kernel/linux.git/tools/perf/util/sort.c, branch v5.9.2

perf util: Fix memory leak of prefix_if_not_in

2020-05-28T13:03:27+00:00

Need to free "str" before return when asprintf() failed to avoid memory leak. Signed-off-by: Xie XiuQi Cc: Alexander Shishkin Cc: Hongbo Yao Cc: Jiri Olsa Cc: Li Bin Cc: Mark Rutland Cc: Namhyung Kim Link: http://lore.kernel.org/lkml/20200521133218.30150-4-liwei391@huawei.com Signed-off-by: Arnaldo Carvalho de Melo

perf symbol: Fix kernel symbol address display

2020-05-05T19:35:32+00:00

Running commands ./perf record -e rb0000 -- find . ./perf report -v reveals symbol names and its addresses. There is a mismatch between kernel symbol and address. Here is an example for kernel symbol check_chain_key: 3.55% find /lib/modules/.../build/vmlinux 0xf11ec v [k] check_chain_key This address is off by 0xff000 as can be seen with: [root@t35lp46 ~]# fgrep check_chain_key /proc/kallsyms 00000000001f00d0 t check_chain_key [root@t35lp46 ~]# objdump -t ~/linux/vmlinux| fgrep check_chain_key 00000000001f00d0 l F .text 00000000000001e8 check_chain_key [root@t35lp46 ~]# This function is located in main memory 0x1f00d0 - 0x1f02b4. It has several entries in the perf data file with the correct address: [root@t35lp46 perf]# ./perf report -D -i perf.data.find-bad | \ fgrep SAMPLE| fgrep 0x1f01ec PERF_RECORD_SAMPLE(IP, 0x1): 22228/22228: 0x1f01ec period: 1300000 addr: 0 PERF_RECORD_SAMPLE(IP, 0x1): 22228/22228: 0x1f01ec period: 1300000 addr: 0 The root cause happens when reading symbol tables during perf report. A long gdb call chain leads to machine__deliver_events perf_evlist__deliver_event perf_evlist__deliver_sample build_id__mark_dso_hits thread__find_map(1) Read correct address from sample entry map__load dso__load Some more functions to end up in .... dso__load_sym. Function dso__load_syms checks for kernel relocation and symbol adjustment for the kernel and results in kernel map adjustment of kernel .text segment address (0x100000 on s390) kernel .text segment offset in file (0x1000 on s390). This results in all kernel symbol addresses to be changed by subtracting 0xff000 (on s390). For the symbol check_chain_key we end up with 0x1f00d0 - 0x100000 + 0x1000 = 0xf11d0 and this address is saved in the perf symbol table. This calculation is also applied by the mapping functions map__mapip() and map__unmapip() to map IP addresses to dso mappings. During perf report processing functions process_sample_event (builtin-report.c) machine__resolve thread__find_map hist_entry_iter_add are called. Function thread__find_map(1) takes the correct sample address and applies the mapping function map__mapip() from the kernel dso and saves the modified address in struct addr_location for further reference. From now on this address is used. Funktion process_sample_event() then calls hist_entry_iter_add() to save the address in member ip of struct hist_entry. When samples are displayed using perf_evlist__tty_browse_hists hists__fprintf hist_entry__fprintf hist_entry__snprintf __hist_entry__snprintf _hist_entry__sym_snprintf() This simply displays the address of the symbol and ignores the dso <-> map mappings done in function thread__find_map. This leads to the address mismatch. Output before: ot@t35lp46 perf]# ./perf report -v | fgrep check_chain_key 3.55% find /lib/modules/5.6.0d-perf+/build/vmlinux 0xf11ec v [k] check_chain_key [root@t35lp46 perf]# Output after: [root@t35lp46 perf]# ./perf report -v | fgrep check_chain_key 3.55% find /lib/modules/5.6.0d-perf+/build/vmlinux 0x1f01ec v [k] check_chain_key [root@t35lp46 perf]# Signed-off-by: Thomas Richter Acked-by: Sumanth Korikkar Cc: Heiko Carstens Cc: Vasily Gorbik Link: http://lore.kernel.org/lkml/20200415070744.59919-1-tmricht@linux.ibm.com Signed-off-by: Arnaldo Carvalho de Melo

perf evsel: Rename perf_evsel{prev,next}() to evsel{prev,next}()

