kernel/tools/perf/util/symbol.c, branch linux-5.2.y

perf annotate: Fix s390 gap between kernel end and module start

2019-08-16T08:10:55Z

commit b9c0a64901d5bdec6eafd38d1dc8fa0e2974fccb upstream. During execution of command 'perf top' the error message: Not enough memory for annotating '__irf_end' symbol!) is emitted from this call sequence: __cmd_top perf_top__mmap_read perf_top__mmap_read_idx perf_event__process_sample hist_entry_iter__add hist_iter__top_callback perf_top__record_precise_ip hist_entry__inc_addr_samples symbol__inc_addr_samples symbol__get_annotation symbol__alloc_hist In this function the size of symbol __irf_end is calculated. The size of a symbol is the difference between its start and end address. When the symbol was read the first time, its start and end was set to: symbol__new: __irf_end 0xe954d0-0xe954d0 which is correct and maps with /proc/kallsyms: root@s8360046:~/linux-4.15.0/tools/perf# fgrep _irf_end /proc/kallsyms 0000000000e954d0 t __irf_end root@s8360046:~/linux-4.15.0/tools/perf# In function symbol__alloc_hist() the end of symbol __irf_end is symbol__alloc_hist sym:__irf_end start:0xe954d0 end:0x3ff80045a8 which is identical with the first module entry in /proc/kallsyms This results in a symbol size of __irf_req for histogram analyses of 70334140059072 bytes and a malloc() for this requested size fails. The root cause of this is function __dso__load_kallsyms() +-> symbols__fixup_end() Function symbols__fixup_end() enlarges the last symbol in the kallsyms map: # fgrep __irf_end /proc/kallsyms 0000000000e954d0 t __irf_end # to the start address of the first module: # cat /proc/kallsyms | sort | egrep ' [tT] ' .... 0000000000e952d0 T __security_initcall_end 0000000000e954d0 T __initramfs_size 0000000000e954d0 t __irf_end 000003ff800045a8 T fc_get_event_number [scsi_transport_fc] 000003ff800045d0 t store_fc_vport_disable [scsi_transport_fc] 000003ff800046a8 T scsi_is_fc_rport [scsi_transport_fc] 000003ff800046d0 t fc_target_setup [scsi_transport_fc] On s390 the kernel is located around memory address 0x200, 0x10000 or 0x100000, depending on linux version. Modules however start some- where around 0x3ff xxxx xxxx. This is different than x86 and produces a large gap for which histogram allocation fails. Fix this by detecting the kernel's last symbol and do no adjustment for it. Introduce a weak function and handle s390 specifics. Reported-by: Klaus Theurich Signed-off-by: Thomas Richter Acked-by: Heiko Carstens Cc: Hendrik Brueckner Cc: Vasily Gorbik Cc: stable@vger.kernel.org Link: http://lkml.kernel.org/r/20190724122703.3996-2-tmricht@linux.ibm.com Signed-off-by: Arnaldo Carvalho de Melo Signed-off-by: Greg Kroah-Hartman

perf symbols: Introduce DSO_BINARY_TYPE__BPF_PROG_INFO

2019-03-19T19:52:07Z

Introduce a new dso type DSO_BINARY_TYPE__BPF_PROG_INFO for BPF programs. In symbol__disassemble(), DSO_BINARY_TYPE__BPF_PROG_INFO dso will call into a new function symbol__disassemble_bpf() in an upcoming patch, where annotation line information is filled based bpf_prog_info and btf saved in given perf_env. Committer notes: Removed the unnamed union with 'bpf_prog' and 'cache' in 'struct dso', to fix this bug when exiting 'perf top': # perf top perf: Segmentation fault -------- backtrace -------- perf[0x5a785a] /lib64/libc.so.6(+0x385bf)[0x7fd68443c5bf] perf(rb_first+0x2b)[0x4d6eeb] perf(dso__delete+0xb7)[0x4dffb7] perf[0x4f9e37] perf(perf_session__delete+0x64)[0x504df4] perf(cmd_top+0x1957)[0x454467] perf[0x4aad18] perf(main+0x61c)[0x42ec7c] /lib64/libc.so.6(__libc_start_main+0xf2)[0x7fd684428412] perf(_start+0x2d)[0x42eead] # # addr2line -fe ~/bin/perf 0x4dffb7 dso_cache__free /home/acme/git/perf/tools/perf/util/dso.c:713 That is trying to access the dso->data.cache, and that is not used with BPF programs, so we end up accessing what is in bpf_prog.first_member, b00m. Signed-off-by: Song Liu Reviewed-by: Jiri Olsa Cc: Alexei Starovoitov Cc: Daniel Borkmann Cc: Namhyung Kim Cc: Peter Zijlstra Cc: Stanislav Fomichev Cc: kernel-team@fb.com Link: http://lkml.kernel.org/r/20190312053051.2690567-13-songliubraving@fb.com [ split from a larger patch ] Signed-off-by: Arnaldo Carvalho de Melo

