Add support for 32bit CI build #782

Krishna-13-cyber · 2024-02-22T12:14:25Z

Addresses #748

codecov · 2024-02-22T12:23:37Z

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 94.71%. Comparing base (8d52a36) to head (52040e8).
Report is 3 commits behind head on master.

Additional details and impacted files

@@           Coverage Diff           @@
##           master     #782   +/-   ##
=======================================
  Coverage   94.71%   94.71%           
=======================================
  Files          49       49           
  Lines        7342     7342           
=======================================
  Hits         6954     6954           
  Misses        388      388

vgvassilev · 2024-02-22T14:20:40Z

Does that work for PRs? I don’t see a build.

Krishna-13-cyber · 2024-02-22T14:39:37Z

Does that work for PRs? I don’t see a build.

Not then, I have enabled it now.

vgvassilev · 2024-02-23T12:19:58Z

.github/workflows/platforms.yml

@@ -0,0 +1,118 @@
+---
+name: 'Platforms'


Suggested change

name: 'Platforms'

name: 'Arch'

vgvassilev · 2024-02-23T12:22:02Z

.github/workflows/platforms.yml

+           - name: "Compile library"
+             run: |
+               mkdir build && cd build
+               cmake -DLLVM_DIR=/usr/lib/ -DCMAKE_BUILD_TYPE=DEBUG -DLLVM_EXTERNAL_LIT="$(which lit)" ../../clad


Suggested change

cmake -DLLVM_DIR=/usr/lib/ -DCMAKE_BUILD_TYPE=DEBUG -DLLVM_EXTERNAL_LIT="$(which lit)" ../../clad

cmake -DLLVM_EXTERNAL_LIT="$(which lit)" ../../clad

vgvassilev · 2024-02-23T12:22:15Z

.github/workflows/platforms.yml

+           - name: "Compile library"
+             run: |
+               mkdir build && cd build
+               cmake -DLLVM_DIR=/usr/lib/ -DCMAKE_BUILD_TYPE=DEBUG -DLLVM_EXTERNAL_LIT="$(which lit)" ../../clad


Suggested change

cmake -DLLVM_DIR=/usr/lib/ -DCMAKE_BUILD_TYPE=DEBUG -DLLVM_EXTERNAL_LIT="$(which lit)" ../../clad

cmake -DLLVM_EXTERNAL_LIT="$(which lit)" ../../clad

vgvassilev · 2024-02-23T12:23:14Z

.github/workflows/platforms.yml

+           - name: "Compile library"
+             run: |
+               mkdir build && cd build
+               cmake -DLLVM_DIR=/usr/lib/ -DCMAKE_BUILD_TYPE=DEBUG -DLLVM_EXTERNAL_LIT="$(which lit)" ../../clad


Suggested change

cmake -DLLVM_DIR=/usr/lib/ -DCMAKE_BUILD_TYPE=DEBUG -DLLVM_EXTERNAL_LIT="$(which lit)" ../../clad

cmake -DLLVM_EXTERNAL_LIT="$(which lit)" ../../clad

Can we extract the step "Compile library" into one place as it seems to repeat across builds. The same for the other common steps. Maybe this could help?

Sorry for the delayed response. Yes, I will look into this and patch it asap.

The patch has been updated but the workflow report says path not found, its not able to fetch this patch vgvassilev/clad/.github/workflows/called.yml@master.
My test-workflow works fine with the same https://github.com/Krishna-13-cyber/clad/actions/runs/8045618661

Can you take a look how the clad main workflow is organized, we want something similar for this PR.

Yes, will do this.
Thanks!

Hi, I have updated with respect to clad convention adhering to matrix strategy.
Thanks!

vgvassilev · 2024-02-27T13:41:11Z

Nice! Now we need to figure out how to suppress the test failures. For this PR maybe it would be fine if we allow failures on armv7 and x86.

vgvassilev · 2024-02-27T15:15:46Z

.github/workflows/arch.yml

+  architecture:
+    strategy:
+      matrix:
+        target: [x86, armv7, aarch64]


Looks like armv7 and aarch64 are considerable slower. Can we disable them?

Yes, they are slower.
We had an intention to target multiple architectures right to get hold of failures/bugs so why should we disable them (due to time constraints).
Can we try any other workaround to fit in those(slower ones) as well?

Krishna-13-cyber · 2024-02-27T15:23:33Z

Nice! Now we need to figure out how to suppress the test failures. For this PR maybe it would be fine if we allow failures on armv7 and x86.

Yes, we can solve the failures in a seperate PR.

vgvassilev · 2024-02-27T15:25:38Z

Nice! Now we need to figure out how to suppress the test failures. For this PR maybe it would be fine if we allow failures on armv7 and x86.

Yes, we can solve the failures in a seperate PR.

Looks like this won't work. Maybe in this PR we should add to each failing test: XFAIL: target={{(i686|i386).*}}. Can you try this and then we can drop the non-functional continue-on-error thing.

