https://bugs.kde.org/show_bug.cgi?id=444545

--- Comment #8 from Peter Hull <[email protected]> ---
I did a little more checking. I reduced the test program somewhat (attached)
I tried putting a breakpoint on memchr both standalone and running under
valgrind, to see what got called. I saw for 'standalone'

(gdb) run
Starting program: /home/peter/Projects/vgtest/a.out 
...
Breakpoint 1, memchr_ifunc () at ../sysdeps/x86_64/multiarch/memchr.c:29
29      libc_ifunc_redirected (__redirect_memchr, memchr, IFUNC_SELECTOR ());
(gdb) step
__memchr_sse2 () at ../sysdeps/x86_64/multiarch/../memchr.S:34

and under Valgrind

$ valgrind --tool=massif --vgdb-stop-at=startup -v ./a.out
(gdb) target remote | /usr/libexec/valgrind/../../bin/vgdb --pid=1734
...
(gdb) br memchr
Breakpoint 1 at 0x401040 (2 locations)
(gdb) cont
Continuing.

Breakpoint 1, memchr_ifunc () at ../sysdeps/x86_64/multiarch/memchr.c:29
29      libc_ifunc_redirected (__redirect_memchr, memchr, IFUNC_SELECTOR ());
(gdb) step
__memchr_avx2 () at ../sysdeps/x86_64/multiarch/memchr-avx2.S:60
60      ENTRY (MEMCHR)

Note that in the first case we're calling into __memchr_sse2 and for the second
it is __memchr_avx2.

Could this be significant?

-- 
You are receiving this mail because:
You are watching all bug changes.

Reply via email to