megaraid_sas: which patches are needed for 2.6.32.y?

megaraid_sas: which patches are needed for 2.6.32.y?

[Jonathan Nieder]

Hi again,

Jonathan Nieder wrote[0]:

> I am working on improving support for newer megaraid controllers in
> Debian squeeze's 2.6.32.y-based kernel.
>
> On top of what's in 2.6.32.y, Debian squeeze currently has the
> following patches applied:
[
 list at [1] below
]
> When Buford Peek needed support for a Dell PERC H710P I provided him
> with the following as a quick set of patches to test:
[
 list at [2]
]
> The result seemed to work ok. :)  But it left out at least one
> important fix:
>
>   f575c5d3ebdc Fix instance access in megasas_reset_timer
>
> It also is not particularly minimal.
>
> After applying all of the above, the following commits still remain
> unapplied:
[
 list at [3]
]
> Questions:
>
> 1) Are any of these important fixes that should be applied to 2.6.32.y
>    so everyone who uses a 2.6.32-based kernel can start benefitting
>    from them?

No reply.  Is this the right address to reach the megaraid_sas
maintainers?  Would you be interested in making sure the
2.6.32.y-based kernels used by various Linux distributors have
up-to-date hardware support and fixes?  What can I do to help make
that happen?

Thanks,
Jonathan

[0] http://bugs.debian.org/666108
[1] Patches on top of 2.6.32.y used to add support for the 9240 family
and similar devices:

>   c35188377f12 Add poll mechanism to megaraid sas driver
>   72c4fd36dc7f add sysfs for AEN polling
>   879111224d07 Add new megaraid SAS 2 controller support to the driver
>   7bebf5c79cb6 allocate the application cmds to sas2 controller
>   0c79e681eef1 Fix the fix for fw hang caused by megaraid sas application

[2] Patches used to add Dell PERC H710P support:

>   81e403ce3c6a infrastructure to get PDs from FW
>   7e8a75f4dfbf Add the support for updating the OS after adding/removing
>                the devices from FW
>   bdc6fb8d69fa add the logical drive list to driver
>   c978684254d1 driver fixed the device update issue
>   f4c9a1317d32 add the IEEE SGE support to SAS2 controller
>   39a985547cbf Add Online Controller Reset to MegaRAID SAS drive
>   0d49016bbab4 Rename megaraid_sas.c to megaraid_sas_base.c
>   80d9da98b403 Add MSI-X support and msix_disable module parameter
>   aeab3fd7b865 make driver PCI legacy I/O port free driver
>   b6d5d8808b4c Use lowest memory bar for SR-IOV VF support
>   cd50ba8ede5c Add struct megasas_instance_template changes
>   9c915a8c99bc Add 9565/9285 specific code

[3] Patches in 3.0.y absent from the above two lists:

>   8d5682532133 tape drive support fix
>   044833b572b9 report system PDs to OS
>   7218df69e360 use the firmware boot timeout when waiting for commands
>   a0b7736828f7 Update version number and documentation
>   780a3762fb92 Zero pad_0 in mfi structure
>   63bad45db1d4 version and documentation update
>   837f5fe89c84 support devices update flag
>   1fd106851698 Add input parameter for max_sectors
>   707e09bd867c Add three times Online controller reset
>   e340c3537239 Version and documentation update
>   3f1530c1e1f7 Update GPL headers.
>   eb1b12377376 Fix failure gotos
>   53ef2bbd2068 Add missing check_and_restore_queue_depth call
>   0a77066acc78 Enable MSI-X before calling megasas_init_fw
>   e1419191d8d3 Call tasklet_schedule for MSI-X
>   66192dfe1e74 Fix probe_one to clear MSI-X flags in kdump
>   f51244058963 Fix megasas_build_dcdb_fusion to not filter by TYPE_DISK
>   eaa3c240de25 Fix megasas_build_dcdb_fusion to use correct LUN field
>   4c598b23807a Add CFG_CLEARED AEN
>   f86c5424b027 Fix tasklet_init call
>   1ac515ef3f2f Fix fault state handling
>   42a8d2b34d10 Fix max_sectors for IEEE SGL
>   ebf054b00b0a Fix iMR OCR support to work correctly
>   00fa2b191b4b Version and Changelog update
>   3f1abce4aba4 Remove MSI-X black list, use MFI_REG_STATE instead
>   70d031f36fa5 Remove un-used function
>   7e70e7336515 Check MFI_REG_STATE.fault.resetAdapter
>   46fd256e0558 Disable interrupts/free_irq() in megasas_shutdown()
>   541f90b7c6df Fix bug where AENs could be lost in probe() and resume()
>   495c5609700e Convert 6,10,12 byte CDB's for FastPath IO
>   3cc6851f9a35 Add 1078 OCR support
>   4f788dce0baf Version and Changelog update

Re: megaraid_sas: which patches are needed for 2.6.32.y?

