FAQ Search Today's Posts Mark Forums Read
» Video Reviews

» Linux Archive

Linux-archive is a website aiming to archive linux email lists and to make them easily accessible for linux users/developers.


» Sponsor

» Partners

» Sponsor

Go Back   Linux Archive > Ubuntu > Ubuntu Kernel Team

 
 
LinkBack Thread Tools
 
Old 01-25-2010, 02:31 PM
Stefan Bader
 
Default Bogus EEPROM checksum and now?

Hi Luis,

ok this was somewhat to be expected when adding the patch to check for the
EEPROM checksum to stable[1]. The question is, is the a workable way to get that
fixed on the adapter or would it not have been better to have at least an option
to allow users to override it with the message that they cannot submit any
bugreports that get listened to?

-Stefan

[1] https://bugs.launchpad.net/ubuntu/+source/linux/+bug/506180?comments=all


--
kernel-team mailing list
kernel-team@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/kernel-team
 
Old 01-25-2010, 06:31 PM
Stefan Bader
 
Default Bogus EEPROM checksum and now?

Stefan Bader wrote:
> Hi Luis,
>
> ok this was somewhat to be expected when adding the patch to check for the
> EEPROM checksum to stable[1]. The question is, is the a workable way to get that
> fixed on the adapter or would it not have been better to have at least an option
> to allow users to override it with the message that they cannot submit any
> bugreports that get listened to?

Note: I overlooked that this specific case was already fixed by a later patch
that did not get into 2.6.31.y. Still generically it would be interesting what
would be the approach if there is someone approaching with a invalid checksum.
Are the EEPROMs to your knowledge simply be upgradeable by vendors (if they
care) or would you think an override acceptable?

-Stefan

>
> [1] https://bugs.launchpad.net/ubuntu/+source/linux/+bug/506180?comments=all
>
>


--
kernel-team mailing list
kernel-team@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/kernel-team
 
Old 01-25-2010, 06:35 PM
"Luis R. Rodriguez"
 
Default Bogus EEPROM checksum and now?

On Mon, Jan 25, 2010 at 07:31:17AM -0800, Stefan Bader wrote:
> Hi Luis,
>
> ok this was somewhat to be expected when adding the patch to check for the
> EEPROM checksum to stable[1]. The question is, is the a workable way to get that
> fixed on the adapter or would it not have been better to have at least an option
> to allow users to override it with the message that they cannot submit any
> bugreports that get listened to?
>
> -Stefan
>
> [1] https://bugs.launchpad.net/ubuntu/+source/linux/+bug/506180?comments=all

EEPROM checksumming should always work and if the card's EEPROM checksum
does not check out the card is likely busted. The software work arounds
to add support for a card with busted EEPROM are just too much if a pain
in the ass to deal with, it would introduce a lot of complexity. In the
best case scenerio a sample EEPROM would need to be provided for each
type of EEPROM supported (or a few) with some default values which we
can gaurantee work well for the card. Note that the cards's EEPROM even
contains calibrate data, and antenna information, so this makes this
really difficult to do, and I would simply advise against it.

Luis

--
kernel-team mailing list
kernel-team@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/kernel-team
 
Old 01-25-2010, 08:08 PM
"Luis R. Rodriguez"
 
Default Bogus EEPROM checksum and now?

On Mon, Jan 25, 2010 at 11:31:07AM -0800, Stefan Bader wrote:
> Stefan Bader wrote:
> > Hi Luis,
> >
> > ok this was somewhat to be expected when adding the patch to check for the
> > EEPROM checksum to stable[1]. The question is, is the a workable way to get that
> > fixed on the adapter or would it not have been better to have at least an option
> > to allow users to override it with the message that they cannot submit any
> > bugreports that get listened to?
>
> Note: I overlooked that this specific case was already fixed by a later patch
> that did not get into 2.6.31.y.

Indeed, also AFAICT I did not push the EEPROM checksum into 2.6.31, was that
a patch you cherry picked in to 2.6.31 for Ubuntu?

> Still generically it would be interesting what
> would be the approach if there is someone approaching with a invalid checksum.

Just tell them their card is busted. There was indeed a bug in the checksum
computation, and I fixed that.

> Are the EEPROMs to your knowledge simply be upgradeable by vendors (if they
> care) or would you think an override acceptable?

They are programmed once, that's it. Its done by the ODMs based on samplying
of a card out of a group. The checksum shall never fail, the other OS drivers
also use this same checksum.

Luis

--
kernel-team mailing list
kernel-team@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/kernel-team
 
Old 01-25-2010, 09:47 PM
"Luis R. Rodriguez"
 
Default Bogus EEPROM checksum and now?

On Mon, Jan 25, 2010 at 11:31 AM, Stefan Bader
<stefan.bader@canonical.com> wrote:
> Stefan Bader wrote:
>> Hi Luis,
>>
>> ok this was somewhat to be expected when adding the patch to check for the
>> EEPROM checksum to stable[1]. The question is, is the a workable way to get that
>> fixed on the adapter or would it not have been better to have at least an option
>> to allow users to override it with the message that they cannot submit any
>> bugreports that get listened to?
>
> Note: I overlooked that this specific case was already fixed by a later patch
> that did not get into 2.6.31.y. Still generically it would be interesting what
> would be the approach if there is someone approaching with a invalid checksum.
> Are the EEPROMs to your knowledge simply be upgradeable by vendors (if they
> care) or would you think an override acceptable?

I replied to these e-mails but from my atheros address and those
didn't get accepted. Can someone ACK them? I'll go sign up with that
address now for future e-mails.

Luis

--
kernel-team mailing list
kernel-team@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/kernel-team
 
Old 01-26-2010, 07:36 AM
Stefan Bader
 
Default Bogus EEPROM checksum and now?

