November 2008 - Devel - libvirt List Archives

[libvirt] [PATCH] fix build for srcdir != builddir
by Guido Günther 25 Nov '08

25 Nov '08

Hi, libvirt_sym.version is looked for at the wrong location. O.k. to apply? Cheers, -- Guido

2 3

[libvirt] [PATCH] Xen Events Updated
by Ben Guthro 25 Nov '08

25 Nov '08

This version of the Xen-Events patch goes a step further than the last version, and now emits the following domain events: STARTED STOPPED ADDED REMOVED This is accomplished by monitoring /etc/xen, and taking advantage of what seemed to be dormant code in the XM driver. By re-enabling the "config cache", we can properly track domains when files come, and go into this directory. configure.in | 13 + include/libvirt/virterror.h | 1 po/POTFILES.in | 1 src/Makefile.am | 3 src/virterror.c | 3 src/xen_inotify.c | 270 +++++++++++++++++++++++++++ src/xen_inotify.h | 32 +++ src/xen_unified.c | 190 +++++++++++++++++++ src/xen_unified.h | 61 ++++++ src/xm_internal.c | 24 +- src/xs_internal.c | 430 +++++++++++++++++++++++++++++++++++++++++++- src/xs_internal.h | 51 +++++ 12 files changed, 1069 insertions(+), 10 deletions(-)

3 15

[libvirt] PATCH: allow to set nvcpus = 0
by Evgeniy Sokolov 25 Nov '08

25 Nov '08

OpenVZ uses all CPUs available in system - by default (number of CPUs did not set) - number of CPUs = 0 Currenty, libvirt don't allow to set nvcpus = 0. Attached patch removes limitation in libvirt set nvcpu = 0, but add it to each driver which allow to set number of virtual CPU. For OpenVZ set default number of CPUs = 0.

2 3

[libvirt] [PATCH] fix python events
by David Lively 25 Nov '08

25 Nov '08

This patch gets python events working again after upstream changes, and make the test implementation properly clean up after itself and implement the new EventImpl API properly. Note that the Python RemoveHandle and RemoveTimeout implementations should return the opaque object registered by the corresponding AddHandle/Timeout calls, in lieu of calling the (C) freefunc. (The binding code will then call freefunc if it's not NULL.) Ignoring this means you'll leak memory in the same way that C RemoveHandle/Timeout leak if they don't (now) call the freefunc. I also moved around some of the error checking code to unclutter (and speed up) the common code paths. For instance, we now check that the virRegisterEventImpl arguments are callable just once (and return failure if they're not), rather than checking them before every call and blithely ignoring them if they're not callable. Dave examples/domain-events/events-python/event-test.py | 29 +-- python/libvir.c | 200+++++++++++++++---- python/libvir.py | 4 python/libvirt_wrap.h | 8 python/types.c | 1 python/virConnect.py | 4 6 files changed, 194 insertions(+), 52 deletions(-)

3 10

[libvirt] [PATCH] libvirt.c: fix virStateActive return value; document ...
by Jim Meyering 25 Nov '08

25 Nov '08

The s/1/-1/ fix induces no semantic change, since the sole use of virStateActive tests solely for nonzero. >From 4bc9713207a2ed6b101e2067f7bba82d1df45987 Mon Sep 17 00:00:00 2001 From: Jim Meyering <meyering(a)redhat.com> Date: Mon, 24 Nov 2008 15:52:55 +0100 Subject: [PATCH] libvirt.c: fix virStateActive return value; document some new functions * src/libvirt.c (virStateActive): Return -1 upon error, not 1, to be consistent with the other virState* functions. (virStateActive, virStateCleanup, virStateReload, virStateActive): Add per-function comments. --- src/libvirt.c | 30 +++++++++++++++++++++++++++++- 1 files changed, 29 insertions(+), 1 deletions(-) diff --git a/src/libvirt.c b/src/libvirt.c index f445333..10835b9 100644 --- a/src/libvirt.c +++ b/src/libvirt.c @@ -636,6 +636,13 @@ virRegisterStateDriver(virStateDriverPtr driver) return virStateDriverTabCount++; } +/** + * virStateInitialize: + * + * Initialize all virtualization drivers. + * + * Return 0 if successful, -1 upon error. + */ int virStateInitialize(void) { int i, ret = 0; @@ -650,6 +657,13 @@ int virStateInitialize(void) { return ret; } +/** + * virStateCleanup: + * + * Run each virtualization driver's cleanup method. + * + * Return 0 if successful, -1 upon error. + */ int virStateCleanup(void) { int i, ret = 0; @@ -661,6 +675,13 @@ int virStateCleanup(void) { return ret; } +/** + * virStateReload + * + * Run each virtualization driver's reload method. + * + * Return 0 if successful, -1 upon error. + */ int virStateReload(void) { int i, ret = 0; @@ -672,13 +693,20 @@ int virStateReload(void) { return ret; } +/** + * virStateActive + * + * Run each virtualization driver's "active" method. + * + * Return 0 if successful, -1 upon error. + */ int virStateActive(void) { int i, ret = 0; for (i = 0 ; i < virStateDriverTabCount ; i++) { if (virStateDriverTab[i]->active && virStateDriverTab[i]->active()) - ret = 1; + ret = -1; } return ret; } -- 1.6.0.4.1021.g4320

