Fix off-by-one in reading cache address from memory
Commit a9ac7c153
properly avoided writing past the end of cache_line,
but in doing so introduced an error as to the first memory address that
should be written. This fix avoids writing past the cache_line by simply
reading the previous memory location.
Fixes the CP/M 68K simulation example and issue #181
This commit is contained in:
parent
f0f3176c78
commit
af713b78e1
1 changed files with 1 additions and 1 deletions
|
@ -269,7 +269,7 @@ static t_stat ReadICache(t_addr tpc)
|
|||
t_stat rc;
|
||||
uint8* mem;
|
||||
|
||||
ASSERT_OKRET(Mem((tpc+CACHE_SIZE)&addrmask,&mem));
|
||||
ASSERT_OKRET(Mem((tpc+CACHE_SIZE-1)&addrmask,&mem));
|
||||
|
||||
/* 68000/08/10 do not like unaligned access */
|
||||
if (cputype < 3 && (tpc & 1)) return STOP_ERRADR;
|
||||
|
|
Loading…
Add table
Reference in a new issue