- Devel - libvirt List Archives

[libvirt] [PATCH 0/3] Allow to override disk geometry.
by Viktor Mihajlovski 17 Aug '12

17 Aug '12

With qemu it is possible to override the geometry (cylinders, heads, sectors) of disks. This series adds a new XML tag for geometry and the related support in the qemu driver. J.B. Joret (3): qemu: Support for Disk Geometry Override qemu: Documentation for Disk Geometry Override qemu: Testcase for Disk Geometry Override docs/formatdomain.html.in | 25 +++++++ docs/schemas/domaincommon.rng | 28 ++++++++ src/conf/domain_conf.c | 71 ++++++++++++++++++++ src/conf/domain_conf.h | 17 +++++ src/libvirt_private.syms | 2 + src/qemu/qemu_command.c | 59 ++++++++++++++++ .../qemuxml2argv-disk-geometry.args | 4 + .../qemuxml2argv-disk-geometry.xml | 26 +++++++ tests/qemuxml2argvtest.c | 2 + 9 files changed, 234 insertions(+), 0 deletions(-) create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-disk-geometry.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-disk-geometry.xml

2 13

[libvirt] [PATCH 0/8] Honour current process label when generating SELinux labels
by Daniel P. Berrange 17 Aug '12

17 Aug '12

This patch series makes a number of changes to the SELinux label generation code. This is intended to make it fully honour the current process label when generating VM labels, so that dynamic label generation works better with custom policies, or confined user accounts.

4 27

[libvirt] QEMU 1.2 Test Day - August 16 2012
by Stefan Hajnoczi 17 Aug '12

17 Aug '12

I have set up the QEMU 1.2 Testing wiki page and suggest August 16 as the Test Day: http://wiki.qemu.org/Planning/1.2/Testing Test Day is an event for QEMU contributors and users to try out the release candidate. QEMU has a large feature matrix that is hard to test by a single entity. On Test Day everyone is encouraged to test their favorite features, host OSes, and guest OSes to make sure that the release candidate is stable. Please add yourself to the http://wiki.qemu.org/Planning/1.2/Testing wiki page! The Test Day is August 16 2012, one day after the planned -rc0 release. On the day, use #qemu IRC on irc.oftc.net to chat about bugs and update the wiki page with your pass/fail results. There are usually -rc1, -rc2, ... follow-up release candidates. Please retest those if you have time, especially if bugs you found are supposed to be resolved. Stefan

1 3

[libvirt] [PATCH] build: fix syntax check during 'make distcheck'
by Eric Blake 17 Aug '12

17 Aug '12

'make distcheck' was failing because a syntax check file, .sc-start-sc_vulnerable_makefile_CVE-2012-3386, got left behind. I traced it to the 'distdir' rule depending on a shortcut syntax-check name rather than the full rule name normally used during 'local-check' from maint.mk. * cfg.mk (distdir): Depend on full rule, not shorthand name. --- Pushing under the build-breaker rule. cfg.mk | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/cfg.mk b/cfg.mk index 1318593..e9138a8 100644 --- a/cfg.mk +++ b/cfg.mk @@ -85,7 +85,7 @@ local-checks-to-skip = \ ifeq ($(filter dist%, $(MAKECMDGOALS)), ) local-checks-to-skip += sc_vulnerable_makefile_CVE-2012-3386 else -distdir: sc_vulnerable_makefile_CVE-2012-3386 +distdir: sc_vulnerable_makefile_CVE-2012-3386.z endif # Files that should never cause syntax check failures. -- 1.7.11.2

