Commit Graph
10 Commits
Author SHA1 Message Date
dok.slade 99ffecd675 Merge JIT instructions when loading 32-bits immediate values :
PowerPC needs 2 instructions to load a 32-bits immediate value in a register (lis + addi; lis + ori). We can merge both instructions into a single SetImmediate().
This commit shares code with r5799, so to prevent another flame war on OSX vs Windows, I added #ifdef around the "bad" lines...


git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@5982 8ced0084-cf51-0410-be5f-012b33b47a6e
2010-07-26 23:06:34 +00:00
dok.slade d340b9497e Commit r5970 part #5: Removed unneeded MOV in branch instructions
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@5977 8ced0084-cf51-0410-be5f-012b33b47a6e
2010-07-25 19:10:50 +00:00
dok.slade 673f18837b Commit r5970 part #4: Merge some loops in PPCAnalyst::Flatten()
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@5976 8ced0084-cf51-0410-be5f-012b33b47a6e
2010-07-25 19:01:32 +00:00
dok.slade 16129f6f6d Commit r5970 part #3.1: Forgot to remove code unused since r5974
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@5975 8ced0084-cf51-0410-be5f-012b33b47a6e
2010-07-25 18:52:02 +00:00
dok.slade c7b4df19d7 Commit r5970 part #3: Inline the call to computeCR: it now costs a few more memory bytes per JITed instruction but it removes the CPU overhead of the CALL. This allowed to remove some unneeded MOV as well.
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@5974 8ced0084-cf51-0410-be5f-012b33b47a6e
2010-07-25 18:44:34 +00:00
dok.slade cba42fab1d Commit r5970 part #2: Jit64::GenerateCarry() don't need a temporary register anymore
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@5973 8ced0084-cf51-0410-be5f-012b33b47a6e
2010-07-25 18:14:48 +00:00
dok.slade 68c8f345d7 Commit r5970 part #1: Fix what seems to be a bug in PowerPC instruction RFI: the wrong bit was cleared in MSR
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@5972 8ced0084-cf51-0410-be5f-012b33b47a6e
2010-07-25 18:03:14 +00:00
dok.slade fa2cfcd089 Revert r5970 before committing again part by part (except modifications about CR union)
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@5971 8ced0084-cf51-0410-be5f-012b33b47a6e
2010-07-25 17:54:03 +00:00
dok.slade cbc66f9467 More performance work on the PowerPC JIT compiler:
* Merge some loops in PPCAnalyst::Flatten()
* Put ppcState.cr and ppcState.cr_fast[] into a single C++ union. This allows quick access to the whole CR register without needing to merge the cr_fast array. The implemented solution assumes the host system is little-endian, but it seems to be already assumed in many places in the code...
* Inline the call to computeCR: it now costs a few more memory bytes per JITed instruction but it removes the CPU overhead of the CALL. This allowed to remove some unneeded MOV as well.
* Jit64::GenerateCarry() don't need a temporary register anymore
* Fix what seems to be a bug in PowerPC instruction RFI: the wrong bit was cleared in MSR

git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@5970 8ced0084-cf51-0410-be5f-012b33b47a6e
2010-07-25 15:37:56 +00:00
dok.slade f8ce075ab1 Commit patch submitted in issue 2951 (PowerPC JIT optimizations)
git-svn-id: https://dolphin-emu.googlecode.com/svn/trunk@5954 8ced0084-cf51-0410-be5f-012b33b47a6e
2010-07-23 19:30:00 +00:00