Subject: 2.0BETA crash in uvm_pdaemon.c
To: None <>
From: Peter Bex <>
List: netbsd-users
Date: 09/27/2004 21:55:36
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

Hi list,

I've been having regular crashes on my 2.0BETA system, always in the same
place.  Now I finally built a gdb-able kernel file, which also crashed.
The backtrace is as follows:

#24 0xc0102c6b in calltrap ()
#25 0xc0267216 in uvmpd_scan () at ../../../../uvm/uvm_pdaemon.c:793
#26 0xc0266b40 in uvm_pageout (arg=3D0xc3790294) at ../../../../uvm/uvm_pda=
(gdb) l uvm_pdaemon.c:793
788              */
789    =20
790             UVMHIST_LOG(pdhist, "  starting 'free' loop",0,0,0,0);
791    =20
792             pages_freed =3D uvmexp.pdfreed;
793             uvmpd_scan_inactive(&uvm.page_inactive);
794             pages_freed =3D uvmexp.pdfreed - pages_freed;
795    =20
796             /*
797              * we have done the scan to get free pages.   now we work o=
n meeting
(gdb) select 25
(gdb) print pages_freed
$1 =3D -1069243456
(gdb) print uvmexp.pdfreed
$2 =3D 871505
(gdb) print uvm.page_inactive
$3 =3D {tqh_first =3D 0xc044a3c0, tqh_last =3D 0xc048a640}
(gdb) print &uvm.page_inactive
$4 =3D (struct pglist *) 0xc035bff4
(gbd) _

What does this mean?  It appears almost as if directly after an assignment
the assigned variables are not the same anymore!

I'm using CVS version 1.59 of uvm_pdaemon.c, this is on:
cpu0: Intel Pentium II (686-class), 350.82 MHz, id 0x652

I hope someone can help me fix this problem.  I can put the kernel and
coredump files online if necessary.

"The process of preparing programs for a digital computer
 is especially attractive, not only because it can be economically
 and scientifically rewarding, but also because it can be an aesthetic
 experience much like composing poetry or music."
							-- Donald Knuth

Content-Type: application/pgp-signature
Content-Disposition: inline

Version: GnuPG v1.2.4 (NetBSD)