Re: freezing and rebooting with vidcontrol -m on
Okay, answering my own question for the sake of the archive, I believe it was an on combination of some of the options SC_something that vidcontrol -m on and my mouse and kernel didnt like. I was using some to save on kernel size and memory used but then decided to start using the computer on the console again so I just commented all those SC feature disabling out and all seems well. Adam wrote: Hi, I don't really know where to go with this issue on my own anymore so im tossing it to the list for suggestions. I had a celeron running in a Abit BX6-2 system running -current approx a month or 1.5 old. Today I moved the disks to a Abit BP6 Motherboard and recompiled the kernel for smp and the other hardware changes I intended to make. Well, when I boot it up it would get down to approx "starting i386 whatever" and the boot would stop, the keyboard wouldnt type (although numlock would work for x amount of time) and you couldnt drop into ddb. Sometimes it would stick in this braindead mode, sometimes if you tried typing things to get it unstuck it would just reboot after a brief total keyboard lockup. No panic, just a reboot. I booted singleuser and cleaned up my rc.conf, disabling some noncritical things, and got the system booting fine. I have a usb logitech mouse by the way. Later I noticed no mouse cursor on my consoles even though I had moused running, so I ran vidcontrol -m on. WHAM. I had found the culprit. Okay so I know what caused it and what not to run ;0 I took the opportunity to cvsup and upgrade the whole system to -current as of today, and to my dismay it still does the same thing. I really wish I could drop into ddb when it happens but it wont let me. If serial might be a better option I could probably dig out a null modem cable... Suggestions Please! To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: freezing and rebooting with vidcontrol -m on
Okay, answering my own question for the sake of the archive, I believe it was an on combination of some of the options SC_something that vidcontrol -m on and my mouse and kernel didnt like. I was using some to save on kernel size and memory used but then decided to start using the computer on the console again so I just commented all those SC feature disabling out and all seems well. Thank you for reporting. Would you tell me which SC_XXX options you used to have in your kernel configuration file? Kazu To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: freezing and rebooting with vidcontrol -m on
I dont recall exactly :( SC_NO_FONT_LOADING was one of them, I may have commented one of the others in the old conf trying to fix it. I have a spare comp with a freebsd install I could play around with it on and see if I can recreate it instead of goofing around with my server that had the problem, fsck'ing 78 gigs is not pretty ;} (Yes I eventually stopped mounting the big drives while testing but I still dont like torturing / and /usr hehe) On Sat, 10 Jun 2000, Kazutaka YOKOTA wrote: Okay, answering my own question for the sake of the archive, I believe it was an on combination of some of the options SC_something that vidcontrol -m on and my mouse and kernel didnt like. I was using some to save on kernel size and memory used but then decided to start using the computer on the console again so I just commented all those SC feature disabling out and all seems well. Thank you for reporting. Would you tell me which SC_XXX options you used to have in your kernel configuration file? Kazu To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: freezing and rebooting with vidcontrol -m on
On Sat, 10 Jun 2000, Matthew Thyer wrote: Are you using power management (presumably so you can poweroff your ATX system) ? You must not use APM with an SMP kernel currently. Remove the lines in your /etc/rc.conf that read: apm_enable="YES" apmd_enable="YES" (Or change them to ="NO"). Or you can update your sources because this problem has just been fixed (but there are suspend/resume problems with SMP still). Been there, did that, suspend even works. Waiting on lusers to logout so I can reboot and try a new kernel, I think i might have been using some console options it didnt like. Adam wrote: Hi, I don't really know where to go with this issue on my own anymore so im tossing it to the list for suggestions. I had a celeron running in a Abit BX6-2 system running -current approx a month or 1.5 old. Today I moved the disks to a Abit BP6 Motherboard and recompiled the kernel for smp and the other hardware changes I intended to make. Well, when I boot it up it would get down to approx "starting i386 whatever" and the boot would stop, the keyboard wouldnt type (although numlock would work for x amount of time) and you couldnt drop into ddb. Sometimes it would stick in this braindead mode, sometimes if you tried typing things to get it unstuck it would just reboot after a brief total keyboard lockup. No panic, just a reboot. I booted singleuser and cleaned up my rc.conf, disabling some noncritical things, and got the system booting fine. I have a usb logitech mouse by the way. Later I noticed no mouse cursor on my consoles even though I had moused running, so I ran vidcontrol -m on. WHAM. I had found the culprit. Okay so I know what caused it and what not to run ;0 I took the opportunity to cvsup and upgrade the whole system to -current as of today, and to my dismay it still does the same thing. I really wish I could drop into ddb when it happens but it wont let me. If serial might be a better option I could probably dig out a null modem cable... Suggestions Please! To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: freezing
