5f0d6f1b24
exactly twice by me) which causes the code at the top of copyin which word-aligns the destination pointer to copy more bytes than necessary, resulting in an alignment fault later on. The bug is difficult to reproduce, but as far as I can tell it seems to be as a result of the condition codes being corrupted, possibly following a page fault caused by the first ldrbt/strb instructions. The subsequent ldr<cc>bt/str<cc>b instructions then _always_ execute. I can't think of any obvious reason why this would happen, though. |
||
---|---|---|
.. | ||
arm | ||
arm32 | ||
conf | ||
footbridge | ||
fpe-arm | ||
include | ||
iomd | ||
ixp12x0 | ||
mainbus | ||
ofw | ||
s3c2xx0 | ||
sa11x0 | ||
xscale | ||
Makefile |