Sounds pretty much like faulty hardware, RAM errors are often the cause of segfaults. Overclocked CPU might be the reason too. Cyrix CPU looks suspicious to me.

Did you try to run meminfo on that system?

-Kastus

On Wed, Mar 14, 2001 at 06:55:39PM -0500, Stan Koper wrote:

...

Hello all,

Nice to see all the wonderful comments about SuSE 7.1, but unfortunately I haven't been able to share in the success. In fact, SuSE 7.0 was a dog, as far as I was concerned, generating frequent segmentation faults all over the place.

But I didn't pay much for 7.1 personal, $19.99 at Best Buy, minus my Shaw's grocery card 10% discount. But it did inspire me to reinstall 7.0, which seemed to go better this time. Unfortunately, I thought I should upgrade the kernel to 2.2.18, and so I did that, with the result that

...

system is now freezing, with a disk error/Kernel OOPS message as described below.

This is an Abit IT5H motherboard, probably version 1.2 or 1.5, with a Cyrix 200 chip. It has 96 megs of memory, a CS4936 sound card, and an EIDE CD-ROM, 24x. I forget what kind of video card, but PCI SVGA. This is

...

gateway machine to my home network, so it has a 3COM 3C509 ISA card as eth0, hooked up to a cable modem, and a Realek 1839 10/100 PCI card as eth1 for the internal connection..

When I first set this machine up as the gateway, it had an IDE hard drive, probably 3 gigs, under a 5.x version of SuSE. Then I upgraded to 6.4, and installed everything on a new 8 gig IDE drive. At that time eth1 was a cheapo NE2000 card. Everything worked fine.

About the time that SuSE 7.0 came out, I was able to purchase a Seagate 9 gig SCSI drive for about $25, and got a Tekram DC-390 for around $30. So what I did was unplug the ide drives, plug in the cable from the SCSI drive to the controller card, go into the bios setup and erase the IDE drive entries (well, run the ID hard drives program), and reboot so that only

Konstantin, Faulty hardware is a possibility, but it doesn't explain why SuSE 6.4 works flawlessley on the same system, from the IDE drives. The CPU is not overclocked, although as I explained in a message to Jerry Kreps, this used to be a windows machine, and we did have some difficulty in the summer time with crashes which I believe were due to an overheated CPU. We bumped the fan up, and the problems went away. Anyway, maybe the drive has gone south, but it's brand new. It was inexpensive because it was on auction at Egghead, and when I first saw it, there were 20 minutes left and the bid price was still the original $4.99 (or thereabouts). I *couldn't* pass up that deal, and by the time the smoke had cleared, the final price was around $25. But it was advertised as brand new, and I don't know if there are any problems with Seagate full height SCSI drives. I will keep hoping that someone on the list can understand what the error message *means*, and guide me accordingly. Thanks, Stan Koper ----- Original Message ----- From: "Konstantin (Kastus) Shchuka" To: Sent: Wednesday, March 14, 2001 11:31 PM Subject: Re: [SLE] 7.1 Won't Install, and Worse the the the

...

SCSI hard drive was up and running.

I installed 7.0 on that drive. Installing was problematic. This was 7.0 professsional, and at one point I even returned the set to the store, because it wasn't reading files on one of the CD's. But the replacement did the same thing. Anyway, I eventually got it installed and working, although with fairly frequent segmentation fault messages, mostly having to do with apache, and probably a few other things. But since I blew that away and reinstalled 7.0, those problems are gone..

But 7.1 doesn't install at all. First, it doesn't recognize the hard drive right off the bat. That drops it out of Yast2 to Yast1. So I install the SCSI module for the tekram card. Then I add the network modules, but only the Realtek card is found, not the 3c509 (not really a problem, 7.0 didn't recognize the 3com card either, it just has to be done later). At that point I'm ready to roll, and I select installation/update. The CD spins, and then stops, and there it sits. Nothing. Nada. Dead.