On Fri, 18 Feb 2000, Adam wrote: I do not mean to throw blame on freebsd at all (unless tons of people also see this) but I had a freeze earlier today. I had several windows open, netscape, and was compiling qt2 with not much disk activity at all when it just froze hard as a rock in all ways I can think of. If it happens again I'll look into ddc and serial console or perhaps recreating on normal console. I dont suspect hardware fault but I wouldn't dismiss it. On Thu, 17 Feb 2000, Joao Pedras wrote: Hello all While making -j 4 buildworld and moving a netscape window, everything frozen. Happens often if do other things while cpu and disk are very active. Happens quite often. Anyone else has noticed this ? I have had a number of freezes, during a rm -rf and a find at the same time, during a cvs checkout, during a fsck on reboot. The fsck caused a message about an "SCB Timeout" I think. I have returned to slightly older 4.0-Current kernel and was going to build a new kernel. I can try to get more information this evening, if anybody is interested. For me the trigger seems to be heavy disk activity. configuration summary: K6-2 333, 80MB, aha2940(I think, adaptec PCI ultra), 4 2GB segate SCSI drives. Brian Beattie| The only problem with [EMAIL PROTECTED] | winning the rat race ... www.aracnet.com/~beattie | in the end you're still a rat To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: freezing
eheh I mean dead frozen, like if I was watching at a screeshot of a X session ]:) Alfred Perlstein wrote: Do you mean dead frozen, as in needs a reboot? or frozen for a second or so? The first one I haven't seen recently, the second I have noticed. -Alfred ^\ /^ O O o00-(_)-00o-- Common sense is the collection of prejudices acquired by age eighteen. -- Albert Einstein - PGP key available upon request or may be cut at http://pedras.webvolution.net/pgpkey.html To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: freezing
* Joao Pedras [EMAIL PROTECTED] [000216 17:50] wrote: Hello all While making -j 4 buildworld and moving a netscape window, everything frozen. Happens often if do other things while cpu and disk are very active. Happens quite often. Anyone else has noticed this ? Do you mean dead frozen, as in needs a reboot? or frozen for a second or so? The first one I haven't seen recently, the second I have noticed. -Alfred To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: freezing
* Joao Pedras [EMAIL PROTECTED] [000216 18:07] wrote: eheh I mean dead frozen, like if I was watching at a screeshot of a X session ]:) Alfred Perlstein wrote: Do you mean dead frozen, as in needs a reboot? or frozen for a second or so? The first one I haven't seen recently, the second I have noticed. -Alfred oy! well i guess you need to grab a serial console and see if you can get a crashdump/traceback. let us know if you find out, also check your machine's cooling devices to make sure they are functioning normally. good luck, -Alfred To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: freezing
Hi, On 17 Feb, Joao Pedras wrote: While making -j 4 buildworld and moving a netscape window, everything frozen. Happens often if do other things while cpu and disk are very active. Happens quite often. Anyone else has noticed this ? I had that problem three times: Once, when I tried using a K6-3 on my Gigabyte GA5-AX Motherboard. They just weren't working together, when I changed back to a K6-2 (both tried at 350 MHz) the problems disappeared. The K6-3 works fine now in another machine (MSI Mainboard or something like that). Another time something like that happened, when my cooler failed. On one of my old Suns I had a similar problem, when one of the SIMMs was damaged. So, get on and check 'ya hardware. Bye, Oliver To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: freezing
On Thu, 17 Feb 2000, Joao Pedras wrote: eheh I mean dead frozen, like if I was watching at a screeshot of a X session ]:) I used to get this on a game, which was trying somehow to load a font that wasn't correct. I found it wasn't really frozen, though, because you *could* telnet in (very slow). Are you *really* sure it's frozen? Do you have a network connection you can test telnet or ping with, or maybe a serial terminal you could hook to it? Alfred Perlstein wrote: Do you mean dead frozen, as in needs a reboot? or frozen for a second or so? The first one I haven't seen recently, the second I have noticed. -Alfred ^\ /^ O O o00-(_)-00o-- Common sense is the collection of prejudices acquired by age eighteen. -- Albert Einstein - PGP key available upon request or may be cut at http://pedras.webvolution.net/pgpkey.html To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message Chuck Robey| Interests include C Java programming, FreeBSD, [EMAIL PROTECTED] | electronics, communications, and signal processing. New Year's Resolution: I will not sphroxify gullible people into looking up fictitious words in the dictionary. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: freezing...
