* Re: Qlogic driver hang
2003-10-02 15:45 ` Qlogic driver hang Chris Worley
@ 2003-10-02 18:16 ` Steven Dake
2003-10-02 18:25 ` Mark Bellon
2003-10-03 5:53 ` Andrew Vasquez
2003-10-03 18:15 ` Andrew Vasquez
2 siblings, 1 reply; 7+ messages in thread
From: Steven Dake @ 2003-10-02 18:16 UTC (permalink / raw)
To: Chris Worley, mbellon; +Cc: linux-scsi
At one point some code was added to enable parity on the chip. This may
be the cause of the hang. Try disabling parity checking in the driver.
On Thu, 2003-10-02 at 08:45, Chris Worley wrote:
> In using the 6.06.00 version of the qlogic driver, the module load gets
> to:
>
> Verifying chip...
>
> And hangs the system. The older driver would print out:
>
> Firmware version: 3.01.13, Driver version 6.01.00
>
> at this point. Any ideas what I'm doing wrong?
>
> Thanks,
>
> Chris
> On Tue, 2003-09-30 at 16:25, Andrew Vasquez wrote:
> > Chris
> >
> > > The tarball from Qlogic was labeled:
> > >
> > > qla2x00-v6.01.00-dist.tgz
> > >
> > > It's the full-height, PCI-X, 133MHz, single port, 2340 (I think).
> > >
> > >
> >
> > Ouch, that's a really old driver (with some known [error handling]
> > problems). Would it be possible to start with something a bit more
> > recent, 6.06.00 has been posted to QLogic's website:
> >
> > http://www.qlogic.com/support/product_resources.asp?id=253
> > http://download.qlogic.com/drivers/13591/qla2x00-v6.06.00-dist.tgz
> >
> > I'd suggest we start from there. If you are still experiencing
> > problems, forward the /var/log/messages file after you enable
> > 'Extended Error Logging' from the BIOS utility. If you feel
> > comfortable changing some lines in the driver source, more descript
> > (helpful) logs can be obtained with the following three changes:
> >
> > o In the qla_settings.h file change the following line from:
> >
> > #define DEBUG_QLA2100 0 /* For Debug of qla2x00 */
> > to
> > #define DEBUG_QLA2100 1 /* For Debug of qla2x00 */
> >
> > o In the qla2x00.h file change the following line from:
> >
> > /* #define QL_DEBUG_LEVEL_2 */ /* Output error msgs to COM1 */
> > to
> > #define QL_DEBUG_LEVEL_2 /* Output error msgs to COM1 */
> >
> > o In the qla2x00.c file change the following line from:
> >
> > static int extended_error_logging = 0; /* 0 = off, 1 = log errors */
> > to
> > static int extended_error_logging = 1; /* 0 = off, 1 = log errors */
> >
> > Recompile & load the driver.
> >
> > Regards,
> > Andrew Vasquez
>
> -
> To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
>
>
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: Qlogic driver hang
2003-10-02 18:16 ` Steven Dake
@ 2003-10-02 18:25 ` Mark Bellon
2003-10-02 19:24 ` Tim Pepper
0 siblings, 1 reply; 7+ messages in thread
From: Mark Bellon @ 2003-10-02 18:25 UTC (permalink / raw)
To: sdake; +Cc: Chris Worley, linux-scsi
>
>
>At one point some code was added to enable parity on the chip. This may
>be the cause of the hang. Try disabling parity checking in the driver.
>
>On Thu, 2003-10-02 at 08:45, Chris Worley wrote:
>
>
>>In using the 6.06.00 version of the qlogic driver, the module load gets
>>to:
>>
>> Verifying chip...
>>
>>And hangs the system. The older driver would print out:
>>
>> Firmware version: 3.01.13, Driver version 6.01.00
>>
>>at this point. Any ideas what I'm doing wrong?
>>
>>
There is a known problem in some hardware implementations of the 2312
and possibly other
chips - parity errors in the firmware RAM will lock up a system. The
last time I looked the public driver did not have a configuration option
to do this. If it's not there let me know and I can dig the lines up for
you to try.
mark
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: Qlogic driver hang
2003-10-02 18:25 ` Mark Bellon
@ 2003-10-02 19:24 ` Tim Pepper
2003-10-02 19:43 ` Mark Bellon
0 siblings, 1 reply; 7+ messages in thread
From: Tim Pepper @ 2003-10-02 19:24 UTC (permalink / raw)
To: Mark Bellon; +Cc: sdake, Chris Worley, linux-scsi
On Thu 02 Oct at 11:25:32 -0700 mbellon@mvista.com done said:
> >
> >
> >At one point some code was added to enable parity on the chip. This may
> >be the cause of the hang. Try disabling parity checking in the driver.
I think too that different bios flash version have different settings for
enabling/disabling parity (and other things) that can make a difference.
The latest on their webpage is 1.34. I understand the firmware which the
driver downloads at load time still uses certain bits from the version
flashed to the card. Somewhere around bios 1.2x (where x was something
high..it's been a while) I had some hangs go away. All the new cards
I've gotten appear to ship with ancient BIOS's and sorely need flashing.
