4, breakpoint = 5, cstate_core = 34, cstate_pkg = 35, i915 = 14, intel_bts = 11, intel_pt = 12, kprobe = 8, msr = 13, power = 36, software = 1, trac> # CACHE info available, use -I to display # time of first sample : 124739.850375 # time of last sample : 124744.855181 # sample duration : 5004.806 ms # sample duration : 5004.806 ms # MEM_TOPOLOGY info available, use -I to display # bpf_prog_info 2: bpf_prog_7cc47bbf07148bfe_hid_tail_call addr 0xffffffffc0000978 size 113 # bpf_prog_info 47: bpf_prog_713a545fe0530ce7_restrict_filesystems addr 0xffffffffc0000748 size 305 # bpf_prog_info 163: bpf_prog_bd834b0730296056 addr 0xffffffffc000df14 size 331 # bpf_prog_info 258: bpf_prog_ee0e253c78993a24_sd_devices addr 0xffffffffc001fc08 size 264 # bpf_prog_info 259: bpf_prog_40ddf486530245f5_sd_devices addr 0xffffffffc00204bc size 318 # bpf_prog_info 260: bpf_prog_6deef7357e7b4530_sd_fw_egress addr 0xffffffffc0020630 size 63 # bpf_prog_info 261: bpf_prog_6deef7357e7b4530_sd_fw_ingress addr 0xffffffffc0020688 size 63 # bpf_prog_info 262: bpf_prog_b37200ab714f0e17_sd_devices addr 0xffffffffc002072c size 110 # bpf_prog_info 263: bpf_prog_b90a282ee45cfed9_sd_devices addr 0xffffffffc00207d8 size 393 # bpf_prog_info 264: bpf_prog_ee0e253c78993a24_sd_devices addr 0xffffffffc002099c size 264 # bpf_prog_info 265: bpf_prog_6deef7357e7b4530_sd_fw_egress addr 0xffffffffc0020ad4 size 63 # bpf_prog_info 266: bpf_prog_6deef7357e7b4530_sd_fw_ingress addr 0xffffffffc0020b50 size 63 # bpf_prog_info 267: bpf_prog_ee0e253c78993a24_sd_devices addr 0xffffffffc002d98c size 264 # bpf_prog_info 268: bpf_prog_be31ae23198a0378_sd_devices addr 0xffffffffc002dac8 size 297 # bpf_prog_info 269: bpf_prog_ccbbf91f3c6979c7_sd_devices addr 0xffffffffc002dc54 size 360 # bpf_prog_info 270: bpf_prog_3a0ef5414c2f6fca_sd_devices addr 0xffffffffc002dde8 size 456 # bpf_prog_info 271: bpf_prog_6deef7357e7b4530_sd_fw_egress addr 0xffffffffc0020bd4 size 63 # bpf_prog_info 272: bpf_prog_6deef7357e7b4530_sd_fw_ingress addr 0xffffffffc00299b4 size 63 # bpf_prog_info 273: bpf_prog_ee0e253c78993a24_sd_devices addr 0xffffffffc002dfd0 size 264 # bpf_prog_info 274: bpf_prog_6deef7357e7b4530_sd_fw_egress addr 0xffffffffc0029a3c size 63 # bpf_prog_info 275: bpf_prog_6deef7357e7b4530_sd_fw_ingress addr 0xffffffffc002d71c size 63 # bpf_prog_info 276: bpf_prog_6deef7357e7b4530_sd_fw_egress addr 0xffffffffc002d7a8 size 63 # bpf_prog_info 277: bpf_prog_6deef7357e7b4530_sd_fw_ingress addr 0xffffffffc002e13c size 63 # bpf_prog_info 278: bpf_prog_6deef7357e7b4530_sd_fw_egress addr 0xffffffffc002e1a8 size 63 # bpf_prog_info 279: bpf_prog_6deef7357e7b4530_sd_fw_ingress addr 0xffffffffc002e234 size 63 # bpf_prog_info 280: bpf_prog_be31ae23198a0378_sd_devices addr 0xffffffffc002e2ac size 297 # bpf_prog_info 281: bpf_prog_6deef7357e7b4530_sd_fw_egress addr 0xffffffffc002e42c size 63 # bpf_prog_info 282: bpf_prog_6deef7357e7b4530_sd_fw_ingress addr 0xffffffffc002e49c size 63 # bpf_prog_info 290: bpf_prog_ee0e253c78993a24_sd_devices addr 0xffffffffc0004b18 size 264 # bpf_prog_info 294: bpf_prog_0b1566e4b83190c5_sd_devices addr 0xffffffffc0004c50 size 360 # bpf_prog_info 295: bpf_prog_ee0e253c78993a24_sd_devices addr 0xffffffffc001cfc8 size 264 # bpf_prog_info 296: bpf_prog_6deef7357e7b4530_sd_fw_egress addr 0xffffffffc0013abc size 63 # bpf_prog_info 297: bpf_prog_6deef7357e7b4530_sd_fw_ingress addr 0xffffffffc0013b24 size 63 # btf info of id 2 # btf info of id 52 # HYBRID_TOPOLOGY info available, use -I to display # cpu_atom pmu capabilities: branches=32, max_precise=3, pmu_name=alderlake_hybrid # cpu_core pmu capabilities: branches=32, max_precise=3, pmu_name=alderlake_hybrid # intel_pt pmu capabilities: topa_multiple_entries=1, psb_cyc=1, single_range_output=1, mtc_periods=249, ip_filtering=1, output_subsys=0, cr3_filtering=1, psb_periods=3f, event_trace=0, cycl> # missing features: TRACING_DATA BRANCH_STACK GROUP_DESC AUXTRACE STAT CLOCKID DIR_FORMAT COMPRESSED CPU_PMU_CAPS CLOCK_DATA # ======== # root@number:~# And then transferring it to a ARM64 machine, a