On 9/16/19 11:24 AM, Darrick J. Wong wrote:
> On Fri, Sep 06, 2019 at 02:26:46PM +0800, yu kuai wrote:
>> In commit d03a2f1b9fa8 ("xfs: include WARN, REPAIR build options in
>> XFS_BUILD_OPTIONS"), Eric pointed out that the XFS_BUILD_OPTIONS string,
>> shown at module init time and in modinfo output, does not currently
>> include all available build options. So, he added in CONFIG_XFS_WARN and
>> CONFIG_XFS_REPAIR. However, this is not enough, add in CONFIG_XFS_QUOTA
>> and CONFIG_XFS_ASSERT_FATAL.
>>
>> Signed-off-by: yu kuai <yukuai3@xxxxxxxxxx>
>> ---
>> fs/xfs/xfs_super.h | 10 ++++++++++
>> 1 file changed, 10 insertions(+)
>>
>> diff --git a/fs/xfs/xfs_super.h b/fs/xfs/xfs_super.h
>> index 763e43d..b552cf6 100644
>> --- a/fs/xfs/xfs_super.h
>> +++ b/fs/xfs/xfs_super.h
>> @@ -11,9 +11,11 @@
>> #ifdef CONFIG_XFS_QUOTA
>> extern int xfs_qm_init(void);
>> extern void xfs_qm_exit(void);
>> +# define XFS_QUOTA_STRING "quota, "
>> #else
>> # define xfs_qm_init() (0)
>> # define xfs_qm_exit() do { } while (0)
>> +# define XFS_QUOTA_STRING
>> #endif
>>
>> #ifdef CONFIG_XFS_POSIX_ACL
>> @@ -50,6 +52,12 @@ extern void xfs_qm_exit(void);
>> # define XFS_WARN_STRING
>> #endif
>>
>> +#ifdef CONFIG_XFS_ASSERT_FATAL
>> +# define XFS_ASSERT_FATAL_STRING "fatal assert, "
>
> /me wonders if the space here will screw up any scripts that try to
> parse the logging string, but OTOH that seems pretty questionable to me.
>
> Also, whatever happened to adding a sysfs file so that scripts (ok let's
> be honest, xfstests) could programmatically figure out the capabilities
> of the running xfs module?
I sent it ([PATCH] xfs: show build options in sysfs) but it was met with some
IMHO nonspecific commentary but never received a review, so was never merged.
-Eric
