)]}' { "log": [ { "commit": "2fe17c1075836b66678ed2a305fd09b6773883aa", "tree": "eb5287be8138686682eef9622872cfc7657e0664", "parents": [ "64c23e86873ee410554d6d1c76b60da47025e96f" ], "author": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Fri Jan 14 13:07:43 2011 +0100" }, "committer": { "name": "Al Viro", "email": "viro@zeniv.linux.org.uk", "time": "Mon Jan 17 02:25:31 2011 -0500" }, "message": "fallocate should be a file operation\n\nCurrently all filesystems except XFS implement fallocate asynchronously,\nwhile XFS forced a commit. Both of these are suboptimal - in case of O_SYNC\nI/O we really want our allocation on disk, especially for the !KEEP_SIZE\ncase where we actually grow the file with user-visible zeroes. On the\nother hand always commiting the transaction is a bad idea for fast-path\nuses of fallocate like for example in recent Samba versions. Given\nthat block allocation is a data plane operation anyway change it from\nan inode operation to a file operation so that we have the file structure\navailable that lets us check for O_SYNC.\n\nThis also includes moving the code around for a few of the filesystems,\nand remove the already unnedded S_ISDIR checks given that we only wire\nup fallocate for regular files.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Al Viro \u003cviro@zeniv.linux.org.uk\u003e\n" }, { "commit": "008d23e4852d78bb2618f2035f8b2110b6a6b968", "tree": "81c88f744f6f3fc84132527c1ddc0b4da410c5e2", "parents": [ "8f685fbda43deccd130d192c9fcef1444649eaca", "bfc672dcf323877228682aff79dff8ecd9f30ff8" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jan 13 10:05:56 2011 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jan 13 10:05:56 2011 -0800" }, "message": "Merge branch \u0027for-next\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/jikos/trivial\n\n* \u0027for-next\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/jikos/trivial: (43 commits)\n Documentation/trace/events.txt: Remove obsolete sched_signal_send.\n writeback: fix global_dirty_limits comment runtime -\u003e real-time\n ppc: fix comment typo singal -\u003e signal\n drivers: fix comment typo diable -\u003e disable.\n m68k: fix comment typo diable -\u003e disable.\n wireless: comment typo fix diable -\u003e disable.\n media: comment typo fix diable -\u003e disable.\n remove doc for obsolete dynamic-printk kernel-parameter\n remove extraneous \u0027is\u0027 from Documentation/iostats.txt\n Fix spelling milisec -\u003e ms in snd_ps3 module parameter description\n Fix spelling mistakes in comments\n Revert conflicting V4L changes\n i7core_edac: fix typos in comments\n mm/rmap.c: fix comment\n sound, ca0106: Fix assignment to \u0027channel\u0027.\n hrtimer: fix a typo in comment\n init/Kconfig: fix typo\n anon_inodes: fix wrong function name in comment\n fix comment typos concerning \"consistent\"\n poll: fix a typo in comment\n ...\n\nFix up trivial conflicts in:\n - drivers/net/wireless/iwlwifi/iwl-core.c (moved to iwl-legacy.c)\n - fs/ext4/ext4.h\n\nAlso fix missed \u0027diabled\u0027 typo in drivers/net/bnx2x/bnx2x.h while at it.\n" }, { "commit": "3889fd57ea3c58209354862523275774fca9db03", "tree": "1e68fb9b2691c0b792a36be7d87c987e8508dac6", "parents": [ "b40971426a837e9dc9c66e1b6bbcb3874eafe4e0" ], "author": { "name": "Jiaying Zhang", "email": "jiayingz@google.com", "time": "Mon Jan 10 12:47:05 2011 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Mon Jan 10 12:47:05 2011 -0500" }, "message": "ext4: flush the i_completed_io_list during ext4_truncate\n\nTed first found the bug when running 2.6.36 kernel with dioread_nolock\nmount option that xfstests #13 complained about wrong file size during fsck.\nHowever, the bug exists in the older kernels as well although it is\nsomehow harder to trigger.\n\nThe problem is that ext4_end_io_work() can happen after we have truncated an\ninode to a smaller size. Then when ext4_end_io_work() calls \next4_convert_unwritten_extents(), we may reallocate some blocks that have \nbeen truncated, so the inode size becomes inconsistent with the allocated\nblocks. \n\nThe following patch flushes the i_completed_io_list during truncate to reduce \nthe risk that some pending end_io requests are executed later and convert \nalready truncated blocks to initialized. \n\nNote that although the fix helps reduce the problem a lot there may still \nbe a race window between vmtruncate() and ext4_end_io_work(). The fundamental\nproblem is that if vmtruncate() is called without either i_mutex or i_alloc_sem\nheld, it can race with an ongoing write request so that the io_end request is\nprocessed later when the corresponding blocks have been truncated.\n\nTed and I have discussed the problem offline and we saw a few ways to fix\nthe race completely:\n\na) We guarantee that i_mutex lock and i_alloc_sem write lock are both hold \nwhenever vmtruncate() is called. The i_mutex lock prevents any new write\nrequests from entering writeback and the i_alloc_sem prevents the race\nfrom ext4_page_mkwrite(). Currently we hold both locks if vmtruncate()\nis called from do_truncate(), which is probably the most common case.\nHowever, there are places where we may call vmtruncate() without holding\neither i_mutex or i_alloc_sem. I would like to ask for other people\u0027s\nopinions on what locks are expected to be held before calling vmtruncate().\nThere seems a disagreement among the callers of that function.\n\nb) We change the ext4 write path so that we change the extent tree to contain \nthe newly allocated blocks and update i_size both at the same time --- when \nthe write of the data blocks is completed.\n\nc) We add some additional locking to synchronize vmtruncate() and \next4_end_io_work(). This approach may have performance implications so we\nneed to be careful.\n\nAll of the above proposals may require more substantial changes, so\nwe may consider to take the following patch as a bandaid.\n\nSigned-off-by: Jiaying Zhang \u003cjiayingz@google.com\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n\n" }, { "commit": "8aefcd557d26d0023a36f9ec5afbf55e59f8f26b", "tree": "e13143306cd64525cddd2cc2513c448275a1d95a", "parents": [ "353eb83c1422c6326eaab30ce044a179c6018169" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Mon Jan 10 12:29:43 2011 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Mon Jan 10 12:29:43 2011 -0500" }, "message": "ext4: dynamically allocate the jbd2_inode in ext4_inode_info as necessary\n\nReplace the jbd2_inode structure (which is 48 bytes) with a pointer\nand only allocate the jbd2_inode when it is needed --- that is, when\nthe file system has a journal present and the inode has been opened\nfor writing. This allows us to further slim down the ext4_inode_info\nstructure.\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "353eb83c1422c6326eaab30ce044a179c6018169", "tree": "fd43e39b344de5ee43d82c90f0affda68a65f2df", "parents": [ "8a2005d3f84457b7d7d8646ab5195341d9e5f06a" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Mon Jan 10 12:18:25 2011 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Mon Jan 10 12:18:25 2011 -0500" }, "message": "ext4: drop i_state_flags on architectures with 64-bit longs\n\nWe can store the dynamic inode state flags in the high bits of\nEXT4_I(inode)-\u003ei_flags, and eliminate i_state_flags. This saves 8\nbytes from the size of ext4_inode_info structure, which when\nmultiplied by the number of the number of in the inode cache, can save\na lot of memory.