Re: buildworld fails with distcc, buildkernel works fine though

["Simon 'corecode' Schubert" <corecode@xxxxxxxxxxxx>]

On 14.10.2004, at 12:42, YONETANI Tomokazu wrote:
> - showing the exact place where your buildworld failed and the actual
>   error messages will give us better information than just saying
>   "some such things" or stuff like that :)

definitely true.

> - the system clock on each node must be in sync or the make program
>   gets screwed.

if you use distcc, you just distribute the compiling, not even the 
preprocessing, so make isn't an issue here.

> - you probably have to install exact same version of OS on each
>   node, same make.conf file, and the same environment variables.
>   diffent configurations produce diffent outputs, so mixing them up
>   will result in unexpected errors.

that's true for CCVER and OBJFORMAT_PATH, all other variables shouldn't 
be used when just compiling

> - I think you'll get better result by distributing sub-makes to remote
>   machines rather than compiler processes, because buildworld involves
>   many other tasks than just compiling C codes, like formatting 
> manpages
>   and documents.

True with regard to speed, tho utterly complex to realize.

Thinking about it: how about

1. distcc'ing the tools stage. this will result in a fresh gcc.
2. distributing the compiler (or all the tools) to all nodes.
3. using this compiler (now guarranteed to be the same everywhere) to 
compile world. hostcc will stay on host, of course.

cheers
  simon

--
/"\
\ /
 \     ASCII Ribbon Campaign
/ \  Against HTML Mail and News

Attachment: PGP.sig
Description: This is a digitally signed message part