Re: Kernel crash on Asus A7N8X-X
On Tuesday 04 March 2008 05:59:59 pm Frédéric PRACA wrote: Hello dear hackers, I own a Asus A7N8X-X motherboard (NForce2 chipset) with a Radeon 9600 video card. After upgrading from 6.3 to 7.0, I launched xorg which crashed the kernel. After looking in the kernel core dump, I found that the agp_nvidia_flush_tlb function of /usr/src/sys/pci/agp_nvidia.c crashed on the line 377. The loop fails from the beginning (when i==0). I commented out the two last loops and it seems to work now but as I didn't understand what is this code for, I'd like to have some explanation about it and want to know if someone got the same problem. The Linux AGP driver has the same code. It appears to be forcing a read of the TLB registers to force prior writes to clear the TLB entries to flush perhaps? I'm not sure why you are getting a panic. What kind of fault did you get? (The original kernel panic messages would be needed.) -- John Baldwin ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Kernel crash on Asus A7N8X-X
On Thursday 06 March 2008 08:31:26 am John Baldwin wrote: On Tuesday 04 March 2008 05:59:59 pm Frédéric PRACA wrote: Hello dear hackers, I own a Asus A7N8X-X motherboard (NForce2 chipset) with a Radeon 9600 video card. After upgrading from 6.3 to 7.0, I launched xorg which crashed the kernel. After looking in the kernel core dump, I found that the agp_nvidia_flush_tlb function of /usr/src/sys/pci/agp_nvidia.c crashed on the line 377. The loop fails from the beginning (when i==0). I commented out the two last loops and it seems to work now but as I didn't understand what is this code for, I'd like to have some explanation about it and want to know if someone got the same problem. The Linux AGP driver has the same code. It appears to be forcing a read of the TLB registers to force prior writes to clear the TLB entries to flush perhaps? I'm not sure why you are getting a panic. What kind of fault did you get? (The original kernel panic messages would be needed.) Actually, it looks like you have a 64MB aperture and with either a 32MB or 64MB aperture this loop runs off the end of the GATT (GATT has 16384 entries * 4 bytes == 64k == 16 pages on x86) so if it dies before it starts the next loop that might explain it. The patch below makes it walk the full GATT reading the first word from each page to force a flush w/o walking off the end of the GATT. Actually, this is what appears to have happened: (gdb) set $start = 0xd4d05000 (ag_virtual) (gdb) set $fva = 3570491392(eva in trap_pfault() frame) (gdb) p ($fva - $start) / 4 $2 = 17408 That's well over your current ag_entries of 16384. Try this patch (note Linux's in-kernel agp driver has the same bug): Index: agp_nvidia.c === RCS file: /host/cvs/usr/cvs/src/sys/dev/agp/agp_nvidia.c,v retrieving revision 1.13 diff -u -r1.13 agp_nvidia.c --- agp_nvidia.c12 Nov 2007 21:51:37 - 1.13 +++ agp_nvidia.c6 Mar 2008 13:37:43 - @@ -347,7 +347,7 @@ struct agp_nvidia_softc *sc; u_int32_t wbc_reg, temp; volatile u_int32_t *ag_virtual; - int i; + int i, pages; sc = (struct agp_nvidia_softc *)device_get_softc(dev); @@ -373,9 +373,10 @@ ag_virtual = (volatile u_int32_t *)sc-gatt-ag_virtual; /* Flush TLB entries. */ - for(i = 0; i 32 + 1; i++) + pages = sc-gatt-ag_entries * sizeof(u_int32_t) / PAGE_SIZE; + for(i = 0; i pages; i++) temp = ag_virtual[i * PAGE_SIZE / sizeof(u_int32_t)]; - for(i = 0; i 32 + 1; i++) + for(i = 0; i pages; i++) temp = ag_virtual[i * PAGE_SIZE / sizeof(u_int32_t)]; return (0); -- John Baldwin ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Kernel crash on Asus A7N8X-X
