Page MenuHomePhabricator

arnab-oss (Arnab Dutta)
User

Projects

User does not belong to any projects.

User Details

User Since
Aug 13 2021, 3:59 AM (67 w, 5 d)

Recent Activity

Thu, Nov 3

arnab-oss updated the diff for D137389: [MLIR] Fix bug in the method constructing semi affine expression from flattened form.

Changed variable names to camel back style

Thu, Nov 3, 11:48 PM · Restricted Project, Restricted Project
arnab-oss added reviewers for D137389: [MLIR] Fix bug in the method constructing semi affine expression from flattened form: springerm, bondhugula, hanchung.
Thu, Nov 3, 11:24 PM · Restricted Project, Restricted Project
arnab-oss requested review of D137389: [MLIR] Fix bug in the method constructing semi affine expression from flattened form.
Thu, Nov 3, 11:22 PM · Restricted Project, Restricted Project

Oct 31 2022

arnab-oss added a comment to D136794: [mlir] Use getAffineExprFromFlatForm for simplifying flatten cases..

This patch fixes the repeating index issue.

Oct 31 2022, 2:08 AM · Restricted Project, Restricted Project
arnab-oss added reviewers for D137062: [MLIR] Fix bug in the method constructing semi affine expression from flattened form: bondhugula, springerm, hanchung, mravishankar.
Oct 31 2022, 2:07 AM · Restricted Project, Restricted Project
arnab-oss requested review of D137062: [MLIR] Fix bug in the method constructing semi affine expression from flattened form.
Oct 31 2022, 2:02 AM · Restricted Project, Restricted Project

Sep 22 2022

arnab-oss retitled D134425: [NFC] Create a AllocLikeOpInterface and make memref::AllocOp, memref::AllocaOp and gpu::AllocOp implement it. from Create a AllocLikeOpInterface and make memref::AllocLikeOp and gpu::AllocOp implement it. to [NFC] Create a AllocLikeOpInterface and make memref::AllocOp, memref::AllocaOp and gpu::AllocOp implement it..
Sep 22 2022, 5:00 AM · Restricted Project, Restricted Project, Restricted Project
arnab-oss updated the diff for D134425: [NFC] Create a AllocLikeOpInterface and make memref::AllocOp, memref::AllocaOp and gpu::AllocOp implement it..

Refactoring

Sep 22 2022, 4:56 AM · Restricted Project, Restricted Project, Restricted Project
arnab-oss updated the diff for D134425: [NFC] Create a AllocLikeOpInterface and make memref::AllocOp, memref::AllocaOp and gpu::AllocOp implement it..

Make gpu.alloc op implement AllocLikeOpInterface

Sep 22 2022, 4:28 AM · Restricted Project, Restricted Project, Restricted Project
arnab-oss requested review of D134425: [NFC] Create a AllocLikeOpInterface and make memref::AllocOp, memref::AllocaOp and gpu::AllocOp implement it..
Sep 22 2022, 4:13 AM · Restricted Project, Restricted Project, Restricted Project

Sep 5 2022

arnab-oss added a comment to D132117: [MLIR] NFC. Introduce mlir::hasEffect and refactor usages dialect util.

LGTM
Thanks for implementing the suggested changes.

Sep 5 2022, 8:23 AM · Restricted Project, Restricted Project

Aug 26 2022

arnab-oss updated the summary of D128986: Fold memref.expand_shape and memref.collapse_shape ops.
Aug 26 2022, 11:16 PM · Restricted Project, Restricted Project
arnab-oss updated the summary of D128986: Fold memref.expand_shape and memref.collapse_shape ops.
Aug 26 2022, 1:29 AM · Restricted Project, Restricted Project
arnab-oss updated the diff for D128986: Fold memref.expand_shape and memref.collapse_shape ops.

Minor NFC changes.

Aug 26 2022, 1:14 AM · Restricted Project, Restricted Project

Aug 23 2022

arnab-oss updated the diff for D128986: Fold memref.expand_shape and memref.collapse_shape ops.

Minor changes.

