cvs commit: ports/x11/nvidia-driver Makefile distinfo pkg-plist ports/x11/nvidia-driver/files 6113-patch-src::nvidia_pci.c 6113-patch-src::nvidia_subr.c pkg-message.in

[Alexey Dokuchaev]

Hello,

Thanks you for detailed report and traces.  I'll look into it.

./danfe

On Thu, Dec 07, 2006 at 01:55:02AM +0100, C?dric Jonas wrote:
> On Wed, 6 Dec 2006 21:46:44 +0000 (UTC)
> Alexey Dokuchaev <danfe at FreeBSD.org> wrote:
> 
> > danfe       2006-12-06 21:46:44 UTC
> > 
> >   FreeBSD ports repository
> > 
> >   Modified files:
> >     x11/nvidia-driver    Makefile distinfo pkg-plist 
> >     x11/nvidia-driver/files pkg-message.in 
> >   Removed files:
> >     x11/nvidia-driver/files 6113-patch-src::nvidia_pci.c 
> >                             6113-patch-src::nvidia_subr.c 
> >   Log:
> >   - Update to most recent version from nVidia, 9631
> >   - As threatened, drop support for FreeBSD 4.X
> >   - Apply minor fix for master sites [*]
> >   - Add last padding line for pkg-message for prettier output
> >   
> >   Tested on:      FreeBSD 6.2-PRERELEASE (i386)
> >   
> >   [*] alepulver@ will take care of master sites in separate commit
> >   
> >   Revision  Changes    Path
> >   1.50      +9 -43     ports/x11/nvidia-driver/Makefile
> >   1.15      +3 -6      ports/x11/nvidia-driver/distinfo
> >   1.2       +0 -12
> > ports/x11/nvidia-driver/files/6113-patch-src::nvidia_pci.c (dead)
> > 1.2       +0 -18
> > ports/x11/nvidia-driver/files/6113-patch-src::nvidia_subr.c (dead)
> > 1.3       +1 -0      ports/x11/nvidia-driver/files/pkg-message.in
> > 1.13      +3 -27     ports/x11/nvidia-driver/pkg-plist
> > _______________________________________________ cvs-ports at freebsd.org
> > mailing list http://lists.freebsd.org/mailman/listinfo/cvs-ports To
> > unsubscribe, send any mail to "cvs-ports-unsubscribe at freebsd.org"
> > 
> > !DSPAM:45774135311535244610024!
> > 
> 
> I got also a panic under 7.0-CURRENT:
> 
> Fatal trap 12: page fault while in kernel mode
> fault virtual address   = 0x6
> fault code              = supervisor write, page not present
> instruction pointer     = 0x20:0xc056a981
> stack pointer           = 0x28:0xe6c4ca4c
> frame pointer           = 0x28:0xe6c4ca60
> code segment            = base 0x0, limit 0xfffff, type 0x1b
>                         = DPL 0, pres 1, def32 1, gran 1
> processor eflags        = interrupt enabled, resume, IOPL = 0
> current process         = 3534 (nvidia-settings)
> trap number             = 12
> panic: page fault
> 
> It appears under many situations: by closing nvidia-settings, by
> closing the WM, etc.
> 
> (kgdb) list *0xc056a981
> 0xc056a981 is in destroy_devl (/usr/src/sys/kern/kern_conf.c:700).
> 695             dev->si_drv2 = 0;
> 696             bzero(&dev->__si_u, sizeof(dev->__si_u));
> 697
> 698             if (!(dev->si_flags & SI_ALIAS)) {
> 699                     /* Remove from cdevsw list */
> 700                     LIST_REMOVE(dev, si_list);
> 701
> 702                     /* If cdevsw has no more struct cdev *'s, clean
> it */ 703                     if (LIST_EMPTY(&csw->d_devs))
> 704                             fini_cdevsw(csw);
> 
> This is the same for me, in each case.
> 
> (kgdb) backtrace
> #0  doadump () at pcpu.h:166
> #1  0xc0593350 in boot (howto=260)
> at /usr/src/sys/kern/kern_shutdown.c:411 #2  0xc05935b0 in panic
> (fmt=0xc0758392 "%s") at /usr/src/sys/kern/kern_shutdown.c:567
> #3  0xc072c49c in trap_fatal (frame=0xe6c4ca0c, eva=6)
>     at /usr/src/sys/i386/i386/trap.c:869
> #4  0xc072c203 in trap_pfault (frame=0xe6c4ca0c, usermode=0, eva=6)
>     at /usr/src/sys/i386/i386/trap.c:778
> #5  0xc072be41 in trap (frame=
>       {tf_fs = 8, tf_es = 40, tf_ds = 40, tf_edi = -995848192, tf_esi =
> 1717986919, tf_ebp = -423310752, tf_isp = -423310792, tf_ebx =
> -995848192, tf_edx = 6, tf_ecx = 0, tf_eax = 0, tf_trapno = 12, tf_err
> = 2, tf_eip = -1068062335, tf_cs = 32, tf_eflags = 590406, tf_esp =
> 600, tf_ss = 400}) at /usr/src/sys/i386/i386/trap.c:463 #6  0xc071cfca
> in calltrap () at /usr/src/sys/i386/i386/exception.s:138 #7  0xc056a981
> in destroy_devl (dev=0xc4a49000) at /usr/src/sys/kern/kern_conf.c:700
> #8  0xc056a9ec in destroy_dev (dev=0xc4a49000)
> at /usr/src/sys/kern/kern_conf.c:721 #9  0xc0c7c868 in ?? () #10
> 0xc4a49000 in ?? () #11 0xc50a4180 in ?? ()
> #12 0xc0f3f340 in ?? ()
> #13 0x00000000 in ?? ()
> #14 0xc4a49000 in ?? ()
> #15 0xc5174754 in ?? ()
> #16 0xc5174754 in ?? ()
> #17 0xe6c4caac in ?? ()
> #18 0xc0569ea3 in giant_close (dev=0xc50a4180, fflag=3, devtype=8192, 
>     td=0xc4a808c0) at /usr/src/sys/kern/kern_conf.c:284
> Previous frame identical to this frame (corrupt stack?)
> 
> PS: I know, the nvidia driver is certified to run only with 6.x
> Releases, but this don't seem to be a nvidia driver issue...?!
> 
> Regards,
> 
> -- 
> C?dric Jonas                                        cedric at decemplex.net
> 
> GPG ID:                                                         30CCFE8D
> GPG Key:                 http://box.decemplex.net/~cedric/cedric.key.asc
> GPG Fingerprint:      CF03 E1FD 9428 1B6B E971  B107 9044 AA99 30CC FE8D
> 
> Jabber-ID:                                          cedric at decemplex.net