The new patch for this bug is in https://reviews.llvm.org/D60811.

Hi @nemanjai,
I saw that this part of the code was introduced by you in commit - https://reviews.llvm.org/D26023
Can you please tell what this piece of code is doing and if at all the assert statement needs to be revisited?

Filed bug at :- https://bugs.llvm.org/show_bug.cgi?id=41177

Can anyone confirm if this has been pushed upstream in the master branch? I pulled latest code for Apple Swift master, and did not see this change in swift-llvm repository.

@jsji any inputs on how to produce a reduced testcase that can show the problem.

@lebedev.ri, no, as explained in above comments. Need some help in resolving this issue.

@jsji can you please let me know how to produce a reduced testcase that can show the problem? Sorry, but I am quite new to this, any help would be greatly helpful.

@dblaikie, anyone you know who can help here?

Sorry for the delayed response.
Yes you are right, the submitted change would not effect the behavior of the program. I think I misread / misinterpreted the condition earlier, took it to be e = ++i, not realising its a simple for loop increment.
Also the error that's being thrown up might be intermittent, since when I made this change from ++i to i++, the build seemed to have passed, though now it fails again.

@compnerd Has this been pushed upstream? I pulled latest code for Apple Swift master, and did not see this change in swift-llvm repository. Please confirm.

Hi, Can someone please review this? Thanks!

Hi, can someone point me the right repo to submit these changes?

Any updates on this one?

Hi @labath , Should I be submitting this patch request directly to "https://github.com/apple/swift-lldb"?

@compnerd - Please let me know once this is committed. Thanks.

@compnerd - I got a confirmation that we can go ahead and merge this. Please go ahead and commit this on my behalf.

@compnerd - I will check this and let you know. In the meanwhile can you please point towards the new license documentation/details?

Sure @compnerd, please go ahead and commit this on my behalf like last time. Thanks in advance!

Any updates on this?

In https://reviews.llvm.org/D55503, we had extended the condition to include PPC.
However, the check required to build LLVM on PPC64LE was missed in SVN r348970.

Extending the condition to include PPC

@nemanjai, any updates on this one?

@nemanjai, please review and let me know if this looks good.

@nemanjai Created a new diff - https://reviews.llvm.org/D56382
Please review.

Hello @nemanjai Any updates on verification of sanitizers on PPC64LE with this PR?

@nemanjai Thanks, please do let me know in case you need anything from my side.

Hello @nemanjai @compnerd, please let me know how to proceed on this further.

@nemanjai, thanks for clarifying! Yes, and separating out PPC64LE from the platform check helps build and execute Swift on PPC64LE. Please suggest.

The Swift 5.0 toolchain on PowerPC64LE also works fine if we remove the "defined(ppc)" check altogether in the original code i.e keep only "!(defined(NetBSD) || defined(OpenBSD)))".
Would that affect other PPC platforms (32/64 and BE/LE PPC other than PowerPC64LE)?

Hi, The changes committed as a part of commit r348970 are different from the original patch which I submitted and which also fixed build issues while building Swift 5.0 toolchain (lldb component). The new patch committed still breaks the swift toolchain build as before:-

Update for PowerPC64LE

@compnerd, thanks for the explanation and accepting this change revision. Yes, I would need someone to commit this on my behalf, as I might not have the rights to do a commit. Please go ahead and do it for me if you can. Thanks in advance!

As per the comments in the code:-

Hi Reviewers,
Can you please review this? This change is resolving a blocker while building swift toolchain on PowerPC64LE and is a bit urgent. Since this is a small change, can you please take this up? Thanks!

Please review the suggested changes and approve/merge if this looks good. Thanks!

Incorporating review comments.

Yes, the swift toolchain successfully gets built with this change.

Previous review (for the swift-llvm GitHub repo): https://github.com/apple/swift-llvm/pull/129

Earlier diff created was using the command "git show HEAD -U999999 > mypatch.patch" as mentioned in "https://llvm.org/docs/Phabricator.html#phabricator-request-review-web".
The updated patch is created using git diff and reflects the correct diff of the changes.

Please review the attached change as this is a blocker for our development tasks.