Skip to contentSkip to conversation

Linux 2.6.22.8

DISCUSSION Greg Kroah-Hartman 3 participants 10 messages Sep 25, 2007 1 message not in the archive

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

We (the -stable team) are announcing the release of the 2.6.22.8 kernel. It contains a single security bugfix for the the sound subsystem. There is potential for local privilege escalation, so all users are certainly encouraged to upgrade.

I'll also be replying to this message with a copy of the patch between 2.6.22.7 and 2.6.22.8

The updated 2.6.22.y git tree can be found at:

        git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-2.6.22.y.git
and can be browsed at the normal kernel.org git web browser:
        http://git.kernel.org/?p=linux/kernel/git/stable/linux-2.6.22.y.git;a=summary

thanks,

greg k-h

--------

 Makefile              |    2 -
 sound/core/memalloc.c |   68 ++++++++++++++++++++++++++++----------------------
 2 files changed, 40 insertions(+), 30 deletions(-)

Summary of changes from v2.6.22.7 to v2.6.22.8
==============================================

Greg Kroah-Hartman (1):

      Linux 2.6.22.8

Takashi Iwai (1):
Convert snd-page-alloc proc file to use seq_file (CVE-2007-4571)

#1 lore raw
Makefile +1 -1
index 12edea0..dc7a45d 100644
@@ -1,7 +1,7 @@
 VERSION = 2
 PATCHLEVEL = 6
 SUBLEVEL = 22
-EXTRAVERSION = .7
+EXTRAVERSION = .8
 NAME = Holy Dancing Manatees, Batman!
 
 # *DOCUMENTATION*
sound/core/memalloc.c +39 -29

140 lines not on this page · raw message

author
#3 lore raw
Quoted 4 lines · Oliver Pinter

Hi!

the queue for 2.6.22.8 (2.6.22.8-rc1) is for the next release or thuse paches droped?

The queue is for the next release, which will be called 2.6.22.9 (hopefully) as a security update caused the 2.6.22.8 release to happen last evening.

Sorry for any possible confusion.

thanks,

greg k-h

#4 lore raw

Re: Linux 2.6.22.8, pata_ali issue

Quoted 2 lines

From my point, sorry for thread hijacking, but I did not have

the upper node anymore.

So, here is something that I think should go into the next 2.6.22. A proper fix is already in 2.6.23-git-du-jour.

/home/build/rt.jengelh.10.2-i386/var/tmp/kernel-source-2.6.22.7-build /usr/src/linux-2.6.22.7-13.1/drivers/ata/pata_ali.c:
In function ■ali_init_chipset■:

/home/build/rt.jengelh.10.2-i386/var/tmp/kernel-source-2.6.22.7-build /usr/src/linux-2.6.22.7-13.1/drivers/ata/pata_ali.c:466:
warning: ■rev■ is used uninitialized in this function

Yes, it is indeed used uninitialized, and that is dangerous, given that I do have a chipset that uses pata_ali.ko and currently runs a 'broken' kernel...it has not exploded yet, though.

I could not backport and create a patch because struct pci_dev does not have the ->revision field in 2.6.22 yet. Ideas?

#6 lore raw

Re: Linux 2.6.22.8, pata_ali issue

In reply to a message not in the archive
Quoted 14 lines · Oliver Pinter

the rev is locali declared:

[snap]

...

static void ali_init_chipset(struct pci_dev *pdev)

{
       u8 rev, tmp;
       struct pci_dev *north, *isa_bridge;

       pci_read_config_byte(pdev, PCI_REVISION_ID, &rev);
Quoted 5 lines

...

[snap]

what's the git id in git tree?

Huh? My top git is 4942de4a0e914f205d351a81873f4f63986bcc3c and it has this:

static void ali_init_chipset(struct pci_dev *pdev)

{                                                                               
        u8 tmp;                                                                 
        struct pci_dev *north, *isa_bridge;                                     
                                                                                
        /*                                                                      
         * The chipset revision selects the driver operations and               
         * mode data.                                                           
         */                                                                     
                                                                                
        if (pdev->revision >= 0x20 && pdev->revision < 0xC2) {                  

No pci_read_config_byte here..

