[PATCH][X86] Improve buildFromShuffleMostly for AVX

[Adam Nemet]

Quentin looked at this and suggested to be slightly less conservative in detecting the lowered idiom.

In this new version, the idiom is also matched if the extract_vector_elt accesses higher elements beyond the first.  I also fixed some formatting issues.

OK to commit?

Adam

On Apr 16, 2014, at 1:03 PM, Adam Nemet <anemet at apple.com> wrote:

> Ping
> 
> On Apr 7, 2014, at 4:09 PM, Adam Nemet <anemet at apple.com> wrote:
> 
>> For a 256-bit BUILD_VECTOR consisting mostly of shuffles of 256-bit vectors,
>> both the BUILD_VECTOR and its operands may need to be legalized in multiple
>> steps.  Consider:
>> 
>> (v8f32 (BUILD_VECTOR (extract_vector_elt (v8f32 %vreg0,) Constant<1>),
>>                    (extract_vector_elt %vreg0, Constant<2>),
>>                    (extract_vector_elt %vreg0, Constant<3>),
>>                    (extract_vector_elt %vreg0, Constant<4>),
>>                    (extract_vector_elt %vreg0, Constant<5>),
>>                    (extract_vector_elt %vreg0, Constant<6>),
>>                    (extract_vector_elt %vreg0, Constant<7>),
>>                    %vreg1))
>> 
>> a. We can't build a 256-bit vector efficiently so, we need to split it into
>> two 128-bit vecs and combine them with VINSERTX128.
>> 
>> b. Operands like (extract_vector_elt (v8f32 %vreg0), Constant<7>) needs to
>> be split into a VEXTRACTX128 and a further extract_vector_elt from the
>> resulting 128-bit vector.
>> 
>> c. The extract_vector_elt from b. is lowered into a shuffle to the first
>> element and a movss.
>> 
>> Depending on the order in which we legalize the BUILD_VECTOR and its
>> operands[1], buildFromShuffleMostly may be faced with:
>> 
>> (v4f32 (BUILD_VECTOR (extract_vector_elt
>>                     (vector_shuffle<1,u,u,u> (extract_subvector %vreg0, Constant<4>), undef),
>>                     Constant<0>),
>>                    (extract_vector_elt
>>                     (vector_shuffle<2,u,u,u> (extract_subvector %vreg0, Constant<4>), undef),
>>                     Constant<0>),
>>                    (extract_vector_elt
>>                     (vector_shuffle<3,u,u,u> (extract_subvector %vreg0, Constant<4>), undef),
>>                     Constant<0>),
>>                    %vreg1))
>> 
>> In order to figure out the underlying vector and their identity we need to see
>> through the shuffles.
>> 
>> [1] Note that the order in which operations and their operands are legalized is
>> only guaranteed in the first iteration of LegalizeDAG.
>> 
>> Fixes <rdar://problem/16296956>
>> 
>> Adam
>> 
> 
>> _______________________________________________
>> llvm-commits mailing list
>> llvm-commits at cs.uiuc.edu
>> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits
> 
> _______________________________________________
> llvm-commits mailing list
> llvm-commits at cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits
-------------- next part --------------
A non-text attachment was scrubbed...
Name: X86-Improve-buildFromShuffleMostly-for-AVX.patch
Type: application/octet-stream
Size: 3986 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20140418/3faeced8/attachment.obj>