I''ve updated extras^H^H^H^H^H^HFedora devel sparse to 0.3 and built it
in
Koji. I have no clue if magic is required to get it into the tag used for
new builds'' dependencies. The following ought to be adequate to tweak
the
kernel spec to use it rather than build it (assuming the old spec crap
wants to stick around conditionally for FC-6 copies).
Index: kernel-2.6.spec
==================================================================RCS file:
/cvs/pkgs/rpms/kernel/devel/kernel-2.6.spec,v
retrieving revision 1.3128
diff -u -b -p -r1.3128 kernel-2.6.spec
--- kernel-2.6.spec 3 May 2007 19:56:54 -0000 1.3128
+++ kernel-2.6.spec 3 May 2007 22:19:56 -0000
@@ -327,8 +355,15 @@ BuildConflicts: rhbuildsys(DiskFree) < 5
Source0: ftp://ftp.kernel.org/pub/linux/kernel/v2.6/linux-%{kversion}.tar.bz2
#Source1: xen-%{xen_hv_cset}.tar.bz2
Source2: Config.mk
+
%define sparsever 0.3
+%if "%fedora" >= "7"
+%define buildsparse 0
+BuildRequires: sparse >= %{sparsever}
+%else
+%define buildsparse 1
Source3: sparse-%{sparsever}.tar.bz2
+%endif
Source10: COPYING.modules
Source11: genkey
@@ -1436,10 +1491,12 @@ find . \( -name "*.orig" -o -name
"*~" \
cd ..
+%if %{buildsparse}
# unpack sparse.
if [ ! -d sparse-%{sparsever} ] ; then
%setup -D -T -q -a 3
fi
+%endif
# Unpack the Xen tarball.
%if %{includexen}
@@ -1462,12 +1519,15 @@ cd xen
%build
%if %{usesparse}
+%if %{buildsparse}
# Build sparse.
perl -p -i -e ''s/-O /-O2 -finline-functions /''
sparse-%{sparsever}/Makefile
make -C sparse-%{sparsever}
-
# Pass these options to kernel builds.
%define sparse_mflags C=1 CHECK=../sparse-%{sparsever}/sparse
+%else
+%define sparse_mflags C=1
+%endif
%endif
#
On Thu, May 03, 2007 at 03:25:04PM -0700, Roland McGrath wrote: > I''ve updated extras^H^H^H^H^H^HFedora devel sparse to 0.3 and built it in > Koji. I have no clue if magic is required to get it into the tag used for > new builds'' dependencies. The following ought to be adequate to tweak the > kernel spec to use it rather than build it (assuming the old spec crap > wants to stick around conditionally for FC-6 copies). heh, I had this on my todo, but you moved quicker than I did :) given we''re not moving FC-6 cvs over to the new repo, and that they already use disjunct spec files, we may as well make it unconditional. I don''t think there''s a great deal of point sparse''ing FC-6, (or even F7 once its out). Having it just a devel thing to check the latest tree seems to be more useful than checking the latest -stable release. comments? other than that, looks fine to me. Dave -- http://www.codemonkey.org.uk
On Thu, 2007-05-03 at 18:30 -0400, Dave Jones wrote:> On Thu, May 03, 2007 at 03:25:04PM -0700, Roland McGrath wrote: > > I''ve updated extras^H^H^H^H^H^HFedora devel sparse to 0.3 and built it in > > Koji. I have no clue if magic is required to get it into the tag used for > > new builds'' dependencies. The following ought to be adequate to tweak the > > kernel spec to use it rather than build it (assuming the old spec crap > > wants to stick around conditionally for FC-6 copies). > > heh, I had this on my todo, but you moved quicker than I did :) > given we''re not moving FC-6 cvs over to the new repo, and that they > already use disjunct spec files, we may as well make it unconditional. > > I don''t think there''s a great deal of point sparse''ing FC-6, > (or even F7 once its out). Having it just a devel thing to > check the latest tree seems to be more useful than checking the > latest -stable release. comments?Don''t we want to make sure than when we push updates to FC-6 that the sparse error logs aren''t getting worse (relative to an earlier FC-6 kernel)? (or have I misunderstood the workflow here?) [snip]
On Thu, May 03, 2007 at 06:48:56PM -0400, David Malcolm wrote: > On Thu, 2007-05-03 at 18:30 -0400, Dave Jones wrote: > > On Thu, May 03, 2007 at 03:25:04PM -0700, Roland McGrath wrote: > > > I''ve updated extras^H^H^H^H^H^HFedora devel sparse to 0.3 and built it in > > > Koji. I have no clue if magic is required to get it into the tag used for > > > new builds'' dependencies. The following ought to be adequate to tweak the > > > kernel spec to use it rather than build it (assuming the old spec crap > > > wants to stick around conditionally for FC-6 copies). > > > > heh, I had this on my todo, but you moved quicker than I did :) > > given we''re not moving FC-6 cvs over to the new repo, and that they > > already use disjunct spec files, we may as well make it unconditional. > > > > I don''t think there''s a great deal of point sparse''ing FC-6, > > (or even F7 once its out). Having it just a devel thing to > > check the latest tree seems to be more useful than checking the > > latest -stable release. comments? > Don''t we want to make sure than when we push updates to FC-6 that the > sparse error logs aren''t getting worse (relative to an earlier FC-6 > kernel)? (or have I misunderstood the workflow here?) Well, if we had infrastructure other than ''eyeballs & diff'' to look over then maybe, but until then, I don''t think it really adds anything other than slightly increasing buildtime. Dave -- http://www.codemonkey.org.uk