[Yum] Segmentation fault
mstenner at ece.arizona.edu
Tue Apr 6 17:03:18 UTC 2004
On Tue, Apr 06, 2004 at 09:47:56AM -0400, seth vidal wrote:
> On Tue, 2004-04-06 at 02:29, Jules Colding wrote:
> > On Mon, 2004-04-05 at 20:47, seth vidal wrote:
> > > On Mon, 2004-04-05 at 07:20, Jules Colding wrote:
> > > > Hi,
> > > >
> > > > I got a SEGV while doing "yum install firefox". The last part of the
> > > > strace output is below. More info available upon request.
> > >
> > > pretty sure this is an rpm error reading in your gpg pubkeys - you might
> > > want to check out red hat bugzilla on this one.
> > Will do, but it worked when I commented a few fedora-us repos out:
> still could be the pubkeys - but try this - clear out the repositories
> that you commented out.
> then uncomment them.
> then run yum again and look for the segfault
SEGV means there is a bug in some C code*. It COULD be in the python
interpreter, in a core python module that contains some C code, or
some "local" python module that contains some C code. Personally, I
would put these in increasing order of likelihood simply due to the
amount of testing/use that each of these options gets.
As yum includes exactly 0 lines of C, there's really no chance this is
actually a yum bug. However, it is obviously occurring when you run
yum, so that seems like the best direction to go to find it. I just
wanted to be clear about where it could be. I agree with Seth that
this is most likely an rpm library issue.
* If you want to be really picky, it would be possible to write a
perfect module to provide direct memory access and manipulation to
the python level. Then, one could use that module incorrectly and
get a SEGV from a bug at the python layer. However, nobody is doing
that here :)
Michael D. Stenner mstenner at ece.arizona.edu
ECE Department, the University of Arizona 520-626-1619
1230 E. Speedway Blvd., Tucson, AZ 85721-0104 ECE 524G
More information about the Yum