Andres Cimmarusti wrote:

> I tested two kernels on Squeeze:
>
> - 3.1.8 from squeeze-backports, which was compiled using gcc 4.4.5 (as
> my own compiled 3.0.x and 3.1.x kernels)
>
> - 3.1.8 from wheezy, compiled using gcc 4.6.2
>
> Both kernels loaded the firmware correctly without any problems. The
> backport comes from the version in testing, so they have the same
> .config file. I've attached both dmesg.
>
> Using wheezy, This same kernel fails to load the firmware everytime,
> as compiled with gcc 4.6.2

Excellent.  So now we've confirmed it's not a kernel problem.

I just noticed the following change upstream:

        commit c3eae82a844b
        Author: Paul Fertser <fercer...@gmail.com>
        Date:   Sat Oct 29 21:52:49 2011 +0400

            Bluetooth: ath3k: output firmware filename when request_firmware 
failed

            This makes it much easier for the users to understand why the driver
            refuses to load when the firmware is unavailable.

which gives me hope that upstream is looking at this kind of thing
(that patch was merged in v3.3-rc1).  This suggests two possible next
steps:

 - Please attach output from "dpkg-query -L firmware-atheros" on a
   squeeze (unaffected) and wheezy (affected) system.

 - Could you try v3.3-rc1 or newer on wheezy?  It works like this:

        # prerequisites
        apt-get install git build-essential

        # get the latest kernel
        git clone \
          git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git

        # or, if you already have a kernel handy:
        cd linux
        git fetch origin

        # check it out
        git checkout origin/master

        # configure it
        cp /boot/config-$(uname -r) .config
        make localmodconfig; # optional: minimize configuration

        # build it
        make deb-pkg; # can add -j<n> for parallel build

        # install and test
        dpkg -i ../<name of package>
        reboot

I don't expect 3.3-rc1 to work any better, but it should give a more
useful error message.

If that doesn't provide any insight, I have a message ready in my
drafts folder asking upstream for more help.

Thanks,
Jonathan



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org

Reply via email to