f2fs: don't set RO when shutting down f2fs
Jaegeuk Kim
maintainerPATCH 1/1 f2fs: don't set RO when shutting down f2fs
Shutdown does not check the error of thaw_super due to readonly, which causes a deadlock like below.
f2fs_ioc_shutdown(F2FS_GOING_DOWN_FULLSYNC) issue_discard_thread
- bdev_freeze
- freeze_super
- f2fs_stop_checkpoint()
- f2fs_handle_critical_error - sb_start_write
- set RO - waiting
- bdev_thaw
- thaw_super_locked
- return -EINVAL, if sb_rdonly()
- f2fs_stop_discard_thread
-> wait for kthread_stop(discard_thread);
--- fs/f2fs/super.c | 11 +++++++++-- 1 file changed, 9 insertions(+), 2 deletions(-)
fs/f2fs/super.c +9 -2
@@ -4135,9 +4135,16 @@ void f2fs_handle_critical_error(struct f2fs_sb_info *sbi, unsigned char reason,if (shutdown)set_sbi_flag (sbi,SBI_IS_SHUTDOWN ); -/* continue filesystem operators if errors=continue */ -if (continue_fs ||f2fs_readonly (sb)) +/* +* Continue filesystem operators if errors=continue. Should not set +* RO by shutdown, since RO bypasses thaw_super which can hang the +* system. +*/ +if (continue_fs ||f2fs_readonly (sb) || + reason ==STOP_CP_REASON_SHUTDOWN ) { +f2fs_warn (sbi,"Stopped filesystem due to readon: %d" , reason);return ; + }f2fs_warn (sbi,"Remounting filesystem read-only" );/*
Signature · 1 line
2.44.0.478.gd926399ef9-goog
Chao Yu
maintainerQuoted 30 lines · Jaegeuk Kim
Shutdown does not check the error of
thaw_superdue to readonly, which causes a deadlock like below.
f2fs_ioc_shutdown(F2FS_GOING_DOWN_FULLSYNC)issue_discard_thread
-bdev_freeze
-freeze_super
-f2fs_stop_checkpoint()
-f2fs_handle_critical_error-sb_start_write
- set RO - waiting
-bdev_thaw
-thaw_super_locked
- return-EINVAL, ifsb_rdonly()
-f2fs_stop_discard_thread
-> wait forkthread_stop(discard_thread);Reported-by "Light Hsieh (謝明燈)" <[email protected]>Signed-off-by Jaegeuk Kim <[email protected]>--- fs/f2fs/super.c | 11 +++++++++-- 1 file changed, 9 insertions(+), 2 deletions(-)fs/f2fs/super.c +0
-1@@ -4135,9 +4135,16 @@ void f2fs_handle_critical_error(struct f2fs_sb_info *sbi, unsigned char reason,if (shutdown)set_sbi_flag (sbi,SBI_IS_SHUTDOWN ); -/* continue filesystem operators if errors=continue */
fs/f2fs/super.c void f2fs_handle_critical_error(struct f2fs_sb_info *sbi, unsigned char reason, · line 4146
Quoted 10 lines
fs/f2fs/super.c +8
-1-if (continue_fs ||f2fs_readonly (sb)) +/* +* Continue filesystem operators if errors=continue. Should not set +* RO by shutdown, since RO bypasses thaw_super which can hang the +* system. +*/ +if (continue_fs ||f2fs_readonly (sb) || + reason ==STOP_CP_REASON_SHUTDOWN ) { +f2fs_warn (sbi,"Stopped filesystem due to readon: %d" , reason);return ;
Do we need to set RO after bdev_thaw() in f2fs_do_shutdown()?