\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "8a2005d3f84457b7d7d8646ab5195341d9e5f06a", "tree": "4d94161103e806435745e1544fb24e6398b2e06b", "parents": [ "b05e6ae58a13b56e3e11882c1fc71948c9b29760" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Mon Jan 10 12:13:42 2011 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Mon Jan 10 12:13:42 2011 -0500" }, "message": "ext4: reorder ext4_inode_info structure elements to remove unneeded padding\n\nBy reordering the elements in the ext4_inode_info structure, we can\nreduce the padding needed on an x86_64 system by 16 bytes.\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "b05e6ae58a13b56e3e11882c1fc71948c9b29760", "tree": "452cbadcbc8091b4db95f917f28b0f9de845dabf", "parents": [ "01f49d0b9d0209dc1194255b11601e4b94447b36" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Mon Jan 10 12:13:26 2011 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Mon Jan 10 12:13:26 2011 -0500" }, "message": "ext4: drop ec_type from the ext4_ext_cache structure\n\nWe can encode the ec_type information by using ee_len \u003d\u003d 0 to denote\nEXT4_EXT_CACHE_NO, ee_start \u003d\u003d 0 to denote EXT4_EXT_CACHE_GAP, and if\nneither is true, then the cache type must be EXT4_EXT_CACHE_EXTENT.\nThis allows us to reduce the size of ext4_ext_inode by another 8\nbytes. (ec_type is 4 bytes, plus another 4 bytes of padding)\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n\n" }, { "commit": "01f49d0b9d0209dc1194255b11601e4b94447b36", "tree": "c3ec53c7725bf7a9188e844d8c8c09432d2be6ae", "parents": [ "f232109773ff5b0c840a6761d74940b9cf0d66ec" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Mon Jan 10 12:13:03 2011 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Mon Jan 10 12:13:03 2011 -0500" }, "message": "ext4: use ext4_lblk_t instead of sector_t for logical blocks\n\nThis fixes a number of places where we used sector_t instead of\next4_lblk_t for logical blocks, which for ext4 are still 32-bit data\ntypes. No point wasting space in the ext4_inode_info structure, and\nrequiring 64-bit arithmetic on 32-bit systems, when it isn\u0027t\nnecessary.\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "f232109773ff5b0c840a6761d74940b9cf0d66ec", "tree": "a32da5ebd1a3e074e9f52107f3bdc23babb6340c", "parents": [ "ad4fb9cafe100a4a9de6e0529015e584d94ac8dc" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Mon Jan 10 12:12:36 2011 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Mon Jan 10 12:12:36 2011 -0500" }, "message": "ext4: replace i_delalloc_reserved_flag with EXT4_STATE_DELALLOC_RESERVED\n\nRemove the short element i_delalloc_reserved_flag from the\next4_inode_info structure and replace it a new bit in i_state_flags.\nSince we have an ext4_inode_info for every ext4 inode cached in the\ninode cache, any savings we can produce here is a very good thing from\na memory utilization perspective.\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "f7c21177af0b32a2cd9ee36189637f0c1f0e1e17", "tree": "ba83d4ddcd4b7e15ff575f0b75013ba9ed62f249", "parents": [ "f9a62d090cf47fae2fe6f6bd8eb9f24482573fd8" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Mon Jan 10 12:10:55 2011 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Mon Jan 10 12:10:55 2011 -0500" }, "message": "ext4: Use ext4_error_file() to print the pathname to the corrupted inode\n\nWhere the file pointer is available, use ext4_error_file() instead of\next4_error_inode().\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "4b7bd364700d9ac8372eff48832062b936d0793b", "tree": "0dbf78c95456a0b02d07fcd473281f04a87e266d", "parents": [ "c0d8768af260e2cbb4bf659ae6094a262c86b085", "90a8a73c06cc32b609a880d48449d7083327e11a" ], "author": { "name": "Jiri Kosina", "email": "jkosina@suse.cz", "time": "Wed Dec 22 18:57:02 2010 +0100" }, "committer": { "name": "Jiri Kosina", "email": "jkosina@suse.cz", "time": "Wed Dec 22 18:57:02 2010 +0100" }, "message": "Merge branch \u0027master\u0027 into for-next\n\nConflicts:\n\tMAINTAINERS\n\tarch/arm/mach-omap2/pm24xx.c\n\tdrivers/scsi/bfa/bfa_fcpim.c\n\nNeeded to update to apply fixes for which the old branch was too\noutdated.