On Thu, Mar 10, 2022 at 3:55 PM Ananyev, Konstantin
<konstantin.anan...@intel.com> wrote:
> > Calls to rte_memcpy for 1 < n < 16 could result in unaligned
> > loads/stores, which is undefined behaviour according to the C
> > standard, and strict aliasing violations.
> >
> > The code was changed to use a packed structure that allows aliasing
> > (using the __may_alias__ attribute) to perform the load/store
> > operations. This results in code that has the same performance as the
> > original code and that is also C standards-compliant.
> >
> > Fixes: d35cc1fe6a7a ("eal/x86: revert select optimized memcpy at run-time")
> > Cc: sta...@dpdk.org
> >
> > Signed-off-by: Luc Pelletier <lucp.at.w...@gmail.com>
> Acked-by: Konstantin Ananyev <konstantin.anan...@intel.com>
> Tested-by: Konstantin Ananyev <konstantin.anan...@intel.com>Thanks, applied. > As a side note, we probably need to check other similar places in DPDK code. What would be the best way to detect those problematic places? I tried UBsan, and it did report some of the issues fixed with this patch. -- David Marchand