Tim
--
*********************************************************
* tpepper@vato dot org * Venimus, Vidimus, *
* http://www.vato.org/~tpepper * Dolavimus *
*********************************************************
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: Qlogic driver hang
2003-10-02 19:24 ` Tim Pepper
@ 2003-10-02 19:43 ` Mark Bellon
0 siblings, 0 replies; 7+ messages in thread
From: Mark Bellon @ 2003-10-02 19:43 UTC (permalink / raw)
To: Tim Pepper; +Cc: sdake, Chris Worley, linux-scsi
>
>
>I think too that different bios flash version have different settings for
>enabling/disabling parity (and other things) that can make a difference.
>The latest on their webpage is 1.34. I understand the firmware which the
>driver downloads at load time still uses certain bits from the version
>flashed to the card. Somewhere around bios 1.2x (where x was something
>high..it's been a while) I had some hangs go away. All the new cards
>I've gotten appear to ship with ancient BIOS's and sorely need flashing.
>
>
The driver supports a compiled in version of the firmware and, at last
look, always uses that regardless of what in out there in non-volatile
storage. The firmware compiled into the 6.06 driver is 3.02.13 for the
23xx. This is a relatively new firmware.
By-the-by, which HBA is being used?
mark
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: Qlogic driver hang
2003-10-02 15:45 ` Qlogic driver hang Chris Worley
2003-10-02 18:16 ` Steven Dake
@ 2003-10-03 5:53 ` Andrew Vasquez
2003-10-03 18:15 ` Andrew Vasquez
2 siblings, 0 replies; 7+ messages in thread
From: Andrew Vasquez @ 2003-10-03 5:53 UTC (permalink / raw)
To: linux-scsi
On Thu, 02 Oct 2003, Chris Worley wrote:
> In using the 6.06.00 version of the qlogic driver, the module load gets
> to:
>
> Verifying chip...
>
> And hangs the system. The older driver would print out:
>
> Firmware version: 3.01.13, Driver version 6.01.00
>
> at this point. Any ideas what I'm doing wrong?
>
So much for forward progress :O
Would it be possible to get the logs mentioned below? I'm rather
curious where exactly (or as close as possible) the hang is occuring.
How 'hard' is the hang? Is the cap-locks key available (light
flashes)? Is the HBA connected to anything at the time of module
load?
> > I'd suggest we start from there. If you are still experiencing
> > problems, forward the /var/log/messages file after you enable
> > 'Extended Error Logging' from the BIOS utility. If you feel
> > comfortable changing some lines in the driver source, more descript
> > (helpful) logs can be obtained with the following three changes:
> >
> > o In the qla_settings.h file change the following line from:
> >
> > #define DEBUG_QLA2100 0 /* For Debug of qla2x00 */
> > to
> > #define DEBUG_QLA2100 1 /* For Debug of qla2x00 */
> >
> > o In the qla2x00.h file change the following line from:
> >
> > /* #define QL_DEBUG_LEVEL_2 */ /* Output error msgs to COM1 */
> > to
> > #define QL_DEBUG_LEVEL_2 /* Output error msgs to COM1 */
> >
> > o In the qla2x00.c file change the following line from:
> >
> > static int extended_error_logging = 0; /* 0 = off, 1 = log errors */
> > to
> > static int extended_error_logging = 1; /* 0 = off, 1 = log errors */
> >
> > Recompile & load the driver.
> >
Regards,
Andrew Vasquez
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: Qlogic driver hang
2003-10-02 15:45 ` Qlogic driver hang Chris Worley
2003-10-02 18:16 ` Steven Dake
2003-10-03 5:53 ` Andrew Vasquez
@ 2003-10-03 18:15 ` Andrew Vasquez
2 siblings, 0 replies; 7+ messages in thread
From: Andrew Vasquez @ 2003-10-03 18:15 UTC (permalink / raw)
To: linux-scsi
On Thu, 02 Oct 2003, Chris Worley wrote:
> In using the 6.06.00 version of the qlogic driver, the module load gets
> to:
>
> Verifying chip...
>
> And hangs the system. The older driver would print out:
>
> Firmware version: 3.01.13, Driver version 6.01.00
>
> at this point. Any ideas what I'm doing wrong?
>
So much for forward progress :O
Would it be possible to get the logs mentioned below? I'm rather
curious where exactly (or as close as possible) the hang is occuring.
How 'hard' is the hang? Is the cap-locks key available (light
flashes)? Is the HBA connected to anything at the time of module
load?
> > I'd suggest we start from there. If you are still experiencing
> > problems, forward the /var/log/messages file after you enable
> > 'Extended Error Logging' from the BIOS utility. If you feel
> > comfortable changing some lines in the driver source, more descript
> > (helpful) logs can be obtained with the following three changes:
> >
> > o In the qla_settings.h file change the following line from:
> >
> > #define DEBUG_QLA2100 0 /* For Debug of qla2x00 */
> > to
> > #define DEBUG_QLA2100 1 /* For Debug of qla2x00 */
> >
> > o In the qla2x00.h file change the following line from:
> >
> > /* #define QL_DEBUG_LEVEL_2 */ /* Output error msgs to COM1 */
> > to
> > #define QL_DEBUG_LEVEL_2 /* Output error msgs to COM1 */
> >
> > o In the qla2x00.c file change the following line from:
> >
> > static int extended_error_logging = 0; /* 0 = off, 1 = log errors */
> > to
> > static int extended_error_logging = 1; /* 0 = off, 1 = log errors */
> >
> > Recompile & load the driver.
> >
Regards,
Andrew Vasquez
^ permalink raw reply [flat|nested] 7+ messages in thread