OK, so rather than try to guess based on x86, lets involve the experts;
I've cc'd the sparclinux list.

Sparc people,

The failure alluded to below occurs on an SMP system, but not on a UP
one.  The symptom appears to indicate the failure of interrupt delivery
on SMP, which is why the MPT card fails to initialise.  How do we debug
this on a sparc?

Thanks,

James


On Wed, 2011-12-14 at 09:31 +0100, Biblioteka UR wrote:
> Praveen,
> 
> Thank you for your response.
> 
> This is Sparc server and here is silo, not grub. So I do not know how 
> well I tried your suggestions.
> 
> root@fire:/boot# ls -l
> total 28506
> lrwxrwxrwx 1 root root        1 Nov  5 08:28 boot -> .
> -rw-r--r-- 1 root root    88515 Nov 14 16:35 config-3.1.0-1-sparc64
> -rw-r--r-- 1 root root    88956 Nov 14 16:47 config-3.1.0-1-sparc64-smp
> lrwxrwxrwx 1 root root        1 Nov  5 08:28 etc -> .
> -rw-r--r-- 1 root root     1024 Aug 26  2010 fd.b
> -rw-r--r-- 1 root root      512 Aug 26  2010 first.b
> -rw-r--r-- 1 root root     1024 Aug 26  2010 generic.b
> -rw-r--r-- 1 root root      692 Aug 26  2010 ieee32.b
> lrwxrwxrwx 1 root root       30 Nov 23 08:13 initrd.img -> 
> initrd.img-3.1.0-1-sparc64-smp
> -rw-r--r-- 1 root root 10201990 Nov 23 08:05 initrd.img-3.1.0-1-sparc64
> -rw-r--r-- 1 root root 10293974 Dec 12 10:34 initrd.img-3.1.0-1-sparc64-smp
> lrwxrwxrwx 1 root root       26 Nov 23 08:04 initrd.img.old -> 
> initrd.img-3.1.0-1-sparc64
> -rw-r--r-- 1 root root     7704 Aug 26  2010 isofs.b
> drwxr-xr-x 2 root root    12288 Nov  5 08:22 lost+found
> -rw-r--r-- 1 root root     7680 Nov  5 08:28 old.b
> -rw-r--r-- 1 root root    78336 Nov  5 08:28 second.b
> -rw-r--r-- 1 root root      199 Dec 14 08:36 silo.conf
> -rw-r--r-- 1 root root    76387 Aug 26  2010 silotftp.b
> -rw-r--r-- 1 root root  1629480 Nov 14 16:35 System.map-3.1.0-1-sparc64
> -rw-r--r-- 1 root root  1676706 Nov 14 16:47 System.map-3.1.0-1-sparc64-smp
> -rw-r--r-- 1 root root      512 Aug 26  2010 ultra.b
> lrwxrwxrwx 1 root root       27 Nov 23 08:13 vmlinuz -> 
> vmlinuz-3.1.0-1-sparc64-smp
> -rw-r--r-- 1 root root  2385979 Nov 14 16:34 vmlinuz-3.1.0-1-sparc64
> -rw-r--r-- 1 root root  2504212 Nov 14 16:46 vmlinuz-3.1.0-1-sparc64-smp
> lrwxrwxrwx 1 root root       23 Nov 23 08:04 vmlinuz.old -> 
> vmlinuz-3.1.0-1-sparc64
> 
> I tried:
> root@fire:/boot# cat silo.conf
> root=/dev/sda2
> partition=1
> default=LinuxOLD
> read-only
> timeout=100
> 
> image=/vmlinuz
>          label=Linux
>          initrd=/initrd.img
>          append="pci=routeirq"
> 
> image=/vmlinuz.old
>          label=LinuxOLD
>          initrd=/initrd.img.old
> 
> and
> 
> root@fire:/boot# cat silo.conf
> root=/dev/sda2
> partition=1
> default=LinuxOLD
> read-only
> timeout=100
> 
> image=/vmlinuz
>          label=Linux
>          initrd=/initrd.img
>          append="irqpoll"
> 
> image=/vmlinuz.old
>          label=LinuxOLD
>          initrd=/initrd.img.old
> 
> 
> These changes have not helped. I have the same error messages.
> 
> Unfortunately I can't collect the kernel messages with smp linux-image. 
> I set in rsyslog.conf
> 
> kern.debug                      -/var/log/kern.debug
> 
> I had already set
> 
> *.=debug;\
>          auth,authpriv.none;\
>          news.none;mail.none     -/var/log/debug
> 
> but there is no log messages from broken boot. I can collect messages 
> only from putty.
> 
> Regards,
> Mariusz
> 
> PS. Sorry for my English. :)
> 
> W dniu 2011-12-13 19:36, Krishnamoorthy, Praveen pisze:
> > Mariusz,
> >
> >>>> [   68.319518] mptbase: ioc0: WARNING - Issuing Reset from
> >> mpt_config!!, doorbell=0x24000000
> >>>> [   69.175505] mptbase: ioc0: Attempting Retry Config request type
> >> 0x3, page 0x, action 0
> >>>> [   84.267524] mptbase: ioc0: WARNING - Issuing Reset from
> >> mpt_config!!, doorbell=0x24000000
> > As Nagalakshmi pointed out, the series of reset happens because the config 
> > request for reading the page header fails. This is the first time the 
> > message queues are used when the card is coming up, therefore taking into 
> > account, that the same driver and same card works perfectly on non-smp 
> > linux kernel,  I am guessing that the config request would have been sent 
> > successfully and the firmware would have processed the request and raised 
> > an interrupt through the IRQ line assigned for this card, it is somehow not 
> > routed to our driver's interrupt service routine. Therefore could you try 
> > the following to check if any of it works?
> >
> > 1. add pci=routeirq to the kernel boot parameters in /boot/grub/menu.lst
> > Eg)         
> > title          Debian XYZ
> > root           (hdX,X)
> > kernel         /boot/vmlinuz-XYZ root=XX ro quiet splash pci=routeirq
> > initrd         /boot/initrd.img-XYZ
> >
> > 2. add irqpoll to the kernel boot parameters in /boot/grub/menu.lst
> > Eg)         
> > title          Debian XYZ
> > root           (hdX,X)
> > kernel         /boot/vmlinuz-XYZ root=XX ro quiet splash irqpoll
> > initrd         /boot/initrd.img-XYZ
> >
> > Regards,
> > Praveen
> >
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
> the body of a message to majord...@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html





-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Reply via email to