Re: [libvirt] [Xen-devel] [libvirt test] 50401: regressions - FAIL
On Tue, 2015-04-14 at 02:27 +0000, osstest service user wrote:
flight 50401 libvirt real [real] http://logs.test-lab.xenproject.org/osstest/logs/50401/
Regressions :-(
Tests which did not succeed and are blocking, including tests which could not be run: build-armhf-libvirt 5 libvirt-build fail REGR. vs. 50368 [...] Per http://logs.test-lab.xenproject.org/osstest/logs/50401/build-armhf-libvirt/5... this is:
qemu/qemu_driver.c: In function 'qemuDomainAddCgroupForThread': qemu/qemu_driver.c:4641:34: error: declaration of 'index' shadows a global declaration [-Werror=shadow] qemu/qemu_driver.c: In function 'qemuDomainHotplugAddPin': qemu/qemu_driver.c:4674:29: error: declaration of 'index' shadows a global declaration [-Werror=shadow] qemu/qemu_driver.c: In function 'qemuDomainHotplugPinThread': qemu/qemu_driver.c:4702:32: error: declaration of 'index' shadows a global declaration [-Werror=shadow] qemu/qemu_driver.c: In function 'qemuDomainDelCgroupForThread': qemu/qemu_driver.c:4733:34: error: declaration of 'index' shadows a global declaration [-Werror=shadow] cc1: all warnings being treated as errors This seems to be a general issue unrelated to Xen.
version targeted for testing: libvirt b487bb810ec95df862e7e80468c8e861ed80b0cb baseline version: libvirt 225aa80246d5e4a9e3a16ebd4c482525045da3db
After a quick glance I don't see a fix post-b487bb810ec9 either in master or on the libvirt list. Looking at the range under test it looks like one or more of John's changes is adding parameters called index, shadowing index(3) from <strings.h>. Ian.
On Tue, Apr 14, 2015 at 10:33:45AM +0100, Ian Campbell wrote:
On Tue, 2015-04-14 at 02:27 +0000, osstest service user wrote:
flight 50401 libvirt real [real] http://logs.test-lab.xenproject.org/osstest/logs/50401/
Regressions :-(
Tests which did not succeed and are blocking, including tests which could not be run: build-armhf-libvirt 5 libvirt-build fail REGR. vs. 50368 [...] Per http://logs.test-lab.xenproject.org/osstest/logs/50401/build-armhf-libvirt/5... this is:
qemu/qemu_driver.c: In function 'qemuDomainAddCgroupForThread': qemu/qemu_driver.c:4641:34: error: declaration of 'index' shadows a global declaration [-Werror=shadow] qemu/qemu_driver.c: In function 'qemuDomainHotplugAddPin': qemu/qemu_driver.c:4674:29: error: declaration of 'index' shadows a global declaration [-Werror=shadow] qemu/qemu_driver.c: In function 'qemuDomainHotplugPinThread': qemu/qemu_driver.c:4702:32: error: declaration of 'index' shadows a global declaration [-Werror=shadow] qemu/qemu_driver.c: In function 'qemuDomainDelCgroupForThread': qemu/qemu_driver.c:4733:34: error: declaration of 'index' shadows a global declaration [-Werror=shadow] cc1: all warnings being treated as errors
This seems to be a general issue unrelated to Xen.
version targeted for testing: libvirt b487bb810ec95df862e7e80468c8e861ed80b0cb baseline version: libvirt 225aa80246d5e4a9e3a16ebd4c482525045da3db
After a quick glance I don't see a fix post-b487bb810ec9 either in master or on the libvirt list.
Looking at the range under test it looks like one or more of John's changes is adding parameters called index, shadowing index(3) from <strings.h>.
Yeah, we've had this problem several times before - we usually just do a s/index/idx/ or similar to address it. Regards, Daniel -- |: http://berrange.com -o- http://www.flickr.com/photos/dberrange/ :| |: http://libvirt.org -o- http://virt-manager.org :| |: http://autobuild.org -o- http://search.cpan.org/~danberr/ :| |: http://entangle-photo.org -o- http://live.gnome.org/gtk-vnc :|
On Tue, 2015-04-14 at 10:37 +0100, Daniel P. Berrange wrote:
On Tue, Apr 14, 2015 at 10:33:45AM +0100, Ian Campbell wrote:
On Tue, 2015-04-14 at 02:27 +0000, osstest service user wrote:
flight 50401 libvirt real [real] http://logs.test-lab.xenproject.org/osstest/logs/50401/
Regressions :-(
Tests which did not succeed and are blocking, including tests which could not be run: build-armhf-libvirt 5 libvirt-build fail REGR. vs. 50368 [...] Per http://logs.test-lab.xenproject.org/osstest/logs/50401/build-armhf-libvirt/5... this is:
qemu/qemu_driver.c: In function 'qemuDomainAddCgroupForThread': qemu/qemu_driver.c:4641:34: error: declaration of 'index' shadows a global declaration [-Werror=shadow] qemu/qemu_driver.c: In function 'qemuDomainHotplugAddPin': qemu/qemu_driver.c:4674:29: error: declaration of 'index' shadows a global declaration [-Werror=shadow] qemu/qemu_driver.c: In function 'qemuDomainHotplugPinThread': qemu/qemu_driver.c:4702:32: error: declaration of 'index' shadows a global declaration [-Werror=shadow] qemu/qemu_driver.c: In function 'qemuDomainDelCgroupForThread': qemu/qemu_driver.c:4733:34: error: declaration of 'index' shadows a global declaration [-Werror=shadow] cc1: all warnings being treated as errors
This seems to be a general issue unrelated to Xen.
version targeted for testing: libvirt b487bb810ec95df862e7e80468c8e861ed80b0cb baseline version: libvirt 225aa80246d5e4a9e3a16ebd4c482525045da3db
After a quick glance I don't see a fix post-b487bb810ec9 either in master or on the libvirt list.