[adam radford]

On Sun, Jun 24, 2012 at 8:13 AM, Jonathan Nieder <jrnieder@gmail.com> wrote:
...
>> Questions:
>>
>> 1) Are any of these important fixes that should be applied to 2.6.32.y
>>    so everyone who uses a 2.6.32-based kernel can start benefitting
>>    from them?

If you have already backported the patches from list 2 to Debian
2.6.32.x kernel, please also backport the patches from list 3.

-Adam
--
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: megaraid_sas: which patches are needed for 2.6.32.y?

[Jonathan Nieder]

adam radford wrote:
> On Sun, Jun 24, 2012 at 8:13 AM, Jonathan Nieder <jrnieder@gmail.com> wrote:

>>> Questions:
>>>
>>> 1) Are any of these important fixes that should be applied to 2.6.32.y
>>>    so everyone who uses a 2.6.32-based kernel can start benefitting
>>>    from them?
>
> If you have already backported the patches from list 2 to Debian
> 2.6.32.x kernel, please also backport the patches from list 3.

Debian's squeeze kernel currently only has the five patches marked [1].

The patches marked [2] have been backported and tested but not applied
(waiting on advice from knowledgeable folks such as you :)).

If someone makes a backport of the patches marked [3] as well, would
you be able to look over or test the combination?  That would be much
appreciated.  If that's not possible, your advice would still be
welcome but we will need to use a more minimal set of changes.

Thanks,
Jonathan
--
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: megaraid_sas: which patches are needed for 2.6.32.y?

[adam radford]

On Thu, Jun 28, 2012 at 1:27 PM, Jonathan Nieder <jrnieder@gmail.com> wrote:
...
> If someone makes a backport of the patches marked [3] as well, would
> you be able to look over or test the combination?  That would be much
> appreciated.  If that's not possible, your advice would still be
> welcome but we will need to use a more minimal set of changes.

Yes, I will be able to look over and/or test the combination.  If you
apply all patches then diff vs. the scsi-misc tip, you should only see
kernel 2.6.32.y specific changes.

-Adam

Re: megaraid_sas: which patches are needed for 2.6.32.y?

[Jonathan Nieder]

Hi,

In June, adam radford wrote:
> On Thu, Jun 28, 2012 at 1:27 PM, Jonathan Nieder <jrnieder@gmail.com> wrote:

>> If someone makes a backport of the patches marked [3] as well, would
>> you be able to look over or test the combination?  That would be much
>> appreciated.  If that's not possible, your advice would still be
>> welcome but we will need to use a more minimal set of changes.
>
> Yes, I will be able to look over and/or test the combination.  If you
> apply all patches then diff vs. the scsi-misc tip, you should only see
> kernel 2.6.32.y specific changes.

Sorry for the long delay.

I've made a backport with all relevant patches I could find from the
3.0.y tree and put the patches and a built package at

  http://alioth.debian.org/~jrnieder-guest/temp/driver-test/

Patches are against 2.6.32.y.  The most notable missing patch is the
BKL conversion --- for Debian squeeze it seems safest to just keep
using the BKL.  I'd be happy for any thoughts you have.

Thanks,
Jonathan

Re: megaraid_sas: which patches are needed for 2.6.32.y?

[Jonathan Nieder]

On Mon, Oct 29, 2012, Jonathan Nieder wrote:
> In June, adam radford wrote:
>> On Thu, Jun 28, 2012 at 1:27 PM, Jonathan Nieder <jrnieder@gmail.com> wrote:

>>> If someone makes a backport of the patches marked [3] as well, would
>>> you be able to look over or test the combination?  That would be much
>>> appreciated.
[...]
>> Yes, I will be able to look over and/or test the combination.
[...]
> I've made a backport with all relevant patches I could find from the
> 3.0.y tree and put the patches and a built package at
>
>   http://alioth.debian.org/~jrnieder-guest/temp/driver-test/
>
> Patches are against 2.6.32.y.

Gentle ping.  LSI folks, would it make sense for Debian to apply this
series to their 2.6.32.y-based kernel?  Would some other format (a git
tree instead of the current "git am"-ready patch series) be easier to
review?

Thanks for your kindness,
Jonathan

Re: megaraid_sas: which patches are needed for 2.6.32.y?

[Jonathan Nieder]

tags 666108 - moreinfo
quit

Torbjørn Thorsen wrote:

> The kernel boots with no problems, the storage controller seems to be
> working nicely.
>
> root@xen14:~# dmesg | grep -i mega
> [    3.340638] megasas: 00.00.05.38-rc1 Wed. May. 11 17:00:00 PDT 2011
> [    3.340742] megasas: 0x1000:0x005b:0x1028:0x1f38: bus 3:slot 0:func 0

Yay, thanks much for checking.

Ben and Dann, what do you think?  Should we apply these patches[1]?

Regards,
Jonathan

[1] http://alioth.debian.org/~jrnieder-guest/temp/driver-test/megaraid-driver-backport.mbox
--
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