3 5

[libvirt] [PATCH] fix --without-xen build
by David Lively 25 Nov '08

25 Nov '08

The last change to tests/Makefile.am broke the build when --without-xen is specified. This patch fixes it ... Dave tests/Makefile.am | 9 +++++++++ 1 file changed, 9 insertions(+)

3 5

[libvirt] [PATCH] * src/virterror.c (virReportErrorHelper): Sync doc to code.
by Jim Meyering 25 Nov '08

25 Nov '08

The dom and net parameters are long gone. Reflect s/msg/fmt/ renaming: >From 8b87344dbe31fe478ac4a679005835e03a329ace Mon Sep 17 00:00:00 2001 From: Jim Meyering <meyering(a)redhat.com> Date: Mon, 24 Nov 2008 15:55:24 +0100 Subject: [PATCH] * src/virterror.c (virReportErrorHelper): Sync doc to code. --- src/virterror.c | 6 ++---- 1 files changed, 2 insertions(+), 4 deletions(-) diff --git a/src/virterror.c b/src/virterror.c index 0723db8..112a19e 100644 --- a/src/virterror.c +++ b/src/virterror.c @@ -747,17 +747,15 @@ virErrorMsg(virErrorNumber error, const char *info) } /** - * virReportErrorHelper + * virReportErrorHelper: * * @conn: the connection to the hypervisor if available - * @dom: the domain if available - * @net: the network if available * @domcode: the virErrorDomain indicating where it's coming from * @errcode: the virErrorNumber code for the error * @filename: Source file error is dispatched from * @funcname: Function error is dispatched from * @linenr: Line number error is dispatched from - * @msg: the message to display/transmit + * @fmt: the format string * @...: extra parameters for the message display * * Helper function to do most of the grunt work for individual driver -- 1.6.0.4.1021.g4320

2 1

[libvirt] [PATCH] tests: virsh-all: new script
by Jim Meyering 22 Nov '08

22 Nov '08

Actually, this was the first step. This tiny script merely invokes all commands blindly, solely to get test coverage, even if they (as most do) fail. Additional tests will be smarter about providing sensible arguments. >From 058681eab07936f4ecbb1ff514528eb921d27074 Mon Sep 17 00:00:00 2001 From: Jim Meyering <meyering(a)redhat.com> Date: Fri, 21 Nov 2008 10:18:43 +0100 Subject: [PATCH] tests: virsh-all: new script * tests/virsh-all: New script. * tests/Makefile.am (test_scripts): Add virsh-all. --- tests/Makefile.am | 1 + tests/virsh-all | 39 +++++++++++++++++++++++++++++++++++++++ 2 files changed, 40 insertions(+), 0 deletions(-) create mode 100755 tests/virsh-all diff --git a/tests/Makefile.am b/tests/Makefile.am index bdbf927..0b4eebd 100644 --- a/tests/Makefile.am +++ b/tests/Makefile.am @@ -57,6 +57,7 @@ test_scripts += \ read-non-seekable \ undefine \ vcpupin + virsh-all endif EXTRA_DIST += $(test_scripts) diff --git a/tests/virsh-all b/tests/virsh-all new file mode 100755 index 0000000..af7e09d --- /dev/null +++ b/tests/virsh-all @@ -0,0 +1,39 @@ +#!/bin/sh +# blindly run each and every command listed by "virsh help" + +# Copyright (C) 2008 Free Software Foundation, Inc. + +# This program is free software: you can redistribute it and/or modify +# it under the terms of the GNU General Public License as published by +# the Free Software Foundation, either version 3 of the License, or +# (at your option) any later version. + +# This program is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +# GNU General Public License for more details. + +# You should have received a copy of the GNU General Public License +# along with this program. If not, see <http://www.gnu.org/licenses/>. + +if test "$VERBOSE" = yes; then + set -x + virsh --version +fi + +. $srcdir/test-lib.sh + +fail=0 + +test_url=test:///default + +virsh -c $test_url help > cmds || framework_failure +cmds=$(sed -n 's/^ $[^ ][^ ]*$ .*/\1/p' cmds) || framework_failure +test -n "$cmds" || framework_failure + +for i in $cmds; do + echo testing $i... 1>&2 + virsh -c $test_url $i < /dev/null +done + +(exit $fail); exit $fail -- 1.6.0.4.1021.g4320

2 2

[libvirt] [PATCH] * gnulib/tests/test-EOVERFLOW: Remove.
by Jim Meyering 22 Nov '08

