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: [patch] Fix PR42664: ICE on compilation of polyhedron benchmarks with -mfma4


On 01/16/2010 12:06 AM, Sebastian Pop wrote:
the attached patch reverts

2009-12-02 Richard Henderson<rth@redhat.com>

         * config/i386/i386.c (ix86_fixup_binary_operands): For FMA4, force
         all operands into registers.

that is no longer needed after the following patches cleaned up
all input operands of FMA4 insns that were changed to
nonimmediate operands:

2009-12-03  Sebastian Pop<sebastian.pop@amd.com>
             Richard Henderson<rth@redhat.com>

         * config/i386/i386-protos.h (ix86_fma4_valid_op_p): Remove.
         * config/i386/i386.c (ix86_fma4_valid_op_p): Remove.
         * config/i386/i386.md: Do not use ix86_fma4_valid_op_p.
         * config/i386/sse.md (fma4_*): Remove alternative with operand 1
         matching a memory access.  Do not use ix86_fma4_valid_op_p.
         (xop_*): Same.
         Do not use ix86_fma4_valid_op_p in FMA4 and XOP splitters.

and

2009-12-07 Sebastian Pop<sebastian.pop@amd.com>

         * config/i386/sse.md: Remove all FMA4 splitters.
         Allow the second operand of FMA4 insns to be a nonimmediate.
         Fix comments punctuation.

The attached patch was tested on amd64-linux with bootstrap and
regression test.  Ok to apply to trunk?

2010-01-08 Harsha Jagasia<harsha.jagasia@amd.com>

PR target/42664
* config/i386/i386.c (ix86_fixup_binary_operands):
Revert FMA4 fixup of operands.

This is OK for mainline.


Thanks,
Uros.


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