Thanks,
Quoted 4 lines
fs/f2fs/super.c +1
-0+ }f2fs_warn (sbi,"Remounting filesystem read-only" );/*
Jaegeuk Kim
maintainerQuoted 34 lines · Chao Yu
Quoted 33 lines · Jaegeuk Kim
Shutdown does not check the error of
thaw_superdue to readonly, which causes a deadlock like below.
f2fs_ioc_shutdown(F2FS_GOING_DOWN_FULLSYNC)issue_discard_thread
-bdev_freeze
-freeze_super
-f2fs_stop_checkpoint()
-f2fs_handle_critical_error-sb_start_write
- set RO - waiting
-bdev_thaw
-thaw_super_locked
- return-EINVAL, ifsb_rdonly()
-f2fs_stop_discard_thread
-> wait forkthread_stop(discard_thread);Reported-by "Light Hsieh (謝明燈)" <[email protected]>Signed-off-by Jaegeuk Kim <[email protected]>--- fs/f2fs/super.c | 11 +++++++++-- 1 file changed, 9 insertions(+), 2 deletions(-)fs/f2fs/super.c +3
-2@@ -4135,9 +4135,16 @@ void f2fs_handle_critical_error(struct f2fs_sb_info *sbi, unsigned char reason,if (shutdown)set_sbi_flag (sbi,SBI_IS_SHUTDOWN ); -/* continue filesystem operators if errors=continue */ -if (continue_fs ||f2fs_readonly (sb)) +/* +* Continue filesystem operators if errors=continue. Should not set +* RO by shutdown, since RO bypasses thaw_super which can hang the
Quoted 8 lines
fs/f2fs/super.c void f2fs_handle_critical_error(struct f2fs_sb_info *sbi, unsigned char reason, · line 4146
Quoted 6 lines
fs/f2fs/super.c +5
-0+* system. +*/ +if (continue_fs ||f2fs_readonly (sb) || + reason ==STOP_CP_REASON_SHUTDOWN ) { +f2fs_warn (sbi,"Stopped filesystem due to readon: %d" , reason);return ;Do we need to set RO after
bdev_thaw()inf2fs_do_shutdown()?
IIRC, shutdown doesn't need to set RO as we stopped the checkpoint. I'm more concerned on any side effect caused by this RO change.
Quoted 6 lines
Thanks,
Quoted 3 lines
fs/f2fs/super.c +1
-0+ }f2fs_warn (sbi,"Remounting filesystem read-only" );/*
Chao Yu
maintainerQuoted 44 lines · Jaegeuk Kim
Quoted 34 lines · Chao Yu
Quoted 33 lines · Jaegeuk Kim
Shutdown does not check the error of
thaw_superdue to readonly, which causes a deadlock like below.
f2fs_ioc_shutdown(F2FS_GOING_DOWN_FULLSYNC)issue_discard_thread
-bdev_freeze
-freeze_super
-f2fs_stop_checkpoint()
-f2fs_handle_critical_error-sb_start_write
- set RO - waiting
-bdev_thaw
-thaw_super_locked
- return-EINVAL, ifsb_rdonly()
-f2fs_stop_discard_thread-> wait for kthread_stop(discard_thread);Reported-by "Light Hsieh (謝明燈)" <[email protected]>Signed-off-by Jaegeuk Kim <[email protected]>--- fs/f2fs/super.c | 11 +++++++++-- 1 file changed, 9 insertions(+), 2 deletions(-)fs/f2fs/super.c +3
-2@@ -4135,9 +4135,16 @@ void f2fs_handle_critical_error(struct f2fs_sb_info *sbi, unsigned char reason,if (shutdown)set_sbi_flag (sbi,SBI_IS_SHUTDOWN ); -/* continue filesystem operators if errors=continue */ -if (continue_fs ||f2fs_readonly (sb)) +/* +* Continue filesystem operators if errors=continue. Should not set +* RO by shutdown, since RO bypasses thaw_super which can hang the Quoted 8 lines
fs/f2fs/super.c void f2fs_handle_critical_error(struct f2fs_sb_info *sbi, unsigned char reason, · line 4146
Quoted 6 lines
fs/f2fs/super.c +5
-0+* system. +*/ +if (continue_fs ||f2fs_readonly (sb) || + reason ==STOP_CP_REASON_SHUTDOWN ) { +f2fs_warn (sbi,"Stopped filesystem due to readon: %d" , reason);return ;Do we need to set RO after
bdev_thaw()inf2fs_do_shutdown()?
