lkml.org 
[lkml]   [2012]   [Mar]   [13]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [PATCH 0/5] Persist printk buffer across reboots.
On Tue, Mar 13, 2012 at 03:10:49PM -0700, Andrew Morton wrote:
> On Tue, 13 Mar 2012 10:08:51 -0700
> Daniel Walker <dwalker@fifo99.com> wrote:
>
> > On Tue, Mar 13, 2012 at 01:36:36AM -0400, Avery Pennarun wrote:
> > > The last patch in this series implements a new CONFIG_PRINTK_PERSIST option
> > > that, when enabled, puts the printk buffer in a well-defined memory location
> > > so that we can keep appending to it after a reboot. The upshot is that,
> > > even after a kernel panic or non-panic hard lockup, on the next boot
> > > userspace will be able to grab the kernel messages leading up to it. It
> > > could then upload the messages to a server (for example) to keep crash
> > > statistics.
> >
> > There's currently driver/mtd/mtdoops.c, fs/pstore/, and
> > drivers/staging/android/ram_console.c that do similar things
> > as this. Did you investigate those for potentially modifying them to add
> > this functionality ? If so what issues did you find?
> >
> > I have a arm MSM G1 with persistent memory at 0x16d00000 size 20000bytes..
> > It's fairly simple you just have to ioremap the memory, but then it's good
> > for writing.. Currently the android ram_console uses this. How would I
> > convert this area for use with your changes?
>
> Yes, and various local implementations which do things such as stuffing
> the log buffer into NVRAM as the kernel is crashing.
>
> I do think we'd need some back-end driver arrangement which will permit
> the use of stores which aren't in addressible mamory.
>
> It's quite the can of worms, but definitely worth doing if we can get
> it approximately correct.

For sure.. I've had at least a couple non-OOPS based crashed I would
have liked to get logs for.

There is also this series,

http://lists.infradead.org/pipermail/kexec/2011-July/005258.html

It seems awkward that pstore is in fs/pstore/ then pstore ends up as the
"back end" where it could just be the whole solution.

Daniel


\
 
 \ /
  Last update: 2012-03-14 03:21    [from the cache]
©2003-2011 Jasper Spaans