Stream: git-wasmtime

Topic: wasmtime / Issue #1943 In correct moves for x64 FLT in ne...

Wasmtime GitHub notifications bot (Jun 29 2020 at 23:39):

jlb6740 opened Issue #1943:

@bnjbvr @julian-seward1

Hi, I believe the recent patch https://github.com/bytecodealliance/wasmtime/pull/1915 introduce different behavior with register allocation. Before the patch a floating point add lowered to this:

Disassembly of 10 bytes:
0: 55 push rbp
1: 48 89 e5 mov rbp, rsp
4: f3 0f 58 c1 addss xmm0, xmm1
8: 5d pop rbp
9: c3 ret

But afterwards it lowered to this:

Disassembly of 18 bytes:
0: 55 push rbp
1: 48 89 e5 mov rbp, rsp
4: f3 0f 10 c8 movss xmm1, xmm0
8: f3 0f 58 c8 addss xmm1, xmm0
c: f3 0f 10 c1 movss xmm0, xmm1
10: 5d pop rbp
11: c3 ret

It is not clear to me what change in this patch introduces these moves nor how to fix them. Any suggestions?

Wasmtime GitHub notifications bot (Jun 30 2020 at 10:01):

bnjbvr commented on Issue #1943:

Weird, Inst::is_move should identify the movss instructions as moves, unless we missed something when rejiggering the float moves.

Wasmtime GitHub notifications bot (Jun 30 2020 at 10:09):

bjorn3 commented on Issue #1943:

asm 4: f3 0f 10 c8 movss xmm1, xmm0 8: f3 0f 58 c8 addss xmm1, xmm0 c: f3 0f 10 c1 movss xmm0, xmm1

This is equivalent to xmm0 = xmm0 + xmm0, not xmm0 = xmm0 + xmm1 like it was before.

Wasmtime GitHub notifications bot (Jun 30 2020 at 10:22):

bjorn3 commented on Issue #1943:

I think I found the problem: https://github.com/bytecodealliance/wasmtime/pull/1915/files#r447576280

Wasmtime GitHub notifications bot (Jun 30 2020 at 10:31):

julian-seward1 commented on Issue #1943:

Yeah, I can believe that explanation. It will have messed up the live ranges, which will have (correctly) defeated the RA's coalescer, hence the extra moves, and the incorrect resulting computation.

Last updated: Nov 22 2024 at 16:03 UTC