Quoted 2 lines
IIRC, shutdown doesn't need to set RO as we stopped the checkpoint. I'm more concerned on any side effect caused by this RO change.
Okay, I just wonder whether we need to follow semantics of errors=remount-ro semantics, but it looks fine since shutdown operation simulated by ioctl could not be treated as an inner critical error,
errors=%s Specify f2fs behavior on critical errors. This supports modes:
"panic", "continue" and "remount-ro", respectively, trigger panic immediately, continue without doing anything, and remount the partition in read-only mode. By default it uses "continue" mode.
Also, it keeps the behavior consistent w/ what we do for errors=panic case.
if (F2FS_OPTION(sbi).errors == MOUNT_ERRORS_PANIC &&
!shutdown && !system_going_down() &&
^^^^^^^^^
!is_sbi_flag_set(sbi, SBI_IS_SHUTDOWN))
panic("F2FS-fs (device %s): panic forced after error\n",
sb->s_id);Thanks,
Quoted 7 lines
Quoted 7 lines
Thanks,
Quoted 3 lines
fs/f2fs/super.c +1
-0+ }f2fs_warn (sbi,"Remounting filesystem read-only" );/*
Quoted 29 lines · Jaegeuk Kim
Shutdown does not check the error of
thaw_superdue to readonly, which causes a deadlock like below.
f2fs_ioc_shutdown(F2FS_GOING_DOWN_FULLSYNC)issue_discard_thread
-bdev_freeze
-freeze_super
-f2fs_stop_checkpoint()
-f2fs_handle_critical_error-sb_start_write
- set RO - waiting
-bdev_thaw
-thaw_super_locked
- return-EINVAL, ifsb_rdonly()
-f2fs_stop_discard_thread
-> wait forkthread_stop(discard_thread);Reported-by "Light Hsieh (謝明燈)" <[email protected]>Signed-off-by Jaegeuk Kim <[email protected]>--- fs/f2fs/super.c | 11 +++++++++-- 1 file changed, 9 insertions(+), 2 deletions(-)fs/f2fs/super.c +0
-0@@ -4135,9 +4135,16 @@ void f2fs_handle_critical_error(struct f2fs_sb_info *sbi, unsigned char reason,if (shutdown)set_sbi_flag (sbi,SBI_IS_SHUTDOWN );
fs/f2fs/super.c void f2fs_handle_critical_error(struct f2fs_sb_info *sbi, unsigned char reason, · line 4144
Quoted 10 lines
fs/f2fs/super.c +7
-2-/* continue filesystem operators if errors=continue */ -if (continue_fs ||f2fs_readonly (sb)) +/* +* Continue filesystem operators if errors=continue. Should not set +* RO by shutdown, since RO bypasses thaw_super which can hang the +* system. +*/ +if (continue_fs ||f2fs_readonly (sb) || + reason ==STOP_CP_REASON_SHUTDOWN ) {
I think we can use "shutdown" variable instead of "reason == STOP_CP_REASON_SHUTDOWN" to be concise.
fs/f2fs/super.c void f2fs_handle_critical_error(struct f2fs_sb_info *sbi, unsigned char reason, · line 4145
Quoted 1 line
fs/f2fs/super.c +1
-0+f2fs_warn (sbi,"Stopped filesystem due to readon: %d" , reason);
readon -> reason?
