fedora kernel utrace updates

Dave Jones davej at redhat.com
Mon Jun 11 21:30:29 UTC 2007


On Mon, Jun 11, 2007 at 05:22:10PM -0400, Chuck Ebbert wrote:
 > On 06/11/2007 04:46 PM, Roland McGrath wrote:
 > > I've started using quilt to maintain backport versions of my utrace patches.
 > > http://people.redhat.com/roland/utrace/ now includes 2.6.21/ (for F-7) and
 > > 2.6.20.11 (for FC-6).  Done this way it is easy for me to plop in the new
 > > utrace/ptrace code from the bleeding edge version when I have fixes, without
 > > repeating the backport work, which is confined to the earlier patches in the
 > > series.  
 > > 
 > > Since quilt doesn't have a handy feature for generating a single patch, and
 > > I know better than to trust combinediff, I'd like to switch the Fedora
 > > update kernels to using the broken out patches separately in the spec file.
 > > If one were actually trying to keep track, this also makes it a little
 > > easier to follow the changed patches in cvs, since there will be less
 > > flutter around parts that aren't actually changing.
 > > 
 > 
 > 'quilt snapshot' can be used to make combined patches, but I prefer
 > the separate patches anyway. For FC6 I'd prefer to do the commits;
 > if you'd notify us when you make a change that would help.

Same goes for F-7 I guess.  For devel however, it's a bit of a pain to
have to disable a dozen ApplyPatch's during the (brief) window
when I rebase, and you haven't rediffed utrace yet (if its trivial,
I've done it myself in the past, but it isn't always the case..)

In more recent times, I actually as part of my daily -git rebasing
have this tied into my scripts..
rm -f linux-2.6-utrace.patch
wget http://people.redhat.com/roland/utrace/2.6-current/linux-2.6-utrace.patch

So roll-up patches are still kinda handy for making that sort of
thing a breeze.

Better yet of course would be upstream integration, making this
whole problem go away.  But then, you knew that :)

	Dave

-- 
http://www.codemonkey.org.uk




More information about the Fedora-kernel-list mailing list