perf report: Implement browsing of individual samples

2019-03-11T19:33:19Z

Now 'perf report' can show whole time periods with 'perf script', but the user still has to find individual samples of interest manually. It would be expensive and complicated to search for the right samples in the whole perf file. Typically users only need to look at a small number of samples for useful analysis. Also the full scripts tend to show samples of all CPUs and all threads mixed up, which can be very confusing on larger systems. Add a new --samples option to save a small random number of samples per hist entry. Use a reservoir sample technique to select a representatve number of samples. Then allow browsing the samples using 'perf script' as part of the hist entry context menu. This automatically adds the right filters, so only the thread or cpu of the sample is displayed. Then we use less' search functionality to directly jump the to the time stamp of the selected sample. It uses different menus for assembler and source display. Assembler needs xed installed and source needs debuginfo. Currently it only supports as many samples as fit on the screen due to some limitations in the slang ui code. Signed-off-by: Andi Kleen Tested-by: Arnaldo Carvalho de Melo Acked-by: Jiri Olsa Link: http://lkml.kernel.org/r/20190311174605.GA29294@tassilo.jf.intel.com Signed-off-by: Arnaldo Carvalho de Melo

perf report: Parse time quantum

2019-03-11T14:56:03Z

Many workloads change over time. 'perf report' currently aggregates the whole time range reported in perf.data. This patch adds an option for a time quantum to quantisize the perf.data over time. This just adds the option, will be used in follow on patches for a time sort key. Signed-off-by: Andi Kleen Cc: Jiri Olsa Cc: Namhyung Kim Link: http://lkml.kernel.org/r/20190305144758.12397-6-andi@firstfloor.org [ Use NSEC_PER_[MU]SEC ] Signed-off-by: Arnaldo Carvalho de Melo

perf report: Support output in nanoseconds

2019-03-11T14:56:02Z

Upcoming changes add timestamp output in perf report. Add a --ns argument similar to perf script to support nanoseconds resolution when needed. Signed-off-by: Andi Kleen Cc: Jiri Olsa Cc: Namhyung Kim Link: http://lkml.kernel.org/r/20190305144758.12397-5-andi@firstfloor.org Signed-off-by: Arnaldo Carvalho de Melo

perf tools: Fix split_kallsyms_for_kcore() for trampoline symbols

2019-02-06T13:00:40Z

Kallsyms symbols do not have a size, so the size becomes the distance to the next symbol. Consequently the recently added trampoline symbols end up with large sizes because the trampolines are some distance from one another and the main kernel map. However, symbols that end outside their map can disrupt the symbol tree because, after mapping, it can appear incorrectly that they overlap other symbols. Add logic to truncate symbol size to the end of the corresponding map. Signed-off-by: Adrian Hunter Acked-by: Jiri Olsa Cc: stable@vger.kernel.org Fixes: d83212d5dd67 ("kallsyms, x86: Export addresses of PTI entry trampolines") Link: http://lkml.kernel.org/r/20190109091835.5570-2-adrian.hunter@intel.com Signed-off-by: Arnaldo Carvalho de Melo

pref tools: Add missing map.h includes

2019-02-06T13:00:38Z

Lots of places get the map.h file indirectly, and since we're going to remove it from machine.h, then those need to include it directly, do it now, before we remove that dep. Cc: Adrian Hunter Cc: Jiri Olsa Cc: Namhyung Kim Link: https://lkml.kernel.org/n/tip-ob8jehdjda8h5jsrv9dqj9tf@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo

perf symbols: Use cached rbtrees

2019-01-25T14:12:10Z

At the cost of an extra pointer, we can avoid the O(logN) cost of finding the first element in the tree (smallest node). Signed-off-by: Davidlohr Bueso Tested-by: Arnaldo Carvalho de Melo Cc: Jiri Olsa Cc: Namhyung Kim Link: http://lkml.kernel.org/r/20181206191819.30182-6-dave@stgolabs.net Signed-off-by: Arnaldo Carvalho de Melo

perf symbols: Add 'arch_cpu_idle' to the list of kernel idle symbols

2019-01-09T19:21:15Z

When testing 'perf top' on a armhf system (32-bit, Orange Pi Zero), I noticed that 'arch_cpu_idle' dominated, add it to the list of idle symbols, so that we can see what is that being done when not idle. Cc: Adrian Hunter Cc: Jiri Olsa Cc: Namhyung Kim Link: https://lkml.kernel.org/n/tip-4q2b5g4p2hrstrhp9t2mrlho@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo

perf symbols: Relax checks on perf-PID.map ownership

2018-12-18T19:17:41Z

Those are simple enough, and usually not produced by root, instead by whatever user is running java, rust, Node.js JIT code that end up generating those /tmp/perf-PID.map for resolution of symbols in the anonymous executable maps. Having to use --force to resolve symbols in 'perf top' is a distraction, as recently I experienced when node.js symbols were not being resolved by 'perf top'. Acked-by: Ingo Molnar Cc: Adrian Hunter Cc: Frederic Weisbecker Cc: Hítalo Silva Cc: Jiri Olsa Cc: Namhyung Kim Cc: Pekka Enberg Cc: Peter Zijlstra Link: https://lkml.kernel.org/n/tip-tk2jgo2v4v2yjuj28axbpppo@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo