Are there any tests planned for this patch? We would have to commit XL binaries, or I think a better approach would be to write C code with a user defined eh fsm table.

LGTM.

Would like you to prune the new tests, then I think this will be ready to land.

I think it makes sense to split the STDC macro changes and TOS_AIX__ into different patches.

Please rework the test update into init-ppc.c

LGTM

Minor suggestion to keep and remap some of the diagnostic cases.

Reworked the prefixes and indentation.

Added ppc64le target.

added some commentary notes for the review.

Modified the tests to work at opt. Replaced initialization of return values with literals to simplify the IR.

Formatted the RUN steps.

I'll format the RUN steps as suggested but with respect to the other suggestions, I prefer to keep the test as-is.

I think the review title "ByVal formal argument support: passing on the stack" implies that it covers the reg/stack split case. I noted you described it better in the review description. I think a slightly more descriptive abstract would be useful for the commit message.

Added MIR expect output for tests that expect calls to memcpy to ensure the memcpy calls are generated independently from caller register setup.

Addressed unnecessary iterator check while looping on expect byval argument registers.

Rebased to account for changes committed in D76902.

My comments are only suggestions/discussion and can be addressed as you see fit on the commit.

I had intended to move the 5-8 byte byval callee cases out of aix64-cc-byval.ll and put them here. I don't want aix64-cc-byval.ll to exist when we're done. If the 5-8 byte tests aren't useful for the callee I'd rather delete them. They were of interest to the caller handling to make sure that we implement the left justification correctly.

Fixed expected output for 64-byte byval arg test.

Added a test for the homogeneous float stuct Sean suggested.

Tweaked the lambda capture on GetLoad.

Discussion ensued post-approval regarding use of CHECK-DAG on the ADJSTACKDOWN lines so I thought I'd post the final version before committing.

Clang format.

Forgot the 64 byte byval test.

I split out some of the suggested test improvements to https://reviews.llvm.org/D76875

Summary:
-Test tidy ups
-FYI that I noted a 32-bit 8 byte aligned by-val (struct S { double d; }) that will need special handling in future
-Requested some comments in the code / tests to explain the suboptimal codegen, which otherwise looks odd.

I still have to update the tests as per Sean's suggestion but I'll post the updated code for now.

Updated code to address comments.

Addressed Sean's comments:
-General test tidy up, remove an unnecessary assertion, tweaked a variable rename.

-Added skipping of 0 size by val args.
-Tidied up tests.

-Add CC_AIX logic to pass over 0 size by-val arguments.
-Tidy the by val tests added.
-Some expected output changes required in aix-cc-byval.ll related to use of ZEXTLOAD now that I've added the run step.

Added the missing run steps to aix-cc-byval.ll

-Use ZEXTLOAD rather than EXTLOAD. This changed expected output because we're able to use rotate left word immediate with mask form instructions.
-Split byval cc tests into new test files.

1. Split byval cc tests into a separate file.
2. Use ZEXTLOAD

One comment update. Should be ready to commit.

Some of the prior test changes have been committed as NFC patches. I've updated the diff and readers can now see the instruction changes for widening arguments to register width before writing to the stack.

I think we need some changes to ensure frame objects are created with their proper size and not word size. i64 is passed as two i32s but we should coalesce these into a single 8-byte object. Like-wise arguments smaller than word size are generating frame objects that are word size.

In D75126#1891669, @ZarkoCA wrote:

It would be better if the test cleanup was done in a separate patch to make it clearer what the expected output of the patch. But that said, it's not difficult to find added testcase and it would be ok with me to leave the modified test cases.

In D74015#1882933, @sfertile wrote:

Patch LGTM as far a formatting/naming/testing etc. C++ specifics is outside my wheelhouse though, so I can't confirm things like the tail padding rules are correct for AIX. Because of that I'm not comfortable being the one to accept the patch.

All minor comments except for the concern on whether or not a truncation node is needed when ValSize < LocSize.

From my perspective, the only issue holding this up is settling on the name. I'd like to hammer that out and get this committed.

I believe the current patch is sound and correct. I had one test suggestion and one suggestion to use a lambda to reuse some of the cr spill logic.

Code semantics look ok. If you can post a revision to address the minor concerns I need to have one more pass-through of the test updates in detail.