Skip to contentSkip to conversation

Linux 2.6.16.58

DISCUSSION Adrian Bunk 6 participants 10 messages Jan 15, 2008 – Jan 18, 2008

lore
Share & tools
lore
#0 lore raw

Security fixes since 2.6.16.57:
- CVE-2007-2525: PPPOE: fix memory leak (local DoS)
- CVE-2007-3731: Handle bogus %cs selector in single-step instruction decoding
- CVE-2007-6417: tmpfs: restore missing clear_highpage

Location:
ftp://ftp.kernel.org/pub/linux/kernel/v2.6/

git tree:
git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-2.6.16.y.git

RSS feed of the git tree:
http://www.kernel.org/git/?p=linux/kernel/git/stable/linux-2.6.16.y.git;a=rss

Changes since 2.6.16.57:

Adrian Bunk (3):

      ipv4/arp.c:arp_process(): remove bogus #ifdef mess
      Linux 2.6.16.58-rc1
      Linux 2.6.16.58

Alan Cox (1):
[SCSI] aacraid: fix security weakness

Ben Collins (1):
[ATM]: Add CPPFLAGS to byteorder.h check

Charles Hardin (1):
[PFKEY]: Sending an SADB_GET responds with an SADB_GET

Chas Williams (1):
[ATM]: [he] initialize lock and tasklet earlier

Evgeniy Polyakov (1):
[PKT_SCHED]: Fix OOPS when removing devices from a teql queuing discipline

Florian Zumbiehl (1):

      PPPOE: fix memory leak (local DoS) (CVE-2007-2525)

Hugh Dickins (1):

      tmpfs: restore missing clear_highpage (CVE-2007-6417)

Jean Delvare (2):
hwmon/lm87: Disable VID when it should be
hwmon/lm87: Fix a division by zero

Michal Schmidt (1):
[PPP_MPPE]: Don't put InterimKey on the stack

Patrick McHardy (2):
[INET_DIAG]: Fix oops in netlink_rcv_skb
[NETLINK]: Fix unicast timeouts

Peter Zijlstra (1):
i386: fixup TRACE_IRQ breakage

Radu Rendec (1):
[PKT_SCHED] CLS_U32: Fix endianness problem with u32 classifier hash masks.

Roland McGrath (1):
Handle bogus %cs selector in single-step instruction decoding (CVE-2007-3731)

YOSHIFUJI Hideaki (1):
[IPV6]: Fix unbalanced socket reference with MSG_CONFIRM.

 Makefile                     |    2 +-
 arch/i386/kernel/entry.S     |    1 -
 arch/i386/kernel/ptrace.c    |   22 +++++++++++++++-------
 arch/i386/kernel/traps.c     |   10 ++++++----
 arch/x86_64/kernel/ptrace.c  |   23 ++++++++++++++++-------
 drivers/atm/Makefile         |    2 +-
 drivers/atm/he.c             |   10 +++++-----
 drivers/hwmon/lm87.c         |    4 ++--
 drivers/net/ppp_mppe.c       |   14 ++++++--------
 drivers/net/pppox.c          |    2 +-
 drivers/scsi/aacraid/linit.c |    4 ++--
 include/linux/netlink.h      |    2 +-
 ipc/mqueue.c                 |    6 ++++--
 mm/shmem.c                   |    5 +++--
 net/ipv4/arp.c               |   19 -------------------
 net/ipv4/inet_diag.c         |   10 +++++++++-
 net/ipv6/raw.c               |    2 +-
 net/key/af_key.c             |    2 +-
 net/netlink/af_netlink.c     |   10 +++++-----
 net/sched/cls_u32.c          |    4 ++--
 net/sched/sch_teql.c         |    3 +++
 21 files changed, 84 insertions(+), 73 deletions(-)
#2 lore raw

Hi,

Quoted 10 lines · Jiri Kosina
Quoted 7 lines · Adrian Bunk

RSS feed of the git tree:
http://www.kernel.org/git/?p=linux/kernel/git/stable/linux-2.6.16.y.git;a=rss
Changes since 2.6.16.57:
Adrian Bunk (3):

      ipv4/arp.c:arp_process(): remove bogus #ifdef mess
      Linux 2.6.16.58-rc1
      Linux 2.6.16.58
Quoted 4 lines

Hi Adrian,

I don't seem to see v2.6.16.58 tag in your git repository -- the HEAD (a8d648b74) is tagged v2.6.16.58-rc1. Is that intended?

Not replying for Adrian, but it happened several times to me too. What sometimes happens when we push Git updates from home is that we don't push all the refs. Sometimes I forget to update the tags, sometimes the master, etc...

Just looked at the git repo itself, it looks like it has not changed since 2008/01/08 :-/

Adrian, maybe your git-push did not succeed ?

Regards,
Willy

#3 lore raw
In reply to Jiri Kosina
Quoted 11 lines · Jiri Kosina
Quoted 7 lines · Adrian Bunk

RSS feed of the git tree:
http://www.kernel.org/git/?p=linux/kernel/git/stable/linux-2.6.16.y.git;a=rss
Changes since 2.6.16.57:
Adrian Bunk (3):

      ipv4/arp.c:arp_process(): remove bogus #ifdef mess
      Linux 2.6.16.58-rc1
      Linux 2.6.16.58

Hi Adrian,

Hi Jiri,

Quoted 2 lines

I don't seem to see v2.6.16.58 tag in your git repository -- the HEAD (a8d648b74) is tagged v2.6.16.58-rc1. Is that intended?

do you insist on hearing the truth ("Sorry, I forgot to push."), or can I tell you "I just wanted to see whether anyone notices." instead? ;-)

Quoted 1 line

Jiri Kosina

cu
Adrian

Signature · 4 lines
       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed
#4 lore raw

Hi Adrian,

Quoted 2 lines · Adrian Bunk

do you insist on hearing the truth ("Sorry, I forgot to push."), or can I tell you "I just wanted to see whether anyone notices." instead? ;-)

:-)

I think we should intentionally do this test once in a while. Two days ago I was informed that 2.6.20.21 (the latest 2.6.20 I pushed 3 months ago) was not clonable nor pullable due to a problem with the HEAD pointing to latest tag instead of latest commit. Since the error was also present in my local tree, I'm certain that it remained there unnoticed for 3 months !

Cheers,
Willy

#5 lore raw
Quoted 6 lines · Willy Tarreau

Hi Adrian,

Quoted 2 lines · Adrian Bunk

do you insist on hearing the truth ("Sorry, I forgot to push."), or can I tell you "I just wanted to see whether anyone notices." instead? ;-)

Quoted 8 lines

:-)

I think we should intentionally do this test once in a while. Two days ago I was informed that 2.6.20.21 (the latest 2.6.20 I pushed 3 months ago) was not clonable nor pullable due to a problem with the HEAD pointing to latest tag instead of latest commit. Since the error was also present in my local tree, I'm certain that it remained there unnoticed for 3 months !

According to git [1] Ralf merged 2.6.20.21 on 2007-10-18 into the 2.6.20 branch of his mips tree, so it must have been somehow possible...

Quoted 2 lines

Cheers,
Willy

cu
Adrian

[1] http://www.linux-mips.org/git?p=linux.git;a=shortlog;h=linux-2.6.20-stable

Signature · 4 lines
       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed
#6 lore raw
Quoted 16 lines · Adrian Bunk
Quoted 6 lines · Willy Tarreau

Hi Adrian,

Quoted 2 lines · Adrian Bunk

do you insist on hearing the truth ("Sorry, I forgot to push."), or can I tell you "I just wanted to see whether anyone notices." instead? ;-)

Quoted 8 lines

:-)

I think we should intentionally do this test once in a while. Two days ago I was informed that 2.6.20.21 (the latest 2.6.20 I pushed 3 months ago) was not clonable nor pullable due to a problem with the HEAD pointing to latest tag instead of latest commit. Since the error was also present in my local tree, I'm certain that it remained there unnoticed for 3 months !

Quoted 2 lines

According to git [1] Ralf merged 2.6.20.21 on 2007-10-18 into the 2.6.20 branch of his mips tree, so it must have been somehow possible...

Maybe he did this locally (Linus showed me that local FS cloning disables a lot of tests). Or maybe it could work for some pulls :-/ Definitely a weird issue anyway!

Cheers,
Willy

#7 lore raw
Quoted 15 lines · Willy Tarreau
Quoted 14 lines · Adrian Bunk
Quoted 12 lines · Willy Tarreau
Quoted 2 lines · Adrian Bunk

do you insist on hearing the truth ("Sorry, I forgot to push."), or can I tell you "I just wanted to see whether anyone notices." instead? ;-)

:-)

I think we should intentionally do this test once in a while. Two days ago I was informed that 2.6.20.21 (the latest 2.6.20 I pushed 3 months ago) was not clonable nor pullable due to a problem with the HEAD pointing to latest tag instead of latest commit. Since the error was also present in my local tree, I'm certain that it remained there unnoticed for 3 months !

Quoted 6 lines
Quoted 2 lines

According to git [1] Ralf merged 2.6.20.21 on 2007-10-18 into the 2.6.20 branch of his mips tree, so it must have been somehow possible...

Maybe he did this locally (Linus showed me that local FS cloning disables a lot of tests). Or maybe it could work for some pulls :-/ Definitely a weird issue anyway!

Err... I managed to do it for thinkpad-acpi's merge tree too, a lot of time ago. So it was somehow possible to get it through the anonymous git server, probably through the tag. What I do know is that I usually send head's up email when I notice something like this, and I find it strange that you didn't get one from me at that time... but I may have forgotten to send it at that time.

Signature · 4 lines
  "One disk to rule them all, One disk to find them. One disk to bring
  them all and in the darkness grind them. In the Land of Redmond
  where the shadows lie." -- The Silicon Valley Tarot
  Henrique Holschuh
#8 lore raw
Quoted 6 lines · Henrique de Moraes Holschuh

Err... I managed to do it for thinkpad-acpi's merge tree too, a lot of time ago. So it was somehow possible to get it through the anonymous git server, probably through the tag. What I do know is that I usually send head's up email when I notice something like this, and I find it strange that you didn't get one from me at that time... but I may have forgotten to send it at that time.

Well, since there are commands that work on this (push and log at least), I believe that some operations could succeed. I *know* that remote cloning was impossible, and it *looked* like pulling too was during my attempts to fix the problem before understanding its root cause. It is possible that I cumulated problems though, if at least two people could pull from it.

Cheers,
Willy

#9 lore raw
Quoted 6 lines · Willy Tarreau

Well, since there are commands that work on this (push and log at least), I believe that some operations could succeed. I *know* that remote cloning was impossible, and it *looked* like pulling too was during my attempts to fix the problem before understanding its root cause. It is possible that I cumulated problems though, if at least two people could pull from it.

Ah, I think I recalled something about the episode. I knew the tag was there (gitweb), and I did have some trouble to get it, but I *think* what worked out was:

git fetch <repo> <tag>

Signature · 4 lines
  "One disk to rule them all, One disk to find them. One disk to bring
  them all and in the darkness grind them. In the Land of Redmond
  where the shadows lie." -- The Silicon Valley Tarot
  Henrique Holschuh