btrfs-progs: Continue checking even we found something wrong in free space cache
PATCH 1/1 btrfs-progs: Continue checking even we found something wrong in free space cache
No need to abort checking, especially for RO check free space cache is meaningless, the errors in fs/extent tree is more interesting for developers.
So continue checking even something in free space cache is wrong.
--- check/main.c | 1 - 1 file changed, 1 deletion(-)
check/main.c +0 -1
@@ -9885,7 +9885,6 @@ int cmd_check(int argc, char **argv)error ("errors found in free space tree" ); else error("errors found in free space cache"); - goto out; }/*
Signature · 1 line
2.18.0
Quoted 17 lines · Qu Wenruo
No need to abort checking, especially for RO check free space cache is meaningless, the errors in fs/extent tree is more interesting for developers.
So continue checking even something in free space cache is wrong.
Reported-by Etienne Champetier <[email protected]>Signed-off-by Qu Wenruo <[email protected]>--- check/main.c | 1 - 1 file changed, 1 deletion(-)check/main.c +0
-0
check/main.c int cmd_check(int argc, char **argv) · line 9887
Quoted 3 lines
check/main.c +0
-0@@ -9885,7 +9885,6 @@ int cmd_check(int argc, char **argv)error ("errors found in free space tree" ); else error("errors found in free space cache");
Please make it a warning and update the message so it says that it will continue despite the errors found.
I noticed that error handling in check_space_cache is a bit fishy, some errors are stored to 'ret' but ignored at the end of the function.
Quoted 13 lines · David Sterba
Quoted 12 lines · Qu Wenruo
No need to abort checking, especially for RO check free space cache is meaningless, the errors in fs/extent tree is more interesting for developers.
So continue checking even something in free space cache is wrong.
Reported-by Etienne Champetier <[email protected]>Signed-off-by Qu Wenruo <[email protected]>--- check/main.c | 1 - 1 file changed, 1 deletion(-)
Quoted 11 lines
check/main.c int cmd_check(int argc, char **argv) · line 9887
Quoted 8 lines
check/main.c +0
-0@@ -9885,7 +9885,6 @@ int cmd_check(int argc, char **argv)error ("errors found in free space tree" ); else error("errors found in free space cache");Please make it a warning and update the message so it says that it will continue despite the errors found.
I'm fine to add warning, but isn't it the expected behavior?
In fact I'm quite surprised that when we found something wrong we just abort checking.
It's never the case for file/extent tree check. We always report all errors we found, not only the first error and exit.
Thanks,
Qu
Quoted 4 lines
I noticed that error handling in
check_space_cacheis a bit fishy, some errors are stored to 'ret' but ignored at the end of the function.
Quoted 28 lines · Qu Wenruo
Quoted 13 lines · David Sterba
Quoted 12 lines · Qu Wenruo
No need to abort checking, especially for RO check free space cache is meaningless, the errors in fs/extent tree is more interesting for developers.
So continue checking even something in free space cache is wrong.
Reported-by Etienne Champetier <[email protected]>Signed-off-by Qu Wenruo <[email protected]>--- check/main.c | 1 - 1 file changed, 1 deletion(-)Quoted 11 lines
check/main.c int cmd_check(int argc, char **argv) · line 9887
Quoted 8 lines
check/main.c +0
-0@@ -9885,7 +9885,6 @@ int cmd_check(int argc, char **argv)error ("errors found in free space tree" ); else error("errors found in free space cache");Please make it a warning and update the message so it says that it will continue despite the errors found.
Quoted 7 lines
I'm fine to add warning, but isn't it the expected behavior?
In fact I'm quite surprised that when we found something wrong we just abort checking.
It's never the case for file/extent tree check. We always report all errors we found, not only the first error and exit.
You're right, in context of 'check' this is the expected behaviour. So let's keep error().