[libvirt] [PATCH] build: fix regression in large file support
https://bugzilla.redhat.com/show_bug.cgi?id=728772 points out that libvirt commit 1c93fbbb pulled in a gnulib regression that broke large file support. * .gnulib: Update to latest, for largefile fix. --- Pushing under the trivial rule. .gnulib | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/.gnulib b/.gnulib index 7f494c7..4470580 160000 --- a/.gnulib +++ b/.gnulib @@ -1 +1 @@ -Subproject commit 7f494c7d725db4a7f3abdef09d4070725487da67 +Subproject commit 4470580881a7b821b52fb5635102ef3e27aa5af4 -- 1.7.4.4
On 08/08/2011 09:53 AM, Eric Blake wrote:
https://bugzilla.redhat.com/show_bug.cgi?id=728772 points out that libvirt commit 1c93fbbb pulled in a gnulib regression that broke large file support.
* .gnulib: Update to latest, for largefile fix. ---
Pushing under the trivial rule.
.gnulib | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/.gnulib b/.gnulib index 7f494c7..4470580 160000 --- a/.gnulib +++ b/.gnulib @@ -1 +1 @@ -Subproject commit 7f494c7d725db4a7f3abdef09d4070725487da67 +Subproject commit 4470580881a7b821b52fb5635102ef3e27aa5af4
For reference: * .gnulib 7f494c7...4470580 (6):
largefile: fix typo that regressed large file support _Noreturn not a separate module openat: Fix warnings and commens when building unlinkat.c on Hurd. Tests for module 'pathmax'. canonicalize-lgpl: Support larger filenames on the Hurd. pathmax: Leave PATH_MAX undefined on the Hurd, and a constant otherwise.
-- Eric Blake eblake@redhat.com +1-801-349-2682 Libvirt virtualization library http://libvirt.org
Hi Eric, On Mon, Aug 08, 2011 at 09:53:39AM -0600, Eric Blake wrote:
https://bugzilla.redhat.com/show_bug.cgi?id=728772 points out that libvirt commit 1c93fbbb pulled in a gnulib regression that broke large file support.
I had a quick look and rerunning this pulls in quite some auto* changes. Any pointers how a minimal patch against 0.9.4 should look like that distributors could apply? Cheers, -- Guido
* .gnulib: Update to latest, for largefile fix. ---
Pushing under the trivial rule.
.gnulib | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/.gnulib b/.gnulib index 7f494c7..4470580 160000 --- a/.gnulib +++ b/.gnulib @@ -1 +1 @@ -Subproject commit 7f494c7d725db4a7f3abdef09d4070725487da67 +Subproject commit 4470580881a7b821b52fb5635102ef3e27aa5af4 -- 1.7.4.4
-- libvir-list mailing list libvir-list@redhat.com https://www.redhat.com/mailman/listinfo/libvir-list
On Mon, Aug 08, 2011 at 10:15:37PM +0200, Guido Günther wrote:
Hi Eric,
On Mon, Aug 08, 2011 at 09:53:39AM -0600, Eric Blake wrote:
https://bugzilla.redhat.com/show_bug.cgi?id=728772 points out that libvirt commit 1c93fbbb pulled in a gnulib regression that broke large file support.
I had a quick look and rerunning this pulls in quite some auto* changes. Any pointers how a minimal patch against 0.9.4 should look like that distributors could apply?
Yes, there is a patch attached to that BZ ticket above that is the minimal workaround distros should apply. 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, Aug 09, 2011 at 10:54:36AM +0100, Daniel P. Berrange wrote:
On Mon, Aug 08, 2011 at 10:15:37PM +0200, Guido Günther wrote:
Hi Eric,
On Mon, Aug 08, 2011 at 09:53:39AM -0600, Eric Blake wrote:
https://bugzilla.redhat.com/show_bug.cgi?id=728772 points out that libvirt commit 1c93fbbb pulled in a gnulib regression that broke large file support.
I had a quick look and rerunning this pulls in quite some auto* changes. Any pointers how a minimal patch against 0.9.4 should look like that distributors could apply?
Yes, there is a patch attached to that BZ ticket above that is the minimal workaround distros should apply. Works great. Thanks, -- Guido
On 08/09/2011 08:37 AM, Guido Günther wrote:
On Tue, Aug 09, 2011 at 10:54:36AM +0100, Daniel P. Berrange wrote:
On Mon, Aug 08, 2011 at 10:15:37PM +0200, Guido Günther wrote:
Hi Eric,
On Mon, Aug 08, 2011 at 09:53:39AM -0600, Eric Blake wrote:
https://bugzilla.redhat.com/show_bug.cgi?id=728772 points out that libvirt commit 1c93fbbb pulled in a gnulib regression that broke large file support.
I had a quick look and rerunning this pulls in quite some auto* changes. Any pointers how a minimal patch against 0.9.4 should look like that distributors could apply?
Yes, there is a patch attached to that BZ ticket above that is the minimal workaround distros should apply. Works great. Thanks,
That still implies re-running the autotools to convert configure.ac changes over to configure. If you want an absolute minimum patch that does not require the autotools during your distro build, and you are just building for Linux, then it should be sufficient to change your build script to use: ./configure CPPFLAGS=-D_FILE_OFFSET_BITS=64 -- Eric Blake eblake@redhat.com +1-801-349-2682 Libvirt virtualization library http://libvirt.org
On Tue, Aug 09, 2011 at 09:24:55AM -0600, Eric Blake wrote:
On 08/09/2011 08:37 AM, Guido Günther wrote:
On Tue, Aug 09, 2011 at 10:54:36AM +0100, Daniel P. Berrange wrote:
On Mon, Aug 08, 2011 at 10:15:37PM +0200, Guido Günther wrote:
Hi Eric,
On Mon, Aug 08, 2011 at 09:53:39AM -0600, Eric Blake wrote:
https://bugzilla.redhat.com/show_bug.cgi?id=728772 points out that libvirt commit 1c93fbbb pulled in a gnulib regression that broke large file support.
I had a quick look and rerunning this pulls in quite some auto* changes. Any pointers how a minimal patch against 0.9.4 should look like that distributors could apply?
Yes, there is a patch attached to that BZ ticket above that is the minimal workaround distros should apply. Works great. Thanks,
That still implies re-running the autotools to convert configure.ac changes over to configure. If you want an absolute minimum patch that does not require the autotools during your distro build, and you are just building for Linux, then it should be sufficient to change your build script to use:
./configure CPPFLAGS=-D_FILE_OFFSET_BITS=64 We also have kFreeBSD and Hurd in the mix so I opted for rerunning configure. Thanks, -- Guido
participants (3)
-
Daniel P. Berrange -
Eric Blake -
Guido Günther