Re: QLogic PTI firmware

2016-07-25 Thread rodent 
On Mon, Jul 25, 2016 at 06:14:28PM -0400, alexmcwhir...@triadic.us wrote:
> [...]
> Does anyone have any input on this? Fixing this would help stabilize a 
> ton of early sparc64 machines, almost everything with a scsi controller 
> around the US2 days use and isp1000 card.

IIRC, the Alphas had a few of these isp chips in SCSI adapters which came with
the machines. You might get two-for-one by fixing this. Perhaps, some of this
might help:

http://netbsd.gw.com/cgi-bin/man-cgi?isp++NetBSD-current

http://cvsweb.netbsd.org/bsdweb.cgi/src/sys/dev/pci/isp_pci.c?rev=1.119=text/x-cvsweb-markup_with_tag=MAIN

Re: QLogic PTI firmware

2016-07-25 Thread alexmcwhirter 
Well i'm bringing this email back from the dead, original thread can be 
found here https://lists.debian.org/debian-sparc/2013/01/msg8.html


In a nutshell the qlogicpti driver isn't handling traffic demands very 
well, the request queue will overflow and the system will panic. Just 
letting a box idle for a few days will result in this issue popping up. 
You can accelerate it by doing a large amount of disk I/O.


In particular, we are hitting this code in qlogicpti.c

toss_command:
printk(KERN_EMERG "qlogicpti%d: request queue overflow\n",
   qpti->qpti_id);

/* Unfortunately, unless you use the new EH code, which
 * we don't, the midlayer will ignore the return value,
 * which is insane.  We pick up the pieces like this.
 */
Cmnd->result = DID_BUS_BUSY;
done(Cmnd);
return 1;
}

Does anyone have any input on this? Fixing this would help stabilize a 
ton of early sparc64 machines, almost everything with a scsi controller 
around the US2 days use and isp1000 card.

Re: Booting an LDOM (can't find kernel)

2016-07-25 Thread Anatoly Pugachev 
On Mon, Jul 25, 2016 at 1:55 PM,   wrote:
> On Mon, Jul 25, 2016 at 06:56:35AM +0200, John Paul Adrian Glaubitz wrote:
>>
>> It's panicking because you specified the wrong root device:
>>
>> [90779.943892] EXT4-fs (vdiska1): mounting ext3 file system using the ext4 
>> subsystem
>> [90779.951847] EXT4-fs (vdiska1): mounted filesystem with ordered data mode. 
>> Opts: (null)
>> [90780.025922] Kernel panic - not syncing: Attempted to kill init! 
>> exitcode=0x0200
>> [90780.025922]
>> [90780.026642] CPU: 0 PID: 1 Comm: init Tainted: GE   
>> 4.5.0-2-sparc64-smp #1 Debian 4.5.2-1
>> [90780.027297] Call Trace:
>> [90780.027511]  [0055777c] panic+0xdc/0x260
>> [90780.027860]  [0046a270] do_exit+0xb30/0xb40
>> [90780.028224]  [0046a330] do_group_exit+0x30/0xc0
>> [90780.029623]  [0046a3dc] SyS_exit_group+0x1c/0x40
>> [90780.030027]  [00406294] linux_sparc_syscall+0x34/0x44
>> [90780.031618] Press Stop-A (L1-A) to return to the boot prom
>> [90780.032028] ---[ end Kernel panic - not syncing: Attempted to kill init! 
>> exitcode=0x0200
>> [90780.032028]
>>
>> You specified vdiska1 which is your boot part:
>>
>> > Kernel command line: root=/dev/vdiska1 ro console=ttyS0,9600,8n1
>>
>> /dev/vdiska1 is the boot partition for SILO, not the root filesystem.
>
> There's only two partitions: vdiska1 which contains everything and vdiska2
> which is swap.
>

please make a small ext2/ext3 partition (i suggest at least 300Mb in
size) at the beginning of disk, so /boot should be your /dev/sda1 ,
then make everything else (root fs, swap, etc...).  Works for me (in
LDOMs and physical servers).

Re: Booting an LDOM (can't find kernel)

2016-07-25 Thread rodent 
On Mon, Jul 25, 2016 at 06:56:35AM +0200, John Paul Adrian Glaubitz wrote:
> 
> It's panicking because you specified the wrong root device:
> 
> [90779.943892] EXT4-fs (vdiska1): mounting ext3 file system using the ext4 
> subsystem
> [90779.951847] EXT4-fs (vdiska1): mounted filesystem with ordered data mode. 
> Opts: (null)
> [90780.025922] Kernel panic - not syncing: Attempted to kill init! 
> exitcode=0x0200
> [90780.025922]
> [90780.026642] CPU: 0 PID: 1 Comm: init Tainted: GE   
> 4.5.0-2-sparc64-smp #1 Debian 4.5.2-1
> [90780.027297] Call Trace:
> [90780.027511]  [0055777c] panic+0xdc/0x260
> [90780.027860]  [0046a270] do_exit+0xb30/0xb40
> [90780.028224]  [0046a330] do_group_exit+0x30/0xc0
> [90780.029623]  [0046a3dc] SyS_exit_group+0x1c/0x40
> [90780.030027]  [00406294] linux_sparc_syscall+0x34/0x44
> [90780.031618] Press Stop-A (L1-A) to return to the boot prom
> [90780.032028] ---[ end Kernel panic - not syncing: Attempted to kill init! 
> exitcode=0x0200
> [90780.032028]
> 
> You specified vdiska1 which is your boot part:
> 
> > Kernel command line: root=/dev/vdiska1 ro console=ttyS0,9600,8n1
> 
> /dev/vdiska1 is the boot partition for SILO, not the root filesystem.

There's only two partitions: vdiska1 which contains everything and vdiska2
which is swap.