So I figured stuck with 7.0, I might as well upgrade the kernel. I used the *rpm for k_i386, not k_dflt. And now this. Any suggestions?

Thanks,

Stan Koper

SCSI disk error: host 0 channel 0 id 0 lun 0 return code = 2504000 scsidisk I/O error: dev 08:05 sector 65704 reiserfs_bread: unable to read dev = 2053 block = 8203 size = 4096 unable to handle kernel paging request at virtual address 13fe6f47 current -> tss.cr3 = 05fbd000, %cr3 = 05fbd00 *pde = 00000000 Oops = 0000 CPU = 0 ELP = 0010 : [<c017ff34>] EFLAGS = 00010286 eax: 848babf3 ebx:c0180342 ecs:e3eab380 edx:13fe6f33 esi: c5fa1d18 edi:c5fald58 edp:c5fa1ca1 esp:c5fa1c8c ds: 0018 es: 0018 ss: 0018 Process init (pid: 1, process nr: 1, stackpage = c5fa1000) Stack: c5eb314c c5b291d0 c3e5c018 00000018 c5fa1ca0 00000001 00000002 0000010a 0a619f7f 000001f4 c3eab380 c5fa1d78 00000003 c0180342 c5eb341c c5b29230 00000003 c5fa1d18 c1fa1d58 c3e5ce40 c5fa1d04 c0180318 c5fa1d18 00000000 Call Trace : [<c0180342>] [<c0180318>] [<c0180f0a2>] [<c0180342>] [<c0180342>] [<c0132617>] [<c01324da>] [<c0132860>] [<c01329f7>] {c01c7f04>] [<c01c8215>] [<c01395ac>] [<c019720a>] [<c01395ac>] [<c0197eda>] [<c0197a85>] [<c01972db>] [<c019729e>] [<c0197b127>] [<c019889e>] [<c013470a>] [<c010a485>] [<c010a344>] Code: 0f b7 4a 14 03 4b 24 09 4f 10 8d 47 0c 50 55 51 52 e8 ce f5

-- To unsubscribe send e-mail to suse-linux-e-unsubscribe@suse.com For additional commands send e-mail to suse-linux-e-help@suse.com Also check the FAQ at http://www.suse.com/support/faq

-- To unsubscribe send e-mail to suse-linux-e-unsubscribe@suse.com For additional commands send e-mail to suse-linux-e-help@suse.com Also check the FAQ at http://www.suse.com/support/faq

On Fri, 16 Mar 2001, Stan Koper said:

Konstantin,

Faulty hardware is a possibility, but it doesn't explain why SuSE 6.4 works flawlessley on the same system, from the IDE drives. The CPU is not overclocked, although as I explained in a message to Jerry Kreps, this used to be a windows machine, and we did have some difficulty in the summer time with crashes which I believe were due to an overheated CPU. We bumped the fan up, and the problems went away.

Specifically what controller are you using? There are 2 or 3 different variations of the Tekram 390. I'm using a 390F for example. Also, what driver are you loading for it? There is also more than one driver that will function. For the 390F, the best choice (according to my interpretation of the kernel docs) is the SYM53c8xx driver. I have also used the NCR53C8XX driver without problems. ---------------------------------------------------------------------- John Karns jkarns@csd.net

You may have a surface damage problem on the HD. Try a low level format via the BIOS util with the Tekram (F2 or F6 when prompted during boot). This should mark unread/unwritable sectors as being bad (provided the number doesn't exceed a limit which is drive specific), thus allowing you to proceed.

---------------------------------------------------------------------- John Karns jkarns@csd.net

On Wed, 14 Mar 2001, Stan Koper said:

...

Hello all,

Nice to see all the wonderful comments about SuSE 7.1, but unfortunately I haven't been able to share in the success. In fact, SuSE 7.0 was a dog, as far as I was concerned, generating frequent segmentation faults all over the place.

