On Thu, Jan 17, 2008 at 04:26:40PM +0900, Hiroyuki Kaguchi wrote:
Currently NIC definition order is reversed.
If we set over 4-NICs for a domain by virt-install,
this causes a network communication problem.
This is because virtual mac to eth relation is lost.
This patch fixes the NIC
On Sat, Jan 19, 2008 at 07:18:03PM +, Daniel P. Berrange wrote:
The referencing counting code for Connect/Domain/Network objects has many
repeated codepaths, not all of which are correct. eg, the virFreeDomain
method forgets to release networks when garbage collecting a virConnectPtr,
and
On Sat, Jan 19, 2008 at 07:37:01PM +, Daniel P. Berrange wrote:
This patch makes virsh use the virFileReadAll API from util.c, instead of
duplicating it. This means we need to export this symbol from the library.
okay +1, but se should try to limit the number of __exports in general.
These struct definitions *intentionally* private.
Oops, I did not notice it...
I revised a patch not to access the struct data directly.
In principle this patch looks good. If no one else objects, I'll commit
this, with a few of my own fixes (below).
Rich.
+
On Sat, Jan 19, 2008 at 06:44:19PM +, Daniel P. Berrange wrote:
The openvz_driver.c file registers an implementation of the networking
driver APIs, which is a no-op.
Even though the networking APIs are implemented in the QEMU driver, this
one impl is *shared* and used across all
Daniel P. Berrange wrote:
On Sun, Jan 20, 2008 at 12:46:07PM +, Richard W.M. Jones wrote:
This is an experimental release of the OCaml libvirt bindings and tools.
The main changes are that we have moved to automatically generating as
many of the C bindings as possible, and this includes
Daniel P. Berrange [EMAIL PROTECTED] wrote:
On Sat, Jan 19, 2008 at 10:48:46PM +0100, Jim Meyering wrote:
I was looking at xend_internal.c and wondered why sexpr_int's
sexpr pointer wasn't const... surely, it *can't* modify that, I thought.
So I made it const, and pulled the thread, which
Atsushi SAKAI wrote:
Hi, Rich
I try to build on current libvirt on Windows.
But failed in attached compilation message.
I would be appreciate If you have any suggestions.
My Instruction is follows.
build cygwinenvironment.
Install gettext-devel, gnutls, gnutls-devel additionally.
install
Katti, Vadiraj (STSD-Openview) wrote:
Hi,
I've some questions related to xen. Can anybody help me regarding the
following questions.
These sound more like Xen questions, but anyway ...
1. Is dom0 a logical system(like any other vm's that are configured on
it) ? If yes, then all xm commands
Daniel P. Berrange wrote:
The openvz_driver.c file registers an implementation of the networking
driver APIs, which is a no-op.
Even though the networking APIs are implemented in the QEMU driver, this
one impl is *shared* and used across all virtualization driver backends.
Thus no other
Daniel P. Berrange wrote:
The current Xen capabilities XML data chooses between /usr/lib/xen/bin/qemu-dm
and /usr/lib64/xen/bin/qemu-dm based on whether the guest arch is 64 bit.
This is incorrect because the path is actually determined based on the host
OS. Only x86_64 uses /usr/lib64, all
Daniel P. Berrange wrote:
The python generator will happily ignore functions it can't handle and
pretend everything completed without error. This leads to the situation
where we add new APis to C library and no one ever notices that they
are missing from the python until months later. This
In reviewing the StorageAPI patches, I saw that some bogus input could
cause a segfault. There's similar code in libvirt.c, so this fixes both.
Note that virsh.c does this, too, but it already has the required guard.
The src/libvirt.c patch applies to the trunk, but isn't a big
deal, since it's
Jim Meyering wrote:
In reviewing the StorageAPI patches, I saw that some bogus input could
cause a segfault. There's similar code in libvirt.c, so this fixes both.
Note that virsh.c does this, too, but it already has the required guard.
The src/libvirt.c patch applies to the trunk, but isn't a
On Mon, Jan 21, 2008 at 04:14:15AM -0500, Daniel Veillard wrote:
On Sun, Jan 20, 2008 at 05:29:06PM +, Daniel P. Berrange wrote:
The python generator will happily ignore functions it can't handle and
pretend everything completed without error. This leads to the situation
where we add
On Mon, Jan 21, 2008 at 11:55:45AM +, Richard W.M. Jones wrote:
Daniel P. Berrange wrote:
The current Xen capabilities XML data chooses between
/usr/lib/xen/bin/qemu-dm
and /usr/lib64/xen/bin/qemu-dm based on whether the guest arch is 64 bit.
