| Message ID | 61bb187a554ff0397665a8898ca2bd56419b6944.1686640441.git.wqu@suse.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | btrfs: verify data checksum during _check_btrfs_filesystem() | expand |
On 2023/6/13 15:14, Qu Wenruo wrote: > By default btrfs only checks the metadata consistency, but sometimes if > we have some corruption in data while the test case doesn't utilize > scrub to verify, or there is some bugs in scrub itself, we will not > detect those problems. > > So here we do one step further by utilizing --check-data-csum option, so > that if there is some data corruption, we can detect them early. > > Signed-off-by: Qu Wenruo <wqu@suse.com> This is a little overkilled, for example btrfs/215 intentionally corrupt some data for testing, thus this would cause false alerts. I'll try to find a way to append options so that test cases can determine if they want checksum verification case by case. Thanks, Qu > --- > common/btrfs | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/common/btrfs b/common/btrfs > index bd4dc31fa5a8..be8ac04cd9a3 100644 > --- a/common/btrfs > +++ b/common/btrfs > @@ -144,7 +144,7 @@ _check_btrfs_filesystem() > rm -f $tmp.qgroup_report > fi > > - $BTRFS_UTIL_PROG check $device >$tmp.fsck 2>&1 > + $BTRFS_UTIL_PROG check --check-data-csum $device >$tmp.fsck 2>&1 > if [ $? -ne 0 ]; then > _log_err "_check_btrfs_filesystem: filesystem on $device is inconsistent" > echo "*** fsck.$FSTYP output ***" >>$seqres.full
diff --git a/common/btrfs b/common/btrfs index bd4dc31fa5a8..be8ac04cd9a3 100644 --- a/common/btrfs +++ b/common/btrfs @@ -144,7 +144,7 @@ _check_btrfs_filesystem() rm -f $tmp.qgroup_report fi - $BTRFS_UTIL_PROG check $device >$tmp.fsck 2>&1 + $BTRFS_UTIL_PROG check --check-data-csum $device >$tmp.fsck 2>&1 if [ $? -ne 0 ]; then _log_err "_check_btrfs_filesystem: filesystem on $device is inconsistent" echo "*** fsck.$FSTYP output ***" >>$seqres.full
By default btrfs only checks the metadata consistency, but sometimes if we have some corruption in data while the test case doesn't utilize scrub to verify, or there is some bugs in scrub itself, we will not detect those problems. So here we do one step further by utilizing --check-data-csum option, so that if there is some data corruption, we can detect them early. Signed-off-by: Qu Wenruo <wqu@suse.com> --- common/btrfs | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)