kernel/linux.git/tools/perf/util/jitdump.c, branch v6.1.168

perf jit: Remove unused struct debug_line_info

2022-10-06T11:03:51+00:00

The struct debug_line_info is never used, remove it. Signed-off-by: Yuan Can Cc: Alexander Shishkin Cc: Jiri Olsa Cc: Mark Rutland Cc: Namhyung Kim Cc: Peter Zijlstra Link: https://lore.kernel.org/linux-perf-users/20220927013931.110475-5-yuancan@huawei.com Signed-off-by: Arnaldo Carvalho de Melo

perf inject jit: Ignore memfd and anonymous mmap events if jitdump present

2022-08-10T13:44:02+00:00

Some processes store jitted code in memfd mappings to avoid having rwx mappings. These processes map the code with a writeable mapping and a read-execute mapping. They write the code using the writeable mapping and then unmap the writeable mapping. All subsequent execution is through the read-execute mapping. perf inject --jit ignores //anon* mappings for each process where a jitdump is present because it expects to inject mmap events for each jitted code range, and said jitted code ranges will overlap with the //anon* mappings. Ignore /memfd: and [anon:* mappings so that jitted code contained in /memfd: and [anon:* mappings is treated the same way as jitted code contained in //anon* mappings. Signed-off-by: Brian Robbins Acked-by: Ian Rogers Cc: Alexander Shishkin Cc: Ingo Molnar Cc: Jiri Olsa Cc: Mark Rutland Cc: Namhyung Kim Cc: Peter Zijlstra Link: https://lore.kernel.org/r/20220805220645.95855-1-brianrob@linux.microsoft.com Signed-off-by: Arnaldo Carvalho de Melo

perf namespaces: Add functions to access nsinfo

2022-02-11T17:31:22+00:00

Having functions to access nsinfo reduces the places where reference counting checking needs to be added. Signed-off-by: Ian Rogers Cc: Adrian Hunter Cc: Alexander Shishkin Cc: Alexey Bayduraev Cc: Andi Kleen Cc: Andrew Morton Cc: André Almeida Cc: Andy Shevchenko Cc: Darren Hart Cc: Davidlohr Bueso Cc: Dmitriy Vyukov Cc: Eric Dumazet Cc: German Gomez Cc: Hao Luo Cc: James Clark Cc: Jin Yao Cc: Jiri Olsa Cc: John Garry Cc: Kajol Jain Cc: Kan Liang Cc: Leo Yan Cc: Madhavan Srinivasan Cc: Mark Rutland Cc: Masami Hiramatsu Cc: Miaoqian Lin Cc: Namhyung Kim Cc: Peter Zijlstra Cc: Riccardo Mancini Cc: Shunsuke Nakamura Cc: Song Liu Cc: Stephane Eranian Cc: Stephen Brennan Cc: Steven Rostedt (VMware) Cc: Thomas Gleixner Cc: Thomas Richter Cc: Yury Norov Link: http://lore.kernel.org/lkml/20220211103415.2737789-14-irogers@google.com Signed-off-by: Arnaldo Carvalho de Melo

perf jit: Let convert_timestamp() to be backwards-compatible

2021-04-29T13:31:00+00:00

Commit d110162cafc80dad ("perf tsc: Support cap_user_time_short for event TIME_CONV") supports the extended parameters for event TIME_CONV, but it broke the backwards compatibility, so any perf data file with old event format fails to convert timestamp. This patch introduces a helper event_contains() to check if an event contains a specific member or not. For the backwards-compatibility, if the event size confirms the extended parameters are supported in the event TIME_CONV, then copies these parameters. Committer notes: To make this compiler backwards compatible add this patch: - struct perf_tsc_conversion tc = { 0 }; + struct perf_tsc_conversion tc = { .time_shift = 0, }; Fixes: d110162cafc8 ("perf tsc: Support cap_user_time_short for event TIME_CONV") Signed-off-by: Leo Yan Acked-by: Adrian Hunter Cc: Alexander Shishkin Cc: Gustavo A. R. Silva Cc: Ingo Molnar Cc: Jiri Olsa Cc: Kan Liang Cc: Mark Rutland Cc: Namhyung Kim Cc: Peter Zijlstra Cc: Steve MacLean Cc: Yonatan Goldschmidt Link: https://lore.kernel.org/r/20210428120915.7123-3-leo.yan@linaro.org Signed-off-by: Arnaldo Carvalho de Melo

perf inject jit: Add namespaces support

2021-02-03T16:10:44+00:00

This patch fixes "perf inject --jit" to properly operate on namespaced/containerized processes: * jitdump files are generated by the process, thus they should be looked up in its mount NS. * DSOs of injected MMAP events will later be looked up in the process mount NS, so write them into its NS. * PIDs & TIDs from jitdump events need to be translated to the PID as seen by "perf record" before written into MMAP events. For a process in a different PID NS, the TID & PID given in the jitdump event are actually ignored; I use the TID & PID of the thread which mmap()ed the jitdump file. This is simplified and won't do for forks of the initial process, if they continue using the same jitdump file. Future patches might improve it. This was tested by recording a NodeJS process running with "--perf-prof", inside a Docker container, and by recording another NodeJS process running in the same namespaces as perf itself, to make sure it's not broken for non-containerized processes. Signed-off-by: Yonatan Goldschmidt Acked-by: Jiri Olsa Acked-by: Namhyung Kim Cc: Alexander Shishkin Cc: Ingo Molnar Cc: Mark Rutland Cc: Peter Zijlstra Cc: Stephane Eranian Link: https://lore.kernel.org/r/20201105015604.1726943-1-yonatan.goldschmidt@granulate.io Signed-off-by: Arnaldo Carvalho de Melo

