aic7xxx 6.2.24 bug report

aic7xxx 6.2.24 bug report

[Willem Riede]

Here's a problem with aic7xxx 6.2.24 in 2.5.53 -- it refuses to handle 
my /dev/sdb. Works just fine with 6.2.4 in 2.5.52. 

This is from my log when booting linux 2.5.53:

Dec 24 15:31:15 fallguy kernel: scsi0 : Adaptec AIC7XXX EISA/VLB/PCI SCSI HBA DRIVER, Rev 6.2.24
Dec 24 15:31:15 fallguy kernel:         <Adaptec aic7895 Ultra SCSI adapter>
Dec 24 15:31:15 fallguy kernel:         aic7895C: Ultra Wide Channel A, SCSI Id=7, 32/253 SCBs
Dec 24 15:31:15 fallguy kernel: 
Dec 24 15:31:15 fallguy kernel: (scsi1:A:1): 40.000MB/s transfers (20.000MHz, offset 8, 16bit)
Dec 24 15:31:15 fallguy kernel: (scsi0:A:3:0): refuses WIDE negotiation.  Using 8bit transfers
Dec 24 15:31:15 fallguy kernel: (scsi0:A:3:0): refuses WIDE negotiation.  Using 8bit transfers
Dec 24 15:31:16 fallguy kernel:   Vendor: OnStream  Model: SC-30             Rev: 1.09
Dec 24 15:31:16 fallguy kernel:   Type:   Sequential-Access                  ANSI SCSI revision: 02
Dec 24 15:31:16 fallguy kernel: (scsi0:A:3:0): refuses WIDE negotiation.  Using 8bit transfers
Dec 24 15:31:16 fallguy kernel: (scsi0:A:3:0): refuses WIDE negotiation.  Using 8bit transfers
Dec 24 15:31:16 fallguy kernel: (scsi0:A:6:0): refuses WIDE negotiation.  Using 8bit transfers
Dec 24 15:31:16 fallguy kernel:   Vendor: MATSHITA  Model: CD-R   CW-7502    Rev: 4.17
Dec 24 15:31:16 fallguy kernel:   Type:   CD-ROM                             ANSI SCSI revision: 02
Dec 24 15:31:16 fallguy kernel: (scsi0:A:6:0): refuses WIDE negotiation.  Using 8bit transfers
Dec 24 15:31:16 fallguy kernel: scsi1 : Adaptec AIC7XXX EISA/VLB/PCI SCSI HBA DRIVER, Rev 6.2.24
Dec 24 15:31:16 fallguy kernel:         <Adaptec aic7895 Ultra SCSI adapter>
Dec 24 15:31:16 fallguy kernel:         aic7895C: Ultra Wide Channel B, SCSI Id=7, 32/253 SCBs
Dec 24 15:31:16 fallguy kernel: 
Dec 24 15:31:16 fallguy kernel: (scsi1:A:0): 40.000MB/s transfers (20.000MHz, offset 8, 16bit)
Dec 24 15:31:16 fallguy kernel:   Vendor: IBM       Model: DNES-318350W      Rev: SA30
Dec 24 15:31:16 fallguy kernel:   Type:   Direct-Access                      ANSI SCSI revision: 03
Dec 24 15:31:16 fallguy kernel: scsi1:A:0:0: Tagged Queuing enabled.  Depth 16
Dec 24 15:31:16 fallguy kernel:   Vendor: IBM       Model: DGHS09U           Rev: 0350
Dec 24 15:31:16 fallguy kernel:   Type:   Direct-Access                      ANSI SCSI revision: 03
Dec 24 15:31:16 fallguy kernel: scsi1:A:1:0: Tagged Queuing enabled.  Depth 16
Dec 24 15:31:17 fallguy kernel: SCSI device sda: drive cache: write through
Dec 24 15:31:17 fallguy kernel: SCSI device sda: 35843670 512-byte hdwr sectors (18352 MB)
Dec 24 15:31:17 fallguy kernel:  sda: sda1 sda2 sda3 sda4
Dec 24 15:31:17 fallguy kernel: Attached scsi disk sda at scsi1, channel 0, id 0, lun 0
Dec 24 15:31:17 fallguy kernel: SCSI device sdb: drive cache: write through
Dec 24 15:31:17 fallguy kernel: SCSI device sdb: 17916240 512-byte hdwr sectors (9173 MB)
Dec 24 15:31:17 fallguy kernel:  sdb:SCSI disk error : host 1 channel 0 id 1 lun 0 return code = 8000002
Dec 24 15:31:17 fallguy kernel: Current sdsdb: sense key Aborted Command
Dec 24 15:31:17 fallguy kernel: Additional sense: Synchronous data transfer error
Dec 24 15:31:17 fallguy kernel: end_request: I/O error, dev sdb, sector 0
Dec 24 15:31:17 fallguy kernel: Buffer I/O error on device sd(8,16), logical block 0
Dec 24 15:31:17 fallguy kernel: SCSI disk error : host 1 channel 0 id 1 lun 0 return code = 8000002
Dec 24 15:31:18 fallguy kernel: Current sdsdb: sense key Aborted Command
Dec 24 15:31:18 fallguy kernel: Additional sense: Synchronous data transfer error
Dec 24 15:31:18 fallguy kernel: end_request: I/O error, dev sdb, sector 0
Dec 24 15:31:18 fallguy kernel: Buffer I/O error on device sd(8,16), logical block 0
Dec 24 15:31:18 fallguy kernel:  unable to read partition table
Dec 24 15:31:18 fallguy kernel: Attached scsi disk sdb at scsi1, channel 0, id 1, lun 0