\n" }, { "commit": "af0b44a1970fed1cda31d2969c99c46ffc515160", "tree": "b382d80df604c0bbd61836c13ae0b6f3020728e5", "parents": [ "cad3f00763dcf9dfc62cbddf4bd714ab5a71a0eb" ], "author": { "name": "Eric Sandeen", "email": "sandeen@redhat.com", "time": "Sun Dec 19 22:10:31 2010 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Sun Dec 19 22:10:31 2010 -0500" }, "message": "ext4: zero out nanosecond timestamps for small inodes\n\nWhen nanosecond timestamp resolution isn\u0027t supported on an ext4\npartition (inode size \u003d 128), stat() appears to be returning\nuninitialized garbage in the nanosecond component of timestamps.\n\nEXT4_INODE_GET_XTIME should zero out tv_nsec when EXT4_FITS_IN_INODE\nevaluates to false.\n\nReported-by: Jordan Russell \u003cjr-list-2010@quo.to\u003e\nSigned-off-by: Eric Sandeen \u003csandeen@redhat.com\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "cad3f00763dcf9dfc62cbddf4bd714ab5a71a0eb", "tree": "02ec7c4959f5506f0b2fe128b1318e257ee6f0e4", "parents": [ "b17b35ec13adfeb0346d4b329110b14adc509327" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Sun Dec 19 22:07:02 2010 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Sun Dec 19 22:07:02 2010 -0500" }, "message": "ext4: optimize ext4_check_dir_entry() with unlikely() annotations\n\nThis function gets called a lot for large directories, and the answer\nis almost always \"no, no, there\u0027s no problem\". This means using\nunlikely() is a good thing.\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "a2595b8aa67011419dae26b47e474f46df902989", "tree": "d44d79256f92f32b7958a04e89907f7e1a755767", "parents": [ "673c610033a8202c037ecd068c7a235495acda17" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Wed Dec 15 20:30:48 2010 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Wed Dec 15 20:30:48 2010 -0500" }, "message": "ext4: Add second mount options field since the s_mount_opt is full up\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "673c610033a8202c037ecd068c7a235495acda17", "tree": "e47ca541bf33f1d43d4ced9b72d6100853896236", "parents": [ "fd8c37eccdda21153298997417144b38b1623196" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Wed Dec 15 20:28:48 2010 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Wed Dec 15 20:28:48 2010 -0500" }, "message": "ext4: Move struct ext4_mount_options from ext4.h to super.c\n\nMove the ext4_mount_options structure definition from ext4.h, since it\nis only used in super.c.\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "fd8c37eccdda21153298997417144b38b1623196", "tree": "313e022d9d82f890e728dce66ce6e115ea3970b1", "parents": [ "b0c3844d8af6b9f3f18f31e1b0502fbefa2166be" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Wed Dec 15 20:26:48 2010 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Wed Dec 15 20:26:48 2010 -0500" }, "message": "ext4: Simplify the usage of clear_opt() and set_opt() macros\n\nChange clear_opt() and set_opt() to take a superblock pointer instead\nof a pointer to EXT4_SB(sb)-\u003es_mount_opt. This makes it easier for us\nto support a second mount option field.