On Thu, 25 Aug 2011 17:45:38 +0200 Johannes Weiner wrote: > On Fri, Aug 26, 2011 at 01:09:38AM +1000, Stephen Rothwell wrote: > > Hi, > > > > On Thu, 25 Aug 2011 16:07:01 +0200 Michal Hocko <mhocko@xxxxxxx> wrote: > > > > > > On Thu 25-08-11 15:51:03, Michal Hocko wrote: > > > > > > > > On Wed 24-08-11 14:09:05, Andrew Morton wrote: > > > > > The mm-of-the-moment snapshot 2011-08-24-14-08 has been uploaded to > > > > > > > > > > http://userweb.kernel.org/~akpm/mmotm/ > > > > > > > > I have just downloaded your tree and cannot quilt it up. I am getting: > > > > [...] > > > > patching file tools/power/cpupower/debug/x86_64/centrino-decode.c > > > > Hunk #1 FAILED at 1. > > > > File tools/power/cpupower/debug/x86_64/centrino-decode.c is not empty after patch, as expected > > > > 1 out of 1 hunk FAILED -- rejects in file tools/power/cpupower/debug/x86_64/centrino-decode.c > > > > patching file tools/power/cpupower/debug/x86_64/powernow-k8-decode.c > > > > Hunk #1 FAILED at 1. > > > > File tools/power/cpupower/debug/x86_64/powernow-k8-decode.c is not empty after patch, as expected > > > > 1 out of 1 hunk FAILED -- rejects in file tools/power/cpupower/debug/x86_64/powernow-k8-decode.c > > > > [...] > > > > patching file virt/kvm/iommu.c > > > > Patch linux-next.patch does not apply (enforce with -f) > > > > > > > > Is this a patch (I am using 2.6.1) issue? The failing hunk looks as > > > > follows: > > > > --- a/tools/power/cpupower/debug/x86_64/centrino-decode.c > > > > +++ /dev/null > > > > @@ -1 +0,0 @@ > > > > -../i386/centrino-decode.c > > > > \ No newline at end of file > > > > > > Isn't this just a special form of git (clever) diff to spare some lines > > > when the file deleted? Or is the patch simply corrupted? > > > Anyway, my patch doesn't cope with that. Any hint what to do about it? > > > > Those files were symlinks and were removed by a commit in linux-next. > > diff/patch does not cope with that. > > You can probably replace `patch' in your $PATH by a wrapper that uses > git-apply, which can deal with them. > > Or you could use git-quiltimport, which uses git-apply, to prepare the > -mmotm base tree in git, on top of which you can continue to work with > quilt. > > I do this with a cron-job automatically, you can find the result here: > > http://git.kernel.org/?p=linux/kernel/git/hannes/linux-mmotm.git;a=summary > > If you want to do it manually, there is sometimes confusing binary > file patch sections in -mmotm, which in turn git-apply can not deal > with, so I use the following uncrapdiff.awk filter on the patches > before import. I thought that I recently saw a reference to an unreleased version of quilt that handles git symlinks. Was I dreaming? or where is it? Thanks. > --- > > # Filter out sections that feature an index line but > # no real diff part that would start with '--- file' > > { > if (HEADER ~ /^diff --git /) { > if ($0 ~ /^index /) { > INDEX=$0 > } else if ($0 ~ /^diff --git /) { > print(HEADER) > HEADER=$0 > } else if (INDEX ~ /^index /) { > if ($0 ~ /^--- /) { > print(HEADER) > print(INDEX) > print($0) > } > HEADER="" > INDEX="" > } else { > HEADER=HEADER "\n" $0 > } > } else if ($0 ~ /^diff --git /) { > HEADER=$0 > } else { > print($0) > } > } > > --- --- ~Randy *** Remember to use Documentation/SubmitChecklist when testing your code *** -- To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html