1 0

[libvirt] distcheck issues on maint and master
by Cole Robinson 17 Aug '12

17 Aug '12

I just pushed a bunch of patches to -maint branches, but both are giving me distcheck errors that seem related to a gnulib update: ERROR: files left in build directory after distclean: ./.sc-start-sc_vulnerable_makefile_CVE-2012-3386 Any hints? I tried to reproduce on master, but it has its own set of issues: IOError: [Errno 13] Permission denied: '../../src/hyperv/hyperv_wmi.generated.h' types_typedef = open_and_print(os.path.join(output_dirname, "esx_vi_types.generated.typedef")) File "../../src/esx/esx_vi_generator.py", line 1492, in open_and_print return open(filename, "wb") IOError: [Errno 13] Permission denied: '../../src/esx/esx_vi_types.generated.typedef' make[3]: *** [.hyperv_wmi_generator.stamp] Error 1 make[3]: *** Waiting for unfinished jobs.... make[3]: *** [.esx_vi_generator.stamp] Error 1 I can get past that with: $ git revert 1bfb47dfe61c3cf9a716db072cbe22f26e980081 [master f5a9a90] Revert "Make ESX & Hyper-V code generator safe with parallel builds" Then I get: GEN check-symfile Can't open perl script "../../src/check-symfile.pl": No such file or directory make[4]: *** [check-symfile] Error 2 Which I fixed with: $ git diff diff --git a/src/Makefile.am b/src/Makefile.am index 5ae84b1..a5223f2 100644 --- a/src/Makefile.am +++ b/src/Makefile.am @@ -331,7 +331,7 @@ else !WITH_REMOTE # re-generated when configured --without-remote. check-protocol: endif -EXTRA_DIST += $(PROTOCOL_STRUCTS) +EXTRA_DIST += $(PROTOCOL_STRUCTS) check-symfile.pl check-local: check-protocol check-symfile .PHONY: check-protocol $(PROTOCOL_STRUCTS:structs=struct) But since this is all black magic to me I have no idea if that's correct. Thanks, Cole

3 6

[libvirt] [PATCH] build: ship stamp files
by Eric Blake 17 Aug '12

17 Aug '12

'make distcheck' fails because the generated ESX and HyperV files are (intentionally) marked read-only, but since the stamp file was missing, make assumes they need to be rebuilt. Shipping the stamp file solves the problem. * src/Makefile.am (EXTRA_DIST): Ship stamp files. --- Pushing under the build-breaker rule. src/Makefile.am | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/src/Makefile.am b/src/Makefile.am index ac5a27f..213bd83 100644 --- a/src/Makefile.am +++ b/src/Makefile.am @@ -930,6 +930,8 @@ BUILT_SOURCES += $(ESX_DRIVER_GENERATED) ESX_GENERATED_STAMP = .esx_vi_generator.stamp +EXTRA_DIST += $(ESX_GENERATED_STAMP) + $(ESX_DRIVER_GENERATED): $(ESX_GENERATED_STAMP) $(ESX_GENERATED_STAMP): $(srcdir)/esx/esx_vi_generator.input \ @@ -956,6 +958,8 @@ BUILT_SOURCES += $(HYPERV_DRIVER_GENERATED) HYPERV_GENERATED_STAMP = .hyperv_wmi_generator.stamp +EXTRA_DIST += $(HYPERV_GENERATED_STAMP) + $(HYPERV_DRIVER_GENERATED): $(HYPERV_GENERATED_STAMP) $(HYPERV_GENERATED_STAMP): $(srcdir)/hyperv/hyperv_wmi_generator.input \ -- 1.7.11.2

1 0

[libvirt] [PATCH 0/2] qemu: support setting vlan tag for <interface
by Laine Stump 16 Aug '12

16 Aug '12

This would have been just one small patch, but there was a deficiency in the low level functions to save and restore pre-existing state of a virtual function. I have tested these patches on a RHEL6 machine with an SR-IOV adapter, and it properly handled the save/set .. [attach-to-guest] .. [detach] .. restore cycle both when the original state of the VF had a vlan tag and not, and when the guest was requesting a vlan tag, and not. (Now I'm wondering if this would also work for macvtap "passthrough" mode when the adapter used is a VF - it looks like the only thing stopping it is that the code to save / set / restore in that mode uses the device name of the vf, rather than referring to it as <physical device> vf <n>; since only VFs understand setting the vlan tag, the lower level code just assumes it can't be done)

3 5

[libvirt] [PATCH v2] qemu: Set swap_hard_limit before hard_limit
by Osier Yang 16 Aug '12

