m_zuckerman (michael zuckerman)
User

Projects

User does not belong to any projects.

User Details

User Since
Jun 17 2015, 1:09 AM (109 w, 1 d)

Recent Activity

Yesterday

m_zuckerman added a comment to D34601: [X86][LLVM]Expanding Supports lowerInterleavedStore() in X86InterleavedAccess..

ping

Wed, Jul 19, 1:23 AM

Mon, Jul 17

m_zuckerman added inline comments to D35354: [X86][LLVM][TD] Arranging Atom family in an inheritance ratio.
Mon, Jul 17, 5:07 AM

Sun, Jul 16

m_zuckerman added a comment to D34601: [X86][LLVM]Expanding Supports lowerInterleavedStore() in X86InterleavedAccess..

Hi Michael, I have one small additional comment. Otherwise, this looks good.

Sun, Jul 16, 1:56 AM

Thu, Jul 13

m_zuckerman created D35354: [X86][LLVM][TD] Arranging Atom family in an inheritance ratio.
Thu, Jul 13, 6:48 AM
m_zuckerman updated the diff for D34601: [X86][LLVM]Expanding Supports lowerInterleavedStore() in X86InterleavedAccess..
Thu, Jul 13, 4:33 AM

Wed, Jul 12

m_zuckerman added inline comments to D34601: [X86][LLVM]Expanding Supports lowerInterleavedStore() in X86InterleavedAccess..
Wed, Jul 12, 9:50 AM
m_zuckerman added inline comments to D34601: [X86][LLVM]Expanding Supports lowerInterleavedStore() in X86InterleavedAccess..
Wed, Jul 12, 2:34 AM
m_zuckerman updated the diff for D34601: [X86][LLVM]Expanding Supports lowerInterleavedStore() in X86InterleavedAccess..
Wed, Jul 12, 2:33 AM

Wed, Jul 5

m_zuckerman added a comment to D34601: [X86][LLVM]Expanding Supports lowerInterleavedStore() in X86InterleavedAccess..
In D34601#799242, @zvi wrote:

Michael, can you please add AVX512 targets to the tests so we ensure that the AVX512 targets are at least as good as the AVX2 targets?

Wed, Jul 5, 7:53 AM

Mon, Jul 3

m_zuckerman added a comment to D34601: [X86][LLVM]Expanding Supports lowerInterleavedStore() in X86InterleavedAccess..

In the new patch I added a test checks the AVX sequence and new if verifies that VF32 is with store instructions.

Mon, Jul 3, 7:46 AM

Sat, Jul 1

m_zuckerman updated the diff for D34601: [X86][LLVM]Expanding Supports lowerInterleavedStore() in X86InterleavedAccess..
Sat, Jul 1, 5:11 AM

Thu, Jun 29

m_zuckerman updated the diff for D34807: [Clang][X86][Goldmont] Adding new target-cpu: Goldmont.
Thu, Jun 29, 6:05 AM
m_zuckerman abandoned D34505: [Clang][X86][Goldmont] Adding new target-cpu: Goldmont .
Thu, Jun 29, 4:51 AM
m_zuckerman updated the diff for D34807: [Clang][X86][Goldmont] Adding new target-cpu: Goldmont.

Remove extra feature -fsgsbase

Thu, Jun 29, 3:07 AM
m_zuckerman created D34807: [Clang][X86][Goldmont] Adding new target-cpu: Goldmont.
Thu, Jun 29, 1:03 AM

Wed, Jun 28

m_zuckerman added a comment to D34601: [X86][LLVM]Expanding Supports lowerInterleavedStore() in X86InterleavedAccess..

Have you tried enabling this also for AVX? (I understand if not, because with the current cost numbers that TTI returns for interleaved accesses on AVX we'll probably determine it's not worth vectorizing... so that may need to come along with an update of getInterleavedMemoryOpCost -- maybe at least a TODO comment is needed).

Wed, Jun 28, 8:38 AM
m_zuckerman updated the diff for D34601: [X86][LLVM]Expanding Supports lowerInterleavedStore() in X86InterleavedAccess..
Wed, Jun 28, 8:30 AM

Tue, Jun 27

m_zuckerman updated the diff for D34504: [LLVM][X86][Goldmont] Adding new target-cpu: Goldmont.
Tue, Jun 27, 4:45 AM
m_zuckerman added a comment to D34504: [LLVM][X86][Goldmont] Adding new target-cpu: Goldmont.