2020-05-05T19:35:31+00:00

As those are 'struct evsel' methods, not part of tools/lib/perf/, aka libperf, to whom the perf_ prefix belongs. Cc: Adrian Hunter Cc: Jiri Olsa Cc: Namhyung Kim Signed-off-by: Arnaldo Carvalho de Melo

perf hist: Add fast path for duplicate entries check

2020-04-18T12:05:01+00:00

Perf checks the duplicate entries in a callchain before adding an entry. However the check is very slow especially with deeper call stack. Almost ~50% elapsed time of perf report is spent on the check when the call stack is always depth of 32. The hist_entry__cmp() is used to compare the new entry with the old entries. It will go through all the available sorts in the sort_list, and call the specific cmp of each sort, which is very slow. Actually, for most cases, there are no duplicate entries in callchain. The symbols are usually different. It's much faster to do a quick check for symbols first. Only do the full cmp when the symbols are exactly the same. The quick check is only to check symbols, not dso. Export _sort__sym_cmp. $ perf record --call-graph lbr ./tchain_edit_64 Without the patch $time perf report --stdio real 0m21.142s user 0m21.110s sys 0m0.033s With the patch $time perf report --stdio real 0m10.977s user 0m10.948s sys 0m0.027s Signed-off-by: Kan Liang Acked-by: Jiri Olsa Cc: Adrian Hunter Cc: Alexey Budankov Cc: Andi Kleen Cc: Mathieu Poirier Cc: Michael Ellerman Cc: Namhyung Kim Cc: Pavel Gerasimov Cc: Peter Zijlstra Cc: Ravi Bangoria Cc: Stephane Eranian Cc: Vitaly Slobodskoy Link: http://lore.kernel.org/lkml/20200319202517.23423-18-kan.liang@linux.intel.com Signed-off-by: Arnaldo Carvalho de Melo

perf report: Add 'cgroup' sort key

2020-04-03T12:37:55+00:00

The cgroup sort key is to show cgroup membership of each task. Currently it shows full path in the cgroupfs (not relative to the root of cgroup namespace) since it'd be more intuitive IMHO. Otherwise root cgroup in different namespaces will all show same name - "/". The cgroup sort key should come before cgroup_id otherwise sort_dimension__add() will match it to cgroup_id as it only matches with the given substring. For example it will look like following. Note that record patch adding --all-cgroups patch will come later. $ perf record -a --namespace --all-cgroups cgtest [ perf record: Woken up 1 times to write data ] [ perf record: Captured and wrote 0.208 MB perf.data (4090 samples) ] $ perf report -s cgroup_id,cgroup,pid ... # Overhead cgroup id (dev/inode) Cgroup Pid:Command # ........ ..................... .......... ............... # 93.96% 0/0x0 / 0:swapper 1.25% 3/0xeffffffb / 278:looper0 0.86% 3/0xf000015f /sub/cgrp1 280:cgtest 0.37% 3/0xf0000160 /sub/cgrp2 281:cgtest 0.34% 3/0xf0000163 /sub/cgrp3 282:cgtest 0.22% 3/0xeffffffb /sub 278:looper0 0.20% 3/0xeffffffb / 280:cgtest 0.15% 3/0xf0000163 /sub/cgrp3 285:looper3 Signed-off-by: Namhyung Kim Cc: Alexander Shishkin Cc: Jiri Olsa Cc: Mark Rutland Cc: Peter Zijlstra Link: http://lore.kernel.org/lkml/20200325124536.2800725-6-namhyung@kernel.org Signed-off-by: Arnaldo Carvalho de Melo

