Re: Fw: two (little) problems wit 2.6.15-git7 one with build, one with acpi

On Sunday 15 January 2006 13:55, Sam Ravnborg wrote:
> On Fri, Jan 13, 2006 at 08:18:03PM +0100, Hemmann, Volker Armin wrote:
> > I tried -git8 last night.
> > make all modules_install install gave me again this error:
> >
> > if [ -r System.map -a -x /sbin/depmod ]; then /sbin/depmod -ae -F
> > System.map 2.6.15-git8; fi
> > WARNING: Couldn't open directory /lib/modules/2.6.15-git8: No such file
> > or directory
> > FATAL: Could not open /lib/modules/2.6.15-git8/modules.dep.temp for
> > writing: No such file or directory
> > make: *** [_modinst_post] Fehler 1
>
> Do you have KBUILD_OUTPUT set in your environment?

no, at least I don't find it, when I type env.

> That may explain it since things are running too much in parallel with
> KBUILT_OUTPUT set - reported by Keith Owns a few days ago.
>
> I have tried the above on my machine and was not able to reproduce it.
>
> Can you please drop me the output with V=1 so I can see what is
> happening. Maybe in private mail if it gets too big.

I tried make all modules_install install V=1 &>/tmp/kernel2.log and found
something interessting.

The kernel-sources directory has to be 'virgin'. As soon as I build a kernel
in it once, it worked without any problems - even after a make clean and make
mrproper.

So the steps to reproduce it are:
unpack 2.6.15, patch it, copy .config from other kernel (this time I took the
one from git8 I built earlier), make oldconfig, make all_modules_install
install:

f [ -r System.map -a -x /sbin/depmod ]; then /sbin/depmod -ae -F System.map
2.
6.15-git7; fi
WARNING: Couldn't open directory /lib/modules/2.6.15-git7: No such file or
direc
tory
FATAL: Could not open /lib/modules/2.6.15-git7/modules.dep.temp for writing:
No
such file or directory
make: *** [_modinst_post] Fehler 1

because kernel2.log is 748kb big., I packed it.

Glück Auf
Volker


Attachment:kernel2.log.bz2
Description: BZip2 compressed data

Attachment: config.bz2
Description: BZip2 compressed data

Relevant Pages

  • Re: Where the code trouble is?
    ... but in which writing at would NOT result in a trap, because the architectures happen to have memory at that location. ... Although *NULL might well not trap, it's in general more likely to, which can help with debugging. ... IMHO the first stage in trying to debug a problem is always to reproduce ...
    (comp.lang.c)
  • Re: MSN CHAT
    ... Closest thing i could reproduce is no more than 5 ... I did it by writing a program that continuously writes gibberish to ...
    (microsoft.public.windowsxp.messenger)
  • Re: 5.3-RELEASE and cut(1) bug
    ... > (good posability of the latter). ... You're writing like you need sleep. ... I can reproduce your error on 5.3B7, where 4.9R seems to work fine. ...
    (comp.unix.bsd.freebsd.misc)
  • RE: Problem processing partition
    ... Yes I can reproduce it, I reproduced it serveral times before posting ... from the writing process. ... And when I reprocess the cube it reads ... > Peter Yang ...
    (microsoft.public.sqlserver.olap)