[PATCH 0/5,RFC] mv643xx_eth fixes for 2.6.27

[PATCH 0/5,RFC] mv643xx_eth fixes for 2.6.27

[Lennert Buytenhek]

Hi,

The following five mv643xx_eth fixes for 2.6.27 since commit
6a55617ed5d1aa62b850de2cf66f5ede2eef4825:
  Linus Torvalds (1):
        Linux v2.6.27-rc4

are available in the git repository at:

  git://git.marvell.com/mv643xx_eth.git for-2.6.27
  http://git.marvell.com/?p=mv643xx_eth.git;a=summary

Lennert Buytenhek (5):
      mv643xx_eth: fix NAPI 'rotting packet' issue
      mv643xx_eth: fix double add_timer() on the receive oom timer
      mv643xx_eth: fix inconsistent lock semantics
      mv643xx_eth: fix NULL pointer dereference in rxq_process()
      mv643xx_eth: enforce multiple-of-8-bytes receive buffer size restriction

 drivers/net/mv643xx_eth.c |   33 +++++++++++++++++++--------------
 1 files changed, 19 insertions(+), 14 deletions(-)

The second patch should fix http://bugzilla.kernel.org/show_bug.cgi?id=11377
-- Gerhard, can you please test if it makes your problem go away?  (It
won't make the out of memory messages go away, but it should fix the
kernel panic.)

There is still one open report from a user of doing ip=1.2.3.4 on the
kernel command line causing the interface to come up in a nonfunctional
state in 2.6.27 (whereas the same thing worked in 2.6.26), which is
still being bisected.


thanks,
Lennert

mv643xx_eth fixes pull request for 2.6.27 (was: Re: [PATCH 0/5,RFC] mv643xx_eth fixes for 2.6.27)

[Lennert Buytenhek]

On Sun, Aug 24, 2008 at 04:03:24AM +0200, Lennert Buytenhek wrote:

> The following five mv643xx_eth fixes for 2.6.27 since commit
> 6a55617ed5d1aa62b850de2cf66f5ede2eef4825:
>   Linus Torvalds (1):
>         Linux v2.6.27-rc4
> 
> are available in the git repository at:
> 
>   git://git.marvell.com/mv643xx_eth.git for-2.6.27
>   http://git.marvell.com/?p=mv643xx_eth.git;a=summary
> 
> Lennert Buytenhek (5):
>       mv643xx_eth: fix NAPI 'rotting packet' issue
>       mv643xx_eth: fix double add_timer() on the receive oom timer
>       mv643xx_eth: fix inconsistent lock semantics
>       mv643xx_eth: fix NULL pointer dereference in rxq_process()
>       mv643xx_eth: enforce multiple-of-8-bytes receive buffer size restriction
> 
>  drivers/net/mv643xx_eth.c |   33 +++++++++++++++++++--------------
>  1 files changed, 19 insertions(+), 14 deletions(-)
> 
> The second patch should fix http://bugzilla.kernel.org/show_bug.cgi?id=11377
> -- Gerhard, can you please test if it makes your problem go away?  (It
> won't make the out of memory messages go away, but it should fix the
> kernel panic.)

The submitter confirmed that patch 2 fixes the original issue (and
once that is fixed, patch 4 will fix the subsequently emerging issue).
I've also done some more testing on the series, and it's probably good
for 2.6.27.  I've tacked on a patch to bump the version number -- can
you please pull?

The following changes since commit 6a55617ed5d1aa62b850de2cf66f5ede2eef4825:
  Linus Torvalds (1):
        Linux v2.6.27-rc4

are available in the git repository at:

  git://git.marvell.com/mv643xx_eth.git for-2.6.27
  http://git.marvell.com/?p=mv643xx_eth.git;a=shortlog;h=for-2.6.27

Lennert Buytenhek (6):
      mv643xx_eth: fix NAPI 'rotting packet' issue
      mv643xx_eth: fix double add_timer() on the receive oom timer
      mv643xx_eth: fix inconsistent lock semantics
      mv643xx_eth: fix NULL pointer dereference in rxq_process()
      mv643xx_eth: enforce multiple-of-8-bytes receive buffer size restriction
      mv643xx_eth: bump version to 1.3

 drivers/net/mv643xx_eth.c |   35 ++++++++++++++++++++---------------
 1 files changed, 20 insertions(+), 15 deletions(-)


thanks,
Lennert

Re: mv643xx_eth fixes pull request for 2.6.27

[Jeff Garzik]

Lennert Buytenhek wrote:
> On Sun, Aug 24, 2008 at 04:03:24AM +0200, Lennert Buytenhek wrote:
> 
>> The following five mv643xx_eth fixes for 2.6.27 since commit
>> 6a55617ed5d1aa62b850de2cf66f5ede2eef4825:
>>   Linus Torvalds (1):
>>         Linux v2.6.27-rc4
>>
>> are available in the git repository at:
>>
>>   git://git.marvell.com/mv643xx_eth.git for-2.6.27
>>   http://git.marvell.com/?p=mv643xx_eth.git;a=summary
>>
>> Lennert Buytenhek (5):
>>       mv643xx_eth: fix NAPI 'rotting packet' issue
>>       mv643xx_eth: fix double add_timer() on the receive oom timer
>>       mv643xx_eth: fix inconsistent lock semantics
>>       mv643xx_eth: fix NULL pointer dereference in rxq_process()
>>       mv643xx_eth: enforce multiple-of-8-bytes receive buffer size restriction
>>
>>  drivers/net/mv643xx_eth.c |   33 +++++++++++++++++++--------------
>>  1 files changed, 19 insertions(+), 14 deletions(-)
>>
>> The second patch should fix http://bugzilla.kernel.org/show_bug.cgi?id=11377
>> -- Gerhard, can you please test if it makes your problem go away?  (It
>> won't make the out of memory messages go away, but it should fix the
>> kernel panic.)
> 
> The submitter confirmed that patch 2 fixes the original issue (and
> once that is fixed, patch 4 will fix the subsequently emerging issue).
> I've also done some more testing on the series, and it's probably good
> for 2.6.27.  I've tacked on a patch to bump the version number -- can
> you please pull?
> 
> The following changes since commit 6a55617ed5d1aa62b850de2cf66f5ede2eef4825:
>   Linus Torvalds (1):
>         Linux v2.6.27-rc4
> 
> are available in the git repository at:
> 
>   git://git.marvell.com/mv643xx_eth.git for-2.6.27
>   http://git.marvell.com/?p=mv643xx_eth.git;a=shortlog;h=for-2.6.27
> 
> Lennert Buytenhek (6):
>       mv643xx_eth: fix NAPI 'rotting packet' issue
>       mv643xx_eth: fix double add_timer() on the receive oom timer
>       mv643xx_eth: fix inconsistent lock semantics
>       mv643xx_eth: fix NULL pointer dereference in rxq_process()
>       mv643xx_eth: enforce multiple-of-8-bytes receive buffer size restriction
>       mv643xx_eth: bump version to 1.3
> 
>  drivers/net/mv643xx_eth.c |   35 ++++++++++++++++++++---------------
>  1 files changed, 20 insertions(+), 15 deletions(-)

pulled