Commit Graph

16 Commits

Author SHA1 Message Date
Miss Islington (bot) 53e7f2400a [3.15] gh-150723: Fix perf jitdump files on macOS (GH-150728) (#150832)
gh-150723: Fix perf jitdump files on macOS (GH-150728)

The perf jitdump format defines the thread id field of the JR_CODE_LOAD
record as a 32-bit value, but on macOS it was declared as a uint64_t
(since pthread_threadid_np() returns a uint64_t). Those extra 8 bytes
plus alignment padding shifted every following field, so parsers reading
the file by the spec misread code_size as the code address and failed to
resolve any Python frames.

Declare thread_id as uint32_t on all platforms and truncate the macOS
thread id when writing the record. The value is only informational.
Symbols are resolved by address, and not thread ids so truncation is
safe here.

* Use mach_absolute_time for macOS jitdump timestamps

On macOS the jitdump file is consumed by profilers such as samply, which
timestamp their samples using mach_absolute_time(). The jitdump events were
stamped with clock_gettime(CLOCK_MONOTONIC), a different clock domain that
keeps advancing while the system is asleep, so the JIT code mappings could be
off by days relative to the samples and no Python frame would resolve. Stamp
jitdump events with mach_absolute_time() on macOS so they share the sampler's
clock domain. Linux continues to use CLOCK_MONOTONIC to stay aligned with perf.

Exercise the -Xperf_jit (jitdump) backend through samply and assert that
Python frames resolve, exercising the binary jitdump path end to end.
Skipped when samply is not installed.
(cherry picked from commit 494f2e3c92)

Co-authored-by: Nazım Can Altınova <canaltinova@gmail.com>
2026-06-03 00:41:24 +00:00
Diego Russo c7b7ca2cd5 GH-126910: Add gdb support for unwinding JIT frames (#146071)
Co-authored-by: Pablo Galindo Salgado <pablogsal@gmail.com>
2026-05-02 13:42:03 +00:00
stratakis 8fe8a94a7c gh-144194: Fix mmap failure check in perf_jit_trampoline.c (#143713)
mmap() returns MAP_FAILED ((void*)-1) on error, not NULL. The current
check never detects mmap failures, so jitdump initialization proceeds
even when the memory mapping fails.
2026-01-28 13:30:17 +00:00
stratakis 77bf4ba732 gh-142779: Initialize reserved field for proper padding (#142780)
The jitdump specification specifies a reserved field for padding.

Initialize it so no garbage data is embedded in the jitdump files.
2026-01-22 18:06:36 +01:00
Donghee Na 14f0b5191a gh-142419: Add mmap.set_name method for user custom annotation (gh-142480) 2025-12-18 23:33:49 +09:00
Donghee Na c4ccaf4b10 gh-141770: Annotate anonymous mmap usage if "-X dev" is used (gh-142079) 2025-12-08 14:47:19 +00:00
Petr Viktorin e41c1ce585 gh-136459: Use platform-specific type in perf_jit_trampoline (GH-137031)
gh-136461 added perf support for macOS, with ifdefs around all changes
except increasing thread_id to 64 bits.
Make that change Apple-specific too.
2025-07-23 10:20:42 +02:00
Nazım Can Altınova a667800558 gh-136459: Add perf trampoline support for macOS (#136461) 2025-07-22 16:47:24 +01:00
Pablo Galindo Salgado 236f733d8f gh-136541: Fix several problems of perf trampolines in x86_64 and aarch64 (#136500)
This commit fixes the following problems:

* The x86_64 trampolines are not preserving frame pointers
* The hardcoded offsets to the code segment from the FDE only worked properly for x64_64
* The CIE data was not following conventions of aarch64
* The eh_frame for aarch64 was not fully correct
2025-07-11 14:32:35 +01:00
stratakis 485b499610 gh-128605: Add branch protections for x86_64 in asm_trampoline.S (#128606)
Apply Intel Control-flow Technology for x86-64 on asm_trampoline.S.

Required for mitigation against return-oriented programming (ROP)
and Call or Jump Oriented Programming (COP/JOP) attacks.

Manual application is required for the assembly files.

See also: https://sourceware.org/annobin/annobin.html/Test-cf-protection.html
2025-06-03 09:09:43 +02:00
Pablo Galindo Salgado 7b1a700231 Heavily comment Python/perf_jit_trampoline.c to improve maintainability (#134527)
Signed-off-by: Pablo Galindo <pablogsal@gmail.com>
2025-05-25 21:37:15 +01:00
Victor Stinner 978e37bb5f gh-131238: Add explicit includes to pycore headers (#131257) 2025-03-17 12:32:43 +01:00
Savannah Ostrowski 5b941e57c7 GH-118844: Fix build failures when combining --disable-gil with --enable-experimental-jit (GH-118935) 2024-05-11 16:19:31 -07:00
Pablo Galindo Salgado e6b213ee3f gh-118518: Ensure that the code padding it's applied (#118654) 2024-05-06 18:20:42 +01:00
Pablo Galindo Salgado aeb36c5bb9 gh-118518: Use the raw syscall directly for gettid (#118592) 2024-05-05 12:37:32 +00:00
Pablo Galindo Salgado 1b22d801b8 gh-118518: Allow perf to work without frame pointers (#112254) 2024-05-05 03:07:29 +02:00