16 Aug '12

Setting hard_limit larger than previous swap_hard_limit must fail, it's not that good if one wants to change the swap_hard_limit and hard_limit together. E.g. % virsh memtune rhel6 hard_limit : 1000000 soft_limit : 1000000 swap_hard_limit: 1000000 % virsh memtune rhel6 --hard-limit 1000020 --soft-limit 1000020 \ --swap-hard-limit 1000020 --live This patch reoder the limits setting to set the swap_hard_limit first, hard_limit then, and soft_limit last if it's greater than current swap_hard_limit. And soft_limit first, hard_limit then, swap_hard_limit last, if not. --- src/qemu/qemu_driver.c | 59 +++++++++++++++++++++++++++++++++++++++++------ 1 files changed, 51 insertions(+), 8 deletions(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 80cfa84..a26d3cd 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -6651,11 +6651,18 @@ qemuDomainSetMemoryParameters(virDomainPtr dom, virDomainDefPtr persistentDef = NULL; virCgroupPtr group = NULL; virDomainObjPtr vm = NULL; + virTypedParameterPtr hard_limit = NULL; + virTypedParameterPtr soft_limit = NULL; + virTypedParameterPtr swap_hard_limit = NULL; + virTypedParameterPtr sorted_params[nparams]; + unsigned long long val = 0; + int ret = -1; int rc; virCheckFlags(VIR_DOMAIN_AFFECT_LIVE | VIR_DOMAIN_AFFECT_CONFIG, -1); + if (virTypedParameterArrayValidate(params, nparams, VIR_DOMAIN_MEMORY_HARD_LIMIT, VIR_TYPED_PARAM_ULLONG, @@ -6694,13 +6701,49 @@ qemuDomainSetMemoryParameters(virDomainPtr dom, } } + for (i = 0; i < nparams; i++) { + if (STREQ(params[i].field, VIR_DOMAIN_MEMORY_HARD_LIMIT)) + hard_limit = &params[i]; + else if (STREQ(params[i].field, VIR_DOMAIN_MEMORY_SOFT_LIMIT)) + soft_limit = &params[i]; + else if (STREQ(params[i].field, VIR_DOMAIN_MEMORY_SWAP_HARD_LIMIT)) + swap_hard_limit = &params[i]; + } + + /* It will fail if hard limit greater than swap hard limit anyway */ + if (swap_hard_limit && + hard_limit && + (hard_limit->value.ul > swap_hard_limit->value.ul)) { + virReportError(VIR_ERR_INVALID_ARG, "%s", + _("hard limit must be lower than swap hard limit")); + goto cleanup; + } + + /* Get current swap hard limit */ + rc = virCgroupGetMemSwapHardLimit(group, &val); + if (rc != 0) { + virReportSystemError(-rc, "%s", + _("unable to get swap hard limit")); + goto cleanup; + } + + if (swap_hard_limit->value.ul > val) { + sorted_params[0] = swap_hard_limit; + sorted_params[1] = hard_limit; + sorted_params[2] = soft_limit; + } else { + sorted_params[0] = soft_limit; + sorted_params[1] = hard_limit; + sorted_params[2] = swap_hard_limit; + } + ret = 0; for (i = 0; i < nparams; i++) { - virTypedParameterPtr param = &params[i]; + virTypedParameterPtr param = sorted_params[i]; if (STREQ(param->field, VIR_DOMAIN_MEMORY_HARD_LIMIT)) { if (flags & VIR_DOMAIN_AFFECT_LIVE) { - rc = virCgroupSetMemoryHardLimit(group, params[i].value.ul); + rc = virCgroupSetMemoryHardLimit(group, sorted_params[i]->value.ul); if (rc != 0) { virReportSystemError(-rc, "%s", _("unable to set memory hard_limit tunable")); @@ -6709,11 +6752,11 @@ qemuDomainSetMemoryParameters(virDomainPtr dom, } if (flags & VIR_DOMAIN_AFFECT_CONFIG) { - persistentDef->mem.hard_limit = params[i].value.ul; + persistentDef->mem.hard_limit = sorted_params[i]->value.ul; } } else if (STREQ(param->field, VIR_DOMAIN_MEMORY_SOFT_LIMIT)) { if (flags & VIR_DOMAIN_AFFECT_LIVE) { - rc = virCgroupSetMemorySoftLimit(group, params[i].value.ul); + rc = virCgroupSetMemorySoftLimit(group, sorted_params[i]->value.ul); if (rc != 0) { virReportSystemError(-rc, "%s", _("unable to set memory soft_limit tunable")); @@ -6722,11 +6765,11 @@ qemuDomainSetMemoryParameters(virDomainPtr dom, } if (flags & VIR_DOMAIN_AFFECT_CONFIG) { - persistentDef->mem.soft_limit = params[i].value.ul; + persistentDef->mem.soft_limit = sorted_params[i]->value.ul; } } else if (STREQ(param->field, VIR_DOMAIN_MEMORY_SWAP_HARD_LIMIT)) { if (flags & VIR_DOMAIN_AFFECT_LIVE) { - rc = virCgroupSetMemSwapHardLimit(group, params[i].value.ul); + rc = virCgroupSetMemSwapHardLimit(group, sorted_params[i]->value.ul); if (rc != 0) { virReportSystemError(-rc, "%s", _("unable to set swap_hard_limit tunable")); @@ -6734,7 +6777,7 @@ qemuDomainSetMemoryParameters(virDomainPtr dom, } } if (flags & VIR_DOMAIN_AFFECT_CONFIG) { - persistentDef->mem.swap_hard_limit = params[i].value.ul; + persistentDef->mem.swap_hard_limit = sorted_params[i]->value.ul; } } } @@ -7011,7 +7054,7 @@ qemuDomainSetNumaParameters(virDomainPtr dom, continue; } - /* Ensure the cpuset string is formated before passing to cgroup */ + /* Ensure the cpuset string is formatted before passing to cgroup */ if (!(nodeset_str = virDomainCpuSetFormat(nodeset, VIR_DOMAIN_CPUMASK_LEN))) { virReportError(VIR_ERR_INTERNAL_ERROR, "%s", -- 1.7.7.3