Selon John Baldwin [EMAIL PROTECTED]: On Thursday 06 March 2008 08:31:26 am John Baldwin wrote: The Linux AGP driver has the same code. It appears to be forcing a read of the TLB registers to force prior writes to clear the TLB entries to flush perhaps? I'm not sure why you are getting a panic. What kind of fault did you get? (The original kernel panic messages would be needed.) Actually, it looks like you have a 64MB aperture and with either a 32MB or 64MB aperture this loop runs off the end of the GATT (GATT has 16384 entries * 4 bytes == 64k == 16 pages on x86) so if it dies before it starts the next loop that might explain it. The patch below makes it walk the full GATT reading the first word from each page to force a flush w/o walking off the end of the GATT. Actually, this is what appears to have happened: (gdb) set $start = 0xd4d05000 (ag_virtual) (gdb) set $fva = 3570491392(eva in trap_pfault() frame) (gdb) p ($fva - $start) / 4 $2 = 17408 That's well over your current ag_entries of 16384. Try this patch (note Linux's in-kernel agp driver has the same bug): Index: agp_nvidia.c === RCS file: /host/cvs/usr/cvs/src/sys/dev/agp/agp_nvidia.c,v retrieving revision 1.13 diff -u -r1.13 agp_nvidia.c --- agp_nvidia.c 12 Nov 2007 21:51:37 - 1.13 +++ agp_nvidia.c 6 Mar 2008 13:37:43 - @@ -347,7 +347,7 @@ struct agp_nvidia_softc *sc; u_int32_t wbc_reg, temp; volatile u_int32_t *ag_virtual; - int i; + int i, pages; sc = (struct agp_nvidia_softc *)device_get_softc(dev); @@ -373,9 +373,10 @@ ag_virtual = (volatile u_int32_t *)sc-gatt-ag_virtual; /* Flush TLB entries. */ - for(i = 0; i 32 + 1; i++) + pages = sc-gatt-ag_entries * sizeof(u_int32_t) / PAGE_SIZE; + for(i = 0; i pages; i++) temp = ag_virtual[i * PAGE_SIZE / sizeof(u_int32_t)]; - for(i = 0; i 32 + 1; i++) + for(i = 0; i pages; i++) temp = ag_virtual[i * PAGE_SIZE / sizeof(u_int32_t)]; return (0); -- John Baldwin Thanks a lot John, this code works. I have been able to launch X w/o crashing the kernel. Fred ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Kernel crash on Asus A7N8X-X
Frédéric PRACA wrote: Hello dear hackers, I own a Asus A7N8X-X motherboard (NForce2 chipset) with a Radeon 9600 video card. After upgrading from 6.3 to 7.0, I launched xorg which crashed the kernel. After looking in the kernel core dump, I found that the agp_nvidia_flush_tlb function of /usr/src/sys/pci/agp_nvidia.c crashed on the line 377. The loop fails from the beginning (when i==0). I commented out the two last loops and it seems to work now but as I didn't understand what is this code for, I'd like to have some explanation about it and want to know if someone got the same problem. Usually it's a good idea to show the data that led to your conclusions (backtraces, etc), not just your conclusions. Sometimes there is more going on than is immediately apparent. Krs ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Kernel crash on Asus A7N8X-X
Selon Kris Kennaway [EMAIL PROTECTED]:
Fr�d�ric PRACA wrote:
Hello dear hackers,
I own a Asus A7N8X-X motherboard (NForce2 chipset) with a Radeon 9600 video
card. After upgrading from 6.3 to 7.0, I launched xorg which crashed the
kernel.
After looking in the kernel core dump, I found that the
agp_nvidia_flush_tlb
function of /usr/src/sys/pci/agp_nvidia.c crashed on the line 377. The loop
fails from the beginning (when i==0). I commented out the two last loops
and it
seems to work now but as I didn't understand what is this code for, I'd
like to
have some explanation about it and want to know if someone got the same
problem.
Usually it's a good idea to show the data that led to your conclusions
(backtraces, etc), not just your conclusions. Sometimes there is more
going on than is immediately apparent.
For sure, sorry.
Here's what I got from kgdb :
[GDB will not be able to debug user-mode threads: /usr/lib/libthread_db.so:
Undefined symbol ps_pglobal_lookup]
GNU gdb 6.1.1 [FreeBSD]
Copyright 2004 Free Software Foundation, Inc.
GDB is free software, covered by the GNU General Public License, and you are
welcome to change it and/or distribute copies of it under certain conditions.
Type show copying to see the conditions.
There is absolutely no warranty for GDB. Type show warranty for details.
This GDB was configured as i386-marcel-freebsd.