Aug 23 2022, 4:47 AM · Restricted Project, Restricted Project
arnab-oss updated the diff for D128986: Fold memref.expand_shape and memref.collapse_shape ops.

Addressed comments by nicolas.

Aug 23 2022, 4:43 AM · Restricted Project, Restricted Project

Aug 22 2022

arnab-oss updated the diff for D128986: Fold memref.expand_shape and memref.collapse_shape ops.

Removed wrong changes in clang-format.sh

Aug 22 2022, 11:13 PM · Restricted Project, Restricted Project
arnab-oss updated the diff for D128986: Fold memref.expand_shape and memref.collapse_shape ops.

Addressed comments.

Aug 22 2022, 5:53 AM · Restricted Project, Restricted Project
arnab-oss added a comment to D128986: Fold memref.expand_shape and memref.collapse_shape ops.

@nicolasvasilache I've made the required changes in the API. Also added tests featuring memref.load/store. However I've not handled folding related to vector transfer ops, so no tests involving those ops.

Aug 22 2022, 3:42 AM · Restricted Project, Restricted Project
arnab-oss updated the diff for D128986: Fold memref.expand_shape and memref.collapse_shape ops.

Addressed comments

Aug 22 2022, 3:39 AM · Restricted Project, Restricted Project
arnab-oss added a comment to D132117: [MLIR] NFC. Introduce mlir::hasEffect and refactor usages dialect util.

I think this change alters the semantics, and won't work if the op has deallocate + some other effects on val. Those type of ops still qualify as dealloc ops.

Aug 22 2022, 1:07 AM · Restricted Project, Restricted Project

Aug 12 2022

arnab-oss added a comment to D128986: Fold memref.expand_shape and memref.collapse_shape ops.

I'm getting the following error:

Aug 12 2022, 4:24 AM · Restricted Project, Restricted Project

Jul 25 2022

arnab-oss added a comment to D128986: Fold memref.expand_shape and memref.collapse_shape ops.

@nicolasvasilache can you please take a look at my comments?

Jul 25 2022, 3:26 AM · Restricted Project, Restricted Project

Jul 18 2022

arnab-oss retitled D128986: Fold memref.expand_shape and memref.collapse_shape ops from Fold memref.expand_shape and memref.collapse_shape ops, when they are the parent op of source value of memref/affine load/store ops. to Fold memref.expand_shape and memref.collapse_shape ops..
Jul 18 2022, 1:03 AM · Restricted Project, Restricted Project

Jul 17 2022

arnab-oss added a comment to D128986: Fold memref.expand_shape and memref.collapse_shape ops.

ping @nicolasvasilache can you please re-review?

Jul 17 2022, 11:34 PM · Restricted Project, Restricted Project

Jul 13 2022

arnab-oss updated the diff for D128986: Fold memref.expand_shape and memref.collapse_shape ops.

Addressd review comments.

Jul 13 2022, 10:13 AM · Restricted Project, Restricted Project
arnab-oss updated the diff for D128986: Fold memref.expand_shape and memref.collapse_shape ops.

Addressed review comments.

Jul 13 2022, 8:27 AM · Restricted Project, Restricted Project

Jul 1 2022

arnab-oss updated the diff for D128986: Fold memref.expand_shape and memref.collapse_shape ops.

Discarding unnecessary changes.

Jul 1 2022, 6:01 AM · Restricted Project, Restricted Project
arnab-oss added reviewers for D128986: Fold memref.expand_shape and memref.collapse_shape ops: bondhugula, dcaballe, nicolasvasilache, mehdi_amini.
Jul 1 2022, 5:54 AM · Restricted Project, Restricted Project
arnab-oss requested review of D128986: Fold memref.expand_shape and memref.collapse_shape ops.
Jul 1 2022, 5:52 AM · Restricted Project, Restricted Project

May 13 2022

arnab-oss updated the diff for D124257: [MLIR][GPU] Add canonicalizer for gpu.memcpy.

Rebased on latest master.

May 13 2022, 1:38 AM · Restricted Project, Restricted Project

May 12 2022

arnab-oss updated the diff for D124257: [MLIR][GPU] Add canonicalizer for gpu.memcpy.