2 1

[libvirt] [PATCH] qemu: Ensure the cpuset is formated as expected before passing to cgroup
by Osier Yang 16 Aug '12

16 Aug '12

The parameter value for cpuset could be in special format like "0-10,^7", which is not recongnized by cgroup. This patch is to ensure the cpuset is formated as expected before passing it to cgroup. As a side effect, after the patch, it parses the cpuset early before cgroup setting, to avoid the rollback if cpuset parsing fails afterwards. --- src/qemu/qemu_driver.c | 77 +++++++++++++++++++++++++---------------------- 1 files changed, 41 insertions(+), 36 deletions(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 369e8ed..fc412a1 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -6952,8 +6952,23 @@ qemuDomainSetNumaParameters(virDomainPtr dom, } } else if (STREQ(param->field, VIR_DOMAIN_NUMA_NODESET)) { int rc; - bool savedmask; - char oldnodemask[VIR_DOMAIN_CPUMASK_LEN]; + char *nodeset = NULL; + char *nodeset_str = NULL; + + if (VIR_ALLOC_N(nodeset, VIR_DOMAIN_CPUMASK_LEN) < 0) { + virReportOOMError(); + ret = -1; + goto cleanup; + }; + + if (virDomainCpuSetParse(params[i].value.s, + 0, nodeset, + VIR_DOMAIN_CPUMASK_LEN) < 0) { + virReportError(VIR_ERR_INTERNAL_ERROR, "%s", + _("Failed to parse nodeset")); + ret = -1; + continue; + } if (flags & VIR_DOMAIN_AFFECT_LIVE) { if (vm->def->numatune.memory.mode != @@ -6961,72 +6976,62 @@ qemuDomainSetNumaParameters(virDomainPtr dom, virReportError(VIR_ERR_OPERATION_INVALID, "%s", _("change of nodeset for running domain " "requires strict numa mode")); + VIR_FREE(nodeset); ret = -1; continue; } - rc = virCgroupSetCpusetMems(group, params[i].value.s); - if (rc != 0) { + + /* Ensure the cpuset string is formated before passing to cgroup */ + if (!(nodeset_str = virDomainCpuSetFormat(nodeset, + VIR_DOMAIN_CPUMASK_LEN))) { + virReportError(VIR_ERR_INTERNAL_ERROR, "%s", + _("Failed to format nodeset")); + VIR_FREE(nodeset); + ret = -1; + continue; + } + + if ((rc = virCgroupSetCpusetMems(group, nodeset_str) != 0)) { virReportSystemError(-rc, "%s", _("unable to set numa tunable")); + VIR_FREE(nodeset); + VIR_FREE(nodeset_str); ret = -1; continue; } + VIR_FREE(nodeset_str); /* update vm->def here so that dumpxml can read the new * values from vm->def. */ - savedmask = false; if (!vm->def->numatune.memory.nodemask) { if (VIR_ALLOC_N(vm->def->numatune.memory.nodemask, VIR_DOMAIN_CPUMASK_LEN) < 0) { virReportOOMError(); + VIR_FREE(nodeset); ret = -1; goto cleanup; } } else { - memcpy(oldnodemask, vm->def->numatune.memory.nodemask, - VIR_DOMAIN_CPUMASK_LEN); - savedmask = true; - } - if (virDomainCpuSetParse(params[i].value.s, - 0, - vm->def->numatune.memory.nodemask, - VIR_DOMAIN_CPUMASK_LEN) < 0) { - if (savedmask) - memcpy(vm->def->numatune.memory.nodemask, - oldnodemask, VIR_DOMAIN_CPUMASK_LEN); - else - VIR_FREE(vm->def->numatune.memory.nodemask); - ret = -1; - continue; + VIR_FREE(vm->def->numatune.memory.nodemask); } + + vm->def->numatune.memory.nodemask = nodeset; } if (flags & VIR_DOMAIN_AFFECT_CONFIG) { - savedmask = false; if (!persistentDef->numatune.memory.nodemask) { if (VIR_ALLOC_N(persistentDef->numatune.memory.nodemask, VIR_DOMAIN_CPUMASK_LEN) < 0) { virReportOOMError(); + VIR_FREE(nodeset); ret = -1; goto cleanup; } } else { - memcpy(oldnodemask, persistentDef->numatune.memory.nodemask, - VIR_DOMAIN_CPUMASK_LEN); - savedmask = true; - } - if (virDomainCpuSetParse(params[i].value.s, - 0, - persistentDef->numatune.memory.nodemask, - VIR_DOMAIN_CPUMASK_LEN) < 0) { - if (savedmask) - memcpy(persistentDef->numatune.memory.nodemask, - oldnodemask, VIR_DOMAIN_CPUMASK_LEN); - else - VIR_FREE(persistentDef->numatune.memory.nodemask); - ret = -1; - continue; + VIR_FREE(persistentDef->numatune.memory.nodemask); } + + persistentDef->numatune.memory.nodemask = nodeset; } } } -- 1.7.7.3

