https://bugs.kde.org/show_bug.cgi?id=474332
Bug ID: 474332
Summary: aligned_alloc under Valgrind returns nullptr when
alignment < sizeof(void *)
Classification: Developer tools
Product: valgrind
Version: unspecified
Platform: Archlinux
OS: Linux
Status: REPORTED
Severity: normal
Priority: NOR
Component: memcheck
Assignee: jsew...@acm.org
Reporter: smj...@gmail.com
Target Milestone: ---
STEPS TO REPRODUCE
1. Compile and run the following C snippet:
#include <stdlib.h>
#include <stdio.h>
int main() {
char *p = aligned_alloc(4, 4);
if(p == NULL) {
printf("Allocation failed!\n");
return 1;
}
printf("Allocation succeeded!\n");
return 0;
}
2. Then run it again under Valgrind
OBSERVED RESULT
When running on its own the allocation succeeds, when running under Valgrind it
fails
EXPECTED RESULT
Both cases succeed.
ADDITIONAL INFORMATION
Looking at the source code the problem seems to be highlighted in the comment
here (and then the subsequent check);
https://sourceware.org/git/?p=valgrind.git;a=blob;f=coregrind/m_replacemalloc/vg_replace_malloc.c;hb=23250889de4e2079ad1ede6874cc824bc9dd92db#l2179
Valgrind expects the alignment to be a multiple of sizeof(void *), but this is
not true in glibc
https://sourceware.org/git/?p=glibc.git;a=blob;f=malloc/malloc.c;hb=a43003ebf674f7af8c4b8d6d1b682244f1a28719#l3548,
which only checks that it's a power of two (and then internally adjusts if it's
smaller than some constant, but it doesn't make it fail).
That limit seems to be for posix_memalign, but as far as I know the C standard
says that a valid alignment is implementation dependent, and I would expect for
Valgrind on Linux to be the same as for glibc.
--
You are receiving this mail because:
You are watching all bug changes.
