On Wed, Apr 26, 2017 at 3:57 AM, Luis R. Rodriguez <mcgrof@xxxxxxxxxx> wrote:
> On Tue, Apr 25, 2017 at 02:07:02PM +0200, Jan Tulak wrote:
>> Ehh, it is not really an issue.. cvtnum is called only on two places
>> in the whole xfsprogs.
>
> Huh, I count 46, and spread all over the place:
>
> mcgrof@ergon ~/devel/xfsprogs-dev (git::libiniconfig-conf)$ git grep " cvtnum("| awk '{print $1}'| sort | uniq
> include/input.h:extern
> include/xfs_multidisk.h:extern
> io/fadvise.c:
> io/madvise.c:
> io/mincore.c:
> io/mmap.c:
> io/pread.c:
> io/prealloc.c:
> io/pwrite.c:
> io/readdir.c:
> io/reflink.c:
> io/resblks.c:
> io/seek.c:
> io/sendfile.c:
> io/sync_file_range.c:
> io/truncate.c:
> mkfs/proto.c:
> mkfs/xfs_mkfs.c:
> quota/edit.c:
>
> So 19 files.
They use a different prototype/implementation. If you change the
prototype of cvtnum in include/xfs_multidisk.h and mkfs
implementation, only one call in mkfs's getnum and in mkfs/proto.c's
getnum complains. There are three implementations of cvtnum in
xfsprogs.
>
>> Changing mkfs/proto.c is just few lines added
>> to this patch and the changes in xfs_mkfs.c do cause few conflicts,
>> but it is nothing terrific, I rebased all my further changes in about
>> three minutes. I pushed it into the git tree, check it now...
>
> Will do...
>
>> And thanks for this patch for the patch. :-)
>
> My pleasure, on second thought if the wrap around change can be a separate
> atomic change that might be worth it.
>
Yeah, I realised that too... I will split it if I don't forget.
Cheers,
Jan
> Luis
--
Jan Tulak
jtulak@xxxxxxxxxx / jan@xxxxxxxx
--
To unsubscribe from this list: send the line "unsubscribe linux-xfs" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