perf report: Print al_addr when symbol is not found

2020-03-23T14:08:29+00:00

For branch mode, if the symbol is not found, it prints the address. For example, 0x0000555eee0365a0 in below output. Overhead Command Source Shared Object Source Symbol Target Symbol 17.55% div libc-2.27.so [.] __random [.] __random 6.11% div div [.] 0x0000555eee0365a0 [.] rand 6.10% div libc-2.27.so [.] rand [.] 0x0000555eee036769 5.80% div libc-2.27.so [.] __random_r [.] __random 5.72% div libc-2.27.so [.] __random [.] __random_r 5.62% div libc-2.27.so [.] __random_r [.] __random_r 5.38% div libc-2.27.so [.] __random [.] rand 4.56% div libc-2.27.so [.] __random [.] __random 4.49% div div [.] 0x0000555eee036779 [.] 0x0000555eee0365ff 4.25% div div [.] 0x0000555eee0365fa [.] 0x0000555eee036760 But it's not very easy to understand what the instructions are in the binary. So this patch uses the al_addr instead. With this patch, the output is Overhead Command Source Shared Object Source Symbol Target Symbol 17.55% div libc-2.27.so [.] __random [.] __random 6.11% div div [.] 0x00000000000005a0 [.] rand 6.10% div libc-2.27.so [.] rand [.] 0x0000000000000769 5.80% div libc-2.27.so [.] __random_r [.] __random 5.72% div libc-2.27.so [.] __random [.] __random_r 5.62% div libc-2.27.so [.] __random_r [.] __random_r 5.38% div libc-2.27.so [.] __random [.] rand 4.56% div libc-2.27.so [.] __random [.] __random 4.49% div div [.] 0x0000000000000779 [.] 0x00000000000005ff 4.25% div div [.] 0x00000000000005fa [.] 0x0000000000000760 Now we can use objdump to dump the object starting from 0x5a0. For example, objdump -d --start-address 0x5a0 div 00000000000005a0 : 5a0: ff 25 2a 0a 20 00 jmpq *0x200a2a(%rip) # 200fd0 <__cxa_finalize@plt+0x200a20> 5a6: 68 02 00 00 00 pushq $0x2 5ab: e9 c0 ff ff ff jmpq 570 ... Committer testing: [root@seventh ~]# perf record -a -b sleep 1 [root@seventh ~]# perf report --header-only | grep cpudesc # cpudesc : Intel(R) Core(TM) i5-7500 CPU @ 3.40GHz [root@seventh ~]# perf evlist -v cycles: size: 120, { sample_period, sample_freq }: 4000, sample_type: IP|TID|TIME|CPU|PERIOD|BRANCH_STACK, read_format: ID, disabled: 1, inherit: 1, mmap: 1, comm: 1, freq: 1, task: 1, precise_ip: 3, sample_id_all: 1, exclude_guest: 1, mmap2: 1, comm_exec: 1, ksymbol: 1, bpf_event: 1, branch_sample_type: ANY [root@seventh ~]# Before: [root@seventh ~]# perf report --stdio --dso libsystemd-shared-241.so | head -20 # To display the perf.data header info, please use --header/--header-only options. # # # Total Lost Samples: 0 # # Samples: 2K of event 'cycles' # Event count (approx.): 2240 # # Overhead Command Source Shared Object Source Symbol Target Symbol Basic Block Cycles # ........ ............... ........................ ...................... ...................... .................. # 0.13% systemd-journal libc-2.29.so [.] cfree@GLIBC_2.2.5 [.] _int_free 1 0.09% systemd libsystemd-shared-241.so [.] 0x00007fe406465c82 [.] 0x00007fe406465d80 1 0.09% systemd libsystemd-shared-241.so [.] 0x00007fe406465ded [.] 0x00007fe406465c30 1 0.09% systemd libsystemd-shared-241.so [.] 0x00007fe406465e4e [.] 