This is incorrect because the path is
Daniel Veillard wrote:
Well, docs/libvirt-api.xml is automatically generated from the headers
(and C modules), but python/libvirt-python-api.xml is manually created
to describe similar input but for the functions where the C binding is done
manually. This allows the generator to plug them at
Daniel Veillard [EMAIL PROTECTED] wrote:
On Mon, Jan 21, 2008 at 01:37:25PM +0100, Jim Meyering wrote:
In reviewing the StorageAPI patches, I saw that some bogus input could
cause a segfault. There's similar code in libvirt.c, so this fixes both.
Note that virsh.c does this, too, but it
On Mon, Jan 21, 2008 at 02:48:58PM +, Daniel P. Berrange wrote:
That I found worrying. What is the added benefit for the loss of
portability? You now reference 2 relatively system specific kind of
data structure, where we used to reference only the xmlMutex one which
was portable.
On Mon, Jan 21, 2008 at 02:57:37PM +, Richard W.M. Jones wrote:
Daniel P. Berrange wrote:
The problem is that we need to use pthread_t / pthread_create APIs for
other parts of libvirt. libxml2 doesn't provide any portability layer
for threads - only for mutexes. Any platform which has
On Sun, Jan 20, 2008 at 12:23:26PM +, Richard W.M. Jones wrote:
Daniel P. Berrange wrote:
This patch makes virsh use the virFileReadAll API from util.c, instead of
duplicating it. This means we need to export this symbol from the library.
+1.
Comitted.
Dan.
--
|=- Red Hat,
On Mon, Jan 21, 2008 at 09:29:12AM -0500, Daniel Veillard wrote:
On Mon, Jan 21, 2008 at 02:26:36PM +, Daniel P. Berrange wrote:
On Mon, Jan 21, 2008 at 09:02:09AM -0500, Daniel Veillard wrote:
On Mon, Jan 21, 2008 at 01:58:07PM +, Daniel P. Berrange wrote:
On Mon, Jan 21, 2008
On Mon, Jan 21, 2008 at 09:12:12AM -0500, Daniel Veillard wrote:
On Mon, Jan 21, 2008 at 02:02:28PM +, Daniel P. Berrange wrote:
On Mon, Jan 21, 2008 at 04:14:15AM -0500, Daniel Veillard wrote:
On Sun, Jan 20, 2008 at 05:29:06PM +, Daniel P. Berrange wrote:
The python generator
Daniel P. Berrange [EMAIL PROTECTED] wrote:
hg qclone -p http://hg.berrange.com/libraries/libvirt--storage \
http://hg.berrange.com/libraries/libvirt--devel
Relative to a snapshot of a few minutes ago, here are two patches:
* src/storage_backend_fs.c
Jim Meyering wrote:
I've been seeing this on rawhide for a few days:
xmconfigtest.c:115: warning: 'old_priv' may be used uninitialized in this
function
+1.
I always compile with --enable-compiler-warnings=fatal. Haven't seen
this, but there again Rawhide wasn't working very well for me so
Hi Dan,
I noticed a few new unnecessary if-before-free tests:
src/storage_backend.c:2591: if (groups[j]) free(groups[j]);
src/storage_backend_iscsi.c:if (devpath) free(devpath);
src/storage_backend_iscsi.c:if (portal) free(portal);
src/storage_backend_iscsi.c:if (session)
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
I spent the weekend attempting to get libvirt working on OS X Leopard
10.5.1. I was able to get it to compile, but not connect to libvirtd
running on either my KVM or Xen servers. Here are the instructions for
getting it to compile:
1.
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Ah, I knew hyper was 64-bit, but it has no definition on OS X. I
thought that long on OS X 10.5.1 was 64 bits in width -- my mistake.
There wre two bugs in the remoteAuthenticate method in libvirt 0.4.0
which
can cause crashes. I'd recommend
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Okay, the autogen.sh fails right off the bat. To get it working you
have to edit it so that it checks the ld version with a '-V' instead
of '--version' (OS X specific). Once you do that and run:
./autogen.sh --without-xen --without-qemu
it
On Mon, Jan 21, 2008 at 05:32:33PM -0600, Schley Andrew Kutz wrote:
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Okay, the autogen.sh fails right off the bat. To get it working you
have to edit it so that it checks the ld version with a '-V' instead
of '--version' (OS X specific). Once
On Mon, Jan 21, 2008 at 05:27:33PM -0600, Schley Andrew Kutz wrote:
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Ah, I knew hyper was 64-bit, but it has no definition on OS X. I
thought that long on OS X 10.5.1 was 64 bits in width -- my mistake.