The partition table doesn't get read.

But the disk is still accepted by the 2.5.52 kernel:

Dec 28 15:14:50 fallguy kernel: scsi0 : Adaptec AIC7XXX EISA/VLB/PCI SCSI HBA DRIVER, Rev 6.2.4
Dec 28 15:14:50 fallguy kernel:         <Adaptec aic7895 Ultra SCSI adapter>
Dec 28 15:14:50 fallguy kernel:         aic7895C: Ultra Wide Channel A, SCSI Id=7, 32/253 SCBs
Dec 28 15:14:50 fallguy kernel: 
Dec 28 15:14:50 fallguy kernel: (scsi0:A:3): 10.000MB/s transfers (10.000MHz, offset 7)
Dec 28 15:14:50 fallguy kernel:   Vendor: OnStream  Model: SC-30             Rev: 1.09
Dec 28 15:14:50 fallguy kernel:   Type:   Sequential-Access                  ANSI SCSI revision: 02
Dec 28 15:14:50 fallguy kernel:   Vendor: MATSHITA  Model: CD-R   CW-7502    Rev: 4.17
Dec 28 15:14:50 fallguy kernel:   Type:   CD-ROM                             ANSI SCSI revision: 02
Dec 28 15:14:50 fallguy kernel: (scsi0:A:6): 10.000MB/s transfers (10.000MHz, offset 8)
Dec 28 15:14:50 fallguy kernel: scsi1 : Adaptec AIC7XXX EISA/VLB/PCI SCSI HBA DRIVER, Rev 6.2.4
Dec 28 15:14:50 fallguy kernel:         <Adaptec aic7895 Ultra SCSI adapter>
Dec 28 15:14:51 fallguy kernel:         aic7895C: Ultra Wide Channel B, SCSI Id=7, 32/253 SCBs
Dec 28 15:14:51 fallguy kernel: 
Dec 28 15:14:51 fallguy kernel: (scsi1:A:0): 40.000MB/s transfers (20.000MHz, offset 8, 16bit)
Dec 28 15:14:51 fallguy kernel:   Vendor: IBM       Model: DNES-318350W      Rev: SA30
Dec 28 15:14:51 fallguy kernel:   Type:   Direct-Access                      ANSI SCSI revision: 03
Dec 28 15:14:51 fallguy kernel: (scsi1:A:1): 40.000MB/s transfers (20.000MHz, offset 8, 16bit)
Dec 28 15:14:51 fallguy kernel:   Vendor: IBM       Model: DGHS09U           Rev: 0350
Dec 28 15:14:51 fallguy kernel:   Type:   Direct-Access                      ANSI SCSI revision: 03
Dec 28 15:14:51 fallguy kernel: scsi1:A:0:0: Tagged Queuing enabled.  Depth 16
Dec 28 15:14:51 fallguy kernel: SCSI device sda: drive cache: write through
Dec 28 15:14:51 fallguy kernel: SCSI device sda: 35843670 512-byte hdwr sectors (18352 MB)
Dec 28 15:14:51 fallguy kernel:  sda: sda1 sda2 sda3 sda4
Dec 28 15:14:51 fallguy kernel: Attached scsi disk sda at scsi1, channel 0, id 0, lun 0
Dec 28 15:14:51 fallguy kernel: scsi1:A:1:0: Tagged Queuing enabled.  Depth 16
Dec 28 15:14:51 fallguy kernel: SCSI device sdb: drive cache: write through
Dec 28 15:14:51 fallguy kernel: SCSI device sdb: 17916240 512-byte hdwr sectors (9173 MB)
Dec 28 15:14:51 fallguy kernel:  sdb: sdb1 sdb2 sdb3
Dec 28 15:14:51 fallguy kernel: Attached scsi disk sdb at scsi1, channel 0, id 1, lun 0

