T0b1/wasmtime
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

x64: Mask shift amounts for small types (#4752)

Browse Source 
* x64: Mask shift amounts for small types

* cranelift: Disable i128 shifts in fuzzer again

They are fixed. But we had a bunch of fuzzgen issues come in, and we don't want to accidentaly mark them as fixed

* cranelift: Avoid masking shifts for 32 and 64 bit cases

* cranelift: Add const shift tests and fix them

* cranelift: Remove const `rotl` cases

Now that `put_masked_in_imm8_gpr` works properly we can simplify rotl/rotr
This commit is contained in:
Afonso Bordado
2022-08-24 18:31:38 +01:00
committed by GitHub
parent 9cb987c678
commit d394edcefe
10 changed files with 1920 additions and 117 deletions
Download Patch File Download Diff File Expand all files Collapse all files

10
cranelift/codegen/src/isa/x64/lower.isle
View File

@@ -793,11 +793,6 @@
(rule (lower (has_type (fits_in_64 ty) (rotl src amt)))
(x64_rotl ty src (put_masked_in_imm8_gpr amt ty)))
(rule (lower (has_type (fits_in_64 ty)
(rotl src (u64_from_iconst amt))))
(x64_rotl ty src
(const_to_type_masked_imm8 amt ty)))
;; `i128`.
@@ -819,11 +814,6 @@
(rule (lower (has_type (fits_in_64 ty) (rotr src amt)))
(x64_rotr ty src (put_masked_in_imm8_gpr amt ty)))
(rule (lower (has_type (fits_in_64 ty)
(rotr src (u64_from_iconst amt))))
(x64_rotr ty src
(const_to_type_masked_imm8 amt ty)))
;; `i128`.