I removed from the list of features two features FeatureSlowDivide64 and FeatureSlowPMULLD. According to the optimization manual ,that Igor added ,this issues were fixed in the GoldMont processor. Regard @RKSimon I agree with you that we want to remove the isglm() until use. But I prefer no touch the IntelGLM since I want to specialize Goldmont from the other ATOM processors .

Tue, Jun 27, 3:32 AM
m_zuckerman updated the diff for D34504: [LLVM][X86][Goldmont] Adding new target-cpu: Goldmont.
Tue, Jun 27, 3:22 AM
m_zuckerman added a comment to D34601: [X86][LLVM]Expanding Supports lowerInterleavedStore() in X86InterleavedAccess..

ping

Tue, Jun 27, 1:49 AM
m_zuckerman updated subscribers of D34504: [LLVM][X86][Goldmont] Adding new target-cpu: Goldmont.
Tue, Jun 27, 1:19 AM

Mon, Jun 26

m_zuckerman added a comment to D34601: [X86][LLVM]Expanding Supports lowerInterleavedStore() in X86InterleavedAccess..

Would it be beneficial to work on a more general solution for the 128-bit subvector issue? Won't 16x16 and 8x32 (as well as all the 512-bit equivalents) still suffer?

Mon, Jun 26, 8:37 AM
m_zuckerman updated the diff for D34601: [X86][LLVM]Expanding Supports lowerInterleavedStore() in X86InterleavedAccess..
Mon, Jun 26, 7:59 AM
m_zuckerman updated the diff for D34504: [LLVM][X86][Goldmont] Adding new target-cpu: Goldmont.
Mon, Jun 26, 1:52 AM

Sun, Jun 25

m_zuckerman added a comment to D34504: [LLVM][X86][Goldmont] Adding new target-cpu: Goldmont.

removing SMAP feature.

Sun, Jun 25, 9:27 AM
m_zuckerman updated the diff for D34504: [LLVM][X86][Goldmont] Adding new target-cpu: Goldmont.
Sun, Jun 25, 9:26 AM
m_zuckerman updated the diff for D34504: [LLVM][X86][Goldmont] Adding new target-cpu: Goldmont.
Sun, Jun 25, 9:12 AM
m_zuckerman abandoned D34603: [X86][LLVM]Adding SMAP feature and test.
Sun, Jun 25, 8:37 AM
m_zuckerman reclaimed D34603: [X86][LLVM]Adding SMAP feature and test.
Sun, Jun 25, 8:37 AM
m_zuckerman abandoned D34603: [X86][LLVM]Adding SMAP feature and test.
Sun, Jun 25, 8:37 AM
m_zuckerman added a comment to D34504: [LLVM][X86][Goldmont] Adding new target-cpu: Goldmont.

Don't add the SMAP feature flag. We should only have feature flags that are used by isel.

Sun, Jun 25, 6:44 AM
m_zuckerman added a dependency for D34504: [LLVM][X86][Goldmont] Adding new target-cpu: Goldmont: D34603: [X86][LLVM]Adding SMAP feature and test.
Sun, Jun 25, 6:33 AM
m_zuckerman added a dependent revision for D34603: [X86][LLVM]Adding SMAP feature and test: D34504: [LLVM][X86][Goldmont] Adding new target-cpu: Goldmont.
Sun, Jun 25, 6:33 AM
m_zuckerman created D34603: [X86][LLVM]Adding SMAP feature and test.
Sun, Jun 25, 6:28 AM
m_zuckerman added a reviewer for D34601: [X86][LLVM]Expanding Supports lowerInterleavedStore() in X86InterleavedAccess.: DavidKreitzer.
Sun, Jun 25, 1:54 AM
m_zuckerman added a dependent revision for D32658: Supports lowerInterleavedStore() in X86InterleavedAccess.: D34601: [X86][LLVM]Expanding Supports lowerInterleavedStore() in X86InterleavedAccess..
Sun, Jun 25, 12:39 AM
m_zuckerman added a dependency for D34601: [X86][LLVM]Expanding Supports lowerInterleavedStore() in X86InterleavedAccess.: D32658: Supports lowerInterleavedStore() in X86InterleavedAccess..
Sun, Jun 25, 12:39 AM
m_zuckerman created D34601: [X86][LLVM]Expanding Supports lowerInterleavedStore() in X86InterleavedAccess..
Sun, Jun 25, 12:38 AM