\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "1449032be17abb69116dbc393f67ceb8bd034f92", "tree": "f62757457241c2fdc14105afc12cb2718f7a2e68", "parents": [ "e8a7e48bb248a1196484d3f8afa53bded2b24e71" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Tue Dec 14 15:27:50 2010 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Tue Dec 14 15:27:50 2010 -0500" }, "message": "ext4: Turn off multiple page-io submission by default\n\nJon Nelson has found a test case which causes postgresql to fail with\nthe error:\n\npsql:t.sql:4: ERROR: invalid page header in block 38269 of relation base/16384/16581\n\nUnder memory pressure, it looks like part of a file can end up getting\nreplaced by zero\u0027s. Until we can figure out the cause, we\u0027ll roll\nback the change and use block_write_full_page() instead of\next4_bio_write_page(). The new, more efficient writing function can\nbe used via the mount option mblk_io_submit, so we can test and fix\nthe new page I/O code.\n\nTo reproduce the problem, install postgres 8.4 or 9.0, and pin enough\nmemory such that the system just at the end of triggering writeback\nbefore running the following sql script:\n\nbegin;\ncreate temporary table foo as select x as a, ARRAY[x] as b FROM\ngenerate_series(1, 10000000 ) AS x;\ncreate index foo_a_idx on foo (a);\ncreate index foo_b_idx on foo USING GIN (b);\nrollback;\n\nIf the temporary table is created on a hard drive partition which is\nencrypted using dm_crypt, then under memory pressure, approximately\n30-40% of the time, pgsql will issue the above failure.\n\nThis patch should fix this problem, and the problem will come back if\nthe file system is mounted with the mblk_io_submit mount option.\n\nReported-by: Jon Nelson \u003cjnelson@jamponi.net\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "83668e7141c7a0aa4035bde94344b81f9cf966ab", "tree": "34d9fd52470b475d6e9d88ece8ca1ba80bf85a42", "parents": [ "f7ad6d2e9201a6e1c9ee6530a291452eb695feb8" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Mon Nov 08 13:45:33 2010 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Mon Nov 08 13:45:33 2010 -0500" }, "message": "ext4: fix potential race when freeing ext4_io_page structures\n\nUse an atomic_t and make sure we don\u0027t free the structure while we\nmight still be submitting I/O for that page.\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "f7ad6d2e9201a6e1c9ee6530a291452eb695feb8", "tree": "006cdcfd487404fb61986e3030d96cb33866755d", "parents": [ "ce7e010aef63dc6b37a2354f7c9f5f4aedb37978" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Mon Nov 08 13:43:33 2010 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Mon Nov 08 13:43:33 2010 -0500" }, "message": "ext4: handle writeback of inodes which are being freed\n\nThe following BUG can occur when an inode which is getting freed when\nit still has dirty pages outstanding, and it gets deleted (in this\nbecause it was the target of a rename). In ordered mode, we need to\nmake sure the data pages are written just in case we crash before the\nrename (or unlink) is committed. If the inode is being freed then\nwhen we try to igrab the inode, we end up tripping the BUG_ON at\nfs/ext4/page-io.c:146.\n\nTo solve this problem, we need to keep track of the number of io\ncallbacks which are pending, and avoid destroying the inode until they\nhave all been completed. That way we don\u0027t have to bump the inode\ncount to keep the inode from being destroyed; an approach which\ndoesn\u0027t work because the count could have already been dropped down to\nzero before the inode writeback has started (at which point we\u0027re not\nallowed to bump the count back up to 1, since it\u0027s already started\ngetting freed).\n\nThanks to Dave Chinner for suggesting this approach, which is also\nused by XFS.\n\n kernel BUG at /scratch_space/linux-2.6/fs/ext4/page-io.c:146!