Krishna-13-cyber · 2024-02-27T15:34:49Z

Nice! Now we need to figure out how to suppress the test failures. For this PR maybe it would be fine if we allow failures on armv7 and x86.

Yes, we can solve the failures in a seperate PR.

Looks like this won't work. Maybe in this PR we should add to each failing test: XFAIL: target={{(i686|i386).*}}. Can you try this and then we can drop the non-functional continue-on-error thing.

I added this continue-on-error for all the parallel jobs to run independently as without this thing other jobs abort due to any failure on any of the jobs present.
Yes, I will try adding this XFAIL: target={{(i686|i386).*}} to failing tests.

Krishna-13-cyber · 2024-02-27T17:48:54Z

Looks like this won't work. Maybe in this PR we should add to each failing test: XFAIL: target={{(i686|i386).*}}. Can you try this and then we can drop the non-functional continue-on-error thing.

The failing tests get suppressed with XFAIL:* and not this XFAIL: target={{(i686|i386).*}}.
Should I use this?

vgvassilev · 2024-02-27T17:59:26Z

Looks like this won't work. Maybe in this PR we should add to each failing test: XFAIL: target={{(i686|i386).*}}. Can you try this and then we can drop the non-functional continue-on-error thing.

The failing tests get suppressed with XFAIL:* and not this XFAIL: target={{(i686|i386).*}}. Should I use this?

No, that will suppress even working tests.

vgvassilev · 2024-02-27T20:25:54Z

target={{(i686|i386).*}}

@Krishna-13-cyber, you need // XFAIL: target={{i586.*}}

vgvassilev

Awesome! LGTM!

Krishna-13-cyber · 2024-02-28T15:16:56Z

Awesome! LGTM!

Thanks!

vgvassilev reviewed Feb 23, 2024

View reviewed changes

Krishna-13-cyber force-pushed the ci-platforms branch from fba4b90 to b279712 Compare February 26, 2024 11:33

vgvassilev reviewed Feb 27, 2024

View reviewed changes

Add support for 32bit CI build

52040e8

Krishna-13-cyber force-pushed the ci-platforms branch from a53f831 to 52040e8 Compare February 28, 2024 03:39

vgvassilev approved these changes Feb 28, 2024

View reviewed changes

vgvassilev merged commit a200944 into vgvassilev:master Feb 28, 2024
81 checks passed

vgvassilev mentioned this pull request Mar 2, 2024

Re-enable tests on i568 #792

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add support for 32bit CI build #782

Add support for 32bit CI build #782

Krishna-13-cyber commented Feb 22, 2024

codecov bot commented Feb 22, 2024 •

edited

Loading

vgvassilev commented Feb 22, 2024

Krishna-13-cyber commented Feb 22, 2024 •

edited

Loading

vgvassilev Feb 23, 2024

vgvassilev Feb 23, 2024

vgvassilev Feb 23, 2024

vgvassilev Feb 23, 2024

Krishna-13-cyber Feb 26, 2024

Krishna-13-cyber Feb 26, 2024

vgvassilev Feb 26, 2024

Krishna-13-cyber Feb 26, 2024

Krishna-13-cyber Feb 27, 2024

vgvassilev commented Feb 27, 2024

vgvassilev Feb 27, 2024

Krishna-13-cyber Feb 27, 2024 •

edited

Loading

Krishna-13-cyber commented Feb 27, 2024

vgvassilev commented Feb 27, 2024

Krishna-13-cyber commented Feb 27, 2024

Krishna-13-cyber commented Feb 27, 2024

vgvassilev commented Feb 27, 2024

vgvassilev commented Feb 27, 2024

vgvassilev left a comment

Krishna-13-cyber commented Feb 28, 2024

	cmake -DLLVM_DIR=/usr/lib/ -DCMAKE_BUILD_TYPE=DEBUG -DLLVM_EXTERNAL_LIT="$(which lit)" ../../clad
	cmake -DLLVM_EXTERNAL_LIT="$(which lit)" ../../clad

Add support for 32bit CI build #782

Add support for 32bit CI build #782

Conversation

Krishna-13-cyber commented Feb 22, 2024

codecov bot commented Feb 22, 2024 • edited Loading

Codecov Report

vgvassilev commented Feb 22, 2024

Krishna-13-cyber commented Feb 22, 2024 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

vgvassilev commented Feb 27, 2024

Choose a reason for hiding this comment

Krishna-13-cyber Feb 27, 2024 • edited Loading

Choose a reason for hiding this comment

Krishna-13-cyber commented Feb 27, 2024

vgvassilev commented Feb 27, 2024

Krishna-13-cyber commented Feb 27, 2024

Krishna-13-cyber commented Feb 27, 2024

vgvassilev commented Feb 27, 2024

vgvassilev commented Feb 27, 2024

vgvassilev left a comment

Choose a reason for hiding this comment

Krishna-13-cyber commented Feb 28, 2024

codecov bot commented Feb 22, 2024 •

edited

Loading

Krishna-13-cyber commented Feb 22, 2024 •

edited

Loading

Krishna-13-cyber Feb 27, 2024 •

edited

Loading