Thu, Jun 22

m_zuckerman added a dependency for D34505: [Clang][X86][Goldmont] Adding new target-cpu: Goldmont : D34504: [LLVM][X86][Goldmont] Adding new target-cpu: Goldmont.
Thu, Jun 22, 3:34 AM
m_zuckerman added a dependent revision for D34504: [LLVM][X86][Goldmont] Adding new target-cpu: Goldmont: D34505: [Clang][X86][Goldmont] Adding new target-cpu: Goldmont .
Thu, Jun 22, 3:34 AM
m_zuckerman created D34505: [Clang][X86][Goldmont] Adding new target-cpu: Goldmont .
Thu, Jun 22, 3:27 AM
m_zuckerman created D34504: [LLVM][X86][Goldmont] Adding new target-cpu: Goldmont.
Thu, Jun 22, 3:20 AM

May 10 2017

m_zuckerman added a comment to D32792: [LLVM][inline-asm] Altmacro string escape character '!'.

Thanks a lot :)

May 10 2017, 5:29 AM
m_zuckerman added a comment to D32792: [LLVM][inline-asm] Altmacro string escape character '!'.

Ping :)

May 10 2017, 2:56 AM

May 7 2017

m_zuckerman added inline comments to D32792: [LLVM][inline-asm] Altmacro string escape character '!'.
May 7 2017, 1:22 AM
m_zuckerman updated the diff for D32792: [LLVM][inline-asm] Altmacro string escape character '!'.
May 7 2017, 1:17 AM

May 4 2017

m_zuckerman added a comment to D32792: [LLVM][inline-asm] Altmacro string escape character '!'.

Ping

May 4 2017, 11:52 AM
m_zuckerman updated the diff for D32701: [LLVM][inline-asm][Altmacor] Altmacro string delimiter '<..>'.

As discussed, I have returned to the <...> implementation without the single quote and I added a negative test.

May 4 2017, 2:02 AM

May 3 2017

m_zuckerman added a comment to D32701: [LLVM][inline-asm][Altmacor] Altmacro string delimiter '<..>'.

No worry :) we love to play with the code

May 3 2017, 7:26 AM
m_zuckerman added a comment to D32701: [LLVM][inline-asm][Altmacor] Altmacro string delimiter '<..>'.

The behaviors are very strange when working with single quotes, so I prefer only to do partial support (strings) or not at all.

May 3 2017, 7:25 AM
m_zuckerman updated the diff for D32701: [LLVM][inline-asm][Altmacor] Altmacro string delimiter '<..>'.
May 3 2017, 7:19 AM
m_zuckerman added inline comments to D32701: [LLVM][inline-asm][Altmacor] Altmacro string delimiter '<..>'.
May 3 2017, 7:07 AM
m_zuckerman created D32792: [LLVM][inline-asm] Altmacro string escape character '!'.
May 3 2017, 4:05 AM
m_zuckerman added a dependent revision for D32701: [LLVM][inline-asm][Altmacor] Altmacro string delimiter '<..>': D32792: [LLVM][inline-asm] Altmacro string escape character '!'.
May 3 2017, 4:05 AM

May 2 2017

m_zuckerman updated the diff for D32701: [LLVM][inline-asm][Altmacor] Altmacro string delimiter '<..>'.
May 2 2017, 11:40 PM

May 1 2017

m_zuckerman updated the diff for D32701: [LLVM][inline-asm][Altmacor] Altmacro string delimiter '<..>'.
May 1 2017, 11:45 PM
m_zuckerman updated the summary of D32701: [LLVM][inline-asm][Altmacor] Altmacro string delimiter '<..>'.
May 1 2017, 8:53 AM
m_zuckerman added a dependent revision for D32526: [LLVM][inline-asm] Altmacro absolute expression '%' feature: D32701: [LLVM][inline-asm][Altmacor] Altmacro string delimiter '<..>'.
May 1 2017, 8:51 AM
m_zuckerman added a dependency for D32701: [LLVM][inline-asm][Altmacor] Altmacro string delimiter '<..>': D32526: [LLVM][inline-asm] Altmacro absolute expression '%' feature.
May 1 2017, 8:51 AM
m_zuckerman retitled D32701: [LLVM][inline-asm][Altmacor] Altmacro string delimiter '<..>' from [LLVM][inline-asm] Altmacro string delimiter '<..>' to [LLVM][inline-asm][Altmacor] Altmacro string delimiter '<..>'.
May 1 2017, 8:49 AM
m_zuckerman updated the summary of D32701: [LLVM][inline-asm][Altmacor] Altmacro string delimiter '<..>'.
May 1 2017, 8:40 AM
m_zuckerman created D32701: [LLVM][inline-asm][Altmacor] Altmacro string delimiter '<..>'.
May 1 2017, 8:38 AM
m_zuckerman abandoned D32523: [LLVM][Inline-ASM] Adding Altmacro directive to LLVM .

