[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]
Re: Another illegal instruction awk problem
>>> (gdb) x/i 0x80196183
>>> 0x80196183 <emodd+120>: movd r9,r0
>> emodd - I think that's one of the often-emulated inestructions,
>> meaning ragge is almost certainly right.
> So then... how should I go about figuring out why the emulated
> instruction didn't get emulated?
Oh, it's not that it didn't get emulated. It was in the process of
being emulated when the fault occurred; that's what the PC inside the
kernel's emodd code indicates.
However, I have trouble seeing how "movd r9,r0" could fault showing
SIGILL. While I don't have the -5 source at ready hand, the 4.0.1
source in question has only one "movd r9,r0" instruction, and its
bsbw getval_byte # get multiplier extension operand
movzbl %r0, -(%sp)
movd %r9, %r0
none of which look the least bit SIGILLy to me.
There is clearly something I don't understand going on here. Possibly
this is a 5.x-vs-4.x difference, or possibly it's something worse.
0x80196183 is nowhere near a page boundary, so it's not some weirdoid
page fault mishandling glitch; besides, the page in question must be
wired and mapped into all processes. I think I've quite possibly
contributed about all the value I can to this discussion, so I think I
should bow out at this point.
/~\ The ASCII Mouse
\ / Ribbon Campaign
X Against HTML mouse%rodents-montreal.org@localhost
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B
Main Index |
Thread Index |