Bug 516090

Summary: Regression : Linux FreeBSD and Darwin: refactor *at syscall dirfd checks
Product: [Developer tools] valgrind Reporter: Fred M <dark_footix>
Component: memcheckAssignee: Paul Floyd <pjfloyd>
Status: RESOLVED FIXED    
Severity: normal    
Priority: NOR    
Version First Reported In: 3.26.0   
Target Milestone: ---   
Platform: Other   
OS: Other   
Latest Commit: Version Fixed/Implemented In:
Sentry Crash Report:
Attachments: fix

Description Fred M 2026-02-16 13:45:41 UTC
Created attachment 189654 [details]
fix

Linux FreeBSD and Darwin: refactor *at syscall dirfd checks

https://sourceware.org/git/?p=valgrind.git;a=commitdiff;h=9f5d3d32b5bd5bc546eea74a8f75e53d4a228519;hp=bbf3cf234eacc97a652437c4f5e626c43b1bc76d


On kernel 4.9 with toolchain gcc 10 ,
there is not statx on 4.9
the fallback mecanism is in error. 


valgrind ls
==30462== Memcheck, a memory error detector
==30462== Copyright (C) 2002-2024, and GNU GPL'd, by Julian Seward et al.
==30462== Using Valgrind-3.26.0 and LibVEX; rerun with -h for copyright info
==30462== Command: ls
==30462== 
ERROR: ld.so: object '/usr/libexec/valgrind/vgpreload_core-arm-linux.so' from LD_PRELOAD cannot be preloaded (cannot stat shared object): ignored.
ERROR: ld.so: object '/usr/libexec/valgrind/vgpreload_memcheck-arm-linux.so' from LD_PRELOAD cannot be preloaded (cannot stat shared object): ignored.
ls: error while loading shared libraries: libcrypt.so.1: cannot stat shared object: Error 9
==30462== 
==30462== HEAP SUMMARY:
==30462==     in use at exit: 0 bytes in 0 blocks
==30462==   total heap usage: 0 allocs, 0 frees, 0 bytes allocated
==30462== 
==30462== All heap blocks were freed -- no leaks are possible
==30462== 
==30462== For lists of detected and suppressed errors, rerun with: -s
==30462== ERROR SUMMARY: 0 errors from 0 contexts (suppressed: 0 from 0)


Fix in attachment.
Comment 1 Paul Floyd 2026-02-16 13:59:48 UTC
I'll push that tonight.
Comment 2 Paul Floyd 2026-02-16 19:49:07 UTC
Thanks for pointing that out. arm is one of the platforms not covered by our CI.

I just did a test on RPi OS 32 bit and I got 2 test build failures, one test hang and

== 783 tests, 26 stderr failures, 4 stdout failures, 0 stderrB failures, 0 stdoutB failures, 0 post failures ==

commit bc26530dcf7d961caec3bf6c4969c21ea55144f9 (HEAD -> master, origin/master, origin/HEAD)
Author: Paul Floyd <pjfloyd@wanadoo.fr>
Date:   Mon Feb 16 19:49:02 2026 +0100

    Bug 516090 - Regression : Linux FreeBSD and Darwin: refactor *at syscall dirfd check