On Mon, Jan 10, 2000 at 02:21:56PM -0800, Alfred Perlstein wrote: * Matthew Dillon [EMAIL PROTECTED] [000110 12:11] wrote: I've reproduced the softupdates ftruncate panic and have a core dump to play with. I don't have a panic, but rather many processes start to get stuck in "getblk", unfortunatly they all don't stem from a common codepath, although the first process that wedged appears to have come through indir_trunc: I just experienced the same thing here on my CURRENT system with Kirk's latest SoftUpdates patches included. i have right now a buildworld stuck in "getblk" on rm ... and a "getblk" hang for Mozilla Tinderbox Build on gcc. The Mozilla Tinderbox build "hang" however seems to be reproducible. I have had two consecutive "hangs" at the same spot: checking for gtk-config... /usr/X11R6/bin/gtk-config which results in a stuck gcc process. I will try to build a kernel with kdb support as soon as my buildworld finishes (if it will finish) and see if I can provide a proper backtrace. -- Pascal Hofstee - [EMAIL PROTECTED] -BEGIN GEEK CODE BLOCK- Version: 3.1 GCS d- s+: a-- C++ UB P+ L- E--- W- N+ o? K- w--- O? M V? PS+ PE Y-- PGP-- t+ 5 X-- R tv+ b+ DI D- G e* h+ r- y+ --END GEEK CODE BLOCK-- To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: freezing...
FWIW, a couple weeks ago, I saw the same problem over NFS when autoconf tried to test for gtk-config... It would pretty consistantly hang... robert. Pascal Hofstee writes: On Mon, Jan 10, 2000 at 02:21:56PM -0800, Alfred Perlstein wrote: * Matthew Dillon [EMAIL PROTECTED] [000110 12:11] wrote: I've reproduced the softupdates ftruncate panic and have a core dump to play with. I don't have a panic, but rather many processes start to get stuck in "getblk", unfortunatly they all don't stem from a common codepath, although the first process that wedged appears to have come through indir_trunc: I just experienced the same thing here on my CURRENT system with Kirk's latest SoftUpdates patches included. i have right now a buildworld stuck in "getblk" on rm ... and a "getblk" hang for Mozilla Tinderbox Build on gcc. The Mozilla Tinderbox build "hang" however seems to be reproducible. I have had two consecutive "hangs" at the same spot: checking for gtk-config... /usr/X11R6/bin/gtk-config which results in a stuck gcc process. I will try to build a kernel with kdb support as soon as my buildworld finishes (if it will finish) and see if I can provide a proper backtrace. -- Pascal Hofstee - [EMAIL PROTECTED] -BEGIN GEEK CODE BLOCK- Version: 3.1 GCS d- s+: a-- C++ UB P+ L- E--- W- N+ o? K- w--- O? M V? PS+ PE Y-- PGP-- t+ 5 X-- R tv+ b+ DI D- G e* h+ r- y+ --END GEEK CODE BLOCK-- To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: freezing...
On Tue, 11 Jan 2000, Pascal Hofstee wrote: I just experienced the same thing here on my CURRENT system with Kirk's latest SoftUpdates patches included. i have right now a buildworld stuck in "getblk" on rm ... and a "getblk" hang for Mozilla Tinderbox Build on gcc. The Mozilla Tinderbox build "hang" however seems to be reproducible. I have had two consecutive "hangs" at the same spot: checking for gtk-config... /usr/X11R6/bin/gtk-config which results in a stuck gcc process. I will try to build a kernel with kdb support as soon as my buildworld finishes (if it will finish) and see if I can provide a proper backtrace. I think this goes away if you manually specify the path to gtk-config. I remember I had this problem some weeks ago while trying to build M12 on -CURRENT. I'm not saying you should need to do this, and it is likely a compiler issue that is causing it, but I did get it to work without modifying code. Later.. Doug To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: freezing...
Hello again No, my problem is not ahc related. This single processor system doesn't have any scsi It's a : PPro200/64MbRAM/8.4QuantumIDE/IntelVS440FX/Pioneer12x(which is not detected after upgraded to -CURRENT)/Digital 21240 (de0)/S3 Virge What I have in common with Christian is that it hangs during high usage, as I mentioned in my first post (e.g. compiling something, buildworld). I performed a buildworld this night with yesterday's code (in a common console) and it worked ok. I ONLY noticed this when having high CPU usage (e.g. compiling mysql port) when in using X. I am not saying this is X related. Common point here seems to be a busy system Joao Quoting "Dave J. Boers" [EMAIL PROTECTED]: On Mon, Jan 10, 2000 at 04:10:37AM +0100, Christian Carstensen wrote: this runs stable for 3 hours now... try a current kernel (checked out 4 hours ago), using a version of the file /usr/src/sys/pci/ahc_pci.c from 2000/01/06. ok, maybe it's not by any means stable, but works better than everything else within the last 24 hours. Actually, I've booted a kernel from just before newyear (28 december) which works _reasonably_ fine (although it's the same kernel that gave me the lockup earlier) with a userland of today. Problem is that the lockups (I think) are ahc-related and my SCSI hard drive did refuse to come online on one or two occasions while booting the system cold... I therefore concluded that it might be a problem with the hardware. Now (with the new kernel) I find the scsi system unstable and I have doubts again. One piece of information might also be useful in this context. After the system lockup I sort of benchmarked the scsi performance by doing "dd if=/dev/zero of=./testfile bs=100 count=128" (actually I varied the blocksize) and got a _very poor_ performance of only 4 Mb/sec (which is usually around 10 to 12 Mb/sec). I isolated my drive to be the only scsi device and I even clocked the SCSI bus down from 20 Mhz to 8 Mhz, but to no avail. In the mean time, I disconnected my scsi hard drive and I am running from my IDE disk. --- Powered by Webvolution Networks To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: freezing...