perf tsc: Support cap_user_time_short for event TIME_CONV

2020-09-22T16:46:40+00:00

The synthesized event TIME_CONV doesn't contain the complete parameters for counters, this will lead to wrong conversion between counter cycles and timestamp. This patch extends event TIME_CONV to record flags 'cap_user_time_zero' which is used to indicate the counter parameters are valid or not, if not will directly return 0 for timestamp calculation. And record the flag 'cap_user_time_short' and its relevant fields 'time_cycles' and 'time_mask' for cycle calibration. Signed-off-by: Leo Yan Cc: Adrian Hunter Cc: Alexander Shishkin Cc: Andi Kleen Cc: Gustavo A. R. Silva Cc: Ian Rogers Cc: Jiri Olsa Cc: John Garry Cc: Kemeng Shi Cc: Mark Rutland Cc: Mathieu Poirier Cc: Namhyung Kim Cc: Nick Gasson Cc: Peter Zijlstra Cc: Remi Bernon Cc: Stephane Eranian Cc: Steve Maclean Cc: Will Deacon Cc: Zou Wei Cc: linux-arm-kernel@lists.infradead.org Link: http://lore.kernel.org/lkml/20200914115311.2201-5-leo.yan@linaro.org Signed-off-by: Arnaldo Carvalho de Melo

perf inject jit: Remove //anon mmap events

2020-07-08T16:51:12+00:00

**perf-.map and jit-.dump designs: When a JIT generates code to be executed, it must allocate memory and mark it executable using an mmap call. *** perf-.map design The perf-.map assumes that any sample recorded in an anonymous memory page is JIT code. It then tries to resolve the symbol name by looking at the process' perf-.map. *** jit-.dump design The jit-.dump mechanism takes a different approach. It requires a JIT to write a `/jit-.dump` file. This file must also be mmapped so that perf inject -jit can find the file. The JIT must also add JIT_CODE_LOAD records for any functions it generates. The records are timestamped using a clock which can be correlated to the perf record clock. After perf record, the `perf inject -jit` pass parses the recording looking for a `/jit-.dump` file. When it finds the file, it parses it and for each JIT_CODE_LOAD record: * creates an elf file `/jitted--.so * injects a new mmap record mapping the new elf file into the process. *** Coexistence design The kernel and perf support both of these mechanisms. We need to make sure perf works on an app supporting either or both of these mechanisms. Both designs rely on mmap records to determine how to resolve an ip address. The mmap records of both techniques by definition overlap. When the JIT compiles a method, it must: * allocate memory (mmap) * add execution privilege (mprotect or mmap. either will generate an mmap event form the kernel to perf) * compile code into memory * add a function record to perf-.map and/or jit-.dump Because the jit-.dump mechanism supports greater capabilities, perf prefers the symbols from jit-.dump. It implements this based on timestamp ordering of events. There is an implicit ASSUMPTION that the JIT_CODE_LOAD record timestamp will be after the // anon mmap event that was generated during memory allocation or adding the execution privilege setting. *** Problems with the ASSUMPTION The ASSUMPTION made in the Coexistence design section above is violated in the following scenario. *** Scenario While a JIT is jitting code it will eventually need to commit more pages and change these pages to executable permissions. Typically the JIT will want these collocated to minimize branch displacements. The kernel will coalesce these anonymous mapping with identical permissions before sending an MMAP event for the new pages. The address range of the new mmap will not be just the most recently mmap pages. It will include the entire coalesced mmap region. See mm/mmap.c unsigned long mmap_region(struct file *file, unsigned long addr, unsigned long len, vm_flags_t vm_flags, unsigned long pgoff, struct list_head *uf) { ... /* * Can we just expand an old mapping? */ ... perf_event_mmap(vma); ... } *** Symptoms The coalesced // anon mmap event will be timestamped after the JIT_CODE_LOAD records. This means it will be used as the most recent mapping for that entire address range. For remaining events it will look at the inferior perf-.map for symbols. If both mechanisms are supported, the symbol will appear twice with different module names. This causes weird behavior in reporting. If only jit-.dump is supported, the symbol will no longer be resolved. ** Implemented solution This patch solves the issue by removing // anon mmap events for any process which has a valid jit-.dump file. It tracks on a per process basis to handle the case where some running apps support jit-.dump, but some only support perf-.map. It adds new assumptions: * // anon mmap events are only required for perf-.map support. * An app that uses jit-.dump, no longer needs perf-.map support. It assumes that any perf-.map info is inferior. *** Details Use thread->priv to store whether a jitdump file has been processed During "perf inject --jit", discard "//anon*" mmap events for any pid which has sucessfully processed a jitdump file. ** Testing: // jitdump case perf record perf inject --jit --input perf.data --output perfjit.data // verify mmap "//anon" events present initially perf script --input perf.data --show-mmap-events | grep '//anon' // verify mmap "//anon" events removed perf script --input perfjit.data --show-mmap-events | grep '//anon' // no jitdump case perf record perf inject --jit --input perf.data --output perfjit.data // verify mmap "//anon" events present initially perf script --input perf.data --show-mmap-events | grep '//anon' // verify mmap "//anon" events not removed perf script --input perfjit.data --show-mmap-events | grep '//anon' ** Repro: This issue was discovered while testing the initial CoreCLR jitdump implementation. https://github.com/dotnet/coreclr/pull/26897. ** Alternate solutions considered These were also briefly considered: * Change kernel to not coalesce mmap regions. * Change kernel reporting of coalesced mmap regions to perf. Only include newly mapped memory. * Only strip parts of // anon mmap events overlapping existing jitted--.so mmap events. Signed-off-by: Steve MacLean Acked-by: Ian Rogers Cc: Alexander Shishkin Cc: Jiri Olsa Cc: Mark Rutland Cc: Namhyung Kim Cc: Peter Zijlstra Cc: Stephane Eranian Link: http://lore.kernel.org/lkml/1590544271-125795-1-git-send-email-steve.maclean@linux.microsoft.com Signed-off-by: Arnaldo Carvalho de Melo