Quoted 14 lines
fs/f2fs/super.c +1
-1return ; + }f2fs_warn (sbi,"Remounting filesystem read-only" );/* -- 2.44.0.478.gd926399ef9-goog
_______________________________________________
Linux-f2fs-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/linux-f2fs-devel
Jaegeuk Kim
maintainerShutdown does not check the error of thaw_super due to readonly, which causes a deadlock like below.
f2fs_ioc_shutdown(F2FS_GOING_DOWN_FULLSYNC) issue_discard_thread
- bdev_freeze
- freeze_super
- f2fs_stop_checkpoint()
- f2fs_handle_critical_error - sb_start_write
- set RO - waiting
- bdev_thaw
- thaw_super_locked
- return -EINVAL, if sb_rdonly()
- f2fs_stop_discard_thread
-> wait for kthread_stop(discard_thread);
---
Change log from v1:
- use better variable
- fix typo
fs/f2fs/super.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-)
fs/f2fs/super.c +8 -2
@@ -4159,9 +4159,15 @@ void f2fs_handle_critical_error(struct f2fs_sb_info *sbi, unsigned char reason,if (shutdown)set_sbi_flag (sbi,SBI_IS_SHUTDOWN ); -/* continue filesystem operators if errors=continue */ -if (continue_fs ||f2fs_readonly (sb)) +/* +* Continue filesystem operators if errors=continue. Should not set +* RO by shutdown, since RO bypasses thaw_super which can hang the +* system. +*/ +if (continue_fs ||f2fs_readonly (sb) || shutdown) { +f2fs_warn (sbi,"Stopped filesystem due to reason: %d" , reason);return ; + }f2fs_warn (sbi,"Remounting filesystem read-only" );/*
Signature · 1 line
2.44.0.478.gd926399ef9-goog
Daeho Jeong
Reviewed-byQuoted 50 lines · Jaegeuk Kim
Shutdown does not check the error of
thaw_superdue to readonly, which causes a deadlock like below.
f2fs_ioc_shutdown(F2FS_GOING_DOWN_FULLSYNC)issue_discard_thread
-bdev_freeze
-freeze_super
-f2fs_stop_checkpoint()
-f2fs_handle_critical_error-sb_start_write
- set RO - waiting
-bdev_thaw
-thaw_super_locked
- return-EINVAL, ifsb_rdonly()
-f2fs_stop_discard_thread
-> wait forkthread_stop(discard_thread);Reported-by "Light Hsieh (謝明燈)" <[email protected]>Signed-off-by Jaegeuk Kim <[email protected]>---
Change log from v1:
- use better variable
- fix typofs/f2fs/super.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-)fs/f2fs/super.c +8
-3@@ -4159,9 +4159,15 @@ void f2fs_handle_critical_error(struct f2fs_sb_info *sbi, unsigned char reason,if (shutdown)set_sbi_flag (sbi,SBI_IS_SHUTDOWN ); -/* continue filesystem operators if errors=continue */ -if (continue_fs ||f2fs_readonly (sb)) +/* +* Continue filesystem operators if errors=continue. Should not set +* RO by shutdown, since RO bypasses thaw_super which can hang the +* system. +*/ +if (continue_fs ||f2fs_readonly (sb) || shutdown) { +f2fs_warn (sbi,"Stopped filesystem due to reason: %d" , reason);return ; + }f2fs_warn (sbi,"Remounting filesystem read-only" );/* --
Quoted 3 lines
2.44.0.478.gd926399ef9-goog
Quoted 5 lines
_______________________________________________
Linux-f2fs-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/linux-f2fs-devel
Quoted 15 lines · Jaegeuk Kim
Shutdown does not check the error of
thaw_superdue to readonly, which causes a deadlock like below.
f2fs_ioc_shutdown(F2FS_GOING_DOWN_FULLSYNC)issue_discard_thread
-bdev_freeze
-freeze_super
-f2fs_stop_checkpoint()
-f2fs_handle_critical_error-sb_start_write
- set RO - waiting
-bdev_thaw
-thaw_super_locked
- return-EINVAL, ifsb_rdonly()
-f2fs_stop_discard_thread
-> wait forkthread_stop(discard_thread);
Quoted 2 lines
Reported-by "Light Hsieh (謝明燈)" <[email protected]>Signed-off-by Jaegeuk Kim <[email protected]>
Thanks,