There is no member named pathname.
(kgdb) bt
#0 doadump () at pcpu.h:195
#1 0xc05b49ac in boot (howto=260) at /usr/src/sys/kern/kern_shutdown.c:409
#2 0xc05b4b74 in panic (fmt=Variable fmt is not available.
) at /usr/src/sys/kern/kern_shutdown.c:563
#3 0xc06f9858 in vm_fault (map=0xc1054000, vaddr=3570491392,
fault_type=Variable fault_type is not available.
)
at /usr/src/sys/vm/vm_fault.c:275
#4 0xc0765be5 in trap_pfault (frame=0xd61f1a48, usermode=0, eva=3570491392)
at /usr/src/sys/i386/i386/trap.c:801
#5 0xc0766502 in trap (frame=0xd61f1a48) at /usr/src/sys/i386/i386/trap.c:490
#6 0xc07545bb in calltrap () at /usr/src/sys/i386/i386/exception.s:139
#7 0xc0776ed7 in agp_nvidia_flush_tlb (dev=0xc2a3b100, offset=-1029891720)
at /usr/src/sys/pci/agp_nvidia.c:377
#8 0xc06bb9dd in agp_generic_bind_memory (dev=0xc2a3b100, mem=0xc2fe7cc0,
offset=0) at agp_if.h:76
#9 0xc06bb001 in agp_bind_memory (dev=0xc2a3b100, handle=0xc2fe7cc0,
offset=0) at agp_if.h:128
#10 0xc04bd1bf in drm_agp_bind_memory (handle=0xc2fe7cc0, start=0)
at /usr/src/sys/dev/drm/drm_agpsupport.c:456
#11 0xc04bd4dd in drm_agp_bind (dev=0xc2aad000, request=0xd61f1b68)
at /usr/src/sys/dev/drm/drm_agpsupport.c:331
#12 0xc04bd591 in drm_agp_bind_ioctl (kdev=0xc2ac3400, cmd=2148033590,
data=0xc2c309d0 �|��, flags=67, p=0xc2ed7a50, filp=0x17497)
at /usr/src/sys/dev/drm/drm_agpsupport.c:348
#13 0xc04c25e8 in drm_ioctl (kdev=0xc2ac3400, cmd=2148033590,
data=0xc2c309d0 �|��, flags=67, p=0xc2ed7a50)
---Type return to continue, or q return to quit---
at /usr/src/sys/dev/drm/drm_drv.c:911
#14 0xc0585ad6 in giant_ioctl (dev=0xc2ac3400, cmd=2148033590,
data=0xc2c309d0 �|��, fflag=67, td=0xc2ed7a50)
at /usr/src/sys/kern/kern_conf.c:349
#15 0xc0552517 in devfs_ioctl_f (fp=0xc2c3fd38, com=2148033590,
data=0xc2c309d0, cred=0xc4e62400, td=0xc2ed7a50)
at /usr/src/sys/fs/devfs/devfs_vnops.c:494
#16 0xc05e9b43 in kern_ioctl (td=0xc2ed7a50, fd=8, com=2148033590,
data=0xc2c309d0 �|��) at file.h:266
#17 0xc05e9ca4 in ioctl (td=0xc2ed7a50, uap=0xd61f1cfc)
at /usr/src/sys/kern/sys_generic.c:570
#18 0xc0765f0a in syscall (frame=0xd61f1d38)
at /usr/src/sys/i386/i386/trap.c:1035
#19 0xc0754620 in Xint0x80_syscall ()
at /usr/src/sys/i386/i386/exception.s:196
#20 0x0033 in ?? ()
Previous frame inner to this frame (corrupt stack?)
(kgdb) frame 7
#7 0xc0776ed7 in agp_nvidia_flush_tlb (dev=0xc2a3b100, offset=-1029891720)
at /usr/src/sys/pci/agp_nvidia.c:377
warning: Source file is more recent than executable.
