Implement a sorted TLB lookup table (big improvement over the linear table scan)
Optimize out the 1C000000 -> FC000000 check at the end of the lookup functions
src/sh4/mmu.c
Optimize out the 1C000000 -> FC000000 check at the end of the lookup functions
src/sh4/mmu.c
Fix (extremely boneheaded) failure to convert pc to physical address before
storing in the translation cache (in other words, the translation cache was
effectively disabled for MMU code). MMU code is now about 3 times faster...
src/sh4/sh4trans.c
storing in the translation cache (in other words, the translation cache was
effectively disabled for MMU code). MMU code is now about 3 times faster...
src/sh4/sh4trans.c
Remove FASTCALL from mem_copy_*, not really helping atm (and sometimes hurting)
src/mem.h
src/sh4/sh4mem.c
src/mem.h
src/sh4/sh4mem.c
Declare mem_copy_* functions as FASTCALL
Split sh4_flush_store_queue into TLB/non-TLB versions, and optimize
slightly based on that
src/mem.h
src/sh4/mmu.c
src/sh4/sh4core.h
src/sh4/sh4mem.c
src/sh4/sh4x86.in
src/test/testsh4x86.c
Split sh4_flush_store_queue into TLB/non-TLB versions, and optimize
slightly based on that
src/mem.h
src/sh4/mmu.c
src/sh4/sh4core.h
src/sh4/sh4mem.c
src/sh4/sh4x86.in
src/test/testsh4x86.c
Add shortcut test for long writes to the store queue (far and away the most popular P4 write)
src/sh4/sh4mem.c
src/sh4/sh4mem.c
Add --enable-profiled configure option for convenience (and enable fastcall only on fully optimized builds)
config.h.in
configure
configure.in
src/lxdream.h
config.h.in
configure
configure.in
src/lxdream.h
Fix x86-64 build (typos et al)
Remove Push/pop ebx - don't really need it and saves adding more target-specific asm
src/sh4/ia64abi.h
src/sh4/sh4x86.in
Remove Push/pop ebx - don't really need it and saves adding more target-specific asm
src/sh4/ia64abi.h
src/sh4/sh4x86.in
.