https://bugs.kde.org/show_bug.cgi?id=385409
--- Comment #56 from Andreas Arnez <ar...@linux.ibm.com> --- (In reply to Julian Seward from comment #54) > [...] > Memcheck's "partial-loads-OK" heuristic, which is now enabled by default, > should remove these complaints. So the question is, why isn't it? See > mc_LOADV_128_or_256_slow in mc_main.c. This heuristic doesn't currently apply because VLL is implemented with up to 16 single-byte loads. I'll attach a patch that performs a single 16-byte load instead. Unless the VLL crosses a 16-byte boundary, that load will aligned as well. -- You are receiving this mail because: You are watching all bug changes.