On Mon, 10 Jan 2000, Joao Pedras wrote: What I have in common with Christian is that it hangs during high usage, as I mentioned in my first post (e.g. compiling something, buildworld). this is funny: the system operates well, even when on heavy load (especially disk load), until i want a little more 8). to become more precisely, a cvs checkout or make world is no problem, if - and that's really interesting - nothing else requests the system's attention. i've had some successfully reproducable crashes when generating much i/o usage (cvs, buildworld), which in fact caused no problem, and then simply starting pine. at the moment, pine opened the mail folder, all my noisy hard disks stopped and it was perfectly quiet apart from cpu fans. -- Christian To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: freezing...
On Mon, Jan 10, 2000 at 03:35:47PM +0100, Christian Carstensen wrote: this is funny: the system operates well, even when on heavy load (especially disk load), until i want a little more 8). to become more precisely, a cvs checkout or make world is no problem, if - and that's really interesting - nothing else requests the system's attention. i've had some successfully reproducable crashes when generating much i/o usage (cvs, buildworld), which in fact caused no problem, and then simply starting pine. at the moment, pine opened the mail folder, all my noisy hard disks stopped and it was perfectly quiet apart from cpu fans. That sounds exactly like the very dead state I found my system in when it hang (I only have one occurence, so far, but my system isn't under very heavy load lately). Make -j 9 buildworld completes fine. The hang seems to have occured during a simple 20 Mb transfer of email data from my ide to my scsi disk (which is a very short but i/o intensive operation). There might have been an incoming email at the same time. I will try an idiotic buildworld -j 30 this evening just to see wether I can hang the system reproducibly. Have to go now, however. Regards, Dave Boers. -- Fatal error: replace user and try again. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: freezing...
In message [EMAIL PROTECTED] , Christian Carstensen writes: On Mon, 10 Jan 2000, Joao Pedras wrote: What I have in common with Christian is that it hangs during high usage, as I mentioned in my first post (e.g. compiling something, buildworld). this is funny: the system operates well, even when on heavy load (especially disk load), until i want a little more 8). I have a make world hanging here: 0 19914 19902 169 -2 0 1276 932 getblk D+p10:00.09 yacc -d -o c-parse.c c-parse.y -- Poul-Henning Kamp FreeBSD coreteam member [EMAIL PROTECTED] "Real hackers run -current on their laptop." FreeBSD -- It will take a long time before progress goes too far! To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: freezing...
I've reproduced the softupdates ftruncate panic and have a core dump to play with. -Matt To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: freezing... update to 1.45 of ffs_softdep.c
:I've reproduced the softupdates ftruncate panic and have a core dump :to play with. : : -Matt Ok, Kirk fixed the softupdates ftruncate panic last night. Make sure you have version 1.45 of sys/contrib/softupdates/ffs_softdep.c I would like to know if the person reporting the getblk lockup (I think it was Poul) sees that problem solved with the vinum fix that Alfred posted in regards to or whether we still have an open issue. -Matt Matthew Dillon [EMAIL PROTECTED] To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: freezing...
* Matthew Dillon [EMAIL PROTECTED] [000110 12:11] wrote:
I've reproduced the softupdates ftruncate panic and have a core dump
to play with.
