Re: Build regressions/improvements in v2.6.36-rc8

From: Stephen Rothwell
Date: Sun Oct 17 2010 - 05:13:30 EST


Hi Geert,

On Sun, 17 Oct 2010 10:01:03 +0200 Geert Uytterhoeven <geert@xxxxxxxxxxxxxx> wrote:
>
> Strange...
>
> $ wget http://kisskb.ellerman.id.au/kisskb/buildresult/3387901/; grep
> inode_set index.html
> --2010-10-17 09:59:35--
> http://kisskb.ellerman.id.au/kisskb/buildresult/3387901/
> Resolving kisskb.ellerman.id.au... 203.10.76.25
> Connecting to kisskb.ellerman.id.au|203.10.76.25|:80... connected.
> HTTP request sent, awaiting response... 200 OK
> Length: 272362 (266K) [text/html]
> Saving to: `index.html'
>
> 100%[=====================================================================>]
> 272,362 67.8K/s in 3.9s
>
> 2010-10-17 09:59:39 (67.8 KB/s) - `index.html' saved [272362/272362]
>
> /scratch/kisskb/src/fs/9p/vfs_inode.c:1267: error: implicit
> declaration of function &#39;inode_setattr&#39;
> $
>
> Yep, it's there.

Ah, light dawns ... kisskb, when it gets a failure, attempts to do a
bisect between the failing kernel and the last kernel that built.
Consequently, the log of the bisect attempts will contain some of the
other errors that occurred since the kernel built correctly. (As an
aside, the bisect script appears not to work very well ... :-( )

You were unlucky here that the m68k build went through a series of
different build problems without getting a good build.

You could try cutting the log off after seeing "make failed, trying to
bisect".
--
Cheers,
Stephen Rothwell sfr@xxxxxxxxxxxxxxxx
http://www.canb.auug.org.au/~sfr/

Attachment: pgp00000.pgp
Description: PGP signature