XDR is a language for defining portable
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Here is all the mentions of xdr_hyper in /usr/include on my box:
[EMAIL PROTECTED]:libvirt-HEAD-20080121$ grep -R xdr_hyper /usr/include
/usr/include/nfs/nfsm_subs.h: txdr_hyper(__tmp64, (NMC)-nmc_ptr); \
/usr/include/nfs/nfsm_subs.h
On Mon, Jan 21, 2008 at 04:50:29PM -0600, Schley Andrew Kutz wrote:
5. Edit '/usr/include/net/if.h' and add the following include:
#include net/if_var.h
#include sys/types.h
//
// EDIT: akutz
//
#include sys/socket.h
This is curious - I'm puzzelled why libvirt source uses
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Fails the same way on Linux (Ubuntu 7.10 Gutsy Gibbon - Server -
x86_64):
./configure: line 35963: syntax error near unexpected token
`libxml-2.0,LIBXML_FOUND=yes'
./configure: line 35963: `
PKG_CHECK_EXISTS(libxml-2.0,LIBXML_FOUND=yes)'
On Mon, Jan 21, 2008 at 05:52:06PM -0600, Schley Andrew Kutz wrote:
I've attached the compile failure I get.
Ok, so your header files suggest Mac OS *does* have xdr_hyper, and this
failure is a link time failure:
Undefined symbols:
_xdr_quad_t, referenced from:
On Mon, Jan 21, 2008 at 05:58:06PM -0600, Schley Andrew Kutz wrote:
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Fails the same way on Linux (Ubuntu 7.10 Gutsy Gibbon - Server -
x86_64):
./configure: line 35963: syntax error near unexpected token
`libxml-2.0,LIBXML_FOUND=yes'
On Tue, Jan 22, 2008 at 12:01:10AM +, Daniel P. Berrange wrote:
On Mon, Jan 21, 2008 at 05:52:06PM -0600, Schley Andrew Kutz wrote:
I've attached the compile failure I get.
Ok, so your header files suggest Mac OS *does* have xdr_hyper, and this
failure is a link time failure:
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Not so much. I added the two headers from http://developer.apple.com/documentation/Darwin/Reference/ManPages/man3/xdr.3.html
to remote_protocol.x and the gen fails:
[EMAIL PROTECTED]:libvirt-0.4.0$ make -C qemud remote_protocol.c
rm -f
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
I was missing pkg-config. But the 'make dist' now fails:
make[3]: Entering directory `/usr/local/src/libvirt/docs/examples/
python'
make[3]: Leaving directory `/usr/local/src/libvirt/docs/examples/python'
make[2]: Leaving directory
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
[EMAIL PROTECTED]:qemud$ make remote_protocol.c
rm -f remote_protocol.c
rpcgen -c -o remote_protocol.c remote_protocol.x
unsigned hyper cpu_time;
^^^
remote_protocol.x, line 144: expected ';'
make: *** [remote_protocol.c]
On Mon, Jan 21, 2008 at 06:18:26PM -0600, Schley Andrew Kutz wrote:
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
[EMAIL PROTECTED]:qemud$ make remote_protocol.c
rm -f remote_protocol.c
rpcgen -c -o remote_protocol.c remote_protocol.x
unsigned hyper cpu_time;
Not it. When I changed all instances of hyper to long earlier it
worked. The hyper is not getting parses. That said, I'll still try
your suggestion when I'm back in front of my keyboard.
--
-a
On Jan 21, 2008, at 6:32 PM, Daniel P. Berrange
[EMAIL PROTECTED] wrote:
On Mon, Jan 21, 2008
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Verified that camelCase does not help.
- --
- -a
On Jan 21, 2008, at 6:46 PM, Schley Andrew Kutz wrote:
Not it. When I changed all instances of hyper to long earlier it
worked. The hyper is not getting parses. That said, I'll still try
your
On Mon, Jan 21, 2008 at 07:27:58PM -0600, Schley Andrew Kutz wrote:
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Nevermind, I missed the date. Although, I do wonder if rpcgen on OS X
supports 'hyper'? I am doing some digging to find out if it does.
The Apple docs definitely suggest their
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
I'll try the latter. However, from an earlier e-mail I pointed out
that I cannot get the 'make dist' to work on Linux. It moans about a
missing help file:
cp: cannot stat `./libvirt.devhelp': No such file or directory
make[2]: *** [distdir]
Hi,
I downloaded libvirt-0.4.0 sources and patched it with
libvirt-windows-20071129-2.patch.
I followed all the instructions from http://libvirt.org/windows.html and
experienced some compilation problems which
Caused me to make the following changes.
1- For some reason Cygwin does not
46 matches
Mail list logo