Re: Problems Building 7.0-Beta3 with -Os

From: Nikos Ntarmos <ntarmos_at_ceid.upatras.gr>
Date: Thu, 6 Dec 2007 21:07:22 +0200
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Thu, Dec 06, 2007 at 04:17:25PM +0100, Jan Lentfer wrote:
> Hi Nikos,
> 
> Zitat von Nikos Ntarmos <ntarmos_at_ceid.upatras.gr>:
> >[...]
> >1 1 0  623384 174992 43861   0   0   0  92   0   0   0 1299  138 347 43 57  0
> >>>>>Compilation break
> >0 0 0   70200 711396 2515   0   0   0 136608   0   9   0 1257  453  281 9 91  0
> >       ^^^^^^ ^^^^^^
> >
> >Funny how 'avm' and 'fre' are lower/higher by almost the same amout your
> >max dataseg is after the compilation break... Do you have a swap
> >partition? If you do, perhaps you should try bumping kern.maxdsiz and
> >kern.maxssiz in /boot/loader.conf and see what happens... (e.g. try with
> >kern.maxdsiz="896M" and kern.maxssiz="256M" with 1G RAM and another 1G
> >swap). Also, what is the value of kern.maxusers?
> 
> 
> With your suggested settings it works now - thank you very much.
> Compilation hasn't totally finished yet, but I got far past the point
> where it broke off before. Still I find this quite strange behavior,
> but this seems to be rather a gcc thing than a FBSD problem, right?

gcc is quite the memory hog with higher optimization levels and the
default data segment size limit is 512M, which proved to be inadequate
for compiling gcc -Os. I wouldn't call this a bug of gcc or FBSD though;
it's rather a system tuning thing...

\n\n
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.4 (FreeBSD)
Comment: Nikos Ntarmos <ntarmos_at_ceid.upatras.gr>

iD8DBQFHWEhqm6J1ac+VFgoRAo1QAJ0a97C0W2ymMyMB49Mw+B+1JYyzugCcCZP8
ocrSS1Jb+4ojMCS2Lpn+yVI=
=+FI6
-----END PGP SIGNATURE-----
Received on Thu Dec 06 2007 - 18:07:26 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:39:23 UTC