First Last Prev Next    No search results available      Search page      Enter new bug
Bug#: 133308
Alias:
Product:
Component:
Status: RESOLVED
Resolution: FIXED
Assigned To: Daniel Drake <dsd@gentoo.org>
Hardware:
OS:
Version:
Priority:
Severity:
Reporter: Jim Bowlin <bowlin@mindspring.com>
Add CC:
CC:
Remove selected CCs
URL:
Summary:
Status Whiteboard:
Keywords:

Filename Description Type Creator Created Size Actions
gentoo-sources.patch diff -u gentoo-sources-2.6.16-r[78].ebuild patch Jim Bowlin 2006-05-14 11:09 0000 1.39 KB Details | Diff
2.6.16.x_ieee1394_v71.patch unbzip2'ed kernel patch (92K) patch Jim Bowlin 2006-05-14 11:10 0000 90.69 KB Details | Diff
Create a New Attachment (proposed patch, testcase, etc.) View All

Bug 133308 depends on: Show dependency tree
Bug 133308 blocks:
Votes: 0    Show votes for this bug    Vote for this bug

Additional Comments: (this is where you put emerge --info)


Not eligible to see or edit group visibility for this bug.






View Bug Activity   |   Format For Printing   |   XML   |   Clone This Bug


Description:   Opened: 2006-05-14 11:04 0000
Hi!

Please find attached a diff file for creating gentoo-sources-2.6.16-r8.ebuild
from gentoo-sources-2.6.16-r7.ebuild and also a (92K!) patch file that fixes
the breakage of some (cheap) firewire HDD devices when upgrading the kernel
from 2.6.15 to 1.6.16.x.  The patch only works for x >= 2.

The bug and links to the patch can be found in this kernel bug report:
http://bugzilla.kernel.org/show_bug.cgi?id=6356

I downloaded the kernel patch from here:
http://me.in-berlin.de/~s5r6/linux1394/updates/2.6.16.x/2.6.16.x_ieee1394_v71.patch.bz2

The kernel bug is marked resolved so I think that means Linus merged it into
the kernel, but it wasn't in the latest stable Gentoo kernel: 2.6.16-r7.

To be on the safe side, I added an "oldfirewire" use flag that needs to be set
in order for the patch to be applied.

------- Comment #1 From Jim Bowlin 2006-05-14 11:09:15 0000 -------
Created an attachment (id=86757) [details]
diff -u gentoo-sources-2.6.16-r[78].ebuild

Here are the changes I made to the ebuild.

------- Comment #2 From Jim Bowlin 2006-05-14 11:10:59 0000 -------
Created an attachment (id=86759) [details]
unbzip2'ed kernel patch (92K)

Here is the actually kernel patch that I downloaded from:
http://me.in-berlin.de/~s5r6/linux1394/updates/2.6.16.x/2.6.16.x_ieee1394_v71.patch.bz2

------- Comment #3 From Daniel Drake 2006-05-14 15:12:36 0000 -------
That's a huge patch which addresses a multitude of things unrelated to your
storage issue. Also, we don't apply patches based on use-flags, see
http://dev.gentoo.org/~dsd/genpatches/faq.htm

However, if you can confirm that the individual patch applied to 2.6.16 solves
the issue then we will include it in genpatches when Linus merges it.

The patch is at:
http://marc.theaimsgroup.com/?l=linux-kernel&m=114579421114791&q=raw

However, Stefan says here that he has a separate patch for 2.6.16:
http://marc.theaimsgroup.com/?l=linux1394-devel&m=114699363506345&w=2

Maybe it is best that we wait for that.

------- Comment #4 From Jim Bowlin 2006-05-14 16:22:05 0000 -------
Thanks for setting me straight on a number of items.

I tried the patch you linked to but it failed on
drivers/ieee1394/ohci1394.c although the 2nd part did work
on sbp2.c. You can see that patch explicitly refers to the 2.6.17 build.

The line of code to change exists but it is in the wrong place.
The patch is looking near line 533 but the code in the file is
on line 2530.

The same patch is in your second link but without the explicit
2.6.17 reference. I went to Stefan's 2.6.16.x patch directory:
http://me.in-berlin.de/~s5r6/linux1394/updates/2.6.16.x/patches/

and I found the same patch:
http://me.in-berlin.de/~s5r6/linux1394/updates/2.6.16.x/patches/60+ohci1394-sbp2-fix-scsi_add_device-failed-with-PL_3507-based-devices.patch

So I then looked through the large patch (that I pointed to originally)
and I found that the short patch you gave me depends upon at least one
earlier patch.

The next step seems to wget all of Stefan's 2.6.16.x patches and grep
around to find which preceding patch(es) is/are needed.  But I am
running out of steam now and may not dig any further.

Current status:  The larger patch did work on 2.6.16 but the smaller
one did not.  It may be possible to find a smaller subset of patches
that will work but I am not up to the task and even if I could get it
to work for my setup I feel confident I would break things for others.

I can understand why you don't want to use the larger patch.  As an
alternative you might want to mark 2.6.16 unstable again or issue an
einfo or ewarn that explains that 2.6.16 breaks firewire support for
devices using the PL-3507 chipset.

Thanks again for your help.  If there is something else you want me to try
out, please let me know.

------- Comment #5 From Daniel Drake 2006-05-14 16:50:41 0000 -------
I linked to the second post because Stefan states that he will submit a
backported patch for 2.6.16 linux-stable once Linus accepts the real fix into
2.6.17. We'll just wait for that to happen, the -stable tree will trickle into
genpatches as usual, or we'll pick it up manually.

------- Comment #6 From Daniel Drake 2006-05-15 12:24:22 0000 -------
Please apply this patch against 2.6.16 and confirm the problem goes away:
http://bugzilla.kernel.org/attachment.cgi?id=8117&action=view

------- Comment #7 From Jim Bowlin 2006-05-16 10:12:40 0000 -------
The new patch works and it fixes the problem.

Thanks.  

------- Comment #8 From Daniel Drake 2006-05-21 08:32:38 0000 -------
Thanks for testing. Fixed in gentoo-sources-2.6.16-r8 (genpatches-2.6.16-10)

First Last Prev Next    No search results available      Search page      Enter new bug