Addressed comments by @bondhugula.

May 12 2022, 11:09 PM · Restricted Project, Restricted Project
arnab-oss updated the diff for D124257: [MLIR][GPU] Add canonicalizer for gpu.memcpy.

Addressed comments by @bondhugula.

May 12 2022, 6:32 AM · Restricted Project, Restricted Project

May 6 2022

arnab-oss updated the diff for D124257: [MLIR][GPU] Add canonicalizer for gpu.memcpy.

Addressed comments by @bondhugula.

May 6 2022, 6:09 AM · Restricted Project, Restricted Project

Apr 29 2022

arnab-oss updated the diff for D124257: [MLIR][GPU] Add canonicalizer for gpu.memcpy.

Addressed comments by @bondhugula.

Apr 29 2022, 4:38 AM · Restricted Project, Restricted Project
arnab-oss updated the diff for D124257: [MLIR][GPU] Add canonicalizer for gpu.memcpy.

Fixed linting errors.

Apr 29 2022, 4:33 AM · Restricted Project, Restricted Project
arnab-oss updated the diff for D124257: [MLIR][GPU] Add canonicalizer for gpu.memcpy.

Added hasSingleEffect() as a part of SideEffectInterface.

Apr 29 2022, 4:32 AM · Restricted Project, Restricted Project

Apr 22 2022

arnab-oss added a reviewer for D124257: [MLIR][GPU] Add canonicalizer for gpu.memcpy: mehdi_amini.
Apr 22 2022, 6:54 AM · Restricted Project, Restricted Project
arnab-oss added reviewers for D124257: [MLIR][GPU] Add canonicalizer for gpu.memcpy: csigg, hanbinyoon, jpienaar.
Apr 22 2022, 6:54 AM · Restricted Project, Restricted Project
arnab-oss requested review of D124257: [MLIR][GPU] Add canonicalizer for gpu.memcpy.
Apr 22 2022, 6:49 AM · Restricted Project, Restricted Project

Apr 21 2022

arnab-oss updated subscribers of D121279: [MLIR][GPU] Add canonicalizer for gpu.memcpy.

Can you please provide the test case for which incorrect IR is generated?
I will provide the fix. Thanks in advance.

Apr 21 2022, 8:57 AM · Restricted Project, Restricted Project

Apr 19 2022

arnab-oss updated the diff for D121279: [MLIR][GPU] Add canonicalizer for gpu.memcpy.

Removed trailing white spaces.

Apr 19 2022, 3:46 AM · Restricted Project, Restricted Project

Apr 18 2022

arnab-oss updated the diff for D121279: [MLIR][GPU] Add canonicalizer for gpu.memcpy.

Fixed build issues.

Apr 18 2022, 4:05 AM · Restricted Project, Restricted Project

Apr 14 2022

arnab-oss updated the diff for D121279: [MLIR][GPU] Add canonicalizer for gpu.memcpy.

Addressed review comments and rebased on latest master.

Apr 14 2022, 5:10 AM · Restricted Project, Restricted Project

Apr 13 2022

arnab-oss updated the diff for D121878: [MLIR][GPU] Add canonicalization patterns for folding simple gpu.wait ops..

Adressed comments by @csigg, @bondhugula

Apr 13 2022, 4:08 AM · Restricted Project, Restricted Project

Mar 23 2022

arnab-oss added a comment to D121878: [MLIR][GPU] Add canonicalization patterns for folding simple gpu.wait ops..

ping @csigg, can you please take look at my comment?

Mar 23 2022, 5:04 AM · Restricted Project, Restricted Project

Mar 16 2022

arnab-oss updated the diff for D121878: [MLIR][GPU] Add canonicalization patterns for folding simple gpu.wait ops..

Resolved build issues.

Mar 16 2022, 9:28 PM · Restricted Project, Restricted Project
arnab-oss added a reviewer for D121878: [MLIR][GPU] Add canonicalization patterns for folding simple gpu.wait ops.: csigg.
Mar 16 2022, 9:11 PM · Restricted Project, Restricted Project
arnab-oss requested review of D121878: [MLIR][GPU] Add canonicalization patterns for folding simple gpu.wait ops..
Mar 16 2022, 9:10 PM · Restricted Project, Restricted Project

Mar 15 2022

arnab-oss updated the diff for D121279: [MLIR][GPU] Add canonicalizer for gpu.memcpy.

Addressed comments by @csigg.

Mar 15 2022, 3:20 AM · Restricted Project, Restricted Project

Mar 12 2022

arnab-oss updated the diff for D121279: [MLIR][GPU] Add canonicalizer for gpu.memcpy.

Addressed comments by @bondhugula

Mar 12 2022, 3:01 AM · Restricted Project, Restricted Project

Mar 10 2022

arnab-oss updated the diff for D121279: [MLIR][GPU] Add canonicalizer for gpu.memcpy.

Addressed comments by @bondhugula

Mar 10 2022, 11:04 PM · Restricted Project, Restricted Project
arnab-oss updated the diff for D121279: [MLIR][GPU] Add canonicalizer for gpu.memcpy.

Addressed comments by @bondhugula

Mar 10 2022, 5:44 AM · Restricted Project, Restricted Project

Mar 9 2022

arnab-oss updated the diff for D121279: [MLIR][GPU] Add canonicalizer for gpu.memcpy.

Addressed review comments by @csigg.

Mar 9 2022, 5:11 AM · Restricted Project, Restricted Project
arnab-oss added a reviewer for D121279: [MLIR][GPU] Add canonicalizer for gpu.memcpy: ftynse.
Mar 9 2022, 4:00 AM · Restricted Project, Restricted Project
arnab-oss updated the diff for D121279: [MLIR][GPU] Add canonicalizer for gpu.memcpy.

Set hasVerifier field in gpu.memcpy removed earlier by mistake.

Mar 9 2022, 3:59 AM · Restricted Project, Restricted Project
arnab-oss requested review of D121279: [MLIR][GPU] Add canonicalizer for gpu.memcpy.
Mar 9 2022, 3:53 AM · Restricted Project, Restricted Project

Jan 19 2022

arnab-oss added reviewers for D117673: [MLIR] Introduce copy removal pass: rriddle, ftynse, wsmoses.
Jan 19 2022, 7:19 AM · Restricted Project
arnab-oss updated the diff for D117669: [SCF] Convert scf.while to scf.for.

Fixed build errors.

Jan 19 2022, 7:18 AM · Restricted Project
arnab-oss added reviewers for D117673: [MLIR] Introduce copy removal pass: bondhugula, mehdi_amini, nicolasvasilache.
Jan 19 2022, 6:20 AM · Restricted Project
arnab-oss updated the summary of D117673: [MLIR] Introduce copy removal pass.
Jan 19 2022, 6:20 AM · Restricted Project
arnab-oss requested review of D117673: [MLIR] Introduce copy removal pass.
Jan 19 2022, 6:18 AM · Restricted Project
arnab-oss added reviewers for D117669: [SCF] Convert scf.while to scf.for: mehdi_amini, rriddle, ftynse.
Jan 19 2022, 6:10 AM · Restricted Project
arnab-oss added a reviewer for D117669: [SCF] Convert scf.while to scf.for: bondhugula.
Jan 19 2022, 6:10 AM · Restricted Project
arnab-oss requested review of D117669: [SCF] Convert scf.while to scf.for.
Jan 19 2022, 6:09 AM · Restricted Project

Jan 6 2022

arnab-oss added a comment to D90333: added support raising scf to affine dialect.

ping @ JianyiCheng, can you please let me know if this revision has been abandoned? I actually require a similar kind of pass.

Jan 6 2022, 8:28 AM · Restricted Project
arnab-oss added a reviewer for D90333: added support raising scf to affine dialect: bondhugula.
Jan 6 2022, 7:50 AM · Restricted Project

Dec 19 2021

arnab-oss updated subscribers of D111233: [mlir][linalg] Retire Linalg ConvOp..

