aacraid adaptec 6805 support?

aacraid adaptec 6805 support?

[Louis-David Mitterrand]

Hi,

I compiled the latest 2.6.38.x with aacraid support but my Adaptec RAID
6805 is not recognized:

http://www.adaptec.com/en-us/support/raid/sas_raid/sas-6805/

When I try building the driver source downloaded from adaptec I get a
compile error:

	CC      drivers/scsi/aacraid/commctrl.o
	drivers/scsi/aacraid/commctrl.c: In function 'open_getadapter_fib':
	drivers/scsi/aacraid/commctrl.c:449:3: error: implicit declaration of function 'init_MUTEX_LOCKED' [-Werror=implicit-function-declaration]
	cc1: some warnings being treated as errors

	make[3]: *** [drivers/scsi/aacraid/commctrl.o] Error 1
	make[2]: *** [drivers/scsi/aacraid] Error 2
	make[1]: *** [drivers/scsi] Error 2
	make: *** [drivers] Error 2

Any idea if that card is supported?

Thanks,

Re: aacraid adaptec 6805 support?

[Emmanuel Florac]

Le Fri, 13 May 2011 16:12:42 +0200
Louis-David Mitterrand <vindex+lists-linux-scsi@apartia.org> écrivait:

> Any idea if that card is supported?
> 

I compiled this driver yesterday with the latest 2.6.35 without
problem. Are you sure you have all the proper kernel includes available
at compilation?

-- 
------------------------------------------------------------------------
Emmanuel Florac     |   Direction technique
                    |   Intellique
                    |	<eflorac@intellique.com>
                    |   +33 1 78 94 84 02
------------------------------------------------------------------------
--
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

Re: aacraid adaptec 6805 support?

[Louis-David Mitterrand]

On Fri, May 13, 2011 at 04:49:22PM +0200, Emmanuel Florac wrote:
> Le Fri, 13 May 2011 16:12:42 +0200
> Louis-David Mitterrand <vindex+lists-linux-scsi@apartia.org> écrivait:
> 
> > Any idea if that card is supported?
> 
> I compiled this driver yesterday with the latest 2.6.35 without
> problem. Are you sure you have all the proper kernel includes available
> at compilation?

Ah, but I am trying with 2.6.38.6 and it seems the driver lags behind
kernel sources.

Thanks for the info
--
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

Re: aacraid adaptec 6805 support?

[Emmanuel Florac]

Le Fri, 13 May 2011 16:54:21 +0200
Louis-David Mitterrand <vindex+lists-linux-scsi@apartia.org> écrivait:

> Ah, but I am trying with 2.6.38.6 and it seems the driver lags behind
> kernel sources.
> 

It looks like this driver version (1.1-7) will be included in 2.6.39
though, so it should compile against the most current kernel. Or
maybe grab the source from the linux-next tree instead ?

-- 
------------------------------------------------------------------------
Emmanuel Florac     |   Direction technique
                    |   Intellique
                    |	<eflorac@intellique.com>
                    |   +33 1 78 94 84 02
------------------------------------------------------------------------
--
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

Re: aacraid adaptec 6805 support?

[Louis-David Mitterrand]

On Fri, May 13, 2011 at 05:24:57PM +0200, Emmanuel Florac wrote:
> Le Fri, 13 May 2011 16:54:21 +0200
> Louis-David Mitterrand <vindex+lists-linux-scsi@apartia.org> écrivait:
> 
> > Ah, but I am trying with 2.6.38.6 and it seems the driver lags behind
> > kernel sources.
> 
> It looks like this driver version (1.1-7) will be included in 2.6.39
> though, so it should compile against the most current kernel. Or
> maybe grab the source from the linux-next tree instead ?

Yep, the driver is included - and works - with 2.6.39-rc7.

Now I have another problem: my 8 sata disks used to work with a
LSI-SAS2008 controller in "no-configuration" mode (meaning I configured
no array on that card, just used the disks for sw raid).