377 temp = ag_virtual[i * PAGE_SIZE / sizeof(u_int32_t)];
(kgdb) list
372
373 ag_virtual = (volatile u_int32_t *)sc-gatt-ag_virtual;
374
375 /* Flush TLB entries. */
376 /*for(i = 0; i 32 + 1; i++)
377 temp = ag_virtual[i * PAGE_SIZE / sizeof(u_int32_t)];
378 for(i = 0; i 32 + 1; i++)
379 temp = ag_virtual[i * PAGE_SIZE / sizeof(u_int32_t)];
380 */
381 return (0);
(kgdb) print i
$1 = 0
(kgdb) print *ag_virtual
$2 = 299405313
(kgdb) print ag_virtual
$3 = (volatile u_int32_t *) 0xd4d05000
(kgdb) print *sc
$4 = {agp = {as_aperture = 0xc2a34dc0, as_aperture_rid = 16,
as_maxmem = 461373440, as_allocated = 8388608,
as_state = AGP_ACQUIRE_KERNEL, as_memory = {tqh_first = 0xc2fe7cc0,
tqh_last = 0xc2fe7cc0}, as_nextid = 2, as_isopen = 0,
as_devnode = 0xc2a3c300, as_lock = {lock_object = {
lo_name = 0xc07d15d1 agp lock, lo_type = 0xc07d15d1 agp lock,
lo_flags = 16973824,
Re: Kernel crash on Asus A7N8X-X
On Tue, Mar 04, 2008 at 11:59:59PM +0100, Frédéric PRACA wrote: Hello dear hackers, I own a Asus A7N8X-X motherboard (NForce2 chipset) with a Radeon 9600 video card. After upgrading from 6.3 to 7.0, I launched xorg which crashed the kernel. After looking in the kernel core dump, I found that the agp_nvidia_flush_tlb function of /usr/src/sys/pci/agp_nvidia.c crashed on the line 377. The loop fails from the beginning (when i==0). I commented out the two last loops and it seems to work now but as I didn't understand what is this code for, I'd like to have some explanation about it and want to know if someone got the same problem. I'm in no way familiar with X. That said: you're using an ATI Radeon card, yet the kernel crashed in agp_nvidia.c. nVidia != ATI. Is it possible that you changed video cards at one point, and you're still using the nVidia AGP driver (loaded via /boot/loader.conf)? dmesg might be useful here. -- | Jeremy Chadwickjdc at parodius.com | | Parodius Networking http://www.parodius.com/ | | UNIX Systems Administrator Mountain View, CA, USA | | Making life hard for others since 1977. PGP: 4BD6C0CB | ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Kernel crash on Asus A7N8X-X
Selon Jeremy Chadwick [EMAIL PROTECTED]: On Tue, Mar 04, 2008 at 11:59:59PM +0100, Frédéric PRACA wrote: Hello dear hackers, I own a Asus A7N8X-X motherboard (NForce2 chipset) with a Radeon 9600 video card. After upgrading from 6.3 to 7.0, I launched xorg which crashed the kernel. After looking in the kernel core dump, I found that the agp_nvidia_flush_tlb function of /usr/src/sys/pci/agp_nvidia.c crashed on the line 377. The loop fails from the beginning (when i==0). I commented out the two last loops and it seems to work now but as I didn't understand what is this code for, I'd like to have some explanation about it and want to know if someone got the same problem. I'm in no way familiar with X. That said: you're using an ATI Radeon card, yet the kernel crashed in agp_nvidia.c. nVidia != ATI. Is it possible that you changed video cards at one point, and you're still using the nVidia AGP driver (loaded via /boot/loader.conf)? No, in fact, agp_nvidia.c is the driver for the AGP bus of the NForce2 motherboard chipset, not the video card. dmesg might be useful here. Why not but I can't copy it for the moment. -- | Jeremy Chadwickjdc at parodius.com | | Parodius Networking http://www.parodius.com/ | | UNIX Systems Administrator Mountain View, CA, USA | | Making life hard for others since 1977. PGP: 4BD6C0CB | ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to [EMAIL PROTECTED]
Kernel crash on Asus A7N8X-X
Hello dear hackers, I own a Asus A7N8X-X motherboard (NForce2 chipset) with a Radeon 9600 video card. After upgrading from 6.3 to 7.0, I launched xorg which crashed the kernel. After looking in the kernel core dump, I found that the agp_nvidia_flush_tlb function of /usr/src/sys/pci/agp_nvidia.c crashed on the line 377. The loop fails from the beginning (when i==0). I commented out the two last loops and it seems to work now but as I didn't understand what is this code for, I'd like to have some explanation about it and want to know if someone got the same problem. Fred ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to [EMAIL PROTECTED]
