fabi/Xenia-Canary
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

alternative mutex impl on windows works but i really can't tell if it helps much. use larger size in deferred_command_list to cut down on resizes in big scenes on m:dur

Browse Source
This commit is contained in:
chss95cs@gmail.com
2022-08-14 10:26:50 -07:00
parent a037bdb2e8
commit c9b2d10e17
5 changed files with 58 additions and 7 deletions
Download Patch File Download Diff File Expand all files Collapse all files

15
src/xenia/gpu/command_processor.cc
View File

@@ -493,13 +493,18 @@ void CommandProcessor::WriteRegister(uint32_t index, uint32_t value) {
// very unlikely. these ORS here are meant to be bitwise ors, so that we do
// not do branching evaluation of the conditions (we will almost always take
// all of the branches)
if (XE_UNLIKELY(
(index - XE_GPU_REG_SCRATCH_REG0 < 8) |
(index == XE_GPU_REG_COHER_STATUS_HOST) |
((index - XE_GPU_REG_DC_LUT_RW_INDEX) <=
(XE_GPU_REG_DC_LUT_30_COLOR - XE_GPU_REG_DC_LUT_RW_INDEX)))) {
unsigned expr = (index - XE_GPU_REG_SCRATCH_REG0 < 8) |
(index == XE_GPU_REG_COHER_STATUS_HOST) |
((index - XE_GPU_REG_DC_LUT_RW_INDEX) <=
(XE_GPU_REG_DC_LUT_30_COLOR - XE_GPU_REG_DC_LUT_RW_INDEX));
//chrispy: reordered for msvc branch probability (assumes if is taken and else is not)
if (XE_LIKELY(expr == 0)) {
} else {
HandleSpecialRegisterWrite(index, value);
}
}
void CommandProcessor::MakeCoherent() {