https://bugs.kde.org/show_bug.cgi?id=386945
--- Comment #54 from Mark Wielaard <m...@klomp.org> ---
To make the ldbrx-unaligned testcase work we can use something like the
following:
diff --git a/memcheck/mc_main.c b/memcheck/mc_main.c
index 3ef7cb913..9a7b04c9c 100644
--- a/memcheck/mc_main.c
+++ b/memcheck/mc_main.c
@@ -1508,6 +1508,10 @@ ULong mc_LOADVn_slow ( Addr a, SizeT nBits, Bool
bigendian )
# if defined(VGA_mips64) && defined(VGABI_N32)
if (szB == VG_WORDSIZE * 2 && VG_IS_WORD_ALIGNED(a)
&& n_addrs_bad < VG_WORDSIZE * 2)
+#elif defined(VGA_ppc64be) || defined(VGA_ppc64le)
+ /* Unaligned loads of words on the same page are OK. */
+ if (szB == VG_WORDSIZE && ((a & 0xfff) <= 0x1000 - szB)
+ && n_addrs_bad < VG_WORDSIZE)
# else
if (szB == VG_WORDSIZE && VG_IS_WORD_ALIGNED(a)
&& n_addrs_bad < VG_WORDSIZE)
The question is whether the extra check for the word not crossing a page
boundary is valid.
Aaron, are there are alignment constraints for the usage of the ldbrx
instruction?
(this would also need new exp files for the memcheck/tests/partial_load_ok and
partial_load_dflt tests that check unaligned word loads produce an warning)
--
You are receiving this mail because:
You are watching all bug changes.
