@paulwalker-arm @sdesmalen I will be picking up this work, I will study this patch and see if I have any questions. Thanks!

Updated last comments.

@dmgreen I fixed the issues that you pointed out. This patch prevents the test case from getting stuck on the exponential explosion of the algorithm. Could you review it? Thanks!

Fixed formatting and add test run line

@dmgreen I think so. The code is indeed exponential. Adding this check will avoid the exponential case, but we will be visiting each instruction both ways, we just avoid visiting repeated ones. IMO it won't affect the analysis result.

@dmgreen @fhahn Sorry about that, the test case I posted was incorrect. I updated the test.c file, and in the process of finding the correct test case for you to be able to reproduce I found out the bug is worse than I expected.

Are there any opinions on this bugfix?

@efriedma Thanks Eli, I will close it, as that other patch already fixed that issue.

Hello David, I have this patch here that I was going to upstream (https://reviews.llvm.org/D80746), but I just realized that it is exactly the same as this one.

In D75197#1894971, @c-rhodes wrote:

Thanks for the patch @dancgr but these intrinsics have already been implemented in D74912. @sdesmalen shared a spreadsheet [1] with the mailing list to coordinate the intrinsics upstreaming, it's probably not as useful now given they're almost all implemented but you might find it useful if you haven't seen it already.

[1] https://docs.google.com/spreadsheets/d/1leug3a0eHhi_bWtKJOR0MeZHvfQLoIPtP-um7AWafRM

Latest commit adds a change to the failing MC tests to be exactly the same as the MC tests for eor3 and bcax.

Will look into that and update this review with the necessary changes.

Merging all instructions into same multiclass.

In D73746#1851678, @kmclaughlin wrote:

Thanks for these changes, @dancgr! I just have a couple of comments on the tests in sve2-mla-unpredicated.ll

Add missing i16 tests.

@sdesmalen Hopefully all comments from D73576 were addressed with this patch.

I will make the changes suggested by Sander in a following patch, joined with the saturating multiply-add long intirnsics.

Remove _m from intrinsics, as it is the default behaviour. Also, add unpredicated patterns for SVE2 smulh, umulh.

I'm not sure on some parts, but I have prepared a major update for this patch that I hope will fix most of @sdesmalen concerns.

Tentative changes to address some comments from the reviewers.

I'm not sure on some parts, but I have prepared a major update for this patch that I hope will fix most of @sdesmalen concerns.

In D72799#1822788, @efriedma wrote:

In D72799#1822599, @dancgr wrote:

In D72799#1822588, @efriedma wrote:

Can we also also add mul patterns for targets that have SVE, but not SVE2?

That instruction is restricted to SVE2. Do we have unpredicated vector mul instructions for SVE targets as well?

We can synthesize a predicate using ptrue. We do this in a few places currently; for example, to implement bswap.

In D72799#1822588, @efriedma wrote:

Can we also also add mul patterns for targets that have SVE, but not SVE2?

Fix whitespace errors and add a couple negative tests for invalid input.

In D72654#1818201, @efriedma wrote:

Do we have any negative tests, for immediates that are outside the range of the sve mul immediate instruction?