Re: 2.6.15-mm2

[Andrew Morton <akpm@osdl.org>]

Reuben Farrelly <reuben-lkml@reub.net> wrote:
>
> On 11/01/2006 6:30 p.m., Andrew Morton wrote:
> > Andrew Morton <akpm@osdl.org> wrote:
> >> Reuben Farrelly <reuben-lkml@reub.net> wrote:
> >>> I'm tempted to see if I can narrow it down to a specific -gitX release, maybe 
> >>>  that would narrow it down to say, 200 or so patches ;-)
> >> If -mm2 plus -mm2's linus.patch does not fail then
> >> http://www.zip.com.au/~akpm/linux/patches/stuff/bisecting-mm-trees.txt will
> >> find the dud patch.
> > 
> > Actually 2.6.15-mm1 would be a better one to do the bisection on: it has
> > all the md- patches separated out.
> 
> I've done some more testing - which may change the suggested approach somewhat..
> 
> 2.6.15-mm1 is OK, I'm running it now, rebooted probably 15 times and it's come 
> up every time.
> 2.6.15-git2 is OK, booted up to completion (tested once).
> 2.6.15-git3 was a dud, bootup hung

Ah.

> 2.6.15- [linus.patch from -mm2, which is basically the same as -git3] won't boot
> 2.6.15-mm2 doesn't boot either, tested many times
> 2.6.15-git6 won't boot
> 2.6.15-git7 got stuck also, same issue
> 
> So some change that went in between -git2 and -git3 seems to have caused it. 
> Nothing from -git3 onwards has ever booted to completion.
> 
> Is there any chance a patch came in, was queued in -mm but was never released in 
> any -mm (1|2) release before being sent to Linus/-gitX?  (in this case, -git3). 

Yes, people sneak stuff in at the last minute.

Neil thinks that an IO got lost.  In the git2->git3 diff we have:

 b/drivers/scsi/Kconfig                         |   10 
 b/drivers/scsi/ahci.c                          |    1 
 b/drivers/scsi/ata_piix.c                      |    5 
 b/drivers/scsi/libata-core.c                   |  145 +
 b/drivers/scsi/libata-scsi.c                   |   48 
 b/drivers/scsi/libata.h                        |    4 
 b/drivers/scsi/sata_mv.c                       |    1 
 b/drivers/scsi/sata_promise.c                  |    1 
 b/drivers/scsi/sata_sil.c                      |    1 
 b/drivers/scsi/sata_sil24.c                    |    1 
 b/drivers/scsi/sata_sx4.c                      |    1 
 b/drivers/scsi/scsi_lib.c                      |   50 
 b/drivers/scsi/scsi_sysfs.c                    |   31 
 b/drivers/scsi/sd.c                            |   85 -
 b/fs/bio.c                                     |   26 

Jens, Jeff: were any of those changes added in the final day or two, not
included in the trees which I pull?


> 
> I'm not sure where this leaves quilt testing.  Would quilt testing just narrow 
> me down to it being the linus.patch in mm which actually caused it? (Which I 
> already know is the source)..

Yes, there's not much point in that.

`git bisect' will find it.

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/