Move resize / zero logic for _thnn_conv_depthwise2d from codegen to native code. by gchanan · Pull Request #37957 · pytorch/pytorch

gchanan · 2020-05-06T20:00:03Z

Stack from ghstack:

Kill resize-ing and zero-ing from codegen. #37958 Kill resize-ing and zero-ing from codegen.
Move resize / zero logic for _thnn_conv_depthwise2d from codegen to native code. #37957 Move resize / zero logic for _thnn_conv_depthwise2d from codegen to native code.
Move _thnn_conv2d resize and zero code from codegen to native code. #37956 Move _thnn_conv2d resize and zero code from codegen to native code.
Move resize logic for bmm from codegen to native code. #37955 Move resize logic for bmm from codegen to native code.
[RESUBMIT] Kill broadcasting from the codegen layer. #37907 [RESUBMIT] Kill broadcasting from the codegen layer.

Differential Revision: D21433212

…ative code. [ghstack-poisoned]

dr-ci · 2020-05-06T20:59:04Z

💊 CI failures summary and remediations

As of commit d376a6a (more details on the Dr. CI page):

2/3 failures possibly* introduced in this PR
- 1/2 non-CircleCI failure(s)
1/3 broken upstream at merge base 91c1505 on May 05 from 2:00pm to 6:02pm PDT (21 commits; 06e1b68 - e3d1c4e)

🕵️ 1 new failure recognized by patterns

The following CI failures do not appear to be due to upstream breakages:

pytorch_linux_backward_compatibility_check_test (1/1)

Step: "Run tests" (full log | diagnosis details | 🔁 rerun)

May 07 16:30:09 The PR is introducing backward incompatible changes to the operator library. Please contact PyTorch team to confirm whether this change is wanted or not.

May 07 16:30:09 processing existing schema:  aten::_has_compatible_shallow_copy_type(Tensor self, Tensor from) -> (bool) 
May 07 16:30:09 processing existing schema:  aten::_unsafe_view(Tensor self, int[] size) -> (Tensor) 
May 07 16:30:09 processing existing schema:  aten::unsqueeze(Tensor(a) self, int dim) -> (Tensor(a)) 
May 07 16:30:09 processing existing schema:  aten::unsqueeze_(Tensor(a!) self, int dim) -> (Tensor(a!)) 
May 07 16:30:09 processing existing schema:  aten::vander(Tensor x, int? N=None, bool increasing=False) -> (Tensor) 
May 07 16:30:09 processing existing schema:  aten::var(Tensor self, bool unbiased=True) -> (Tensor) 
May 07 16:30:09 processing existing schema:  aten::var.dim(Tensor self, int[1] dim, bool unbiased=True, bool keepdim=False) -> (Tensor) 
May 07 16:30:09 processing existing schema:  aten::var.out(Tensor self, int[1] dim, bool unbiased=True, bool keepdim=False, *, Tensor(a!) out) -> (Tensor(a!)) 
May 07 16:30:09 processing existing schema:  aten::var.names_dim(Tensor self, str[1] dim, bool unbiased=True, bool keepdim=False) -> (Tensor) 
May 07 16:30:09 processing existing schema:  aten::var.names_out(Tensor self, str[1] dim, bool unbiased=True, bool keepdim=False, *, Tensor(a!) out) -> (Tensor(a!)) 
May 07 16:30:09 The PR is introducing backward incompatible changes to the operator library. Please contact PyTorch team to confirm whether this change is wanted or not.  
May 07 16:30:09  
May 07 16:30:09 Broken ops: [ 
May 07 16:30:09 	aten::quantized_instance_norm(Tensor input, Tensor? weight, Tensor? bias, float eps, float output_scale, int output_zero_point) -> (Tensor) 
May 07 16:30:09 	quantized::group_norm(Tensor input, int num_groups, Tensor weight, Tensor bias, float eps, float output_scale, int output_zero_point) -> (Tensor) 
May 07 16:30:09 	quantized::instance_norm(Tensor input, Tensor weight, Tensor bias, float eps, float output_scale, int output_zero_point) -> (Tensor) 
May 07 16:30:09 	quantized::conv_unpack(__torch__.torch.classes.quantized.Conv2dPackedParamsBase packed_weights) -> (Tensor unpacked_weights, Tensor? B_origin) 
May 07 16:30:09 	aten::quantized_group_norm(Tensor input, int num_groups, Tensor? weight, Tensor? bias, float eps, float output_scale, int output_zero_point) -> (Tensor) 
May 07 16:30:09 ] 
May 07 16:30:09 + cleanup 
May 07 16:30:09 + retcode=1

🚧 1 fixed upstream failure:

These were probably caused by upstream breakages that were already fixed.

