Description of problem: The 3.19-rc5 and 3.19-rc6 upstream kernel patches contain a hunk that creates a symbolic link from arch/arm64/boot/dts/include/dt-bindings to include/dt-bindings. The hunk is below: diff --git a/arch/arm64/boot/dts/include/dt-bindings b/arch/arm64/boot/dts/include/dt-bindings new file mode 120000 index 000000000000..08c00e4972fa --- /dev/null +++ b/arch/arm64/boot/dts/include/dt-bindings @@ -0,0 +1 @@ +../../../../../include/dt-bindings \ No newline at end of file On my local machine running F21 (patch-2.7.1-9.fc21), this applies fine. In the 3.19-rc5 build in koji, it also applied fine and the kernel built correctly: http://koji.fedoraproject.org/koji/buildinfo?buildID=604938 When I went to build the 3.19-rc6 build this morning, it built correctly on my local machine however it failed in koji with: + '[' '!' -f /builddir/build/SOURCES/patch-3.19-rc6.xz ']' + case "$patch" in + unxz + patch -p1 -F1 -s symbolic link target '../../../../../include/dt-bindings' is invalid error: Bad exit status from /var/tmp/rpm-tmp.mWE3ZL (%prep) http://koji.fedoraproject.org/koji/taskinfo?taskID=8722450 Looking at root.log for each respective koj build I see that patch-2.7.1-9.fc22 was used for the 3.19-rc5 build, and patch-2.7.3-1.fc22 was used for the failing 3.19-rc6 build. Seems something that worked fine before is now failing with a patch update. Version-Release number of selected component (if applicable): patch-2.7.3-1.fc22 How reproducible: Assuming always, but I've only tried one koji build. I will kick off a local mockbuild that should bring in the updated patch into the buildroot as well and see if that fails. Steps to Reproduce: 1. fedpkg clone kernel on a rawhide machine with above patch version installed 2. fedpkg prep 3. Actual results: Fails to apply symbolic link hunk in patch-3.19-rc6.xz Expected results: Patch applies as before. Additional info:
Preventing ".." components in symbolic links is intentional in order to fix CVE-2015-1196. https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=775901#13
Maybe WONTFIX was too fast :-) It looks like this patch regression is, erhm, patched in patch 2.7.4. See http://www.gossamer-threads.com/lists/linux/kernel/2099059#2099059
Yes, it's what I spent last week on. Sorry, I'd forgotten to update this. Fixed in: patch-2.7.4-1.fc20 https://admin.fedoraproject.org/updates/FEDORA-2015-1165 patch-2.7.4-1.fc21 https://admin.fedoraproject.org/updates/patch-2.7.4-1.fc21 patch-2.7.4-1.fc22
This bug appears to have been reported against 'rawhide' during the Fedora 22 development cycle. Changing version to '22'. More information and reason for this action is here: https://fedoraproject.org/wiki/Fedora_Program_Management/HouseKeeping/Fedora22