From 53b691e4d85f8a442f14ecf4b3bf0b17d607fb2b Mon Sep 17 00:00:00 2001 Message-ID: <53b691e4d85f8a442f14ecf4b3bf0b17d607fb2b.1720800605.git.jdenemar@redhat.com> From: Jiri Denemark Date: Thu, 11 Jul 2024 13:49:09 +0200 Subject: [PATCH] qemu: Don't leave beingDestroyed=true on inactive domain Recent commit v10.4.0-87-gd9935a5c4f made a reasonable change to only reset beingDestroyed back to false when vm->def->id is reset to make sure other code can detect a domain is (about to become) inactive. It even added a comment saying any caller of qemuProcessBeginStopJob is supposed to call qemuProcessStop to clear beingDestroyed. But not every caller really does so because they first call qemuProcessBeginStopJob and then check whether a domain is still running. If not the qemuProcessStop call is skipped leaving beingDestroyed=true. In case of a persistent domain this may block incoming migrations of such domain as the migration code would think the domain died unexpectedly (even though it's still running). The qemuProcessBeginStopJob function is a wrapper around virDomainObjBeginJob, but virDomainObjEndJob was used directly for cleanup. This patch introduces a new qemuProcessEndStopJob wrapper around virDomainObjEndJob to properly undo everything qemuProcessBeginStopJob did. https://issues.redhat.com/browse/RHEL-43309 Signed-off-by: Jiri Denemark Reviewed-by: Peter Krempa (cherry picked from commit bec903cae84c21850d47a1b4d3ab57ca81189519) Signed-off-by: Jiri Denemark --- src/qemu/qemu_driver.c | 4 ++-- src/qemu/qemu_process.c | 20 ++++++++++++++++---- src/qemu/qemu_process.h | 1 + 3 files changed, 19 insertions(+), 6 deletions(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index fc1704f4fc..d9073b2154 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -2115,7 +2115,7 @@ qemuDomainDestroyFlags(virDomainPtr dom, endjob: if (ret == 0) qemuDomainRemoveInactive(driver, vm, 0, false); - virDomainObjEndJob(vm); + qemuProcessEndStopJob(vm); cleanup: virDomainObjEndAPI(&vm); @@ -3901,7 +3901,7 @@ processMonitorEOFEvent(virQEMUDriver *driver, endjob: qemuDomainRemoveInactive(driver, vm, 0, false); - virDomainObjEndJob(vm); + qemuProcessEndStopJob(vm); } diff --git a/src/qemu/qemu_process.c b/src/qemu/qemu_process.c index b9b6ccf1de..bea42d38c6 100644 --- a/src/qemu/qemu_process.c +++ b/src/qemu/qemu_process.c @@ -8422,7 +8422,8 @@ qemuProcessKill(virDomainObj *vm, unsigned int flags) * qemuProcessBeginStopJob: * * Stop all current jobs by killing the domain and start a new one for - * qemuProcessStop. + * qemuProcessStop. The caller has to make sure qemuProcessEndStopJob is + * called to properly cleanup the job. */ int qemuProcessBeginStopJob(virDomainObj *vm, @@ -8449,8 +8450,9 @@ qemuProcessBeginStopJob(virDomainObj *vm, goto error; /* priv->beingDestroyed is deliberately left set to 'true' here. Caller - * is supposed to call qemuProcessStop, which will reset it after - * 'vm->def->id' is set to -1 */ + * is supposed to call qemuProcessStop (which will reset it after + * 'vm->def->id' is set to -1) and/or qemuProcessEndStopJob to do proper + * cleanup. */ return 0; error: @@ -8459,6 +8461,16 @@ qemuProcessBeginStopJob(virDomainObj *vm, } +void +qemuProcessEndStopJob(virDomainObj *vm) +{ + if (!virDomainObjIsActive(vm)) + QEMU_DOMAIN_PRIVATE(vm)->beingDestroyed = false; + + virDomainObjEndJob(vm); +} + + void qemuProcessStop(virQEMUDriver *driver, virDomainObj *vm, virDomainShutoffReason reason, @@ -8801,7 +8813,7 @@ qemuProcessAutoDestroy(virDomainObj *dom, qemuDomainRemoveInactive(driver, dom, 0, false); - virDomainObjEndJob(dom); + qemuProcessEndStopJob(dom); virObjectEventStateQueue(driver->domainEventState, event); } diff --git a/src/qemu/qemu_process.h b/src/qemu/qemu_process.h index c1ea949215..cb67bfcd2d 100644 --- a/src/qemu/qemu_process.h +++ b/src/qemu/qemu_process.h @@ -169,6 +169,7 @@ typedef enum { int qemuProcessBeginStopJob(virDomainObj *vm, virDomainJob job, bool forceKill); +void qemuProcessEndStopJob(virDomainObj *vm); void qemuProcessStop(virQEMUDriver *driver, virDomainObj *vm, virDomainShutoffReason reason, -- 2.45.2