Should we only be doing this when CET protections are enabled?

LGTM

LGTM

Update more check lines to fully check the clobbers

LGTM

In D80485#2284943, @dmgreen wrote:

@craig.topper @amyk @nemanjai Hello!
Do we want to keep the isMulhCheaperThanMulShift hook, or change it to a isOperationLegalOrCustom call, like most other transforms would use? If the hook is useful on some targets I can override it under MVE simply enough, let me know.

In D87884#2287015, @yubing wrote:

In D87884#2286958, @craig.topper wrote:

In D87884#2286918, @yubing wrote:

With avx512f, the cost SK_Select(v32i16 or v64i8) shoulde be 3(vmovdqa64 + vpternlogq)

The moves probably don't really count since they can be eliminated during register renaming. So only the vpternlog executes.

Eh, Craig, why it has relationship with register renaming? I thought, vternlog's third operand should be provided by a vmovdqa64.
Besides, we can observe the following asm for v32i16's SK_Select:

vmovdqa64       .LCPI0_0(%rip), %zmm0   # zmm0 = [0,0,0,65535,65535,65535,65535,65535,65535,65535,65535,65535,65535,65535,65535,65535,65535,65535,65535,65535,65535,65535,65535,65535,65535,65535,65535,65535,65535,65535,65535,65535]
vpternlogq      $202, 144(%rbp), %zmm4, %zmm0

In D87884#2286918, @yubing wrote:

With avx512f, the cost SK_Select(v32i16 or v64i8) shoulde be 3(vmovdqa64 + vpternlogq)

In D87888#2285878, @rnk wrote:

Honestly, I forget exactly what the memory clobber does beyond the "sideeffect" marker. I would expect LLVM to model these just as external function calls that could read or write memory that is passed to them.

@rnk should this inline asm have a memory clobber on it?

We can do several optimizations for PDEP using computeKnownBits and SimplifyDemandedBits

In D87888#2281360, @RKSimon wrote:

Do we have sufficient backend test coverage for this?

Check my logic

LGTM

ping

Ping

LGTM

Add mir test cases to show change to memory operand

Longer term should we have a variadic version of this like builtin_shufflevector or builtin_convertvector that can handle any reduction?

LGTM

Use isGuaranteedToTransferExecutionToSuccessor. Need to be put together negative tests.

Ensure that the div/rem are the start of their basic blocks so they are guaranteed to execute.

In D87555#2269381, @nikic wrote:

Doesn't this need some guaranteed-to-transfer checks for all instructions we're moving across?

This was committed in c195ae2f003261f2c25f569b07ae556dee57f17d

In D87555#2269355, @lebedev.ri wrote:

Can we use DominatorTree::findNearestCommonDominator()?
If not, please add a negative test, where it's unsafe to do so.