2 2

[libvirt] [PATCHv2] build: avoid warnings on older gcc
by Eric Blake 16 Aug '12

16 Aug '12

The previous patch pulled in a newer version of stat-time.h from gnulib, which tries to suppress some stupid gcc warnings (see <http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54113>) by exploiting a feature of newer gcc. But we still aim to compile on RHEL 6.3, with gcc 4.4.6 (not to mention even older devel platforms like RHEL 5), which fails to build thanks to our development -Werror. It took me a while to figure out how to make our set of warnings smaller on older gcc without losing the benefit of the warnings when using newer gcc (such as the one on Fedora 17), but this should do the trick. * m4/virt-compile-warnings.m4 (LIBVIRT_COMPILE_WARNINGS): Avoid warnings that gnulib can't silence on older gcc. --- v2: Use AC_CACHE_CHECK, add -Werror, and test further. Pushing under the build-breaker rule. m4/virt-compile-warnings.m4 | 18 ++++++++++++++++++ 1 file changed, 18 insertions(+) diff --git a/m4/virt-compile-warnings.m4 b/m4/virt-compile-warnings.m4 index 9dee000..26f6134 100644 --- a/m4/virt-compile-warnings.m4 +++ b/m4/virt-compile-warnings.m4 @@ -58,6 +58,24 @@ AC_DEFUN([LIBVIRT_COMPILE_WARNINGS],[ # Gnulib's stat-time.h violates this dontwarn="$dontwarn -Waggregate-return" + # Gnulib uses '#pragma GCC diagnostic push' to silence some + # warnings, but older gcc doesn't support this. + AC_CACHE_CHECK([whether pragma GCC diagnostic push works], + [lv_cv_gcc_pragma_push_works], [ + save_CFLAGS=$CFLAGS + CFLAGS='-Wunknown-pragmas -Werror' + AC_COMPILE_IFELSE([AC_LANG_PROGRAM([[ + #pragma GCC diagnostic push + #pragma GCC diagnostic pop + ]])], + [lv_cv_gcc_pragma_push_works=yes], + [lv_cv_gcc_pragma_push_works=no]) + CFLAGS=$save_CFLAGS]) + if test $lv_cv_gcc_pragma_push_works = no; then + dontwarn="$dontwarn -Wmissing-prototypes" + dontwarn="$dontwarn -Wmissing-declarations" + fi + # We might fundamentally need some of these disabled forever, but # ideally we'd turn many of them on dontwarn="$dontwarn -Wfloat-equal" -- 1.7.11.2

3 2