\n Call Trace:\n [\u003cffffffff811075b1\u003e] ext4_bio_write_page+0x172/0x307\n [\u003cffffffff811033a7\u003e] mpage_da_submit_io+0x2f9/0x37b\n [\u003cffffffff811068d7\u003e] mpage_da_map_and_submit+0x2cc/0x2e2\n [\u003cffffffff811069b3\u003e] mpage_add_bh_to_extent+0xc6/0xd5\n [\u003cffffffff81106c66\u003e] write_cache_pages_da+0x2a4/0x3ac\n [\u003cffffffff81107044\u003e] ext4_da_writepages+0x2d6/0x44d\n [\u003cffffffff81087910\u003e] do_writepages+0x1c/0x25\n [\u003cffffffff810810a4\u003e] __filemap_fdatawrite_range+0x4b/0x4d\n [\u003cffffffff810815f5\u003e] filemap_fdatawrite_range+0xe/0x10\n [\u003cffffffff81122a2e\u003e] jbd2_journal_begin_ordered_truncate+0x7b/0xa2\n [\u003cffffffff8110615d\u003e] ext4_evict_inode+0x57/0x24c\n [\u003cffffffff810c14a3\u003e] evict+0x22/0x92\n [\u003cffffffff810c1a3d\u003e] iput+0x212/0x249\n [\u003cffffffff810bdf16\u003e] dentry_iput+0xa1/0xb9\n [\u003cffffffff810bdf6b\u003e] d_kill+0x3d/0x5d\n [\u003cffffffff810be613\u003e] dput+0x13a/0x147\n [\u003cffffffff810b990d\u003e] sys_renameat+0x1b5/0x258\n [\u003cffffffff81145f71\u003e] ? _atomic_dec_and_lock+0x2d/0x4c\n [\u003cffffffff810b2950\u003e] ? cp_new_stat+0xde/0xea\n [\u003cffffffff810b29c1\u003e] ? sys_newlstat+0x2d/0x38\n [\u003cffffffff810b99c6\u003e] sys_rename+0x16/0x18\n [\u003cffffffff81002a2b\u003e] system_call_fastpath+0x16/0x1b\n\nReported-by: Nick Bowler \u003cnbowler@elliptictech.com\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\nTested-by: Nick Bowler \u003cnbowler@elliptictech.com\u003e\n" }, { "commit": "b595076a180a56d1bb170e6eceda6eb9d76f4cd3", "tree": "bc01ec7283808013e0b8ce7713fd6fc40f810429", "parents": [ "6aaccece1c483f189f76f1282b3984ff4c7ecb0a" ], "author": { "name": "Uwe Kleine-König", "email": "u.kleine-koenig@pengutronix.de", "time": "Mon Nov 01 15:38:34 2010 -0400" }, "committer": { "name": "Jiri Kosina", "email": "jkosina@suse.cz", "time": "Mon Nov 01 15:38:34 2010 -0400" }, "message": "tree-wide: fix comment/printk typos\n\n\"gadget\", \"through\", \"command\", \"maintain\", \"maintain\", \"controller\", \"address\",\n\"between\", \"initiali[zs]e\", \"instead\", \"function\", \"select\", \"already\",\n\"equal\", \"access\", \"management\", \"hierarchy\", \"registration\", \"interest\",\n\"relative\", \"memory\", \"offset\", \"already\",\n\nSigned-off-by: Uwe Kleine-König \u003cu.kleine-koenig@pengutronix.de\u003e\nSigned-off-by: Jiri Kosina \u003cjkosina@suse.cz\u003e\n" }, { "commit": "eee4adc709afe40d8c02fa154c63dbeb55d911e3", "tree": "e65d8235eb7cf745a1d9dc2ac9ab1c62bdd45758", "parents": [ "61d08673de1fe68bfba86203258377bf39f234b6" ], "author": { "name": "Eric Sandeen", "email": "sandeen@redhat.com", "time": "Wed Oct 27 21:30:15 2010 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Wed Oct 27 21:30:15 2010 -0400" }, "message": "ext4: move ext4_mb_{get,put}_buddy_cache_lock and make them static\n\nThese functions are only used within fs/ext4/mballoc.c, so move them\nso they are used after they are defined, and then make them be static.\n\nSigned-off-by: Eric Sandeen \u003csandeen@redhat.com\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "61d08673de1fe68bfba86203258377bf39f234b6", "tree": "5cb32998c69eca626df54b50cb01e50a83c4db81", "parents": [ "4a873a472b3bbcfd425d7ae210afdec28c04e2e5" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Wed Oct 27 21:30:15 2010 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Wed Oct 27 21:30:15 2010 -0400" }, "message": "ext4: rename mark_bitmap_end() to ext4_mark_bitmap_end()\n\nFix a namespace leak from fs/ext4\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n\n\n" }, { "commit": "4a873a472b3bbcfd425d7ae210afdec28c04e2e5", "tree": "88aa333bfafcc5303b7a3074865f51f67ad836b9", "parents": [ "bf89d16f6ef5389f1b9d8046e838ec87b9b3f8b9" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Wed Oct 27 21:30:14 2010 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Wed Oct 27 21:30:14 2010 -0400" }, "message": "ext4: move flush_completed_IO to fs/ext4/fsync.c and make it static\n\nFix a namespace leak by moving the function to the file where it is\nused and making it static.