perf tools: Replace zero-length array with flexible-array

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

The current codebase makes use of the zero-length array language extension to the C90 standard, but the preferred mechanism to declare variable-length types such as these ones is a flexible array member[1][2], introduced in C99: struct foo { int stuff; struct boo array[]; }; By making use of the mechanism above, we will get a compiler warning in case the flexible array does not occur last in the structure, which will help us prevent some kind of undefined behavior bugs from being inadvertently introduced[3] to the codebase from now on. Also, notice that, dynamic memory allocations won't be affected by this change: "Flexible array members have incomplete type, and so the sizeof operator may not be applied. As a quirk of the original implementation of zero-length arrays, sizeof evaluates to zero."[1] sizeof(flexible-array-member) triggers a warning because flexible array members have incomplete type[1]. There are some instances of code in which the sizeof operator is being incorrectly/erroneously applied to zero-length arrays and the result is zero. Such instances may be hiding some bugs. So, this work (flexible-array member conversions) will also help to get completely rid of those sorts of issues. This issue was found with the help of Coccinelle. [1] https://gcc.gnu.org/onlinedocs/gcc/Zero-Length.html [2] https://github.com/KSPP/linux/issues/21 [3] commit 76497732932f ("cxgb3/l2t: Fix undefined behaviour") Signed-off-by: Gustavo A. R. Silva Cc: Adrian Hunter Cc: Alexander Shishkin Cc: Gustavo A. R. Silva Cc: Ian Rogers Cc: Jiri Olsa Cc: Mark Rutland Cc: Namhyung Kim Cc: Peter Zijlstra Link: http://lore.kernel.org/lkml/20200515172926.GA31976@embeddedor Signed-off-by: Arnaldo Carvalho de Melo

perf inject jit: Fix JIT_CODE_MOVE filename

2019-09-30T20:29:49+00:00

During perf inject --jit, JIT_CODE_MOVE records were injecting MMAP records with an incorrect filename. Specifically it was missing the ".so" suffix. Further the JIT_CODE_LOAD record were silently truncating the jr->load.code_index field to 32 bits before generating the filename. Make both records emit the same filename based on the full 64 bit code_index field. Fixes: 9b07e27f88b9 ("perf inject: Add jitdump mmap injection support") Cc: stable@vger.kernel.org # v4.6+ Signed-off-by: Steve MacLean Acked-by: Jiri Olsa Cc: Alexander Shishkin Cc: Andi Kleen Cc: Brian Robbins Cc: Davidlohr Bueso Cc: Eric Saint-Etienne Cc: John Keeping Cc: John Salem Cc: Leo Yan Cc: Mark Rutland Cc: Namhyung Kim Cc: Peter Zijlstra Cc: Song Liu Cc: Stephane Eranian Cc: Tom McDonald Link: http://lore.kernel.org/lkml/BN8PR21MB1362FF8F127B31DBF4121528F7800@BN8PR21MB1362.namprd21.prod.outlook.com Signed-off-by: Arnaldo Carvalho de Melo

libperf: Add perf_evlist__first()/last() functions

2019-09-25T12:51:48+00:00

Add perf_evlist__first()/last() functions to libperf, as internal functions and rename perf's origins to evlist__first/last. Signed-off-by: Jiri Olsa Cc: Alexander Shishkin Cc: Michael Petlan Cc: Namhyung Kim Cc: Peter Zijlstra Link: http://lore.kernel.org/lkml/20190913132355.21634-29-jolsa@kernel.org Signed-off-by: Arnaldo Carvalho de Melo