[llvm-bugs] [Bug 6246] SLP vectorization when performing scalar operations on vector elements

2017-01-08 Thread via llvm-bugs 
https://llvm.org/bugs/show_bug.cgi?id=6246

Simon Pilgrim  changed:

   What|Removed |Added

 Status|REOPENED|RESOLVED
 Resolution|--- |FIXED

--- Comment #12 from Simon Pilgrim  ---
Fixed at rL289043 - it seems to have stuck this time.

-- 
You are receiving this mail because:
You are on the CC list for the bug.
___
llvm-bugs mailing list
llvm-bugs@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs

[llvm-bugs] [Bug 6246] SLP vectorization when performing scalar operations on vector elements

2016-12-02 Thread via llvm-bugs 
https://llvm.org/bugs/show_bug.cgi?id=6246

Michael Kuperstein  changed:

   What|Removed |Added

 Status|RESOLVED|REOPENED
 Resolution|FIXED   |---

--- Comment #11 from Michael Kuperstein  ---
Reverted again in r288508, reopening.

-- 
You are receiving this mail because:
You are on the CC list for the bug.
___
llvm-bugs mailing list
llvm-bugs@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs

[llvm-bugs] [Bug 6246] SLP vectorization when performing scalar operations on vector elements

2016-12-02 Thread via llvm-bugs 
https://llvm.org/bugs/show_bug.cgi?id=6246

Alexey Bataev  changed:

   What|Removed |Added

 Status|REOPENED|RESOLVED
 Resolution|--- |FIXED

--- Comment #10 from Alexey Bataev  ---
Now, I think, it is finally fixed in r288497

-- 
You are receiving this mail because:
You are on the CC list for the bug.
___
llvm-bugs mailing list
llvm-bugs@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs

[llvm-bugs] [Bug 6246] SLP vectorization when performing scalar operations on vector elements

2016-12-01 Thread via llvm-bugs 
https://llvm.org/bugs/show_bug.cgi?id=6246

Michael Kuperstein  changed:

   What|Removed |Added

 Status|RESOLVED|REOPENED
 Resolution|FIXED   |---

--- Comment #8 from Michael Kuperstein  ---
r288412 was reverted in r288431, reopening.

-- 
You are receiving this mail because:
You are on the CC list for the bug.
___
llvm-bugs mailing list
llvm-bugs@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs

[llvm-bugs] [Bug 6246] SLP vectorization when performing scalar operations on vector elements

2016-12-01 Thread via llvm-bugs 
https://llvm.org/bugs/show_bug.cgi?id=6246

Alexey Bataev  changed:

   What|Removed |Added

 Status|REOPENED|RESOLVED
 Resolution|--- |FIXED

--- Comment #7 from Alexey Bataev  ---
Must be fixed in r288412

-- 
You are receiving this mail because:
You are on the CC list for the bug.
___
llvm-bugs mailing list
llvm-bugs@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs

[llvm-bugs] [Bug 6246] SLP vectorization when performing scalar operations on vector elements

2016-11-12 Thread via llvm-bugs 
https://llvm.org/bugs/show_bug.cgi?id=6246

Simon Pilgrim  changed:

   What|Removed |Added

 Status|RESOLVED|REOPENED
 CC||mku...@google.com,
   ||spatel+l...@rotateright.com
 Resolution|FIXED   |---

--- Comment #6 from Simon Pilgrim  ---
Reopening this. If we look at the equivalent c source then it appears that the
vectorizer isn't working for non 128-bit vectors:

#include 

__m128 _vadd128(__m128 a, __m128 b) {
 return _mm_setr_ps(
   a[0] + b[0],
   a[1] + b[1],
   a[2] + b[2],
   a[3] + b[3]);
}

__m256 _vadd256(__m256 a, __m256 b) {
 return _mm256_setr_ps(
   a[0] + b[0],
   a[1] + b[1],
   a[2] + b[2],
   a[3] + b[3],
   a[4] + b[4],
   a[5] + b[5],
   a[6] + b[6],
   a[7] + b[7]);
}

clang -S -O3 pr6246.c -march=btver2 -o - -emit-llvm

; Function Attrs: norecurse nounwind readnone ssp uwtable
define <4 x float> @_vadd128(<4 x float> %a, <4 x float> %b) local_unnamed_addr
#0 {
entry:
  %0 = fadd <4 x float> %a, %b
  ret <4 x float> %0
}

; Function Attrs: norecurse nounwind readnone ssp uwtable
define <8 x float> @_vadd256(<8 x float> %a, <8 x float> %b) local_unnamed_addr
#0 {
entry:
  %vecext = extractelement <8 x float> %a, i32 0
  %vecext1 = extractelement <8 x float> %b, i32 0
  %add = fadd float %vecext, %vecext1
  %vecext2 = extractelement <8 x float> %a, i32 1
  %vecext3 = extractelement <8 x float> %b, i32 1
  %add4 = fadd float %vecext2, %vecext3
  %vecext5 = extractelement <8 x float> %a, i32 2
  %vecext6 = extractelement <8 x float> %b, i32 2
  %add7 = fadd float %vecext5, %vecext6
  %vecext8 = extractelement <8 x float> %a, i32 3
  %vecext9 = extractelement <8 x float> %b, i32 3
  %add10 = fadd float %vecext8, %vecext9
  %vecext11 = extractelement <8 x float> %a, i32 4
  %vecext12 = extractelement <8 x float> %b, i32 4
  %add13 = fadd float %vecext11, %vecext12
  %vecext14 = extractelement <8 x float> %a, i32 5
  %vecext15 = extractelement <8 x float> %b, i32 5
  %add16 = fadd float %vecext14, %vecext15
  %vecext17 = extractelement <8 x float> %a, i32 6
  %vecext18 = extractelement <8 x float> %b, i32 6
  %add19 = fadd float %vecext17, %vecext18
  %vecext20 = extractelement <8 x float> %a, i32 7
  %vecext21 = extractelement <8 x float> %b, i32 7
  %add22 = fadd float %vecext20, %vecext21
  %vecinit.i = insertelement <8 x float> undef, float %add, i32 0
  %vecinit1.i = insertelement <8 x float> %vecinit.i, float %add4, i32 1
  %vecinit2.i = insertelement <8 x float> %vecinit1.i, float %add7, i32 2
  %vecinit3.i = insertelement <8 x float> %vecinit2.i, float %add10, i32 3
  %vecinit4.i = insertelement <8 x float> %vecinit3.i, float %add13, i32 4
  %vecinit5.i = insertelement <8 x float> %vecinit4.i, float %add16, i32 5
  %vecinit6.i = insertelement <8 x float> %vecinit5.i, float %add19, i32 6
  %vecinit7.i = insertelement <8 x float> %vecinit6.i, float %add22, i32 7
  ret <8 x float> %vecinit7.i
}

-- 
You are receiving this mail because:
You are on the CC list for the bug.
___
llvm-bugs mailing list
llvm-bugs@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs

[llvm-bugs] [Bug 6246] SLP vectorization when performing scalar operations on vector elements

2016-10-13 Thread via llvm-bugs 
https://llvm.org/bugs/show_bug.cgi?id=6246

Alexey Bataev  changed:

   What|Removed |Added

 Status|NEW |RESOLVED
 CC||a.bat...@hotmail.com
 Resolution|--- |FIXED

-- 
You are receiving this mail because:
You are on the CC list for the bug.
___
llvm-bugs mailing list
llvm-bugs@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs