Retro/pcem
Watch
1
0
Fork
You've already forked pcem
 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Rework self modifying code handling.

Browse source 
Code blocks now have a varying level of granularity in the page and dirty masks.
Most blocks have 64-byte granularity, but blocks that are repeatedly modified
drop to 1-byte granularity. This reduces the maximum size of the affected block
significantly, but reduces recompilation due to data located too close to code.

If the block is still marked as dirty, then it is recompiled without any
immediate instruction parameters being baked into the recompiled code, instead
being fetched from the RAM array as needed. This severely reduces recompilation
rates on some SMC-heavy games, eg Duke Nukem 3D, System Shock, Screamer etc,
giving a major speedup.
This commit is contained in:
SarahW 2019-03-16 14:53:26 +00:00
commit ab124776a0
36 changed files with 1719 additions and 233 deletions
Download patch file Download diff file Expand all files Collapse all files

26
src/codegen_ops_helpers.c Normal file
View file

@ -0,0 +1,26 @@
#include "ibm.h"
#include "x86.h"
#include "386_common.h"
#include "codegen.h"
#include "codegen_ir_defs.h"
#include "codegen_reg.h"
#include "codegen_ops_helpers.h"
void LOAD_IMMEDIATE_FROM_RAM_16_unaligned(codeblock_t *block, ir_data_t *ir, int dest_reg, uint32_t addr)
{
/*Word access that crosses two pages. Perform reads from both pages, shift and combine*/
uop_MOVZX_REG_PTR_8(ir, IREG_temp3_W, get_ram_ptr(addr));
uop_MOVZX_REG_PTR_8(ir, dest_reg, get_ram_ptr(addr+1));
uop_SHL_IMM(ir, IREG_temp3_W, IREG_temp3_W, 8);
uop_OR(ir, dest_reg, dest_reg, IREG_temp3_W);
}
void LOAD_IMMEDIATE_FROM_RAM_32_unaligned(codeblock_t *block, ir_data_t *ir, int dest_reg, uint32_t addr)
{
/*Dword access that crosses two pages. Perform reads from both pages, shift and combine*/
uop_MOV_REG_PTR(ir, dest_reg, get_ram_ptr(addr & ~3));
uop_MOV_REG_PTR(ir, IREG_temp3, get_ram_ptr((addr + 4) & ~3));
uop_SHR_IMM(ir, dest_reg, dest_reg, (addr & 3) * 8);
uop_SHL_IMM(ir, IREG_temp3, IREG_temp3, (4 - (addr & 3)) * 8);
uop_OR(ir, dest_reg, dest_reg, IREG_temp3);
}