Let me know if you want me to provide more information, or try something.

Regards, Willem Riede.

Re: aic7xxx 6.2.24 bug report

[Justin T. Gibbs]

> Here's a problem with aic7xxx 6.2.24 in 2.5.53 -- it refuses to handle 
> my /dev/sdb. Works just fine with 6.2.4 in 2.5.52. 
> 
> This is from my log when booting linux 2.5.53:

Can you retest enabling the debug features of the driver and setting
the debug mask to 0x63?  Something like, assuming you are using a
module:

options aic7xxx aic7xxx=verbose.debug:0x63

--
Justin

Re: aic7xxx 6.2.24 bug report

[Willem Riede]

On 2002.12.28 15:48 Justin T. Gibbs wrote:
> Can you retest enabling the debug features of the driver and setting
> the debug mask to 0x63?  Something like, assuming you are using a
> module:
> 
> options aic7xxx aic7xxx=verbose.debug:0x63
> 
What am I doing wrong?

Loading aic7xxx module
Parameter aic7xxx length must be 1-1 characters
aic7xxx: `verbose,debug:0x63' invalid for parameter `aic7xxx'


Thanks, Willem Riede.

Re: aic7xxx 6.2.24 bug report

[Justin T. Gibbs]

> On 2002.12.28 15:48 Justin T. Gibbs wrote:
>> Can you retest enabling the debug features of the driver and setting
>> the debug mask to 0x63?  Something like, assuming you are using a
>> module:
>> 
>> options aic7xxx aic7xxx=verbose.debug:0x63
				  ^<= '.' not ','

>> 
> What am I doing wrong?

You're using characters that the shell might expand.  If
you quote the option string, the choice of ',' or '.' shouldn't
matter.

--
Justin

Re: aic7xxx 6.2.24 bug report

[Willem Riede]

On 2002.12.28 19:35 Justin T. Gibbs wrote:
> > 
> > I'm particularly surprised by the line
> > 
> > Parameter aic7xxx length must be 1-1 characters
> > 
> > which hints that the options are not correctly coded?
> 
> This is due to the shell magic used during initrds.  Did you
> try quoting this options?
> 
> options aic7xxx 'aic7xxx=verbose.debug:0x63'
> 
I have now, without change. I had earlier quoted just "aic7xxx=verbose.debug:0x63"
and tried only aic7xxx=verbose to avoid special characters, all with the same
effect.

One thought just occured to me, the modutils have been in flux, I'll see if
there is a newer version than I'm using now.

Regards, Willem Riede.

Re: aic7xxx 6.2.24 bug report

[Douglas Gilbert]

Willem Riede wrote:
> On 2002.12.28 19:35 Justin T. Gibbs wrote:
> 
>>>I'm particularly surprised by the line
>>>
>>>Parameter aic7xxx length must be 1-1 characters
>>>
>>>which hints that the options are not correctly coded?
>>
>>This is due to the shell magic used during initrds.  Did you
>>try quoting this options?
>>
>>options aic7xxx 'aic7xxx=verbose.debug:0x63'
>>
> 
> I have now, without change. I had earlier quoted just "aic7xxx=verbose.debug:0x63"
> and tried only aic7xxx=verbose to avoid special characters, all with the same
> effect.
> 
> One thought just occured to me, the modutils have been in flux, I'll see if
> there is a newer version than I'm using now.

This seems to be the latest module-init-tools:
http://www.[COUNTRY].kernel.org/pub/linux/kernel/people/rusty/modules/module-init-tools-0.9.6.tar.gz

Announced in lkml:
http://marc.theaimsgroup.com/?l=linux-kernel&m=104089668930417&w=2

BTW Module parameters seem to start working again in lk 2.5.52
(with module-init-tools-0.9.3).

Doug Gilbert

Re: aic7xxx 6.2.24 bug report

[Willem Riede]

On 2002.12.28 21:30 Douglas Gilbert wrote:
> Willem Riede wrote:
> > On 2002.12.28 19:35 Justin T. Gibbs wrote:
> > 
> >>>I'm particularly surprised by the line
> >>>
> >>>Parameter aic7xxx length must be 1-1 characters
> >>>
> >>>which hints that the options are not correctly coded?
> >>
> >>This is due to the shell magic used during initrds.  Did you
> >>try quoting this options?
> >>
> >>options aic7xxx 'aic7xxx=verbose.debug:0x63'
> >>
> > 
> > I have now, without change. I had earlier quoted just "aic7xxx=verbose.debug:0x63"
> > and tried only aic7xxx=verbose to avoid special characters, all with the same
> > effect.
> > 
> > One thought just occured to me, the modutils have been in flux, I'll see if
> > there is a newer version than I'm using now.
> 
> This seems to be the latest module-init-tools:
> http://www.[COUNTRY].kernel.org/pub/linux/kernel/people/rusty/modules/module-init-tools-0.9.6.tar.gz
> 
Actually, I found 0.9.7 on ftp.kernel.org, but it didn't help :-(

Could the trouble perhaps be with:

Red Hat nash version 3.4.28 starting

it's from their 8.0 release.

Thanks, Willem Riede.

Re: aic7xxx 6.2.24 bug report

[Doug Ledford]

On Sun, Dec 29, 2002 at 09:17:29AM -0500, Willem Riede wrote:
> 
> Red Hat nash version 3.4.28 starting
> 
> it's from their 8.0 release.

I'm not sure, but it could be.  I think the current initrd still uses the 
available insmod instead of trying to do the syscalls internally, so that 
shouldn't be a problem.  But, nash doesn't do command line arguments 
exactly identical to bash so I had to play with the command line argument 
escaping in the past to get it to work right in nash.  But, try adding 
ash.static to your initrd, then as the last line in linuxrc call 
ash.static and it will drop you to a shell prompt where you can attempt to 
perform the insmod by hand and see if you can get it to work before the 
kernel panics for lack of a root device.

-- 
  Doug Ledford <dledford@redhat.com>     919-754-3700 x44233
         Red Hat, Inc. 
         1801 Varsity Dr.
         Raleigh, NC 27606