[LLVMbugs] [Bug 21138] New vector shuffle lowering produces an insertf128 + blend instead of a single insertf128

bugzilla-daemon at llvm.org bugzilla-daemon at llvm.org
Sun Oct 5 04:44:22 PDT 2014


Chandler Carruth <chandlerc at gmail.com> changed:

           What    |Removed                     |Added
             Status|NEW                         |RESOLVED
                 CC|                            |chandlerc at gmail.com
         Resolution|---                         |FIXED

--- Comment #1 from Chandler Carruth <chandlerc at gmail.com> ---
Shuld be fixed in r219079. This "regressed" some test cases in the regression
test suite, but only in as much as it returned our generated code for them to
the bad generated code of the old shuffle lowering.

The fix for that is independent and I'm working on it separately. It doesn't
impact the shuffle lowering strategy. See the commit log for details.

You are receiving this mail because:
You are on the CC list for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-bugs/attachments/20141005/18ff68b9/attachment.html>

More information about the llvm-bugs mailing list