Initially Linalg::ConvOp had padding attribute, but the current set of convolution ops don't have them. Is there anyway I can define convolution with padding with the current Linalg Dialect?
CC - @bondhugula , @gysit , @nicolasvasilache

Dec 19 2021, 11:14 PM · Restricted Project

Nov 26 2021

arnab-oss updated the diff for D114238: [MLIR] Avoid creation of buggy affine maps when incorrect values of number of dimensions and number of symbols are provided..

Added doc comments for the getMaxDimAndSymbol().

Nov 26 2021, 11:00 AM · Restricted Project

Nov 25 2021

arnab-oss updated the diff for D114238: [MLIR] Avoid creation of buggy affine maps when incorrect values of number of dimensions and number of symbols are provided..

Rebased on latest master.

Nov 25 2021, 1:54 PM · Restricted Project
arnab-oss updated the diff for D114240: [MLIR] Prevent creation of buggy affine map after linearizing collapsed dimensions of source map.

Rebased on latest master.

Nov 25 2021, 1:45 PM · Restricted Project
arnab-oss updated the diff for D114240: [MLIR] Prevent creation of buggy affine map after linearizing collapsed dimensions of source map.

Removed extra parenthesis as suggested by @bondhugula.

Nov 25 2021, 9:52 AM · Restricted Project
arnab-oss added a comment to D114238: [MLIR] Avoid creation of buggy affine maps when incorrect values of number of dimensions and number of symbols are provided..

@nicolasvasilache, @bondhugula and @rriddle, I have addressed all your comments. Please take a look.

Nov 25 2021, 9:26 AM · Restricted Project
arnab-oss updated the diff for D114238: [MLIR] Avoid creation of buggy affine maps when incorrect values of number of dimensions and number of symbols are provided..

Addressed comments by @nicolasvasilache and @bondhugula

Nov 25 2021, 9:25 AM · Restricted Project
arnab-oss added a comment to D114240: [MLIR] Prevent creation of buggy affine map after linearizing collapsed dimensions of source map.

ping @nicolasvasilache, I have addressed your comments. You can take a look.

Nov 25 2021, 8:43 AM · Restricted Project
arnab-oss updated the diff for D114240: [MLIR] Prevent creation of buggy affine map after linearizing collapsed dimensions of source map.

Addressed comments by @nicolasvasilache and @bondhugula.

Nov 25 2021, 5:35 AM · Restricted Project

Nov 21 2021

arnab-oss added a comment to D114238: [MLIR] Avoid creation of buggy affine maps when incorrect values of number of dimensions and number of symbols are provided..

ping @rriddle and @nicolasvasilache can you please take a look at the replies of your respective comments.

Nov 21 2021, 11:07 PM · Restricted Project
arnab-oss updated the diff for D114238: [MLIR] Avoid creation of buggy affine maps when incorrect values of number of dimensions and number of symbols are provided..

Removed unnecessary #include<algorithm>.

Nov 21 2021, 11:06 PM · Restricted Project
arnab-oss updated the diff for D114238: [MLIR] Avoid creation of buggy affine maps when incorrect values of number of dimensions and number of symbols are provided..

Addressed comments.

Nov 21 2021, 11:01 PM · Restricted Project

Nov 20 2021

arnab-oss added a reviewer for D114240: [MLIR] Prevent creation of buggy affine map after linearizing collapsed dimensions of source map: rriddle.
Nov 20 2021, 7:27 AM · Restricted Project
arnab-oss updated the diff for D114240: [MLIR] Prevent creation of buggy affine map after linearizing collapsed dimensions of source map.

Addressed comments by @bondhugula.

Nov 20 2021, 7:24 AM · Restricted Project

Nov 19 2021

arnab-oss added a reviewer for D114240: [MLIR] Prevent creation of buggy affine map after linearizing collapsed dimensions of source map: bondhugula.
Nov 19 2021, 6:03 AM · Restricted Project
arnab-oss requested review of D114240: [MLIR] Prevent creation of buggy affine map after linearizing collapsed dimensions of source map.
Nov 19 2021, 6:00 AM · Restricted Project
arnab-oss added reviewers for D114238: [MLIR] Avoid creation of buggy affine maps when incorrect values of number of dimensions and number of symbols are provided.: nicolasvasilache, bondhugula.
Nov 19 2021, 5:51 AM · Restricted Project
arnab-oss requested review of D114238: [MLIR] Avoid creation of buggy affine maps when incorrect values of number of dimensions and number of symbols are provided..
Nov 19 2021, 5:50 AM · Restricted Project
arnab-oss updated the diff for D114237: [MLIR] Avoid creation of buggy affine maps while replacing dimension and symbol.