Was combine with the D32523

May 1 2017, 7:35 AM
m_zuckerman added a comment to D32526: [LLVM][inline-asm] Altmacro absolute expression '%' feature.

If it's ok with you and you think the review is ready, please mark LGTM

May 1 2017, 3:54 AM
m_zuckerman updated the diff for D32526: [LLVM][inline-asm] Altmacro absolute expression '%' feature.

merge two reviews to one D32526 and D32523

May 1 2017, 3:28 AM
m_zuckerman added a comment to D32526: [LLVM][inline-asm] Altmacro absolute expression '%' feature.

Can you merge D32523 with this one?

We don't usually add functionality without purpose/tests, and this is the functionality (and tests) D32523 needs. :)

Otherwise, looks good. Are you going to implement the remaining functionality?

May 1 2017, 3:15 AM
m_zuckerman added a comment to D32526: [LLVM][inline-asm] Altmacro absolute expression '%' feature.

Ping

May 1 2017, 1:37 AM

Apr 27 2017

m_zuckerman updated the diff for D32526: [LLVM][inline-asm] Altmacro absolute expression '%' feature.
Apr 27 2017, 4:34 AM
m_zuckerman updated the diff for D32526: [LLVM][inline-asm] Altmacro absolute expression '%' feature.
Apr 27 2017, 4:17 AM

Apr 26 2017

m_zuckerman updated the diff for D32526: [LLVM][inline-asm] Altmacro absolute expression '%' feature.
Apr 26 2017, 7:42 AM
m_zuckerman updated the diff for D32526: [LLVM][inline-asm] Altmacro absolute expression '%' feature.
Apr 26 2017, 3:57 AM
m_zuckerman added reviewers for D32526: [LLVM][inline-asm] Altmacro absolute expression '%' feature: grosbach, rengolin, rnk.
Apr 26 2017, 3:40 AM
m_zuckerman updated the summary of D32526: [LLVM][inline-asm] Altmacro absolute expression '%' feature.
Apr 26 2017, 2:44 AM
m_zuckerman updated subscribers of D32523: [LLVM][Inline-ASM] Adding Altmacro directive to LLVM .
Apr 26 2017, 2:43 AM
m_zuckerman updated subscribers of D32526: [LLVM][inline-asm] Altmacro absolute expression '%' feature.
Apr 26 2017, 2:42 AM
m_zuckerman added a dependency for D32526: [LLVM][inline-asm] Altmacro absolute expression '%' feature: D32523: [LLVM][Inline-ASM] Adding Altmacro directive to LLVM .
Apr 26 2017, 2:42 AM
m_zuckerman added a dependent revision for D32523: [LLVM][Inline-ASM] Adding Altmacro directive to LLVM : D32526: [LLVM][inline-asm] Altmacro absolute expression '%' feature.
Apr 26 2017, 2:42 AM
m_zuckerman created D32526: [LLVM][inline-asm] Altmacro absolute expression '%' feature.
Apr 26 2017, 2:40 AM
m_zuckerman added reviewers for D32523: [LLVM][Inline-ASM] Adding Altmacro directive to LLVM : grosbach, rnk, rengolin.
Apr 26 2017, 1:33 AM
m_zuckerman created D32523: [LLVM][Inline-ASM] Adding Altmacro directive to LLVM .
Apr 26 2017, 12:23 AM

Apr 25 2017

m_zuckerman abandoned D12399: Microsoft compatibility – add support for “relaxation” of memory operands in inline assembly. .
Apr 25 2017, 11:55 PM
m_zuckerman abandoned D31391: [X86][Clang] Converting __mm{|256|512}_movm_epi{8|16|32|64} LLVMIR call into generic intrinsics..
Apr 25 2017, 11:54 PM

Apr 15 2017

m_zuckerman added a comment to D31398: [X86][X86 intrinsics]Folding cmp(sub(a,b),0) into cmp(a,b) optimization.

