multifd: Unconditionally unregister yank function
PATCH 1/1 multifd: Unconditionally unregister yank function
Unconditionally unregister yank function in multifd_load_cleanup(). If it is not unregistered here, it will leak and cause a crash in yank_unregister_instance(). Now if the ioc is still in use afterwards, it will only lead to qemu not being able to recover from a hang related to that ioc.
After checking the code, i am pretty sure that ref is always 1 when arriving here. So all this currently does is remove the unneeded check.
---
This is similar to Peter Xu's 39675ffffb3394d44b880d083a214c5e44786170 "migration: Move the yank unregister of channel_close out" in that it removes the "OBJECT(p->c)->ref == 1" hack. So it makes sense for 6.1 so these patches are together.
migration/multifd.c | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-)
migration/multifd.c +1 -4
@@ -987,10 +987,7 @@ int multifd_load_cleanup(Error **errp)for (i =0 ; i <migrate_multifd_channels (); i++) {MultiFDRecvParams *p = &multifd_recv_state->params[i]; -if (OBJECT (p->c)->ref ==1 ) { -migration_ioc_unregister_yank (p->c); - } - +migration_ioc_unregister_yank (p->c);object_unref (OBJECT (p->c)); p->c =NULL ;qemu_mutex_destroy (&p->mutex);
Signature · 1 line
2.32.0
Quoted 3 lines · Lukas Straub
Unconditionally unregister yank function in
multifd_load_cleanup(). If it is not unregistered here, it will leak and cause a crash inyank_unregister_instance().
Curious whether there was a crash somewhere that you hit?
Quoted 7 lines
Now if the ioc is still in use
afterwards, it will only lead to qemu not being able to recover from a hang related to that ioc.After checking the code, i am pretty sure that ref is always 1 when arriving here. So all this currently does is remove the unneeded check.
Thanks for checking and removing the 2nd ref==1. I wanted to do that but I didn't dare before I look more closely or test more.
The patch looks good to me, it's just that I am not sure whether it suites 6.1 material as we've just released rc2 today. Maybe 6.2 is more suitable?
Quoted 31 lines
Signed-off-by Lukas Straub <[email protected]>---
This is similar to Peter Xu's
39675ffffb3394d44b880d083a214c5e44786170"migration: Move the yank unregister ofchannel_closeout" in that it removes the "OBJECT(p->c)->ref == 1" hack. So it makes sense for 6.1 so these patches are together.migration/multifd.c | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-)migration/multifd.c +1
-4@@ -987,10 +987,7 @@ int multifd_load_cleanup(Error **errp)for (i =0 ; i <migrate_multifd_channels (); i++) {MultiFDRecvParams *p = &multifd_recv_state->params[i]; -if (OBJECT (p->c)->ref ==1 ) { -migration_ioc_unregister_yank (p->c); - } - +migration_ioc_unregister_yank (p->c);object_unref (OBJECT (p->c)); p->c =NULL ;qemu_mutex_destroy (&p->mutex);Signature · 1 line
2.32.0
Signature · 1 line
Peter Xu
Quoted 6 lines · On Wed, 4 Aug 2021 15:39:55 -0400 Peter Xu <[email protected]> wrote:
Quoted 3 lines · Lukas Straub
Unconditionally unregister yank function in
multifd_load_cleanup(). If it is not unregistered here, it will leak and cause a crash inyank_unregister_instance().Curious whether there was a crash somewhere that you hit?
No, I just noticed this while working on a different patch.
Quoted 8 lines
Quoted 7 lines
Now if the ioc is still in use
afterwards, it will only lead to qemu not being able to recover from a hang related to that ioc.After checking the code, i am pretty sure that ref is always 1 when arriving here. So all this currently does is remove the unneeded check.
Quoted 5 lines
Thanks for checking and removing the 2nd ref==1. I wanted to do that but I didn't dare before I look more closely or test more.
The patch looks good to me, it's just that I am not sure whether it suites 6.1 material as we've just released rc2 today. Maybe 6.2 is more suitable?
Yeah, if my assessment of the code is correct it's more of a cleanup.
Quoted 34 lines
Quoted 31 lines
Signed-off-by Lukas Straub <[email protected]>---
This is similar to Peter Xu's
39675ffffb3394d44b880d083a214c5e44786170"migration: Move the yank unregister ofchannel_closeout" in that it removes the "OBJECT(p->c)->ref == 1" hack. So it makes sense for 6.1 so these patches are together.migration/multifd.c | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-)migration/multifd.c +1
-4@@ -987,10 +987,7 @@ int multifd_load_cleanup(Error **errp)for (i =0 ; i <migrate_multifd_channels (); i++) {MultiFDRecvParams *p = &multifd_recv_state->params[i]; -if (OBJECT (p->c)->ref ==1 ) { -migration_ioc_unregister_yank (p->c); - } - +migration_ioc_unregister_yank (p->c);object_unref (OBJECT (p->c)); p->c =NULL ;qemu_mutex_destroy (&p->mutex);Signature · 1 line
2.32.0
Quoted 28 lines · On Wed, 4 Aug 2021 21:26:32 +0200 Lukas Straub <[email protected]> wrote:
Unconditionally unregister yank function in
multifd_load_cleanup(). If it is not unregistered here, it will leak and cause a crash inyank_unregister_instance(). Now if the ioc is still in use afterwards, it will only lead to qemu not being able to recover from a hang related to that ioc.After checking the code, i am pretty sure that ref is always 1 when arriving here. So all this currently does is remove the unneeded check.
Signed-off-by Lukas Straub <[email protected]>---
This is similar to Peter Xu's
39675ffffb3394d44b880d083a214c5e44786170"migration: Move the yank unregister ofchannel_closeout" in that it removes the "OBJECT(p->c)->ref == 1" hack. So it makes sense for 6.1 so these patches are together.migration/multifd.c | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-)migration/multifd.c +0
-0@@ -987,10 +987,7 @@ int multifd_load_cleanup(Error **errp)for (i =0 ; i <migrate_multifd_channels (); i++) {
migration/multifd.c int multifd_load_cleanup(Error **errp) · line 993
Quoted 10 lines
migration/multifd.c +1
-4MultiFDRecvParams *p = &multifd_recv_state->params[i]; -if (OBJECT (p->c)->ref ==1 ) { -migration_ioc_unregister_yank (p->c); - } - +migration_ioc_unregister_yank (p->c);object_unref (OBJECT (p->c)); p->c =NULL ;qemu_mutex_destroy (&p->mutex);
ping...
Quoted 11 lines · Lukas Straub
Unconditionally unregister yank function in
multifd_load_cleanup(). If it is not unregistered here, it will leak and cause a crash inyank_unregister_instance(). Now if the ioc is still in use afterwards, it will only lead to qemu not being able to recover from a hang related to that ioc.After checking the code, i am pretty sure that ref is always 1 when arriving here. So all this currently does is remove the unneeded check.
Signed-off-by Lukas Straub <[email protected]>