I don't have a panic, but rather many processes start to get stuck
in "getblk", unfortunatly they all don't stem from a common codepath,
although the first process that wedged appears to have come through
indir_trunc:
1442 d8ed2700 d93a 1001 1422 1422 004006 3 ranlib getblk cc50a85(kgdb)
proc 0xd8ed2700
(kgdb) bt
#0 mi_switch () at machine/globals.h:117
#1 0xc015d869 in tsleep (ident=0xcc50a850, priority=0x14, wmesg=0xc0292de1 "getblk",
timo=0x0)
at ../../kern/kern_synch.c:437
#2 0xc0155b06 in acquire (lkp=0xcc50a850, extflags=0x20, wanted=0x600) at
../../kern/kern_lock.c:147
#3 0xc0155dbc in lockmgr (lkp=0xcc50a850, flags=0x10022, interlkp=0xc030ba74,
p=0xd8ed2700)
at ../../kern/kern_lock.c:362
#4 0xc017e73e in getblk (vp=0xd66dab20, blkno=0xe565a0, size=0x2000, slpflag=0x0,
slptimeo=0x0) at ../../sys/buf.h:294
#5 0xc017c72a in bread (vp=0xd66dab20, blkno=0xe565a0, size=0x2000, cred=0x0,
bpp=0xd93a1bb8)
at ../../kern/vfs_bio.c:498
#6 0xc02015af in indir_trunc (ip=0xd93a1bfc, dbn=0xe565a0, level=0x0, lbn=0xc,
countp=0xd93a1bec)
at ../../ufs/ffs/ffs_softdep.c:2040
#7 0xc02013f5 in handle_workitem_freeblocks (freeblks=0xc2315300) at
../../ufs/ffs/ffs_softdep.c:1961
#8 0xc0200e40 in softdep_setup_freeblocks (ip=0xc23ca200, length=0x0) at
../../ufs/ffs/ffs_softdep.c:1668
#9 0xc01fead6 in ffs_truncate (vp=0xd9072860, length=0x0, flags=0x0, cred=0x0,
p=0xd8ed2700)
at ../../ufs/ffs/ffs_inode.c:195
#10 0xc0209276 in ufs_inactive (ap=0xd93a1eb8) at ../../ufs/ufs/ufs_inode.c:84
#11 0xc020e3bd in ufs_vnoperate (ap=0xd93a1eb8) at ../../ufs/ufs/ufs_vnops.c:2283
#12 0xc0185145 in vput (vp=0xd9072860) at vnode_if.h:794
#13 0xc0188739 in unlink (p=0xd8ed2700, uap=0xd93a1f80) at
../../kern/vfs_syscalls.c:1421
#14 0xc026f7ba in syscall (frame={tf_fs = 0x2f, tf_es = 0x2f, tf_ds = 0x2f, tf_edi =
0x808b300, tf_esi = 0x8076170,
tf_ebp = 0xbfbffa04, tf_isp = 0xd93a1fd4, tf_ebx = 0x8076160, tf_edx =
0x8074060, tf_ecx = 0x0, tf_eax = 0xa,
tf_trapno = 0x7, tf_err = 0x2, tf_eip = 0x180c41cc, tf_cs = 0x1f, tf_eflags =
0x293, tf_esp = 0xbfbff9c8,
tf_ss = 0x2f}) at ../../i386/i386/trap.c:1057
after that it starts spreading:
1478 d8ed3480 d8f32000 1001 1477 1478 004006 3 zsh getblk cc50b148
1477 d8ed3a80 d8f24000 1001 369 1477 004106 3 rxvt getblk cc50b148
1473 d8ed2a00 d8fbf000 1001 1466 1473 004006 3 cvs getblk cc4a39e8
with varied tracebacks:
(kgdb) proc 0xd8ed3480
(kgdb) bt
#0 mi_switch () at machine/globals.h:117
#1 0xc015d869 in tsleep (ident=0xcc50b148, priority=0x14, wmesg=0xc0292de1 "getblk",
timo=0x0)
at ../../kern/kern_synch.c:437
#2 0xc0155b06 in acquire (lkp=0xcc50b148, extflags=0x20, wanted=0x600) at
../../kern/kern_lock.c:147
#3 0xc0155dbc in lockmgr (lkp=0xcc50b148, flags=0x10022, interlkp=0xc030ba74,
p=0xd8ed3480)
at ../../kern/kern_lock.c:362
#4 0xc017e73e in getblk (vp=0xd66dcf60, blkno=0x20090, size=0x2000, slpflag=0x0,
slptimeo=0x0) at ../../sys/buf.h:294
#5 0xc017c72a in bread (vp=0xd66dcf60, blkno=0x20090, size=0x2000, cred=0x0,
bpp=0xd8f33dec)
at ../../kern/vfs_bio.c:498
#6 0xc01fe869 in ffs_update (vp=0xd8e880e0, waitfor=0x0) at
../../ufs/ffs/ffs_inode.c:98
#7 0xc02092ca in ufs_inactive (ap=0xd8f33e40) at ../../ufs/ufs/ufs_inode.c:92
#8 0xc020e3bd in ufs_vnoperate (ap=0xd8f33e40) at ../../ufs/ufs/ufs_vnops.c:2283
#9 0xc018507d in vrele (vp=0xd8e880e0) at vnode_if.h:794
#10 0xc018bbb3 in vn_close (vp=0xd8e880e0, flags=0x1, cred=0xc2243c80, p=0xd8ed3480)
at ../../kern/vfs_vnops.c:232
#11 0xc018c414 in vn_closefile (fp=0xc1fb6000, p=0xd8ed3480) at
../../kern/vfs_vnops.c:630
#12 0xc01526fd in fdrop (fp=0xc1fb6000, p=0xd8ed3480) at ../../sys/file.h:211
#13 0xc015263b in closef (fp=0xc1fb6000, p=0xd8ed3480) at
../../kern/kern_descrip.c:1003
#14 0xc0151d98 in close (p=0xd8ed3480, uap=0xd8f33f80) at ../../kern/kern_descrip.c:507
#15 0xc026f7ba in syscall (frame={tf_fs = 0x2f, tf_es = 0x2f, tf_ds = 0x2f, tf_edi =
0x80a9000, tf_esi = 0x1,
tf_ebp = 0xbfbff9c8, tf_isp = 0xd8f33fd4, tf_ebx = 0x1816d3c4, tf_edx =
0x80a9138, tf_ecx = 0x2, tf_eax = 0x6,
tf_trapno = 0xc, tf_err = 0x2, tf_eip = 0x1812a22c, tf_cs = 0x1f, tf_eflags =
0x283, tf_esp = 0xbfbff98c,
tf_ss = 0x2f}) at ../../i386/i386/trap.c:1057
(kgdb) proc 0xd8ed3a80