Looking at the range under test it looks like one or more of John's changes is adding parameters called index, shadowing index(3) from <strings.h>.
Yeah, we've had this problem several times before - we usually just do a s/index/idx/ or similar to address it.
I see this is now fixed in libvirt.git#master, thanks. However, I would just comment that contrary to the commit message, I don't think there is anything the Xen build has done which caused this, I think it's down to the LIBVIRT_COMPILE_WARNINGS macro which ends up enabling Wshadow if it was available on the system which ran autogen and on the compiling system. Our builds run on Debian Wheezy, which IIRC uses gcc 4.4 which isn't unusual... Ian.
On Wed, Apr 15, 2015 at 11:17:29AM +0100, Ian Campbell wrote:
On Tue, 2015-04-14 at 10:37 +0100, Daniel P. Berrange wrote:
On Tue, Apr 14, 2015 at 10:33:45AM +0100, Ian Campbell wrote:
On Tue, 2015-04-14 at 02:27 +0000, osstest service user wrote:
flight 50401 libvirt real [real] http://logs.test-lab.xenproject.org/osstest/logs/50401/
Regressions :-(
Tests which did not succeed and are blocking, including tests which could not be run: build-armhf-libvirt 5 libvirt-build fail REGR. vs. 50368 [...] Per http://logs.test-lab.xenproject.org/osstest/logs/50401/build-armhf-libvirt/5... this is:
qemu/qemu_driver.c: In function 'qemuDomainAddCgroupForThread': qemu/qemu_driver.c:4641:34: error: declaration of 'index' shadows a global declaration [-Werror=shadow] qemu/qemu_driver.c: In function 'qemuDomainHotplugAddPin': qemu/qemu_driver.c:4674:29: error: declaration of 'index' shadows a global declaration [-Werror=shadow] qemu/qemu_driver.c: In function 'qemuDomainHotplugPinThread': qemu/qemu_driver.c:4702:32: error: declaration of 'index' shadows a global declaration [-Werror=shadow] qemu/qemu_driver.c: In function 'qemuDomainDelCgroupForThread': qemu/qemu_driver.c:4733:34: error: declaration of 'index' shadows a global declaration [-Werror=shadow] cc1: all warnings being treated as errors
This seems to be a general issue unrelated to Xen.
version targeted for testing: libvirt b487bb810ec95df862e7e80468c8e861ed80b0cb baseline version: libvirt 225aa80246d5e4a9e3a16ebd4c482525045da3db
After a quick glance I don't see a fix post-b487bb810ec9 either in master or on the libvirt list.
Looking at the range under test it looks like one or more of John's changes is adding parameters called index, shadowing index(3) from <strings.h>.
Yeah, we've had this problem several times before - we usually just do a s/index/idx/ or similar to address it.
I see this is now fixed in libvirt.git#master, thanks.
However, I would just comment that contrary to the commit message, I don't think there is anything the Xen build has done which caused this, I think it's down to the LIBVIRT_COMPILE_WARNINGS macro which ends up enabling Wshadow if it was available on the system which ran autogen and on the compiling system.
Our builds run on Debian Wheezy, which IIRC uses gcc 4.4 which isn't unusual...
Yeah, there is nothing Xen specific about the problem - it is entirely down to the build toolchain & compiler options. Regards, Daniel -- |: http://berrange.com -o- http://www.flickr.com/photos/dberrange/ :| |: http://libvirt.org -o- http://virt-manager.org :| |: http://autobuild.org -o- http://search.cpan.org/~danberr/ :| |: http://entangle-photo.org -o- http://live.gnome.org/gtk-vnc :|
On Wed, 2015-04-15 at 14:16 +0100, Daniel P. Berrange wrote:
Yeah, there is nothing Xen specific about the problem - it is entirely down to the build toolchain & compiler options.
FYI our bisector has now tripped over another related problem, http://lists.xen.org/archives/html/xen-devel/2015-04/msg01900.html http://lists.xen.org/archives/html/xen-devel/2015-04/msg01968.html in particular: http://logs.test-lab.xenproject.org/osstest/logs/50460/build-armhf-libvirt/5...: qemu/qemu_domain.c: In function 'qemuDomainSupportsBlockJobs': qemu/qemu_domain.c:3067:11: error: declaration of 'sync' shadows a global declaration [-Werror=shadow] cc1: all warnings being treated as errors Essentially another instance of the same class of issue, in this case arising from: commit 1eccac1d2da7bbe97e1df25fd0ddac6e71b0794a Author: Peter Krempa <pkrempa@redhat.com> Date: Tue Mar 31 17:29:35 2015 +0200 qemu: domain: Add helper to check block job support Given the code in autoconf to enable essentially all warnings I'm a bit surprised -Wshadow isn't being enabled for everyone -- I thought it was quite a long standing warning. I'm also not sure why we are having a sudden rash of these now. But perhaps it is worth adding Wshadow to the set of excluded options until either it is more widely supported or the libvirt code base has been audited? I can send such a patch if you think that's desirable. Ian.
participants (2)
-
Daniel P. Berrange -
Ian Campbell