Thanks

Apr 15 2017, 12:26 PM
m_zuckerman updated the diff for D31398: [X86][X86 intrinsics]Folding cmp(sub(a,b),0) into cmp(a,b) optimization.

I agree with you that your code is more elegant.

Apr 15 2017, 6:05 AM

Apr 13 2017

m_zuckerman updated the diff for D31398: [X86][X86 intrinsics]Folding cmp(sub(a,b),0) into cmp(a,b) optimization.

Hi,
I did a small modification on the code since I am going to retire from the canonical compare representation review.
I added to the code the ability to fold cmp(0, fsub(a,b)) additnal to the reviewrd cmp(fsub(a,b),0)

Apr 13 2017, 10:25 AM

Apr 9 2017

m_zuckerman added a comment to D31398: [X86][X86 intrinsics]Folding cmp(sub(a,b),0) into cmp(a,b) optimization.

Ping

Apr 9 2017, 9:00 AM
m_zuckerman added a comment to D31396: [X86][LLVM][Canonical Compare Intrinsics] Creating a canonical representation for X86 CMP intrinsics.

Ping

Apr 9 2017, 8:59 AM

Apr 5 2017

m_zuckerman updated the diff for D31398: [X86][X86 intrinsics]Folding cmp(sub(a,b),0) into cmp(a,b) optimization.
Apr 5 2017, 4:38 AM
m_zuckerman updated the diff for D31396: [X86][LLVM][Canonical Compare Intrinsics] Creating a canonical representation for X86 CMP intrinsics.
Apr 5 2017, 4:37 AM
m_zuckerman updated the diff for D31396: [X86][LLVM][Canonical Compare Intrinsics] Creating a canonical representation for X86 CMP intrinsics.
Apr 5 2017, 4:35 AM
m_zuckerman updated the diff for D31396: [X86][LLVM][Canonical Compare Intrinsics] Creating a canonical representation for X86 CMP intrinsics.
Apr 5 2017, 3:10 AM

Apr 4 2017

m_zuckerman added an edge to rL299431: [X86][Clang] Converting __mm{|256|512}_movm_epi{8|16|32|64} LLVMIR call into…: D31394: [X86][Clang] Converting __mm{|256|512}_movm_epi{8|16|32|64} LLVMIR call into generic intrinsics..
Apr 4 2017, 6:48 AM
m_zuckerman added 1 commit(s) for D31394: [X86][Clang] Converting __mm{|256|512}_movm_epi{8|16|32|64} LLVMIR call into generic intrinsics.: rL299431: [X86][Clang] Converting __mm{|256|512}_movm_epi{8|16|32|64} LLVMIR call into….
Apr 4 2017, 6:48 AM
m_zuckerman closed D31394: [X86][Clang] Converting __mm{|256|512}_movm_epi{8|16|32|64} LLVMIR call into generic intrinsics..
Apr 4 2017, 6:46 AM

Apr 3 2017

m_zuckerman added a comment to D31394: [X86][Clang] Converting __mm{|256|512}_movm_epi{8|16|32|64} LLVMIR call into generic intrinsics..

Thanks :)

Apr 3 2017, 2:12 PM
m_zuckerman updated the diff for D31394: [X86][Clang] Converting __mm{|256|512}_movm_epi{8|16|32|64} LLVMIR call into generic intrinsics..
Apr 3 2017, 2:04 PM
m_zuckerman added inline comments to D31394: [X86][Clang] Converting __mm{|256|512}_movm_epi{8|16|32|64} LLVMIR call into generic intrinsics..
Apr 3 2017, 12:29 PM
m_zuckerman updated the diff for D31394: [X86][Clang] Converting __mm{|256|512}_movm_epi{8|16|32|64} LLVMIR call into generic intrinsics..
Apr 3 2017, 12:29 PM
m_zuckerman added a comment to D31394: [X86][Clang] Converting __mm{|256|512}_movm_epi{8|16|32|64} LLVMIR call into generic intrinsics..

Ping

Apr 3 2017, 9:06 AM