[cfe-commits] Alignment of return from __builtin_alloca

David Sehr sehr at google.com
Thu Jun 7 15:31:06 PDT 2012

Just out of curiosity, how do those x86 platforms handle
stack-allocated SSE2 variables?  Do they just always use the unaligned
variants of the instructions?
This was the reason that the Linux x86 (32 and 64) ABIs now align
stacks 0mod16, I thought.


On Thu, Jun 7, 2012 at 3:20 PM, Eli Friedman <eli.friedman at gmail.com> wrote:
> On Thu, Jun 7, 2012 at 3:08 PM, David Sehr <sehr at google.com> wrote:
>> Please disregard my last comment.  Implementing that will be a bit
>> challenging, as stackAlign is a back end value rather than a clang
>> TargetInfo value. Are there targets where stackAlign < SuitableAlign?
> Feel free to add a StackAlign member to TargetInfo if you need it...
> it's already part of the DescriptionString (specified with "S128"
> etc.).
> And yes, there are targets where StackAlign < SuitableAlign... many
> x86-32 targets have SuitableAlign == 128 and StackAlign == 32.
> -Eli

More information about the cfe-commits mailing list