Date: Wed, 05 Jul 2000 11:58:17 +0800 (SGT) From: Gregory Hosler <gregory.hosler@eno.ericsson.se> To: FENG Peirong <pfeng@krdl.org.sg> Cc: AIC7xxx@FreeBSD.ORG Subject: Re: Help needed! Message-ID: <XFMail.000705115817.gregory.hosler@eno.ericsson.se> In-Reply-To: <3962A511.821885A6@krdl.org.sg>
next in thread | previous in thread | raw e-mail | index | archive | help
On 05-Jul-00 FENG Peirong wrote:
> Gregory Hosler wrote:
>>
>> On 04-Jul-00 Feng Peirong wrote:
>> > Hi,
>> >
>> > Thanks for your detailed information. And I don't know that only
>> > 2 of 3 connectors can be used at the same time. Thanks a lot.
>> >
>> > Now I've already removed the internal 68 pin cable and the hard
>> > disk ST318436LW, only remain the ST34501N over the 50 pin cable
>> > with TE and TP on. The other end of the cable is the 50 pin ->
>> > 68 pin external port, which is connected with a Yamaha CRW4260
>> > CD-R writer with terminated switch ON. The problem remains!
>> >
>> > After disconnecting the external Yamaha CRW4260 and reboot, the
>> > problem still remains!!
>>
>> try to isolate the problem. for example, use only the main (bootable)
>> disk - remove the secondary disk, the cdrom, unplug the 2nd scsi connector.
>> make sure there is only one, properly terminated, device on that scsi
>> channel.
>>
>> does the corruption happen ? if yes, I'd beging looking at ram, cpu,
>> and other issues.
>
> This is what I am doing, see the description above. Now I only left one
> SCSI drive in system, no secondary SCSI drive, no SCSI CDROM. But when
> computing md5sum of a redhat ISO file, the result digest always changes.
>
> To make it more clear to you, I'd better describe my current system:
> 1. Redhat 6.2 on IDE disk - /dev/hda
> on IDE disk, the md5sum is always correct.
> 2. Windows 2000 Professional on SCSI disk - /dev/sda
> the file system on /dev/sda1 is FAT, where I held the md5sum
> SCSI test under Linux.
Is there any ext2fs partition on that scsi disk, and if yes, do you have
the md5sum problem when the iso image is on the ext2fs partition (I
am testing the idea that your problem might be related to the FAT fs)
To test the scsi disk (and eliminate the filesystem as a potential
possibility), try something like the following:
dd if=/dev/hda1 bs=1024 count=1000 | md5sum
replacing "hda1" with your sda partition (it looks like it would be
sda1 from your description). This will do a raw disk read, totally
bypassing the filesystem. The above reads the 1st 1000 blocks, change
count to a larger number for a stronger test. iterate to see if md5sum
give the same number twice. You can also try /dev/sda (which will read
the raw disk from the start of the disk, as opposed to the start of a
particular partition.)
I am curious as to what results you see on this.
-Greg
>
> When I boot the system using Windows 2000 and compute the md5sum of the
> ISO file on C: (after copy the ISO file again), the result is correct.
>
>> how new is the system ? was it working before ? what was the configuration
>> when it was working before (and specifically what changes were made) ?
>
> This is the first time I install Linux on this machine, which is bought
> in 1998. I expanded the memory to 256M and one more SCSI disk, then one
> more IDE disk after the md5sum test on SCSI disk failed under Linux.
>
> What I am wondering is that why the md5sum test on IDE disk under Linux
> and on SCSI disk under Windows 2000 are OK, but the test on SCSI disk
> under Linux fails.
>
> Peirong
>
>
> To Unsubscribe: send mail to majordomo@FreeBSD.org
> with "unsubscribe aic7xxx" in the body of the message
----------------------------------
E-Mail: Gregory Hosler <gregory.hosler@eno.ericsson.se>
Date: 05-Jul-00
Time: 11:47:45
If Bill Gates had a dime for every time a Windows box crashed...
...Oh, wait a minute, he already does.
----------------------------------
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe aic7xxx" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?XFMail.000705115817.gregory.hosler>