EHCI schedule patch for Hardy/Netbooks

[Stefan Bader]

Gustavo Sverzut Barbieri wrote:
> Hello guys,
> 

Hi Gustavo,

> I asked in #kernel and rtg asked me to mail you, so here it is:
> 
> Bug (private):
>     https://bugs.edge.launchpad.net/payson/+bug/309717

If we want to discuss this into the direction from Alan and also if this should 
get into Hardy as a SRU, we will need a public bug for it.

> Patch (backport from mainline 2.6.26, with changes listed below): 
>     http://launchpadlibrarian.net/25896046/ehci-sched.patch
> Mainline patch:
>    http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff_plain;h=b40e43fcc532fa44a375a37d592e32cd0d50fe7a
> Extra References (comment on above bug):
>    [1] https://bugs.edge.launchpad.net/payson/+bug/309717/comments/15
>    [2] https://lists.berlios.de/pipermail/linux-uvc-devel/2008-
> March/003294.html
>    [3] http://www.mail-archive.com/linux-uvc-
> devel at lists.berlios.de/msg02733.html
>    
> 
> Please consider this backport patch to Hardy as it will fix problems not only 
> on this Toshiba, but other usb video devices and even other usb devices.
> 
> The difference from mainline is the second half of the patch, which did not 
> apply given differences between .24 and .26 kernel. As I'm no expert in EHCI 
> and as the original patch [2] author (Alan Stern) did the simplified patch 
> before, I suppose that it is okay, but may need further investigation, maybe 
> ask Alan.
> 

I would not say I really understand the exact implications but from the code 
and comments this looks more like an optimization than something urgently required.

Would we like to get the whole patch, this would depend on one more patch 
(attached). That one too looks like optimizing away unnecessary workload and 
might also had been improving the situation which the other fix addresses.
The big downside is that this is USB, not some rarely used driver. Any change 
we make might have an unexpected side effect. And from the past we know there 
is a good chance to find those not before the moment the changes go into updates.

So from my view, the risk is already present with just the partial upstream 
patch. If that already fixes the seen problem and we did not find other 
problems with it, I would rather go with that.

Stefan

-- 

When all other means of communication fail, try words!


-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0001-USB-ehci-completes-high-speed-ISO-URBs-sooner.patch
Type: text/x-diff
Size: 4177 bytes
Desc: not available
URL: <https://lists.ubuntu.com/archives/kernel-team/attachments/20090424/159b979a/attachment.patch>