(kgdb) bt
#0 mi_switch () at machine/globals.h:117
#1 0xc015d869 in tsleep (ident=0xcc50b148, priority=0x14, wmesg=0xc0292de1 "getblk",
timo=0x0)
at ../../kern/kern_synch.c:437
#2 0xc0155b06 in acquire (lkp=0xcc50b148, extflags=0x20, wanted=0x600) at
../../kern/kern_lock.c:147
#3 0xc0155dbc in lockmgr (lkp=0xcc50b148, flags=0x10022, interlkp=0xc030ba74,
p=0xd8ed3a80)
at ../../kern/kern_lock.c:362
#4 0xc017e73e in getblk (vp=0xd66dcf60, blkno=0x20090,
Re: freezing... update to 1.45 of ffs_softdep.c
* Matthew Dillon [EMAIL PROTECTED] [000110 14:09] wrote: :I've reproduced the softupdates ftruncate panic and have a core dump :to play with. : : -Matt Ok, Kirk fixed the softupdates ftruncate panic last night. Make sure you have version 1.45 of sys/contrib/softupdates/ffs_softdep.c I would like to know if the person reporting the getblk lockup (I think it was Poul) sees that problem solved with the vinum fix that Alfred posted in regards to or whether we still have an open issue. still a problem, just got snagged a few moments ago, there's a traceback already on the way. :) -Alfred To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: freezing... update to 1.45 of ffs_softdep.c
In message [EMAIL PROTECTED], Matthew Dillon writes: I would like to know if the person reporting the getblk lockup (I think it was Poul) sees that problem solved with the vinum fix that Alfred posted in regards to or whether we still have an open issue. I don't use vinum and -current as cvsup'ed right now has the problem. -- Poul-Henning Kamp FreeBSD coreteam member [EMAIL PROTECTED] "Real hackers run -current on their laptop." FreeBSD -- It will take a long time before progress goes too far! To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: freezing... update to 1.45 of ffs_softdep.c
: I would like to know if the person reporting the getblk lockup (I think : it was Poul) sees that problem solved with the vinum fix that Alfred : posted in regards to or whether we still have an open issue. : :still a problem, just got snagged a few moments ago, there's a traceback :already on the way. :) : :-Alfred Is this backed by the ata driver too? If so, if either you or Poul could backoff to the wd driver and see if the problem continues, I would appreciate it. -Matt Matthew Dillon [EMAIL PROTECTED] To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: freezing... update to 1.45 of ffs_softdep.c
::still a problem, just got snagged a few moments ago, there's a traceback ::already on the way. :) :: ::-Alfred : :Is this backed by the ata driver too? : :If so, if either you or Poul could backoff to the wd driver and see :if the problem continues, I would appreciate it. I was able to reproduce it. Never mind re: trying to use the 'wd' driver. It occurs for me w/ SCSI. It looks like a softupdates problem for sure. -Matt Matthew Dillon [EMAIL PROTECTED] To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: freezing... update to 1.45 of ffs_softdep.c
* Matthew Dillon [EMAIL PROTECTED] [000110 15:13] wrote: ::still a problem, just got snagged a few moments ago, there's a traceback ::already on the way. :) :: ::-Alfred : :Is this backed by the ata driver too? : :If so, if either you or Poul could backoff to the wd driver and see :if the problem continues, I would appreciate it. I was able to reproduce it. Never mind re: trying to use the 'wd' driver. It occurs for me w/ SCSI. It looks like a softupdates problem for sure. ugh, what fun this is, yes, i'm using vinum, softupdates, nfs and ata, but da for the vinum backing. I think i'm one of the bio bug posterboys. Anyhow, it seems that the problem happens because somewhere softupdates (or something else perhaps) LK_KERNPROCs the buffer and never lets go. I've been trying to compile with DEBUG_LOCKS for nearly half an hour but i'm having difficulty with vinum because i think it's intent on giving me a hard time w/ matching itself and the kernel and the userland. argh.. :) -Alfred To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: freezing... update to 1.45 of ffs_softdep.c
The following patch should fix the problem temporarily until
Kirk can dig into it and figure out what went wrong.
