Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

FEXCore: Fallback to the memcpy slow path for overlaps within 32 bytes #3501

Merged
merged 3 commits into from
Mar 21, 2024
Merged

FEXCore: Fallback to the memcpy slow path for overlaps within 32 bytes #3501

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
94fecb9
FEXCore: Remove needless alignment checks for the mem{cpy,set} fastpath
bylaws Mar 20, 2024
d490cb1
FEXCore: Fallback to the memcpy slow path for overlaps within 32 bytes
bylaws Mar 20, 2024
12fb26f
Update InstCountCI
bylaws Mar 20, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
15 changes: 7 additions & 8 deletions FEXCore/Source/Interface/Core/JIT/Arm64/MemoryOps.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1814,10 +1814,6 @@ DEF_OP(MemSet) {
ARMEmitter::ForwardLabel AgainInternal128Exit{};
ARMEmitter::BackwardLabel AgainInternal128{};

// Fallback to byte by byte loop if not 4 byte aligned
and_(ARMEmitter::Size::i64Bit, TMP4, TMP2, 0x3);
cbnz(ARMEmitter::Size::i64Bit, TMP4, &AgainInternal);

if (Direction == -1) {
sub(ARMEmitter::Size::i64Bit, TMP2, TMP2, 32 - Size);
}
Expand Down Expand Up @@ -2127,15 +2123,18 @@ DEF_OP(MemCpy) {
cbz(ARMEmitter::Size::i64Bit, TMP1, &DoneInternal);

if (!Op->IsAtomic) {
ARMEmitter::ForwardLabel AbsPos{};
ARMEmitter::ForwardLabel AgainInternal256Exit{};
ARMEmitter::ForwardLabel AgainInternal128Exit{};
ARMEmitter::BackwardLabel AgainInternal128{};
ARMEmitter::BackwardLabel AgainInternal256{};

// Fallback to byte by byte loop if either of start/end are not 4 byte aligned
orr(ARMEmitter::Size::i64Bit, TMP4, TMP2, TMP3);
and_(ARMEmitter::Size::i64Bit, TMP4, TMP4, 0x3);
cbnz(ARMEmitter::Size::i64Bit, TMP4, &AgainInternal);
sub(ARMEmitter::Size::i64Bit, TMP4, TMP2, TMP3);
tbz(TMP4, 63, &AbsPos);
neg(ARMEmitter::Size::i64Bit, TMP4, TMP4);
Bind(&AbsPos);
sub(ARMEmitter::Size::i64Bit, TMP4, TMP4, 32);
tbnz(TMP4, 63, &AgainInternal);

if (Direction == -1) {
sub(ARMEmitter::Size::i64Bit, TMP2, TMP2, 32 - Size);
Expand Down
Loading
Loading