\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "1f109d5a17b438c4a54cbf6fd87a249e3d72fb21", "tree": "f58c6fd431975bf900b502b80122e175065da657", "parents": [ "5dabfc78dcedbe46cb2e4872dde448de3cec2979" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Wed Oct 27 21:30:14 2010 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Wed Oct 27 21:30:14 2010 -0400" }, "message": "ext4: make various ext4 functions be static\n\nThese functions have no need to be exported beyond file context.\n\nNo functions needed to be moved for this commit; just some function\ndeclarations changed to be static and removed from header files.\n\n(A similar patch was submitted by Eric Sandeen, but I wanted to handle\ncode movement in separate patches to make sure code changes didn\u0027t\naccidentally get dropped.)\n\nSigned-off-by: Eric Sandeen \u003csandeen@redhat.com\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "5dabfc78dcedbe46cb2e4872dde448de3cec2979", "tree": "804ef3e76289978ef6690c2b8f379a14a60b15f0", "parents": [ "7f93cff90fa9be6ed45f6189e136153d1d8631b0" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Wed Oct 27 21:30:14 2010 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Wed Oct 27 21:30:14 2010 -0400" }, "message": "ext4: rename {exit,init}_ext4_*() to ext4_{exit,init}_*()\n\nThis is a cleanup to avoid namespace leaks out of fs/ext4\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "7360d1731e5dc78aec867e65e55f9fb58782b5fe", "tree": "2cc0d139ec129c19150ef481d6adb9b0fd947c89", "parents": [ "367a51a339020ba4d9edb0ce0f21d65bd50b00c9" ], "author": { "name": "Lukas Czerner", "email": "lczerner@redhat.com", "time": "Wed Oct 27 21:30:12 2010 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Wed Oct 27 21:30:12 2010 -0400" }, "message": "ext4: Add batched discard support for ext4\n\nWalk through allocation groups and trim all free extents. It can be\ninvoked through FITRIM ioctl on the file system. The main idea is to\nprovide a way to trim the whole file system if needed, since some SSD\u0027s\nmay suffer from performance loss after the whole device was filled (it\ndoes not mean that fs is full!).\n\nIt search for free extents in allocation groups specified by Byte range\nstart -\u003e start+len. When the free extent is within this range, blocks\nare marked as used and then trimmed. Afterwards these blocks are marked\nas free in per-group bitmap.\n\nSince fstrim is a long operation it is good to have an ability to\ninterrupt it by a signal. This was added by Dmitry Monakhov.\nThanks Dimitry.\n\nSigned-off-by: Lukas Czerner \u003clczerner@redhat.com\u003e\nSigned-off-by: Dmitry Monakhov \u003cdmonakhov@openvz.org\u003e\nReviewed-by: Jan Kara \u003cjack@suse.cz\u003e\nReviewed-by: Dmitry Monakhov \u003cdmonakhov@openvz.org\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "bd2d0210cf22f2bd0cef72eb97cf94fc7d31d8cc", "tree": "f0d1902b7ff4294114614cc706855c3d6b131f73", "parents": [ "1de3e3df917459422cb2aecac440febc8879d410" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Wed Oct 27 21:30:10 2010 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Wed Oct 27 21:30:10 2010 -0400" }, "message": "ext4: use bio layer instead of buffer layer in mpage_da_submit_io\n\nCall the block I/O layer directly instad of going through the buffer\nlayer. This should give us much better performance and scalability,\nas well as lowering our CPU utilization when doing buffered writeback.