But I didn't pay much for 7.1 personal, $19.99 at Best Buy, minus my Shaw's grocery card 10% discount. But it did inspire me to reinstall 7.0, which seemed to go better this time. Unfortunately, I thought I should upgrade the kernel to 2.2.18, and so I did that, with the result that the system is now freezing, with a disk error/Kernel OOPS message as described below.

This is an Abit IT5H motherboard, probably version 1.2 or 1.5, with a Cyrix 200 chip. It has 96 megs of memory, a CS4936 sound card, and an EIDE CD-ROM, 24x. I forget what kind of video card, but PCI SVGA. This is

...

gateway machine to my home network, so it has a 3COM 3C509 ISA card as eth0, hooked up to a cable modem, and a Realek 1839 10/100 PCI card as eth1 for the internal connection..

When I first set this machine up as the gateway, it had an IDE hard drive, probably 3 gigs, under a 5.x version of SuSE. Then I upgraded to 6.4, and installed everything on a new 8 gig IDE drive. At that time eth1 was a cheapo NE2000 card. Everything worked fine.

About the time that SuSE 7.0 came out, I was able to purchase a Seagate 9 gig SCSI drive for about $25, and got a Tekram DC-390 for around $30. So what I did was unplug the ide drives, plug in the cable from the SCSI drive to the controller card, go into the bios setup and erase the IDE drive entries (well, run the ID hard drives program), and reboot so that only

John, That thought occurred to me. I don't know why it would crop up now, however. I did a low level format when I first got the drive, about the middle of last year. Things can change, of course, but it's over 90 days... I once asked on the list "what is a segmentation fault?", and I think the answer was that it had to do with problems which occur in installing the software. That seemed consistent with what I was seeing at the time. But as I noted, since I blew away the original 7.0 install and re-did it, the segmentation faults went away. The crash/freeze problem did not start until I upgraded the kernel to 2.2.18. The problem with 7.1 not installing started while the system was still in the "old" 7.0 mode, continued into the "new" 7.0, and is still the case. But 7.1 will install on my old AMD 486 clone machine (although the default with office takes up more than the 2.8 gigs on that machine's hard drive). I have no concerns about re-doing the low-level format, and starting over again with 7.1 and then, if that fails, with 7.0. I'd just like to make sure that it's necessary. If someone can read the "error" message in my original post (that I so laboriously hand transcribed from the screen), dissect it, and tell me what it really means, I'd be very grateful. Thanks, Stan Koper ----- Original Message ----- From: "John Karns" To: "SuSE Linux" Sent: Thursday, March 15, 2001 9:57 AM Subject: Re: [SLE] 7.1 Won't Install, and Worse the the

...

SCSI hard drive was up and running.

I installed 7.0 on that drive. Installing was problematic. This was 7.0 professsional, and at one point I even returned the set to the store, because it wasn't reading files on one of the CD's. But the replacement did the same thing. Anyway, I eventually got it installed and working, although with fairly frequent segmentation fault messages, mostly having to do with apache, and probably a few other things. But since I blew that away and reinstalled 7.0, those problems are gone..

But 7.1 doesn't install at all. First, it doesn't recognize the hard drive right off the bat. That drops it out of Yast2 to Yast1. So I install the SCSI module for the tekram card. Then I add the network modules, but only the Realtek card is found, not the 3c509 (not really a problem, 7.0 didn't recognize the 3com card either, it just has to be done later). At that point I'm ready to roll, and I select installation/update. The CD spins, and then stops, and there it sits. Nothing. Nada. Dead.

So I figured stuck with 7.0, I might as well upgrade the kernel. I used the *rpm for k_i386, not k_dflt. And now this. Any suggestions?

Thanks,

Stan Koper

SCSI disk error: host 0 channel 0 id 0 lun 0 return code = 2504000 scsidisk I/O error: dev 08:05 sector 65704 reiserfs_bread: unable to read dev = 2053 block = 8203 size = 4096

-- To unsubscribe send e-mail to suse-linux-e-unsubscribe@suse.com For additional commands send e-mail to suse-linux-e-help@suse.com Also check the FAQ at http://www.suse.com/support/faq