Libre Computer RK3399-PC: root@number:~# scp perf.all-number-20231219-104854.tar.bz2 acme@192.168.86.114:. acme@192.168.86.114's password: perf.all-number-20231219-104854.tar.bz2 100% 145MB 85.4MB/s 00:01 root@number:~# root@number:~# ssh acme@192.168.86.114 acme@192.168.86.114's password: Welcome to Ubuntu 23.04 (GNU/Linux 6.1.68-12200-g1c40dda3081e aarch64) * Documentation: https://help.ubuntu.com * Management: https://landscape.canonical.com * Support: https://ubuntu.com/advantage Last login: Tue Dec 19 14:53:18 2023 from 192.168.86.42 acme@roc-rk3399-pc:~$ tar xvf perf.all-number-20231219-104854.tar.bz2 && tar xvf perf.symbols.tar.bz2 -C ~/.debug perf.data perf.symbols.tar.bz2 .build-id/ad/acc227f470409213308050b71f664322e2956c [kernel.kallsyms]/adacc227f470409213308050b71f664322e2956c/ [kernel.kallsyms]/adacc227f470409213308050b71f664322e2956c/kallsyms [kernel.kallsyms]/adacc227f470409213308050b71f664322e2956c/probes .build-id/76/c91f4d62baa06bb52e07e20aba36d21a8f9797 usr/lib64/libz.so.1.2.13/76c91f4d62baa06bb52e07e20aba36d21a8f9797/ <SNIP> .build-id/09/d7e96bc1e3f599d15ca28b36959124b2d74410 usr/lib64/librpm_sequoia.so.1/09d7e96bc1e3f599d15ca28b36959124b2d74410/ usr/lib64/librpm_sequoia.so.1/09d7e96bc1e3f599d15ca28b36959124b2d74410/elf usr/lib64/librpm_sequoia.so.1/09d7e96bc1e3f599d15ca28b36959124b2d74410/probes acme@roc-rk3399-pc:~$ acme@roc-rk3399-pc:~$ perf report --stdio | head -40 # To display the perf.data header info, please use --header/--header-only options. # # Total Lost Samples: 0 # # Samples: 6K of event 'cpu_atom/cycles/P' # Event count (approx.): 4519946621 # # Overhead Command Shared Object Symbol # ........ ............... .............................................. ......................................................................................................................................................... # 1.73% swapper [kernel.kallsyms] [k] intel_idle 1.43% sh [kernel.kallsyms] [k] next_uptodate_folio 0.94% make ld-linux-x86-64.so.2 [.] do_lookup_x 0.90% sh ld-linux-x86-64.so.2 [.] do_lookup_x 0.82% sh [kernel.kallsyms] [k] perf_event_mmap_output 0.74% sh [kernel.kallsyms] [k] filemap_map_pages 0.72% sh ld-linux-x86-64.so.2 [.] _dl_relocate_object 0.69% cc1 [kernel.kallsyms] [k] clear_page_erms 0.61% sh [kernel.kallsyms] [k] unmap_page_range 0.56% swapper [kernel.kallsyms] [k] poll_idle 0.52% cc1 ld-linux-x86-64.so.2 [.] do_lookup_x 0.47% make ld-linux-x86-64.so.2 [.] _dl_relocate_object 0.44% cc1 cc1 [.] make_node(tree_code) 0.43% sh [kernel.kallsyms] [k] native_irq_return_iret 0.38% sh libc.so.6 [.] _int_malloc 0.38% cc1 cc1 [.] decl_attributes(tree_node**, tree_node*, int, tree_node*) 0.38% sh [kernel.kallsyms] [k] clear_page_erms 0.37% cc1 cc1 [.] ht_lookup_with_hash(ht*, unsigned char const*, unsigned long, unsigned int, ht_lookup_option) 0.37% make [kernel.kallsyms] [k] perf_event_mmap_output 0.37% make ld-linux-x86-64.so.2 [.] _dl_lookup_symbol_x 0.35% sh [kernel.kallsyms] [k] _compound_head 0.35% make make [.] hash_find_slot 0.33% sh libc.so.6 [.] __strlen_avx2 0.33% cc1 cc1 [.] ggc_internal_alloc(unsigned long, void (*)(void*), unsigned long, unsigned long) 0.33% sh [kernel.kallsyms] [k] perf_iterate_ctx 0.31% make make [.] jhash_string 0.31% sh [kernel.kallsyms] [k] page_remove_rmap 0.30% cc1 libc.so.6 [.] _int_malloc 0.30% make libc.so.6 [.] _int_malloc acme@roc-rk3399-pc:~$ Signed-off-by: -
Arnaldo Carvalho de Melo authored
To pick up fixes that went thru perf-tools for v6.7 and to get in sync with upstream to check for drift in the copies of headers, etc. Signed-off-by: -
Ian Rogers authored
Switch thread__prepare_access from loop macro maps__for_each_entry to maps__for_each_map function that takes a callback. The function holds the maps lock, which should be held during iteration. Signed-off-by:
-
Ian Rogers authored
Switch perf_event__synthesize_modules from loop macro maps__for_each_entry to maps__for_each_map function that takes a callback. The function holds the maps lock, which should be held during iteration. Signed-off-by:
-