Please rebase on the viable/strict branch (expand for instructions)

Since your merge base is older than viable/strict, run these commands:

git fetch https://github.com/pytorch/pytorch viable/strict
git rebase FETCH_HEAD

Check out the recency history of this "viable master" tracking branch.

pytorch_macos_10_13_py3_test on May 05 from 2:00pm to 6:02pm PDT (21 commits; 06e1b68 - e3d1c4e)
- 🔁 rerun

Extra GitHub checks: 1 failed

Failed: GitHub Actions - clang-tidy

This comment was automatically generated by Dr. CI (expand for details).

Follow this link to opt-out of these comments for your Pull Requests.

Please report bugs/suggestions on the GitHub issue tracker.

See how this bot performed.

This comment has been revised 15 times.

bhosmer · 2020-05-06T21:34:11Z


        # grad_weight and grad_bias need to be resized and zeroed
-        if arg['name'] == 'grad_weight' and base['name'] != '_thnn_conv2d':
+        if arg['name'] == 'grad_weight' and base['name'] != '_thnn_conv2d' and base['name'] != '_thnn_conv_depthwise2d':


reupping earlier note - personally I'd rather have a list of names that's getting checked here, but not a blocker if you either like it better this way or just want to blast ahead

yep, good suggestion. I knew ahead of time (from checking the generated code) that there were only two cases and I was going to remove them in this stack so it didn't seem worth it.

…odegen to native code." Differential Revision: [D21433212](https://our.internmc.facebook.com/intern/diff/D21433212) [ghstack-poisoned]

facebook-github-bot · 2020-05-08T02:26:07Z

@gchanan merged this pull request in dc25190.

…ative code. (pytorch#37957) Summary: Pull Request resolved: pytorch#37957 Test Plan: Imported from OSS Differential Revision: D21433212 Pulled By: gchanan fbshipit-source-id: fb431d5cf06afe2bb87fa2d73e15046f9a8d044d

Move resize / zero logic for _thnn_conv_depthwise2d from codegen to n…

a456cf5

…ative code. [ghstack-poisoned]

This was referenced May 6, 2020

Move resize logic for bmm from codegen to native code. #37955

Closed

Move _thnn_conv2d resize and zero code from codegen to native code. #37956

Closed

Kill resize-ing and zero-ing from codegen. #37958

Closed

gchanan requested a review from bhosmer May 6, 2020 20:13

bhosmer approved these changes May 6, 2020

View reviewed changes

gchanan added 2 commits May 7, 2020 07:33

Update on "Move resize / zero logic for _thnn_conv_depthwise2d from c…

9625895

…odegen to native code." Differential Revision: [D21433212](https://our.internmc.facebook.com/intern/diff/D21433212) [ghstack-poisoned]

Update on "Move resize / zero logic for _thnn_conv_depthwise2d from c…

d376a6a

…odegen to native code." Differential Revision: [D21433212](https://our.internmc.facebook.com/intern/diff/D21433212) [ghstack-poisoned]

facebook-github-bot closed this in dc25190 May 7, 2020

facebook-github-bot added the merged label May 8, 2020

mruberry added the Merged label Oct 28, 2020

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Move resize / zero logic for _thnn_conv_depthwise2d from codegen to native code.#37957

Move resize / zero logic for _thnn_conv_depthwise2d from codegen to native code.#37957
gchanan wants to merge 3 commits intogh/gchanan/269/basefrom
gh/gchanan/269/head

gchanan commented May 6, 2020 •

edited

Loading

Uh oh!

dr-ci Bot commented May 6, 2020 •

edited

Loading

Uh oh!

bhosmer May 6, 2020

Uh oh!

gchanan May 7, 2020

Uh oh!

facebook-github-bot commented May 8, 2020

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

Conversation

gchanan commented May 6, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

dr-ci Bot commented May 6, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

💊 CI failures summary and remediations

🕵️ 1 new failure recognized by patterns

pytorch_linux_backward_compatibility_check_test (1/1)

🚧 1 fixed upstream failure:

Extra GitHub checks: 1 failed

Uh oh!

bhosmer May 6, 2020

Choose a reason for hiding this comment

Uh oh!

gchanan May 7, 2020

Choose a reason for hiding this comment

Uh oh!

facebook-github-bot commented May 8, 2020

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

gchanan commented May 6, 2020 •

edited

Loading

dr-ci Bot commented May 6, 2020 •

edited

Loading