rpm.archscore's influence on smart

Eric MSP Veith eveith at wwweb-library.net
Tue Mar 23 07:26:03 PDT 2010


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Heya,

On Tuesday 23 March 2010, Anders F Björklund <afb at algonet.se> wrote:
> That's an rpm-python bug then, it sounds like ?
> 
> Or does "rpm.archscore" work, and smart crash ?

Jeff, Anders, you're right, it seems to be rpm-python:

- ---%<---
basileia:~# cat /etc/rpm/platform
i686-slackware-linux
i586-slackware-linux
- --->%---

Should be ok, shouldn't it?

- ---%<---
basileia:~# python -c 'import rpm; print rpm.archscore("i586")'
2
Segmentation fault (core dumped)
- --->%---

I'll provide the core dump if it helps somebody, but it's 980K big and I 
guess the ML will reject it.

But it seems I'm misusing the mechanism here, since you do not seem to have 
any problems with it?

However, that still doesn't solve this issue: Why keep the rpm-md packages 
popping up, even if I don't have /etc/rpm/platform (except if that file 
exists and contains "i686-slackware-linux"), but the rpm-dir won't?

On Tuesday 23 March 2010, Jeff Johnson <n3npq at mac.com> wrote:
> Removing /etc/rpm/platform causes a fallback to the older
> mechanism to use arch from uname(2) fields. You will
> have much better control over archscore if you
> add a known CPU-VENDOR-OS string rather than depending
> on whatever comes from uname(2).

Does "known" mean "hardcoded in the RPM5 sources anywhere"? Since then I'll 
have to come up with some patch for me since I'm about to switch to a 
triplet nobody knows as of now. :-)

Thanks.

			Eric
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)

iEYEARECAAYFAkuoz3wACgkQhS0drJ3goJJ5FwCfQb9V2C+lQG/SnS4auiSrUVFC
fpIAn1wJVC605Li5YOLnbF03NpwdBeyZ
=LPuE
-----END PGP SIGNATURE-----



More information about the Smart mailing list