Now when I connect these same disks to the Adaptec controller my first
two /dev/mdX (0 and 1) devices work fine but /dev/md2 doesn't:

	May 13 17:14:37 zenon kernel: md: sde3 does not have a valid v1.2 superblock, no
t importing!

On each disk the third partition /dev/sdX3 is the last and largest but
it seems the Adaptec controller doesn't "see" the end of the disk where
the md superblock is? Does that make sense?

When booting the controller lists all disks in "legacy" mode. What does
that mean vs JBOD?

Thanks,
--
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

Re: aacraid adaptec 6805 support? [sorry if double post...]

[Emmanuel Florac]

Le Fri, 13 May 2011 17:39:22 +0200
Louis-David Mitterrand <vindex+lists-linux-scsi@apartia.org> écrivait:

> On each disk the third partition /dev/sdX3 is the last and largest but
> it seems the Adaptec controller doesn't "see" the end of the disk
> where the md superblock is? Does that make sense?

Not to me, but I've seen weirder things from Adaptec firmwares :)
Are the drives configured as JBOD, or something else? You may segment
the drives (for instance if their sizes differ slightly), so maybe some
of them appear as volume segments instead of JBOD?

> When booting the controller lists all disks in "legacy" mode. What
> does that mean vs JBOD?

I have no idea. However I've seen adapters from different builders
partly (and wrongly) recognize each other on-disk headers. I've
recently moved a bunch of disks from an LSI 9200 4i4e to an Adaptec
6445, but I don't remember of any particular weirdness.

-- 
------------------------------------------------------------------------
Emmanuel Florac     |   Direction technique
                    |   Intellique
                    |	<eflorac@intellique.com>
                    |   +33 1 78 94 84 02
------------------------------------------------------------------------
--
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

Re: aacraid adaptec 6805 support?

[Louis-David Mitterrand]

On Fri, May 13, 2011 at 06:33:56PM +0200, Emmanuel Florac wrote:
> Le Fri, 13 May 2011 17:39:22 +0200
> Louis-David Mitterrand <vindex+lists-linux-scsi@apartia.org> écrivait:
> 
> > On each disk the third partition /dev/sdX3 is the last and largest but
> > it seems the Adaptec controller doesn't "see" the end of the disk
> > where the md superblock is? Does that make sense?
> 
> Not to me, but I've seen weirder things from Adaptec firmwares :)
> Are the drives configured as JBOD, or something else? You may segment
> the drives (for instance if their sizes differ slightly), so maybe some
> of them appear as volume segments instead of JBOD?

Actually I didn't configure the Adaptec controller at all, just let it
boot with disks transfered from the LSI controller. So the disks are not
JBOD, in the adaptec sense at least, which requires "initializing"
(meaning wiping) disks. I thought JBOD meant "just use the disks as they
are" but that is in fact what adaptec calls "legacy" mode:

	May 13 17:09:57 zenon kernel: scsi 0:0:0:0: Direct-Access     Adaptec  6805 Legacy      V1.0 PQ: 0 ANSI: 2

And in that "legacy" mode the disk geometry appears slightly smaller to
the OS than with the LSI controller, hence my problem with /dev/md2. And
/dev/md2 is a huge backup partition that would be very inconvenient to
wipe and re-create (xfs doesn't support shrinking alas, so I can' shrink
the partition).

Anyway thanks for your input,
--
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

Re: aacraid adaptec 6805 support?

[Emmanuel Florac]

Le Sat, 14 May 2011 08:25:33 +0200 vous écriviez:

> And in that "legacy" mode the disk geometry appears slightly smaller
> to the OS than with the LSI controller, hence my problem
> with /dev/md2.

Yes, some controllers reserve out a part of the drive to allow for
using slightly different drives to repair the array. Unfortunately that
parameter may or may not be user-tunable.

-- 
------------------------------------------------------------------------
Emmanuel Florac     |   Direction technique
                    |   Intellique
                    |	<eflorac@intellique.com>
                    |   +33 1 78 94 84 02
------------------------------------------------------------------------
--
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