bjorn3 opened PR #2158 from x64_fixes
to main
:
I am not sure how to implement
sshr
andushr
for small integers.Fixes #2121
bjorn3 updated PR #2158 from x64_fixes
to main
:
I am not sure how to implement
sshr
andushr
for small integers.Fixes #2121
bjorn3 updated PR #2158 from x64_fixes
to main
:
I am not sure how to implement
sshr
andushr
for small integers.Fixes #2121
bjorn3 updated PR #2158 from x64_fixes
to main
:
I am not sure how to implement
sshr
andushr
for small integers.Fixes #2121
bjorn3 updated PR #2158 from x64_fixes
to main
:
I am not sure how to implement
sshr
andushr
for small integers.Fixes #2121
bjorn3 updated PR #2158 from x64_fixes
to main
:
I am not sure how to implement
sshr
andushr
for small integers.Fixes #2121
bjorn3 updated PR #2158 from x64_fixes
to main
:
I am not sure how to implement
sshr
andushr
for small integers.Fixes #2121
bnjbvr requested bnjbvr for a review on PR #2158.
MaxGraey submitted PR Review.
MaxGraey created PR Review Comment:
Is it sign extention from small ints to
i64
, right? Perhaps I'm missing something, that it should look like this I guess:I8 => (((value as i64) << 56) >> 56) as u64, I16 => (((value as i64) << 48) >> 48) as u64, I32 => (((value as i64) << 32) >> 32) as u64,
bjorn3 submitted PR Review.
bjorn3 created PR Review Comment:
Indeed. Good catch! Thanks!
bjorn3 updated PR #2158 from x64_fixes
to main
:
I am not sure how to implement
sshr
andushr
for small integers.Fixes #2121
bnjbvr submitted PR Review.
bnjbvr submitted PR Review.
bnjbvr created PR Review Comment:
We should reuse the
SignExtendRaxRdx
instruction, because this new instruction you added and the former have the same semantics: sign-extend the "data" (originally on x86,rax
is the "data" (Accumulator, hence A) register up to 32 bits, and when more is required, therdx
contains the rest of the Data (hence D in name)). Maybe renameSignExtendRaxRdx
toSignExtendData
, and make the size anOption
to indicate al->ah? (or use 0 for al->ah sign-extension?) Or we could have aSignExtendOpSize
enum that contains all the valid sizes for this instruction?
bnjbvr created PR Review Comment:
nit: I think
bitwise not
would be more colloquial here
bnjbvr created PR Review Comment:
nit: can you please add a bit more details in the unreachable message:
"unhandled output type for shift/rotates: {}"
?
bjorn3 updated PR #2158 from x64_fixes
to main
:
I am not sure how to implement
sshr
andushr
for small integers.Fixes #2121
bjorn3 updated PR #2158 from x64_fixes
to main
:
I am not sure how to implement
sshr
andushr
for small integers.Fixes #2121
bjorn3 updated PR #2158 from x64_fixes
to main
:
I am not sure how to implement
sshr
andushr
for small integers.Fixes #2121
bjorn3 updated PR #2158 from x64_fixes
to main
:
I am not sure how to implement
sshr
andushr
for small integers.Fixes #2121
bnjbvr merged PR #2158.
Last updated: Dec 23 2024 at 13:07 UTC