This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: x86/MMX/SSE shift by immediate



On Dec 12, 2003, at 4:41 PM, Stuart Hastings wrote:


This looks like a typo that crept into several similar vector instruction patterns. Currently, MMX/SSE shifts by an immediate value ICE.

The patch below passes bootstrap and make check on Darwin/x86, no regressions.

Two testcases are offered; both will ICE the current TOT. One is small and intuitive; the other attempts to test every MMX/SSE/SSE2 shift-like instruction that accepts an immediate shift count (including shuffle instructions). It may be "too large," and need breaking up; comments solicited.

O.K. to commit to trunk ? Do you want either or both testcases ?

Drat. I forgot the ChangeLog again:


2003-12-08 Stuart Hastings <stuart@apple.com>

* config/i386/i386.md: Typos in MMX/SSE immediate shifts.

stuart hastings
Apple Computer


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]