Message-ID: <20131018044854.7339.48457.stgit@birch.djwong.org>
Date:	Thu, 17 Oct 2013 21:48:54 -0700
From:	"Darrick J. Wong" <darrick.wong@...cle.com>
To:	tytso@....edu, darrick.wong@...cle.com
Cc:	linux-ext4@...r.kernel.org
Subject: [PATCH v2 00/25] e2fsprogs patchbomb 10/2013

Well, here we go again.  This is the same patchbomb from a couple of
weeks ago, minus the patches that Ted has already accepted, plus
several fixes to resize2fs that weren't ready back then, and a few
other fixes that migrated into the other patches.  This series is
against -next.

Ted, since you've accepted patches into -pu, do you want me to send
patches against -pu as well?  Or put more bluntly, what are your
thoughts about revert-and-replace of patches in -pu?  Patches 2, 6,
11, 23, and 24 have changed significantly since 9/30.

The first eight patches fix miscellaneous errors: #1 stops dirent
iteration after we successfully link an inode into a directory.  #2
fixes a bug that prohibited us from specifyinng a 64bit superblock
number when opening an FS.  #3 prohibits running mke2fs with -E
resize= and meta_bg.  #4 causes users of the badblocks code to reject
64bit block numbers.  #5 fixes shift overflows errors when punching
the end of non-extent files.  #6 refactors all the tests for whether
or not we need to set the LARGE_FILE feature (because someone goofed
earlier).  #7 fixes a problem wherein mkfs ignored non-4096 blocksize
directives in the config file on a device larger than 2^32KB.  #8
cleans up some code in debugfs.

The next two patches fix some 64bit truncation bugs.

Regarding next five patches, I turned on bigalloc and found a number
of bugs relating to the fact that block_alloc_stats2() takes a block
number but operates on clusters.  I've fixed up all the allocation
errors that I found.  I also decided to make the quota code use
ext2fs_punch rather than try to correct its behavior wrt bigalloc.
There was also a bug wherein the requirement that 64-bit bitmaps be
enabled (via EXT2_FLAG_64BITS) for bigalloc filesystems.  There's also
a patch to reduce the e2fsck output verbosity when there are block
bitmap errors.  Note that #11 has been refactored significantly.

The next patch provides the ability to toggle the 64bit feature on any
ext4 filesystem.

The four patches after that fix various resize2fs bugs with bigalloc.

The next two patches fix bugs with metadata_csum.  There's a patch to
fix up some code to test if checksums are enabled instead of a
GDT_CSUM open-code.  Finally, there's a patch to resize2fs to rewrite
checksums of inodes that were relocated.

The next two patches add the ability to edit extended attributes and
add a fuse2fs driver for e2fsprogs.  I admit that the xattr editing
functions clash with the inline_data patches, though sadly, the inline
data patches don't provide an API to access EAs in a separate EA
block.  The fuse driver should work with the latest versions of Linux
fuse (2.9.2) and osxfuse (2.6.1).  I've been using the fuse driver to
test e2fsprogs functionality, which is how I came across most of the
bugs fixed above.  Both of these patches (#23 and #24) have received
fixes since the 9/30 posting.

The final patch adds my metadata checksum test program to the tests/
directory, along with a new metadata_check target to run a quick
check.  It includes substitute mount/umount commands for use with
fuse2fs.

For fuse2fs, I think it'd be useful to reintroduce journal replay too.
(Or cheat and call e2fsck -E journal_only...)  Also, fuse2fs doesn't
yet know how to read or write ACLs yet.

I've tested these e2fsprogs changes against the -next branch as of a
few days ago.  These days, I use a 2GB ramdisk and a 20T "disk" I
constructed out of dm-snapshot to test in an x64 VM.

Comments and questions are, as always, welcome.

--D
--
To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Powered by blists - more mailing lists