#7 lore raw

Re: Linux 2.6.22.8, pata_ali issue

my git top is:

commit 0dc661f882011b941b8b4d4dac84d7a0371f7a7b

Author: Greg Kroah-Hartman <[email protected]>
Date:   Mon Sep 24 23:05:13 2007 -0700
    Linux 2.6.22.8
Quoted 50 lines · Jan Engelhardt
Quoted 14 lines · Oliver Pinter

the rev is locali declared:

[snap]

...

static void ali_init_chipset(struct pci_dev *pdev)

{
       u8 rev, tmp;
       struct pci_dev *north, *isa_bridge;

       pci_read_config_byte(pdev, PCI_REVISION_ID, &rev);
Quoted 5 lines

...

[snap]

what's the git id in git tree?

Huh? My top git is 4942de4a0e914f205d351a81873f4f63986bcc3c and it has this:

static void ali_init_chipset(struct pci_dev *pdev)

{

        u8 tmp;

        struct pci_dev *north, *isa_bridge;


        /*

         * The chipset revision selects the driver operations and

         * mode data.

         */


        if (pdev->revision >= 0x20 && pdev->revision < 0xC2) {

No pci_read_config_byte here..

Signature · 2 lines
Thanks,
Oliver
#8 lore raw

Re: Linux 2.6.22.8, pata_ali issue

hey, your's tree is 2.6-git and not 2.6.22.8 ;)

Quoted 66 lines · Oliver Pinter

my git top is:

commit 0dc661f882011b941b8b4d4dac84d7a0371f7a7b

Author: Greg Kroah-Hartman <[email protected]>
Date:   Mon Sep 24 23:05:13 2007 -0700
    Linux 2.6.22.8
Quoted 47 lines · Jan Engelhardt
Quoted 14 lines · Oliver Pinter

the rev is locali declared:

[snap]

...

static void ali_init_chipset(struct pci_dev *pdev)

{
       u8 rev, tmp;
       struct pci_dev *north, *isa_bridge;

       pci_read_config_byte(pdev, PCI_REVISION_ID, &rev);
Quoted 5 lines

...

[snap]

what's the git id in git tree?

Huh? My top git is 4942de4a0e914f205d351a81873f4f63986bcc3c and it has this:

static void ali_init_chipset(struct pci_dev *pdev)

{

        u8 tmp;

        struct pci_dev *north, *isa_bridge;


        /*

         * The chipset revision selects the driver operations and

         * mode data.

         */


        if (pdev->revision >= 0x20 && pdev->revision < 0xC2) {
Quoted 3 lines

No pci_read_config_byte here..

--
Thanks,
Oliver

Signature · 2 lines
Thanks,
Oliver
author
#9 lore raw

Re: Linux 2.6.22.8, pata_ali issue

In reply to Jan Engelhardt
Quoted 4 lines · Jan Engelhardt

From my point, sorry for thread hijacking, but I did not have the upper node anymore.

How hard is it to start a new message? Me thinks you need a new email client :)

Quoted 2 lines

So, here is something that I think should go into the next 2.6.22. A proper fix is already in 2.6.23-git-du-jour.

Feel free to send the patch, with the git commit id of the upstream commit to the [email protected] address and we will be glad to review it that way.

thanks,

greg k-h

#10 lore raw

Re: Linux 2.6.22.8, pata_ali issue

Quoted 8 lines · Greg KH
Quoted 2 lines · Jan Engelhardt

So, here is something that I think should go into the next 2.6.22. A proper fix is already in 2.6.23-git-du-jour.

Feel free to send the patch, with the git commit id of the upstream commit to the [email protected] address and we will be glad to review it that way.

'right, it is already in the mainline kernel.
So it's again the suse kernel src rpm that has it screwed up... another story. Here's your culprit:

libata-update-libata-to-libata-dev-upstream-5ddf.patch
* Replaces ali_init_chipset() with some new bits, essentially it removes the call to pci_read_config_byte() since the revision is now found in struct pci_dev->revision Basically 2.6.23 stuff.

libata-fix-up-build-after-upstream-update.patch
* changes it back to the 2.6.22 api, so that it compiles again. But only that.
It FORGETS to add pci_read_config_byte() back.

=== libata-fix-up-build-after-upstream-update.patch

Quoted 1 line

From 54ee7cd26c5e438dc15d13b60c9f8a54ca05e3fb Mon Sep 17 00:00:00 2001

From: Tejun Heo <[email protected]>
Date: Fri, 3 Aug 2007 02:21:51 +0900
Subject: [PATCH] libata: fix up build after upstream update
References: 288078

===

The bug is still present in current suse kotd. >:-(

Please fix it there. A cc will go to opensuse-kernel@,
but other than that, I am in the wrong mood right now to do anything.