Fixed documentation @bondhugula .

Nov 19 2021, 5:45 AM · Restricted Project
arnab-oss added a reviewer for D114237: [MLIR] Avoid creation of buggy affine maps while replacing dimension and symbol: bondhugula.
Nov 19 2021, 5:31 AM · Restricted Project
arnab-oss requested review of D114237: [MLIR] Avoid creation of buggy affine maps while replacing dimension and symbol.
Nov 19 2021, 5:30 AM · Restricted Project
arnab-oss added a comment to D112985: [MLIR] Simplify Semi-affine expressions by rule based matching and replacing "expr - q * (expr floordiv q)" with "expr mod q" expression..

ping @dcaballe and @springerm , can you please take a look and review this? This is a continuation of semi affine expression simplification PR.

Nov 19 2021, 4:52 AM · Restricted Project
arnab-oss updated the diff for D112985: [MLIR] Simplify Semi-affine expressions by rule based matching and replacing "expr - q * (expr floordiv q)" with "expr mod q" expression..

Addressed all comments by @bondhugula

Nov 19 2021, 4:51 AM · Restricted Project
arnab-oss updated the diff for D112985: [MLIR] Simplify Semi-affine expressions by rule based matching and replacing "expr - q * (expr floordiv q)" with "expr mod q" expression..

Added test cases.

Nov 19 2021, 4:40 AM · Restricted Project

Nov 13 2021

arnab-oss added a comment to D112808: [MLIR] Simplify Semi Affine Expressions Using flattening.

ping @springerm and @dcaballe , can you please let me know whether the revision is ok or if any more changes are required to be made? Thanks in advance!

Nov 13 2021, 7:38 PM · Restricted Project

Nov 12 2021

arnab-oss updated the diff for D112808: [MLIR] Simplify Semi Affine Expressions Using flattening.

Added CHECK-LABEL in unit tests.

Nov 12 2021, 5:12 AM · Restricted Project

Nov 10 2021

arnab-oss added a comment to D112808: [MLIR] Simplify Semi Affine Expressions Using flattening.

Fixed a bug in the loop iterating over localExprs vector. Addressed comments
by @springerm.

Could you add a test case that exercises that part of the code? (if not already)

Nov 10 2021, 11:37 PM · Restricted Project
arnab-oss updated the diff for D112808: [MLIR] Simplify Semi Affine Expressions Using flattening.

Added an example which is giving correct output now after the previous update bug fix.

Nov 10 2021, 11:37 PM · Restricted Project

Nov 9 2021

arnab-oss updated the diff for D112808: [MLIR] Simplify Semi Affine Expressions Using flattening.

Fixed a bug in the loop iterating over localExprs vector. Addressed comments
by @springerm.

Nov 9 2021, 4:27 AM · Restricted Project

Nov 7 2021

arnab-oss updated the diff for D112808: [MLIR] Simplify Semi Affine Expressions Using flattening.

Addressed all the comments by @bondhugula and @dcaballe.

Nov 7 2021, 10:27 AM · Restricted Project
arnab-oss updated the diff for D112808: [MLIR] Simplify Semi Affine Expressions Using flattening.

Removed unwanted changes from clang-format.sh

Nov 7 2021, 12:16 AM · Restricted Project
arnab-oss updated the diff for D112808: [MLIR] Simplify Semi Affine Expressions Using flattening.

Added an example showing the contents of indices, indexToExprMap and coefficients.

Nov 7 2021, 12:10 AM · Restricted Project

Nov 6 2021

arnab-oss updated the diff for D112808: [MLIR] Simplify Semi Affine Expressions Using flattening.

Addressed all the comments by springerm.

Nov 6 2021, 11:53 PM · Restricted Project