Luis R. Rodriguez wrote:
> On Mon, Jan 25, 2010 at 11:31:07AM -0800, Stefan Bader wrote:
>> Stefan Bader wrote:
>>> Hi Luis,
>>>
>>> ok this was somewhat to be expected when adding the patch to check for the
>>> EEPROM checksum to stable[1]. The question is, is the a workable way to get that
>>> fixed on the adapter or would it not have been better to have at least an option
>>> to allow users to override it with the message that they cannot submit any
>>> bugreports that get listened to?
>> Note: I overlooked that this specific case was already fixed by a later patch
>> that did not get into 2.6.31.y.
>
> Indeed, also AFAICT I did not push the EEPROM checksum into 2.6.31, was that
> a patch you cherry picked in to 2.6.31 for Ubuntu?

I picked commit 359207c687cc8f4f9845c8dadd0d6dabad44e584
Author: Luis R. Rodriguez <lrodriguez@atheros.com>
Date: Mon Jan 4 10:40:39 2010 -0500

ath5k: Fix eeprom checksum check for custom sized eeproms

which was mentioned in the bug report and also seems to be the only ath5k patch
between adding the checksum code and HEAD that seems to be related. If I missed
another one, please tell.


>> Still generically it would be interesting what
>> would be the approach if there is someone approaching with a invalid checksum.
>
> Just tell them their card is busted. There was indeed a bug in the checksum
> computation, and I fixed that.
>
>> Are the EEPROMs to your knowledge simply be upgradeable by vendors (if they
>> care) or would you think an override acceptable?
>
> They are programmed once, that's it. Its done by the ODMs based on samplying
> of a card out of a group. The checksum shall never fail, the other OS drivers
> also use this same checksum.

Thanks a lot for the explanations. Its good to know that.

> Luis


--
kernel-team mailing list
kernel-team@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/kernel-team
 
Old 01-26-2010, 07:37 AM
Stefan Bader
 
Default Bogus EEPROM checksum and now?

Luis R. Rodriguez wrote:
> On Mon, Jan 25, 2010 at 11:31 AM, Stefan Bader
> <stefan.bader@canonical.com> wrote:
>> Stefan Bader wrote:
>>> Hi Luis,
>>>
>>> ok this was somewhat to be expected when adding the patch to check for the
>>> EEPROM checksum to stable[1]. The question is, is the a workable way to get that
>>> fixed on the adapter or would it not have been better to have at least an option
>>> to allow users to override it with the message that they cannot submit any
>>> bugreports that get listened to?
>> Note: I overlooked that this specific case was already fixed by a later patch
>> that did not get into 2.6.31.y. Still generically it would be interesting what
>> would be the approach if there is someone approaching with a invalid checksum.
>> Are the EEPROMs to your knowledge simply be upgradeable by vendors (if they
>> care) or would you think an override acceptable?
>
> I replied to these e-mails but from my atheros address and those
> didn't get accepted. Can someone ACK them? I'll go sign up with that
> address now for future e-mails.
>
> Luis

Thanks, I'll add an ACK from you as well.

Stefan

--
kernel-team mailing list
kernel-team@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/kernel-team
 
Old 01-26-2010, 06:17 PM
"Luis R. Rodriguez"
 
Default Bogus EEPROM checksum and now?

On Tue, Jan 26, 2010 at 12:36:56AM -0800, Stefan Bader wrote:
> Luis R. Rodriguez wrote:
> > On Mon, Jan 25, 2010 at 11:31:07AM -0800, Stefan Bader wrote:
> >> Stefan Bader wrote:
> >>> Hi Luis,
> >>>
> >>> ok this was somewhat to be expected when adding the patch to check for the
> >>> EEPROM checksum to stable[1]. The question is, is the a workable way to get that
> >>> fixed on the adapter or would it not have been better to have at least an option
> >>> to allow users to override it with the message that they cannot submit any
> >>> bugreports that get listened to?
> >> Note: I overlooked that this specific case was already fixed by a later patch
> >> that did not get into 2.6.31.y.
> >
> > Indeed, also AFAICT I did not push the EEPROM checksum into 2.6.31, was that
> > a patch you cherry picked in to 2.6.31 for Ubuntu?
>
> I picked commit 359207c687cc8f4f9845c8dadd0d6dabad44e584
> Author: Luis R. Rodriguez <lrodriguez@atheros.com>
> Date: Mon Jan 4 10:40:39 2010 -0500
>
> ath5k: Fix eeprom checksum check for custom sized eeproms
>
> which was mentioned in the bug report and also seems to be the only ath5k patch
> between adding the checksum code and HEAD that seems to be related. If I missed
> another one, please tell.

Nope that's all, that's the only other fix needed.

> >> Still generically it would be interesting what
> >> would be the approach if there is someone approaching with a invalid checksum.
> >
> > Just tell them their card is busted. There was indeed a bug in the checksum
> > computation, and I fixed that.
> >
> >> Are the EEPROMs to your knowledge simply be upgradeable by vendors (if they
> >> care) or would you think an override acceptable?
> >
> > They are programmed once, that's it. Its done by the ODMs based on samplying
> > of a card out of a group. The checksum shall never fail, the other OS drivers
> > also use this same checksum.
>
> Thanks a lot for the explanations. Its good to know that.

No problem.

Luis

--
kernel-team mailing list
kernel-team@lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/kernel-team
 

Thread Tools




All times are GMT. The time now is 01:07 PM.

VBulletin, Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
Content Relevant URLs by vBSEO ©2007, Crawlability, Inc.
Copyright 2007 - 2008, www.linux-archive.org