kexec refuses to boot latest -mm

[Dhaval Giani]

On Thu, Jan 03, 2008 at 10:42:00PM +0100, Rafael J. Wysocki wrote:
> On Thursday, 3 of January 2008, Dhaval Giani wrote:
> > On Mon, Dec 31, 2007 at 10:08:43AM -0500, Vivek Goyal wrote:
> > > On Sat, Dec 29, 2007 at 11:11:13AM +0530, Dhaval Giani wrote:
> > > > On Fri, Dec 28, 2007 at 09:27:39AM -0500, Vivek Goyal wrote:
> > > > > On Fri, Dec 28, 2007 at 06:15:32PM +0530, Dhaval Giani wrote:
> > > > > > Hi Vivek,
> > > > > > 
> > > > > > I can't seem to get the latest -mm (2.6.24-rc6-mm1) to boot with
> > > > > > kexec. It just gets stuck at "Starting new kernel".
> > > > > > 
> > > > > > It does boot normally when booted as the first kernel.
> > > > > > 
> > > > > > Any hints debugging? (x86 architecture)
> > > > > 
> > > > > I generally try few things.
> > > > > 
> > > > > - Specify earlyprintk= parameter for second kernel and see if control
> > > > >   is reaching to second kernel.
> > > > > 
> > > > > - Otherwise specify --console-serial parameter on "kexec -l" commandline
> > > > >   and it should display a message "I am in purgatory" on serial console.
> > > > >   This will just mean that control has reached at least till purgatory.
> > > > > 
> > > > 
> > > > Ok, so it reaches till here. I get "I'm in purgatory" on the console.
> > > 
> > > One more thing. Is 2.6.24-rc6 working properly?
> > > 
> > 
> > 2.6.24-rc5 boots, so does 2.6.24-rc5-mm1. 2.6.24-rc6 does not boot, nor
> > does 2.6.24-rc6-mm1. Its a regression.
> 
> Added to the list of reported regressions as:
> http://bugzilla.kernel.org/show_bug.cgi?id=9682
> 

I've no clue what I managed to do wrong last night, probably tried to boot
2.6.24-rc6-mm1 thinking it was 2.6.24-rc6. 2.6.24-rc6 boots, but -mm
does not. Sorry for the noise.

With the help of the outbs I have figured out that it is dying somewhere
in setup_memory(). Will look deeper into it.

-- 
regards,
Dhaval