[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#626851: linux-image-2.6.38-2-amd64: after last update fails to boot

Rares Aioanei wrote:

> Here's the config; I tried the good config already with vanilla and
> it works ok, hence my idea that it's more of a config issue rather
> than a kernel bug per se.

The simplest way to check would be to try with a stock configuration
file (like the one from /boot/config-2.6.38-2-amd64).

Looking through the config file you sent, I see

| Linux/x86_64 2.6.38.5 Kernel Configuration

which makes it seem you were building v2.6.38.5.  Is that right?  If
so, it would not be unexpected that it failed to boot, since the stock
Debian 2.6.38-5 package failed, too, right?  The idea if you got the
same message as started the bug would be to run "git bisect bad" and
then build the next kernel after that.  But:

Other notable bits from the .config file:

. CONFIG_EXPERT=y makes it easy to build a kernel that doesn't boot;
  it should be easier to debug with it off.

. "# CONFIG_BLK_DEV_SD is not set" means it can't read your disk ---
  libata presents disks as SCSI to the kernel.
--- Begin Message --- 
On Thu, 26 May 2011 10:36:24 -0500
Jonathan Nieder <jrnieder@gmail.com> wrote:

> Hi,
> 
> Rares Aioanei wrote:
> 
> > Sorry, but after bisecting the results are roughly the same (roughly as in
> > the error message is now "Giving up waiting for root device"). I'm puzzled as
> > to what to do next. If I can help any further, I will do it gladly.
> 
> To be clear, do you mean that you got this result with the version
> halfway between v2.6.38.4 and v2.6.38.5?  What does the .config file
> look like --- could you send it?  (Sorry for the tedious questions.)
> 
> "Giving up waiting for the root device" can mean many things, but it often
> means the hard disk driver didn't load.  Maybe that means an (unrelated)
> localmodconfig bug; trying with a known-good .config file (like so:
> 
> 	cp /path/to/my/good/config/file .config
> 	make oldconfig
> 	make deb-pkg
> 	cd ..
> 	dpkg -i ...
> 	reboot
> 
> ) could be a good idea.
> 
> If this *was* a known-good .config file, I'd suggest running
> 
> 	git bisect bad
> 
> and building and testing the next version it suggests.
> 
> Sorry for the lack of clarity.
> Jonathan
Here's the config; I tried the good config already with vanilla and it works ok, hence my idea that it's more of a config issue rather than a kernel bug per se.

Cheers, 

-- 
Rares Aioanei

Attachment: .config
Description: Binary data


--- End Message ---
Reply to: