@azharudd I don't believe I have commit access, could you land this for me?

Updated the diff to

• limit changes to clang-stage1-RA
• only skip-triggering new jobs
• Added option to override node config in job config (defaults to green-dragon-02)

https://github.com/llvm/llvm-lnt/commit/0476dede37838e2f765f50dd26734cfced9d90d9

Merged in with 0476dede37838e2f765f50dd26734cfced9d90d9

Ping for further feedback (or commit)

Hi @thakis , I think the following patch caused the MSVC based buildbots to fail (see https://lab.llvm.org/buildbot/#/builders/123/builds/8915)

@rgal no I don't think there is anything stopping this. Feel free to commit this for me (sorry for the delayed response)

@lhames Ping for review

Hi @RKSimon ,
First, let me apologize as I completely forgot the original bug was assigned to me. Thank you for tackling this.

use shared_ptrs

@lhames When you say "original fix" do you mean the fix where I moved DefaultMMapper to be a member of SectionMemoryManager?

@lhames Sorry for the broken build, I didn't realize clang tests would be affected here.

@lhames update made. Thank you for the feedback/insight. If this looks good, could you go ahead and commit this for me (and possibly https://reviews.llvm.org/D107084 as well)

@lhames Ping to commit this if you're able :)

@lhames ping for update

@lhames You are correct! I wasn't familiar with ManagedStatic but that does fix it. Do you prefer that approach to what's here? If so, I'll go ahead and update the diff. Originally I thought it made sense to have DefaultMMapperInstance be an instance variable of the SectionMemoryManager since it's the only place it's used, but perhaps not since I would have to move the corresponding class definition into the SectionMemoryManager as I did here

Hi @lhames , thanks for the review. Could I have you commit this for me? I do not have commit access

Hi @ayermolo , looks like

llvm/test/DebugInfo/X86/symbolize-debug-fission-single.test

Looks like this commit caused some test failures on linux build machines. Buildbots for reference:
http://lab.llvm.org:8011/#/builders/94/builds/2238
http://lab.llvm.org:8011/#/builders/105/builds/5805

Hi @Tyker , I believe this patch is causing the following test to fail: clang/test/CodeGen/builtin-align-assumption.c

Hi @int3 , It looks like either this or https://reviews.llvm.org/D95265 is causing linking issues on Windows due to some MachO changes. Could I have you take a look?

@dyung I don't have commit access, would you mind commiting this for me ?

• Fix formatting

• Converted to range-based for loop

@rnk Thanks for the suggestions, I've updated it. If it looks good on your end, yes, I would appreciate it if you could commit this for me, I don't have commit access.

@rnk Sorry for the delay, I moved the file path length handling accordingly

• moved handling of long paths to createGraphFilename()

@rnk

So, if I want clang to overwrite some file, all I have to do is pass clang -mllvm -view-dag-combine1-dags and I can create a file with a name from the input source? That seems dangerous.

Ping requesting review

• removed static declaration from header file
• cleaned up code according to clang-tidy specifications

I've attached a reproducible.zip folder. To reproduce this bug, unzip the folder and compile multiply.cpp to LLVM IR and then run

@RKSimon Thank you for the review, would you be able to commit this for me? I don't have commit access

Thank you for the review @spatel . I do not have commit access, could you possibly commit this for me?

@RKSimon I removed the unnecessary braces. Good catch

@spatel My mistake! I have run clang-format on the patch and the diff should be cleaned up

Ran clang-format on the patch

@rupprecht Thanks for the explanation, that makes perfect sense. This LGTM (but note that I'm fairly new to the binutils)

@rupprecht With most of the lit test adjustments here the only change was the ordering, which makes sense. Could you help me understand why the addition of the various [1-9] style regex expressions were also added to certain tests? Are those just ensuring more verbose matches with FileCheck?

@rupprecht A bit of a beginner to LLVM, but why add the suppression for DynamicRelocations when --disassemble is provided?

Thanks for the review all. I don't have commit access, @grimar could you commit this for me?

@grimar my apologies. I reuploaded the patch with more context according to the documentation you linked! Also, made the changes suggested via the comments above

Restored check-empty check as suggested by @ychen

Relevant BugZilla Bug: https://bugs.llvm.org/show_bug.cgi?id=41830

@thakis thanks for the review. I don't have commit access, would you mind committing this for me?

@stella.stamenova These are the changes from our previous discussion regarding the usage of %T

@stella.stamenova @thakis Thanks for the feedback all, I will go ahead and make this change to the way I'm representing directories, open a new diff for review, and add you as a reviewer.

@stella.stamenova thanks for the review. I don't have commit access, would you mind committing this for me?

@stella.stamenova Thanks for the input, what about now?

@stella.stamenova Good suggestion, I think you are right. I have updated the diff to use %t, that way any end-users can control the way their lit test work in a read-only mount by simply altering the test_exec_root in their config

@cfe-commits Re-pinging this group to request a review on this patch

@cfe-commits ping requesting a review on this patch