fsck.gfs2: Support for checking gfs1 file systems

Hi Bob,

Doing an apply-and-build test on this patchset found a few trivial issues:

Firstly, "git am" found these whitespace issues:

Applying: fsck.gfs2: split function check_leaf_blks to make it /home/andy/src/gfs2-utils/.git/rebase-apply/patch:16: trailing whitespace. static int warn_and_patch(struct gfs2_inode *ip, uint64_t *leaf_no,

Applying: libgfs2: move gfs1 functions from edit to libgfs2 /home/andy/src/gfs2-utils/.git/rebase-apply/patch:89: new blank line
at EOF.
+

And then, when building with gcc 4.6, the patchset introduces the following warnings:

In file included from eattr.c:9:0:
fsck.h: In function ?is_dir?:
fsck.h:145:2: warning: implicit declaration of function ?S_ISDIR?
[-Wimplicit-function-declaration]

In file included from fs_recovery.c:12:0:
fsck.h: In function ?is_dir?:
fsck.h:145:2: warning: implicit declaration of function ?S_ISDIR?
[-Wimplicit-function-declaration]

initialize.c: In function ?reconstruct_single_journal?:
initialize.c:1286:2: warning: implicit declaration of function ?time?
[-Wimplicit-function-declaration]
initialize.c: In function ?initialize?:
initialize.c:630:5: warning: ?ret? may be used uninitialized in this
function [-Wuninitialized]
initialize.c:617:7: note: ?ret? was declared here

In file included from inode_hash.c:11:0:
fsck.h: In function ?is_dir?:
fsck.h:145:2: warning: implicit declaration of function ?S_ISDIR?
[-Wimplicit-function-declaration]

In file included from link.c:12:0:
fsck.h: In function ?is_dir?:
fsck.h:145:2: warning: implicit declaration of function ?S_ISDIR?
[-Wimplicit-function-declaration]

In file included from main.c:18:0:
fsck.h: In function ?is_dir?:
fsck.h:145:2: warning: implicit declaration of function ?S_ISDIR?
[-Wimplicit-function-declaration]

metawalk.c: In function ?check_leaf_blks?:
metawalk.c:622:6: warning: ?okay_to_fix? may be used uninitialized in
this function [-Wuninitialized]
metawalk.c:477:6: note: ?okay_to_fix? was declared here

pass1b.c: In function ?handle_dup_blk?:
pass1b.c:543:26: warning: variable ?ref_in_invalid_inode? set but not
used [-Wunused-but-set-variable]

pass1.c: In function ?pass1?:
pass1.c:1567:11: warning: variable ?offset? set but not used
[-Wunused-but-set-variable]
pass1.c:1566:11: warning: variable ?blk_count? set but not used
[-Wunused-but-set-variable]

In file included from pass1c.c:11:0:
fsck.h: In function ?is_dir?:
fsck.h:145:2: warning: implicit declaration of function ?S_ISDIR?
[-Wimplicit-function-declaration]

In file included from pass3.c:13:0:
fsck.h: In function ?is_dir?:
fsck.h:145:2: warning: implicit declaration of function ?S_ISDIR?
[-Wimplicit-function-declaration]

In file included from pass4.c:10:0:
fsck.h: In function ?is_dir?:
fsck.h:145:2: warning: implicit declaration of function ?S_ISDIR?
[-Wimplicit-function-declaration]

In file included from pass5.c:11:0:
fsck.h: In function ?is_dir?:
fsck.h:145:2: warning: implicit declaration of function ?S_ISDIR?
[-Wimplicit-function-declaration]

In file included from rgrepair.c:15:0:
fsck.h: In function ?is_dir?:
fsck.h:145:2: warning: implicit declaration of function ?S_ISDIR?
[-Wimplicit-function-declaration]

In file included from fs_bits.h:5:0,

                   from util.c:14:

fsck.h: In function ?is_dir?:
fsck.h:145:2: warning: implicit declaration of function ?S_ISDIR?
[-Wimplicit-function-declaration]

Also I noticed that the subject lines in your patch emails were wrapped as this caused git am to chop the ends off the commit shortlogs.

Otherwise this looks like a really promising patchset, nice job :)

Andy