spec2000/186 passes now -- thanks.

Hi @aeubanks, we turned up a corner case where jump threading is creating a self-referential GEP:

%tmp27 = getelementptr inbounds i8, i8* %tmp27, i64 1

Thank you for the quick response! Much appreciated.

@dmgreen, we found that DSE is removing a non dead store in a loop with multiple backedges. File "dse-double-loop.ll" attached. The 2nd store postdominates the 1st one, but it doesn't run on every iteration, so the 1st one is still needed. We worked around it by checking numBackEdges > 1 in addition to the irreducibility check. Would you mind taking a look? I hope bugpoint did not over-reduce the test case; it's derived from "real" code in spec2000 186.crafty.

Thanks for the review comments! Could I please ask someone for help making the commit? I don't have write access.

fixed case in variable name

Our internal test passes---thank you for the fix.

Ah, I missed your comment. I'll pull your fix and retest.

Return true to trigger the simplifier.

Thanks for the consensus. New patch uploaded.

Good clarification: InstCombine vs. InstSimplify.
Merged and reduced test, full diff included.
If the patch is OK, could I ask you one more favor: to commit it? I do not have commit access to the project.

Updated to run InstCombine before the vector transformations. The insert-binop-with-constant test needed a few changes. InstructionSimplify checks for constant divisors with 0/undef elements and undef's the entire result.

If running IC is cheap, that's a good solution. I will put up another patch. Thanks.
The IR looks strange, but it's unfortunately showing up in a few real test cases in our LLVM branch, after we've made some mods to the unroller and vectorizer.

FWIW, dropping the attribute LGTM. The heap/stack promotion is still a work in progress, is that correct? At least, I don't see it happening in trunk.

Restoring latest patch.

@mkazantsev, would you mind taking one more look at this patch? One significant fix was made since your last review, to handle duplicate edges from SwitchInsts. The patch has now been rebased to latest (which hasn't changed it much). Thanks!

Did not include the reverted part of the code (the important part :) ) in the new patch. Added separate testcase for switch-backedge. Thanks for the fix suggestions.

Keep the InLoopPredecessors vector in the original order.

Missed the sort call -- thanks.

Remove duplicate blocks from the list of exit predecessors. This happens when the predecessor inst is a switch with duplicate targets. Avoids block double-processing and maintains the metadata consistency assertion.

Sorry for the trouble. I'll figure out what happened.

Thank you for the thorough review. I have one more request to the reviewers on this list: could someone give a hand to commit the code? I don't have write access to the repo yet.

Previous patch was null, this is the right one.

Rewrote comments, fixed formatting, simplified test and added checks for specific metadata

One more favor to ask: Could I please have some help committing this fix? I'm new to the project and don't have write access yet...

Thanks! I've made the suggested test change.