-Matt
Index: sys/contrib//softupdates/ffs_softdep.c
===
RCS file: /FreeBSD/FreeBSD-CVS/src/sys/contrib/softupdates/ffs_softdep.c,v
retrieving revision 1.45
diff -u -r1.45 ffs_softdep.c
--- sys/contrib//softupdates/ffs_softdep.c 2000/01/10 08:39:03 1.45
+++ sys/contrib//softupdates/ffs_softdep.c 2000/01/11 00:20:55
@@ -1663,10 +1663,13 @@
* still have a bitmap dependency, then the inode has never been
* written to disk, so we can process the freeblks immediately.
*/
+#if 0
if ((inodedep-id_state DEPCOMPLETE) == 0) {
FREE_LOCK(lk);
handle_workitem_freeblocks(freeblks);
- } else {
+ } else
+#endif
+ {
WORKLIST_INSERT(inodedep-id_bufwait, freeblks-fb_list);
FREE_LOCK(lk);
}
To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message
Re: freezing... update to 1.45 of ffs_softdep.c
In article [EMAIL PROTECTED] you wrote: In message [EMAIL PROTECTED], Matthew Dillon writes: I would like to know if the person reporting the getblk lockup (I think it was Poul) sees that problem solved with the vinum fix that Alfred posted in regards to or whether we still have an open issue. I don't use vinum and -current as cvsup'ed right now has the problem. I just saw the same: two processes got frozen in getblk (tin and mutt). No makeworld, no activity on the system at all -- just reading news and mail. After a minute or so the whole system froze. It's with version 1.46 of ffs_softdep.c -- Poul-Henning Kamp FreeBSD coreteam member [EMAIL PROTECTED] "Real hackers run -current on their laptop." FreeBSD -- It will take a long time before progress goes too far! --- "Nuclear war can ruin your whole compile." -- Karl Lehenbauer To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: freezing...
mine's a single processor system PPro200 on Intel VS440FX w/64Mb RAM it just... freezes! my smp system is running -stable, yet! Christian Carstensen wrote: On Mon, 10 Jan 2000, Joao Pedras wrote: Just wanted to ask if anyone else experiences hangs when just compiling something, while using X ? I know this is quite vague but I don't see any other pattern than this. I tested with different window managers, compiling different stuff and sometimes this misteriously happens. This hardware has always been reliable (was?) same on my machine. my hardware has not been unstable, but since 3 or 4 days ago, i see my system hanging without _any_ message, keyboard freezed, etc.pp. not only when running x11. seems to be some problem with the current kernel version. sorry, i probably can't do much debugging work in this case, as i don't see any error message. the kernel is not even dropping me to ddb. btw, this is a 2 processor pII and i haven't tried disabling SMP yet. -- Christian ^\ /^ O O o00-(_)-00o-- Excellent time to become a missing person. - PGP key available upon request or may be cut at http://pedras.webvolution.net/pgpkey.html To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: freezing...
ok, last mail before going to bed, it just... freezes! maybe, that's not interesting at all, but every time the error occured on my system, there was much disk usage (buildworld, cron scripts, ... netscape ;)). -- Christian To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: freezing...
On Mon, Jan 10, 2000 at 12:43:49AM -, Joao Pedras wrote: it just... freezes! Can any of you tell me wether you have SCSI in your system (the ahc driver, perhaps)? I too had one of these strange lockups today (see earlier post) and it seems SCSI related. The lockup occured during an I/O operation. Now I cvsupped and I find the ahc driver panic-ing on boot time with a parity error as well. I was actually on the verge of concluding that it had to be a hardware error, but doubts begin to raise again... Regards, Dave Boers. -- [EMAIL PROTECTED] To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: freezing...