22 Nov '08

Back on October 28, I mistakenly added a binary. This adds the requisite .cvsignore entry and updates the corresponding .gitignore file as well as two stale ones. >From 30a853d37fc2102a7e4009b93867209e318014a1 Mon Sep 17 00:00:00 2001 From: Jim Meyering <meyering(a)redhat.com> Date: Fri, 21 Nov 2008 14:15:15 +0100 Subject: [PATCH] * gnulib/tests/test-EOVERFLOW: Remove. * gnulib/tests/.cvsignore: Add test-EOVERFLOW, then... Run "make sync-vcs-ignore-files" to Update .gitignore files. --- examples/domain-events/events-c/.gitignore | 2 ++ gnulib/tests/.cvsignore | 1 + gnulib/tests/.gitignore | 1 + gnulib/tests/test-EOVERFLOW | Bin 6680 -> 0 bytes src/.gitignore | 1 + 5 files changed, 5 insertions(+), 0 deletions(-) delete mode 100755 gnulib/tests/test-EOVERFLOW diff --git a/examples/domain-events/events-c/.gitignore b/examples/domain-events/events-c/.gitignore index 7b8079b..ed9ac4d 100644 --- a/examples/domain-events/events-c/.gitignore +++ b/examples/domain-events/events-c/.gitignore @@ -1,3 +1,5 @@ +Makefile +Makefile.in *.exe .deps .libs diff --git a/gnulib/tests/.cvsignore b/gnulib/tests/.cvsignore index 6ac20bc..04eb4a0 100644 --- a/gnulib/tests/.cvsignore +++ b/gnulib/tests/.cvsignore @@ -3,6 +3,7 @@ Makefile Makefile.in sys +test-EOVERFLOW test-alloca-opt test-arpa_inet test-c-ctype diff --git a/gnulib/tests/.gitignore b/gnulib/tests/.gitignore index 6ac20bc..04eb4a0 100644 --- a/gnulib/tests/.gitignore +++ b/gnulib/tests/.gitignore @@ -3,6 +3,7 @@ Makefile Makefile.in sys +test-EOVERFLOW test-alloca-opt test-arpa_inet test-c-ctype diff --git a/gnulib/tests/test-EOVERFLOW b/gnulib/tests/test-EOVERFLOW deleted file mode 100755 index ea3a8a27290fea9dd71beab276a9180b71a4847b..0000000000000000000000000000000000000000 GIT binary patch literal 0 HcmV?d00001 literal 6680 zcmcIoU2Ggz6+Zjp*eP!8n3%YUQAX-Ua*#aT7{_sdz}SxM*|3d`<3ON=*^PH>Z?)dF zcE(`?MYV7XTDfkGctPT!72*K|4|zdU0*OQ>MBym~ghCNeg+!SCpj3cr9@^!cJLlLl ... zau)bXmrZ?BcK<8j@D;CN`0sM9)EIV@%NMN$$IfAtfv$5-$(qd;a|K&wUuHtx8&go~ Q6T0rYn6Ht&D`jQ!Z=>`m&;S4c diff --git a/src/.gitignore b/src/.gitignore index 4aff461..cef59e0 100644 --- a/src/.gitignore +++ b/src/.gitignore @@ -15,3 +15,4 @@ libvirt_parthelper libvirt_lxc virsh-net-edit.c virsh-pool-edit.c +libvirt_sym.version -- 1.6.0.4.1021.g4320

2 1

[libvirt] --with-xen=foo
by Ben Guthro 21 Nov '08

21 Nov '08

So, when I merged my changes up yesterday to some of the developments going on, it would seem that one of the changesets I pulled in had a bit of a reduction in configure flexibility. I'm not yet sure what changeset caused it...but I used to be able to specify an external xen tree with the following line in a Makefile... ./autogen.sh --prefix=/usr --with-xen=$(XEN_SRC)/dist/install/usr This now seems to fail to find that xen tree, as I error out with: gcc -DHAVE_CONFIG_H -I. -I. -I.. -I../gnulib/lib -I../gnulib/lib -I../include -I../include -I../qemud -I/usr/include/libxml2 -DBINDIR=\"/usr/libexec\" -DSBINDIR=\"/usr/sbin\" -DSYSCONF_DIR=\"/etc\" -DLOCALEBASEDIR=\"/usr/share/locale\" -DLOCAL_STATE_DIR=\"/var\" -DGETTEXT_PACKAGE=\"libvirt\" -Wall -Wformat -Wformat-security -Wmissing-prototypes -Wnested-externs -Wpointer-arith -Wextra -Wshadow -Wcast-align -Wwrite-strings -Waggregate-return -Wstrict-prototypes -Winline -Wredundant-decls -Wno-sign-compare -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fasynchronous-unwind-tables -g -O2 -MT stats_linux.lo -MD -MP -MF .deps/stats_linux.Tpo -c stats_linux.c -fPIC -DPIC -o .libs/stats_linux.o stats_linux.c:24:16: error: xs.h: No such file or directory stats_linux.c: In function 'check_bd_connected': stats_linux.c:154: warning: implicit declaration of function 'xs_read' stats_linux.c:154: warning: nested extern declaration of 'xs_read' stats_linux.c:154: warning: assignment makes pointer from integer without a cast I assume this has something to do with Daniel B's changes to configure scripts recently. Could you give me a pointer as to where I should be looking? Ben

2 7