Date: Thu, 11 Apr 2002 13:26:59 -0400 (EDT) From: John Baldwin <jhb@FreeBSD.org> To: alpha@FreeBSD.org Subject: Hmm, corrupted pte panic? Message-ID: <XFMail.20020411132659.jhb@FreeBSD.org>
index | next in thread | raw e-mail
I was doing a buildworld -j 8 on the DS20 using a test kernel with some smpng
patches (the patches in question are at
www.freebsd.org/~jhb/patches/smpng.patch, there is nothing in there that goes
near pmap or VM though AFAICT). I got this weird panic during the buildworld:
FreeBSD/alpha (sema.baldwin.cx) (ttyd0)
login: panic: pmap_emulate_reference(0xfffffe001e789238, 0x1600551b0, 1, 0): pa
0x0 not managed
cpuid = 0; panic
Stopped at Debugger+0x34: zapnot v0,#0xf,a0 <v0=0x0,a0=0x6>
db> t
Debugger() at Debugger+0x34
panic() at panic+0x188
pmap_emulate_reference() at pmap_emulate_reference+0x154
trap() at trap+0x49c
XentMM() at XentMM+0x2c
--- memory management fault (from ipl 0) ---
--- user mode ---
db>
For this panic, the relevant code is:
pa = pmap_pte_pa(pte);
KASSERT((*pte & PG_MANAGED) != 0,
("pmap_emulate_reference(%p, 0x%lx, %d, %d): pa 0x%lx not managed",
curthread, v, user, write, pa));
Now, since we now from above that FOR and FOE were set (it was a FOR or FOE
trap that caused the pmap_emulate_reference()) and that PG_MANAGED and the PFN
were 0. I'm wondering if the pte got trashed to 0 out from under me?
Hmm, I wish I could get a cross-built i386 gdb to analyze alpha kernels. :-/
Is there anything in binutils that analyzes stabs so I could try to take that
curthread pointer, locate curthread->td_proc->p_vmspace->vm_map.pmap and then
start looking at the pte's to see which ones (if any) are 0?
--
John Baldwin <jhb@FreeBSD.org> <>< http://www.FreeBSD.org/~jhb/
"Power Users Use the Power to Serve!" - http://www.FreeBSD.org/
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-alpha" in the body of the message
help
Want to link to this message? Use this
URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?XFMail.20020411132659.jhb>