0x00007fe406465de0 1 0.09% systemd-journal systemd-journald [.] free@plt [.] cfree@GLIBC_2.2.5 1 0.09% systemd-journal libc-2.29.so [.] _int_free [.] _int_free 18 0.09% systemd-journal libc-2.29.so [.] _int_free [.] _int_free 2 0.04% systemd libsystemd-shared-241.so [.] bus_resolve@plt [.] bus_resolve 204 0.04% systemd libsystemd-shared-241.so [.] getpid_cached@plt [.] getpid_cached 7 [root@seventh ~]# After: [root@seventh ~]# perf report --stdio --dso libsystemd-shared-241.so | head -20 # To display the perf.data header info, please use --header/--header-only options. # # # Total Lost Samples: 0 # # Samples: 2K of event 'cycles' # Event count (approx.): 2240 # # Overhead Command Source Shared Object Source Symbol Target Symbol Basic Block Cycles # ........ ............... ........................ ...................... ...................... .................. # 0.13% systemd-journal libc-2.29.so [.] cfree@GLIBC_2.2.5 [.] _int_free 1 0.09% systemd libsystemd-shared-241.so [.] 0x00000000000f7c82 [.] 0x00000000000f7d80 1 0.09% systemd libsystemd-shared-241.so [.] 0x00000000000f7ded [.] 0x00000000000f7c30 1 0.09% systemd libsystemd-shared-241.so [.] 0x00000000000f7e4e [.] 0x00000000000f7de0 1 0.09% systemd-journal systemd-journald [.] free@plt [.] cfree@GLIBC_2.2.5 1 0.09% systemd-journal libc-2.29.so [.] _int_free [.] _int_free 18 0.09% systemd-journal libc-2.29.so [.] _int_free [.] _int_free 2 0.04% systemd libsystemd-shared-241.so [.] bus_resolve@plt [.] bus_resolve 204 0.04% systemd libsystemd-shared-241.so [.] getpid_cached@plt [.] getpid_cached 7 [root@seventh ~]# Lets use -v to get full paths and then try objdump on the unresolved address: [root@seventh ~]# perf report -v --stdio --dso libsystemd-shared-241.so |& grep libsystemd-shared-241.so | tail -1 0.04% systemd-journal /usr/lib/systemd/libsystemd-shared-241.so 0x80c1a B [.] 0x0000000000080c1a 0x80a95 B [.] 0x0000000000080a95 61 [root@seventh ~]# [root@seventh ~]# objdump -d --start-address 0x00000000000f7d80 /usr/lib/systemd/libsystemd-shared-241.so | head -20 /usr/lib/systemd/libsystemd-shared-241.so: file format elf64-x86-64 Disassembly of section .text: 00000000000f7d80 : f7d80: 41 39 11 cmp %edx,(%r9) f7d83: 0f 84 ff fe ff ff je f7c88 f7d89: 4c 8d 05 97 09 0c 00 lea 0xc0997(%rip),%r8 # 1b8727 f7d90: b9 49 00 00 00 mov $0x49,%ecx f7d95: 48 8d 15 c9 f5 0b 00 lea 0xbf5c9(%rip),%rdx # 1b7365 f7d9c: 31 ff xor %edi,%edi f7d9e: 48 8d 35 9b ff 0b 00 lea 0xbff9b(%rip),%rsi # 1b7d40 f7da5: e8 a6 d6 f4 ff callq 45450 f7daa: 66 0f 1f 44 00 00 nopw 0x0(%rax,%rax,1) f7db0: 41 56 push %r14 f7db2: 41 55 push %r13 f7db4: 41 54 push %r12 f7db6: 55 push %rbp [root@seventh ~]# If we tried the the reported address before this patch: [root@seventh ~]# objdump -d --start-address 0x00007fe406465d80 /usr/lib/systemd/libsystemd-shared-241.so | head -20 /usr/lib/systemd/libsystemd-shared-241.so: file format elf64-x86-64 [root@seventh ~]# Signed-off-by: Jin Yao Tested-by: Arnaldo Carvalho de Melo Tested-by: Ravi Bangoria Acked-by: Jiri Olsa Cc: Alexander Shishkin Cc: Andi Kleen Cc: Kan Liang Cc: Peter Zijlstra Link: http://lore.kernel.org/lkml/20200227043939.4403-2-yao.jin@linux.intel.com Signed-off-by: Arnaldo Carvalho de Melo

