Host Protected Area unconditional disable

classic Classic list List threaded Threaded
5 messages Options
Reply | Threaded
Open this post in threaded view
|

Host Protected Area unconditional disable

Andy Whitcroft-3
Traditionally we have turned off Host Protected Area (HPA) on drives
when it was detected.  Once we had installs in the field in those mode
we were then unable to turn HPA back on without breaking older installs
which used the extra space.  The flip side has always been that once
HPA has been turned off a power cycle is needed to turn it back on.
This can severely confuse another operating system when it finds the
disk has changed size.  In rare cases it can cause RAID cards to drop
RAID sets on the floor thinking they are corrupt.  All bad.

We have long had a request to turn off the unconditional disabling of
HPA [1].  However as mentioned above this would render those who had
made use of the protected space in their partitions unbootable.  The
solution seemed to be to try and work without disabling HPA and only
disable it if partition access requires it.

It seems that in v2.6.35 a series of patches were committed to mainline
which handle this situation in exactly the suggested manner, the following
commits are those which extended the existing IDE functionality to
ATA drives:

    libata: implement on-demand HPA unlocking
    libata: use the enlarged capacity after late HPA unlock
    SCSI: implement sd_unlock_native_capacity()

It therefore seems that from v2.6.35 we should be able to drop our force
disable of HPA in the commit below.  

  commit eaea8ccbd3adf0b36942ae834eaa825094772c95
  Author: Scott James Remnant <[hidden email]>
  Date: Tue Mar 3 14:20:01 2009 +0000

    UBUNTU: SAUCE: (no-up) libata: Ignore HPA by default.

I am therefore proposing we drop this latter patch for Natty and see who
screams (I am hoping there will be none).  We might want to get a call
for testing out on this feature for natty-alpha-2.

Comments?

-apw

[1] https://bugs.launchpad.net/ubuntu/+source/linux/+bug/380138

--
kernel-team mailing list
[hidden email]
https://lists.ubuntu.com/mailman/listinfo/kernel-team
Reply | Threaded
Open this post in threaded view
|

Re: Host Protected Area unconditional disable

Scott James Remnant-2
I only authored the patch that did the module defaults change in the
kernel rather than as a modprobe.d file.  Part of my reasoning for doing
this was making the fact we have a different default irritate the kernel
team enough that they'd eventually get around to a solution that meant
we didn't have to :p

I'm glad to see my cunning plan worked.

Scott

On 22/12/10 11:29, Andy Whitcroft wrote:

> Traditionally we have turned off Host Protected Area (HPA) on drives
> when it was detected.  Once we had installs in the field in those mode
> we were then unable to turn HPA back on without breaking older installs
> which used the extra space.  The flip side has always been that once
> HPA has been turned off a power cycle is needed to turn it back on.
> This can severely confuse another operating system when it finds the
> disk has changed size.  In rare cases it can cause RAID cards to drop
> RAID sets on the floor thinking they are corrupt.  All bad.
>
> We have long had a request to turn off the unconditional disabling of
> HPA [1].  However as mentioned above this would render those who had
> made use of the protected space in their partitions unbootable.  The
> solution seemed to be to try and work without disabling HPA and only
> disable it if partition access requires it.
>
> It seems that in v2.6.35 a series of patches were committed to mainline
> which handle this situation in exactly the suggested manner, the following
> commits are those which extended the existing IDE functionality to
> ATA drives:
>
>      libata: implement on-demand HPA unlocking
>      libata: use the enlarged capacity after late HPA unlock
>      SCSI: implement sd_unlock_native_capacity()
>
> It therefore seems that from v2.6.35 we should be able to drop our force
> disable of HPA in the commit below.
>
>    commit eaea8ccbd3adf0b36942ae834eaa825094772c95
>    Author: Scott James Remnant<[hidden email]>
>    Date: Tue Mar 3 14:20:01 2009 +0000
>
>      UBUNTU: SAUCE: (no-up) libata: Ignore HPA by default.
>
> I am therefore proposing we drop this latter patch for Natty and see who
> screams (I am hoping there will be none).  We might want to get a call
> for testing out on this feature for natty-alpha-2.
>
> Comments?
>
> -apw
>
> [1] https://bugs.launchpad.net/ubuntu/+source/linux/+bug/380138


--
kernel-team mailing list
[hidden email]
https://lists.ubuntu.com/mailman/listinfo/kernel-team
Reply | Threaded
Open this post in threaded view
|

Re: Host Protected Area unconditional disable

Ben Hutchings-3
In reply to this post by Andy Whitcroft-3
On Wed, 2010-12-22 at 11:29 +0000, Andy Whitcroft wrote:
[...]

> It seems that in v2.6.35 a series of patches were committed to mainline
> which handle this situation in exactly the suggested manner, the following
> commits are those which extended the existing IDE functionality to
> ATA drives:
>
>     libata: implement on-demand HPA unlocking
>     libata: use the enlarged capacity after late HPA unlock
>     SCSI: implement sd_unlock_native_capacity()
>
> It therefore seems that from v2.6.35 we should be able to drop our force
> disable of HPA in the commit below.  
[...]
> Comments?

Tejun Heo kindly implemented these changes after I reported that this
had become a problem when switching from IDE to libata.  I backported
them to Debian's 2.6.32 kernel and so far we've had no more bug reports.

Ben.

--
Ben Hutchings
Once a job is fouled up, anything done to improve it makes it worse.

--
kernel-team mailing list
[hidden email]
https://lists.ubuntu.com/mailman/listinfo/kernel-team

signature.asc (845 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

Re: Host Protected Area unconditional disable

Andy Whitcroft-3
On Wed, Dec 22, 2010 at 03:20:14PM +0000, Ben Hutchings wrote:

> Tejun Heo kindly implemented these changes after I reported that this
> had become a problem when switching from IDE to libata.  I backported
> them to Debian's 2.6.32 kernel and so far we've had no more bug reports.

Indeed qudos to you both for getting this resolved.

:)

-apw

--
kernel-team mailing list
[hidden email]
https://lists.ubuntu.com/mailman/listinfo/kernel-team
Reply | Threaded
Open this post in threaded view
|

Re: Host Protected Area unconditional disable

Tim Gardner-2
On 12/22/2010 08:27 AM, Andy Whitcroft wrote:

> On Wed, Dec 22, 2010 at 03:20:14PM +0000, Ben Hutchings wrote:
>
>> Tejun Heo kindly implemented these changes after I reported that this
>> had become a problem when switching from IDE to libata.  I backported
>> them to Debian's 2.6.32 kernel and so far we've had no more bug reports.
>
> Indeed qudos to you both for getting this resolved.
>
> :)
>
> -apw
>

reverted and pushed

--
Tim Gardner [hidden email]

--
kernel-team mailing list
[hidden email]
https://lists.ubuntu.com/mailman/listinfo/kernel-team