On Mon, 10 Jan 2000, Dave J. Boers wrote: Can any of you tell me wether you have SCSI in your system (the ahc driver, perhaps)? I too had one of these strange lockups today (see earlier post) and it seems SCSI related. The lockup occured during an I/O operation. Now I cvsupped and I find the ahc driver panic-ing on boot time with a parity error as well. welcome to the club. world's getting more amazing with every checkout ;) -- Christian To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: freezing...
ok, this runs stable for 3 hours now... try a current kernel (checked out 4 hours ago), using a version of the file /usr/src/sys/pci/ahc_pci.c from 2000/01/06. ok, maybe it's not by any means stable, but works better than everything else within the last 24 hours. -- Christian To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: freezing...
this runs stable for 3 hours now... try a current kernel (checked out 4 hours ago), using a version of the file /usr/src/sys/pci/ahc_pci.c from 2000/01/06. ok, maybe it's not by any means stable, but works better than everything else within the last 24 hours. grrr, after 10 mins fsck... i was wrong. Problem is that the lockups (I think) are ahc-related and my SCSI hard drive did refuse to come online on one or two occasions while booting the system cold... I therefore concluded that it might be a problem with the hardware. Now (with the new kernel) I find the scsi system unstable and I have doubts again. i'm using a ahc 7895 onboard on a tyan thunder 100. i've seen my cdrom occasionally not coming up until power cycling. is this a known bug with that controller? One piece of information might also be useful in this context. After the system lockup I sort of benchmarked the scsi performance by doing "dd if=/dev/zero of=./testfile bs=100 count=128" (actually I varied the blocksize) and got a _very poor_ performance of only 4 Mb/sec (which is usually around 10 to 12 Mb/sec). I isolated my drive to be the only scsi device and I even clocked the SCSI bus down from 20 Mhz to 8 Mhz, but to no avail. scsi performance is very bad on my system for a long time now. operating on many small files, for example, my IBM DCAS-34330W is 100% busy with only 1.2 MB/s. i don't know why, but i suspect vinum in combination with some old scsi devices to cause that. -- Christian To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: freezing...
On Mon, Jan 10, 2000 at 04:10:37AM +0100, Christian Carstensen wrote: this runs stable for 3 hours now... try a current kernel (checked out 4 hours ago), using a version of the file /usr/src/sys/pci/ahc_pci.c from 2000/01/06. ok, maybe it's not by any means stable, but works better than everything else within the last 24 hours. Actually, I've booted a kernel from just before newyear (28 december) which works _reasonably_ fine (although it's the same kernel that gave me the lockup earlier) with a userland of today. Problem is that the lockups (I think) are ahc-related and my SCSI hard drive did refuse to come online on one or two occasions while booting the system cold... I therefore concluded that it might be a problem with the hardware. Now (with the new kernel) I find the scsi system unstable and I have doubts again. One piece of information might also be useful in this context. After the system lockup I sort of benchmarked the scsi performance by doing "dd if=/dev/zero of=./testfile bs=100 count=128" (actually I varied the blocksize) and got a _very poor_ performance of only 4 Mb/sec (which is usually around 10 to 12 Mb/sec). I isolated my drive to be the only scsi device and I even clocked the SCSI bus down from 20 Mhz to 8 Mhz, but to no avail. In the mean time, I disconnected my scsi hard drive and I am running from my IDE disk. Regards, Dave Boers. -- [EMAIL PROTECTED] To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
Re: freezing...
On Mon, Jan 10, 2000 at 04:48:08AM +0100, Christian Carstensen wrote: i'm using a ahc 7895 onboard on a tyan thunder 100. i've seen my cdrom occasionally not coming up until power cycling. is this a known bug with that controller? Mine is an AIC 7860 (i.e. 2940UA). I see my hard disk failing to come up occasionally, but only after I turn my power on after the system had been turned completely off (which is not often). scsi performance is very bad on my system for a long time now. operating on many small files, for example, my IBM DCAS-34330W is 100% busy with only 1.2 MB/s. i don't know why, but i suspect vinum in combination with some old scsi devices to cause that. What I'm going to say here may be complete crap, but it has been on my mind for some time now, so I may just as well speak up with my suspicions. I have read on several occasions that people find -current performing very poorly lately. I recall someone comparing -current's network performance to that of 3.4 and finding almost a factor 2 difference. I myself have found the following facts lately: 1) scsi performance is down to 4 Mb/sec MAX, writing simply zero's (this used to be around 12 Mb/sec) 2) network performance (NIC = 3COM 3C905B-TX 100Mbit) is 5 Mb/sec MAX instead of the 8 Mb/sec I'm used to. 3) IDE UDMA66 performance is down to 12 Mb/sec writing zero's (this used to be near 16 Mb/sec). 4) buildworlds are at an all time slow. I haven't been able to compile a current in less than an hour for many weeks now. This used to be more like 40 minutes or so (make buildworld -j 9 on my smp system). Add to this the strange lockups we've been having recently that somehow seem i/o or load dependent. I'm no expert on this at all, but it seems legitimate to conclude that something is seriously wrong somewhere. It seems to me that this should _definitely_ get sorted out to the bottom before release time. Regards, Dave Boers. -- [EMAIL PROTECTED] To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-current" in the body of the message