perf report/top: Improve toggle callchain menu option

2020-01-06T14:46:10+00:00

Taking into account the current status of the callchain, i.e. if folded, show "Expand", otherwise "Collapse", also show the name of the entry that will be affected and mention the hotkeys for expanding/collapsing all callchains below the main entry, the one that appears with/without callchains. Reviewed-by: Jiri Olsa Cc: Adrian Hunter Cc: Andi Kleen Cc: Jin Yao Cc: Kan Liang Cc: Linus Torvalds Cc: Namhyung Kim Link: https://lkml.kernel.org/n/tip-03arm6poo8463k5tfcfp7gkk@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo

perf report: Make -F more strict like -s

2019-12-04T15:32:40+00:00

Currently -F allows branch-mode / mem-mode fields with -F even when perf report is not running in that mode. Don't allow that. Suggested-by: Arnaldo Carvalho de Melo Signed-off-by: Ravi Bangoria Acked-by: Jiri Olsa Cc: Alexander Shishkin Cc: Andi Kleen Cc: Jin Yao Cc: Kan Liang Cc: Mark Rutland Cc: Namhyung Kim Link: http://lore.kernel.org/lkml/20191114132213.5419-3-ravi.bangoria@linux.ibm.com Signed-off-by: Arnaldo Carvalho de Melo

perf report/top TUI: Replace pr_err() with ui__error()

2019-12-04T15:27:14+00:00

pr_err() in TUI mode does not print anyting on the screen and just quits. Replace such pr_err() with ui__error(). Before: $ perf report -s + $ After: $ perf report -s + ┌─Error:────────────────┐ │Invalid --sort key: `+'│ │ │ │Press any key... │ └───────────────────────┘ Signed-off-by: Ravi Bangoria Acked-by: Jiri Olsa Tested-by: Arnaldo Carvalho de Melo Cc: Alexander Shishkin Cc: Andi Kleen Cc: Jin Yao Cc: Kan Liang Cc: Mark Rutland Cc: Namhyung Kim Link: http://lore.kernel.org/lkml/20191114132213.5419-2-ravi.bangoria@linux.ibm.com Signed-off-by: Arnaldo Carvalho de Melo

perf dso: Move dso_id from 'struct map' to 'struct dso'

2019-11-19T22:12:26+00:00

And take it into account when looking up DSOs when we have the dso_id fields obtained from somewhere, like from PERF_RECORD_MMAP2 records. Instances of struct map pointing to the same DSO pathname but with anything in dso_id different are in fact different DSOs, so better have different 'struct dso' instances to reflect that. At some point we may want to get copies of the contents of the different objects if we want to do correct annotation or other analysis. With this we get 'struct map' 24 bytes leaner: $ pahole -C map ~/bin/perf struct map { union { struct rb_node rb_node __attribute__((__aligned__(8))); /* 0 24 */ struct list_head node; /* 0 16 */ } __attribute__((__aligned__(8))); /* 0 24 */ u64 start; /* 24 8 */ u64 end; /* 32 8 */ _Bool erange_warned:1; /* 40: 0 1 */ _Bool priv:1; /* 40: 1 1 */ /* XXX 6 bits hole, try to pack */ /* XXX 3 bytes hole, try to pack */ u32 prot; /* 44 4 */ u64 pgoff; /* 48 8 */ u64 reloc; /* 56 8 */ /* --- cacheline 1 boundary (64 bytes) --- */ u64 (*map_ip)(struct map *, u64); /* 64 8 */ u64 (*unmap_ip)(struct map *, u64); /* 72 8 */ struct dso * dso; /* 80 8 */ refcount_t refcnt; /* 88 4 */ u32 flags; /* 92 4 */ /* size: 96, cachelines: 2, members: 13 */ /* sum members: 92, holes: 1, sum holes: 3 */ /* sum bitfield members: 2 bits, bit holes: 1, sum bit holes: 6 bits */ /* forced alignments: 1 */ /* last cacheline: 32 bytes */ } __attribute__((__aligned__(8))); $ Cc: Adrian Hunter Cc: Andi Kleen Cc: Jiri Olsa Cc: Namhyung Kim Link: https://lkml.kernel.org/n/tip-g4hxxmraplo7wfjmk384mfsb@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo