2003-06-21 03:57:11

by Perez-Gonzalez, Inaky

[permalink] [raw]
Subject: AIC7(censored) card gone wild?


Hi Justin, list ...

I have a 2xPIII 797 Mhz with an AIC-7899P U160/m; I have
been running it under 2.5.66 since it was released, always
using the AIC7XXXX driver (new one).

However, suddenly something weird happened; since one week
ago, I get panics (in the serial console) like
the one attached (milikk.panic.txt) always caused or having
an rsync process as current (rsync is used for backup).

I also noticed that at about the same time I started to
get got those panics, I get the following when booting
the kernel:

scsi0 : Adaptec AIC7XXX EISA/VLB/PCI SCSI HBA DRIVER, Rev 6.2.28
<Adaptec aic7899 Ultra160 SCSI adapter>
aic7899: Ultra160 Wide Channel A, SCSI Id=7, 32/253 SCBs

(scsi0:A:0): 160.000MB/s transfers (80.000MHz DT, offset 127, 16bit)
Vendor: QUANTUM Model: ATLAS10K2-TY184L Rev: DA40
Type: Direct-Access ANSI SCSI revision: 03
scsi0:A:0:0: Tagged Queuing enabled. Depth 8
(scsi0:A:6): 160.000MB/s transfers (80.000MHz DT, offset 127, 16bit)
Vendor: QUANTUM Model: ATLAS10K2-TY367L Rev: DDD6
Type: Direct-Access ANSI SCSI revision: 03
scsi0:A:6:0: Tagged Queuing enabled. Depth 8
scsi1 : Adaptec AIC7XXX EISA/VLB/PCI SCSI HBA DRIVER, Rev 6.2.28
<Adaptec aic7899 Ultra160 SCSI adapter>
aic7899: Ultra160 Wide Channel B, SCSI Id=7, 32/253 SCBs

scsi1: Unexpected busfree while idle
SEQADDR == 0x31
scsi1:0:4:0: Attempting to queue an ABORT message
>>>>>>>>>>>>>>>>>> Dump Card State Begins <<<<<<<<<<<<<<<<<
scsi1: Dumping Card State while idle, at SEQADDR 0x9
Card was paused
ACCUM = 0x2, SINDEX = 0x48, DINDEX = 0xe4, ARG_2 = 0x0
HCNT = 0x0 SCBPTR = 0x0
SCSIPHASE[0x0] SCSISIGI[0x0] ERROR[0x0] SCSIBUSL[0x0]
LASTPHASE[0x1] SCSISEQ[0x12] SBLKCTL[0x6] SCSIRATE[0x0]
SEQCTL[0x10] SEQ_FLAGS[0xc0] SSTAT0[0x0] SSTAT1[0x0]
SSTAT2[0x0] SSTAT3[0x0] SIMODE0[0x8] SIMODE1[0xa4]
..... <SNIPPED- full version @ attached milikk.dump> ...

So I wonder, what does that error mean? SCSI1 has attached a
CDRW (Sony Yamaha CDRW 8/4/24) but now it doesn't show up
anymore (and so, I cannot get the model). .

Could it mean by SCSI Adapter is hosed? or my CDRW drive?
or something else? I see this same behavior with 2.5.72,
although it gets stuck in "scsi1: Unexpected busfree while
idle\nSEQADDR == 0x31", repeats it, goes up to 0x32,
back to 0x31 again, repeats 0x31 ...

Removing the CDRW helps (no more message) so I tend to assume
the CDRW is dead; but I am still waiting to see a panic again.

Thanks,

I?aky P?rez-Gonz?lez -- Not speaking for Intel -- all opinions are my own
(and my fault)

<<milikk.dump>>


Attachments:
milikk.panic.txt (880.00 B)
milikk.dump (16.88 kB)
Download all attachments

2003-06-21 09:33:49

by Matthias Andree

[permalink] [raw]
Subject: Re: AIC7(censored) card gone wild?

On Fri, 20 Jun 2003, Perez-Gonzalez, Inaky wrote:

> So I wonder, what does that error mean? SCSI1 has attached a
> CDRW (Sony Yamaha CDRW 8/4/24) but now it doesn't show up
> anymore (and so, I cannot get the model). .

The first step towards finding that out is power cycling (shut down,
switch off for a minute, then start up again) or physically
disconnecting the Yamaha drive (if it's Yamaha).

I've seen Adaptecs fuss and fight with Yamahas more than once --
although in Linux 2.2 and early 2.4 times -- and Yamahas have the nasty
habit of locking up until the next power cycle when something goes
wrong.

> Could it mean by SCSI Adapter is hosed? or my CDRW drive?

It might be either, I'd suspect the CDRW first unless I had information
that suggests otherwise.

Try to find out.

--
Matthias Andree

2003-06-21 20:51:22

by Zack Gilburd

[permalink] [raw]
Subject: Re: AIC7(censored) card gone wild?

On Saturday 21 June 2003 02:47, Matthias Andree wrote:
> On Fri, 20 Jun 2003, Perez-Gonzalez, Inaky wrote:
> > So I wonder, what does that error mean? SCSI1 has attached a
> > CDRW (Sony Yamaha CDRW 8/4/24) but now it doesn't show up
> > anymore (and so, I cannot get the model). .
>
> The first step towards finding that out is power cycling (shut down,
> switch off for a minute, then start up again) or physically
> disconnecting the Yamaha drive (if it's Yamaha).
>
> I've seen Adaptecs fuss and fight with Yamahas more than once --
> although in Linux 2.2 and early 2.4 times -- and Yamahas have the nasty
> habit of locking up until the next power cycle when something goes
> wrong.
>
> > Could it mean by SCSI Adapter is hosed? or my CDRW drive?
>
> It might be either, I'd suspect the CDRW first unless I had information
> that suggests otherwise.
>
> Try to find out.

The aic7xxx has been driving me crazy in 2.5.7x... Something got changed and
now the card will not work for me. I've posted the errors to this ML, but
noone replied.. You can dig for the message, if you'd like. It's the only
other time I've posted to the LKML.

--
Zack Gilburd
http://tehunlose.com



Attachments:
(No filename) (0.00 B)
signed data
(No filename) (189.00 B)
signature
Download all attachments

2003-06-21 22:03:02

by Justin T. Gibbs

[permalink] [raw]
Subject: Re: AIC7(censored) card gone wild?

> The aic7xxx has been driving me crazy in 2.5.7x... Something got changed and
> now the card will not work for me. I've posted the errors to this ML, but
> noone replied.. You can dig for the message, if you'd like. It's the only
> other time I've posted to the LKML.

The messages I have seen are for drivers version 6.2.8 under 2.4.21-ac1.
If you have additional output for different kernels or drivers, please
send them to me directly.

--
Justin

2003-06-22 01:14:12

by Justin T. Gibbs

[permalink] [raw]
Subject: Re: AIC7(censored) card gone wild?

> Hi Justin, list ...
>
> I have a 2xPIII 797 Mhz with an AIC-7899P U160/m; I have
> been running it under 2.5.66 since it was released, always
> using the AIC7XXXX driver (new one).
>
> However, suddenly something weird happened; since one week
> ago, I get panics (in the serial console) like
> the one attached (milikk.panic.txt) always caused or having
> an rsync process as current (rsync is used for backup).
>
> I also noticed that at about the same time I started to
> get got those panics, I get the following when booting
> the kernel:

I'm not sure what may have changed in your configuration to
make these problems start, but there is one recent fix that
may apply to your problem. I do not know if this will apply
to the 6.2.28 driver version you are using since it is based
on the latest driver release from here:

http://people.FreeBSD.org/~gibbs/linux/SRC/

Here's the patch:

--
Justin

==== //depot/aic7xxx/aic7xxx/aic7xxx.c#134 (ktext) ====

@@ -37,7 +37,7 @@
* IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
* POSSIBILITY OF SUCH DAMAGES.
*
- * $Id: //depot/aic7xxx/aic7xxx/aic7xxx.c#133 $
+ * $Id: //depot/aic7xxx/aic7xxx/aic7xxx.c#134 $
*
* $FreeBSD$
*/
@@ -1469,7 +1469,7 @@
* current connection, so we must
* leave it on while single stepping.
*/
- ahc_outb(ahc, SIMODE1, ENBUSFREE);
+ ahc_outb(ahc, SIMODE1, simode1 & ENBUSFREE);
else
ahc_outb(ahc, SIMODE1, 0);
ahc_outb(ahc, CLRINT, CLRSCSIINT);

2003-06-22 03:47:26

by Perez-Gonzalez, Inaky

[permalink] [raw]
Subject: RE: AIC7(censored) card gone wild?

> From: Matthias Andree [mailto:[email protected]]
> On Fri, 20 Jun 2003, Perez-Gonzalez, Inaky wrote:
>
> > So I wonder, what does that error mean? SCSI1 has attached a
> > CDRW (Sony Yamaha CDRW 8/4/24) but now it doesn't show up
> > anymore (and so, I cannot get the model). .
>
> The first step towards finding that out is power cycling (shut down,
> switch off for a minute, then start up again) or physically
> disconnecting the Yamaha drive (if it's Yamaha).

And clean up and check all of the fans, cables, connections,
connect, disconnect ... nothing changes. Something is hosed
up. The three years w/o downtime are biting back now.

> I've seen Adaptecs fuss and fight with Yamahas more than once --
> although in Linux 2.2 and early 2.4 times -- and Yamahas have the nasty
> habit of locking up until the next power cycle when something goes
> wrong.

Yummy ... wonder if it happens the same with their motorbikes.

> > Could it mean by SCSI Adapter is hosed? or my CDRW drive?
>
> It might be either, I'd suspect the CDRW first unless I had information
> that suggests otherwise.

The panics don't go away, so I am afraid I have something getting
to warm in there, on top of the CD burner being burned. God I
hate this things happening ...

Thanks,

I?aky P?rez-Gonz?lez -- Not speaking for Intel -- all opinions are my own (and my fault)

2003-06-22 08:59:45

by Matthias Andree

[permalink] [raw]
Subject: Re: AIC7(censored) card gone wild?

[Yamaha Firmware winds up CD-writer]
> Yummy ... wonder if it happens the same with their motorbikes.

Haven't heard of similar troubles with their motor bikes or stereo
equipment (except they claim higher electrical power output for their
active speakers than the input of the respective power supply is rated.
I wonder how they do _that_.)

> The panics don't go away, so I am afraid I have something getting
> to warm in there, on top of the CD burner being burned. God I
> hate this things happening ...

The Yamaha writers I had - AFAIR - were rated to a max. of 40?C (104?F;
313 K) and didn't last very long.

--
Matthias Andree

2003-06-22 23:32:19

by jw schultz

[permalink] [raw]
Subject: Re: AIC7(censored) card gone wild?

On Sat, Jun 21, 2003 at 09:01:08PM -0700, Perez-Gonzalez, Inaky wrote:
> > From: Matthias Andree [mailto:[email protected]]
> > On Fri, 20 Jun 2003, Perez-Gonzalez, Inaky wrote:
> >
> > > So I wonder, what does that error mean? SCSI1 has attached a
> > > CDRW (Sony Yamaha CDRW 8/4/24) but now it doesn't show up
> > > anymore (and so, I cannot get the model). .
> >
> > The first step towards finding that out is power cycling (shut down,
> > switch off for a minute, then start up again) or physically
> > disconnecting the Yamaha drive (if it's Yamaha).
>
> And clean up and check all of the fans, cables, connections,
> connect, disconnect ... nothing changes. Something is hosed
> up. The three years w/o downtime are biting back now.

You may want to replace the power supply. Most PC power
supplies are junk and some time after a year or two start
drifting out of spec. Out of spec power supplies are
responsible for more hardware failures than probably
anything else.


--
________________________________________________________________
J.W. Schultz Pegasystems Technologies
email address: [email protected]

Remember Cernan and Schmitt