\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n\n" }, { "commit": "640e9396566a1e1f52f2db294755a23f1e62cc97", "tree": "013b0028ceff606beb1290cc9f42704f2a94ab3d", "parents": [ "c999af2b347a55174f702702e0df814d05ef5491" ], "author": { "name": "Eric Sandeen", "email": "sandeen@redhat.com", "time": "Wed Oct 27 21:30:08 2010 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Wed Oct 27 21:30:08 2010 -0400" }, "message": "ext4: remove unused ext4_sb_info members\n\nNot that these take up a lot of room, but the structure is long enough\nas it is, and there\u0027s no need to confuse people with these various\nundocumented \u0026 unused structure members...\n\nSigned-off-by: Eric Sandeen \u003csandeen@redaht.com\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "e0d10bfa91b0a089a9e2c378b5c42f4e96171d95", "tree": "254d6b0b7d7ca2dfd817171d8d641c1a648e9c46", "parents": [ "c41303ced67c4ebf51bf2e7d0f139155e09e0939" ], "author": { "name": "Toshiyuki Okajima", "email": "toshi.okajima@jp.fujitsu.com", "time": "Wed Oct 27 21:30:06 2010 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Wed Oct 27 21:30:06 2010 -0400" }, "message": "ext4: improve llseek error handling for overly large seek offsets\n\nThe llseek system call should return EINVAL if passed a seek offset\nwhich results in a write error. What this maximum offset should be\ndepends on whether or not the huge_file file system feature is set,\nand whether or not the file is extent based or not.\n\n\nIf the file has no \"EXT4_EXTENTS_FL\" flag, the maximum size which can be \nwritten (write systemcall) is different from the maximum size which can be \nsought (lseek systemcall).\n\nFor example, the following 2 cases demonstrates the differences\nbetween the maximum size which can be written, versus the seek offset\nallowed by the llseek system call:\n\n#1: mkfs.ext3 \u003cdev\u003e; mount -t ext4 \u003cdev\u003e\n#2: mkfs.ext3 \u003cdev\u003e; tune2fs -Oextent,huge_file \u003cdev\u003e; mount -t ext4 \u003cdev\u003e\n\nTable. the max file size which we can write or seek\n at each filesystem feature tuning and file flag setting\n+\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d+\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d+\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d+\n| \\ File flag| | |\n| \\ | !EXT4_EXTENTS_FL | EXT4_EXTETNS_FL |\n|case \\| | |\n+------------+-------------------------------+-------------------------------+\n| #1 | write: 2194719883264 | write: -------------- |\n| | seek: 2199023251456 | seek: -------------- |\n+------------+-------------------------------+-------------------------------+\n| #2 | write: 4402345721856 | write: 17592186044415 |\n| | seek: 17592186044415 | seek: 17592186044415 |\n+------------+-------------------------------+-------------------------------+\n\nThe differences exist because ext4 has 2 maxbytes which are sb-\u003es_maxbytes\n(\u003d extent-mapped maxbytes) and EXT4_SB(sb)-\u003es_bitmap_maxbytes (\u003d block-mapped \nmaxbytes). Although generic_file_llseek uses only extent-mapped maxbytes.\n(llseek of ext4_file_operations is generic_file_llseek which uses\nsb-\u003es_maxbytes.)\n\nTherefore we create ext4 llseek function which uses 2 maxbytes.\n\nThe new own function originates from generic_file_llseek().\nIf the file flag, \"EXT4_EXTENTS_FL\" is not set, the function alters \ninode-\u003ei_sb-\u003es_maxbytes into EXT4_SB(inode-\u003ei_sb)-\u003es_bitmap_maxbytes.\n\nSigned-off-by: Toshiyuki Okajima \u003ctoshi.okajima@jp.fujitsu.com\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\nCc: Andreas Dilger \u003cadilger.kernel@dilger.ca\u003e\n" }, { "commit": "857ac889cce8a486d47874db4d2f9620e7e9e5de", "tree": "12401895197d819fcbf2335244d91259f4640aa2", "parents": [ "bfff68738f1cb5c93dab1114634cea02aae9e7ba" ], "author": { "name": "Lukas Czerner", "email": "lczerner@redhat.com", "time": "Wed Oct 27 21:30:05 2010 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Wed Oct 27 21:30:05 2010 -0400" }, "message": "ext4: add interface to advertise ext4 features in sysfs\n\nUser-space should have the opportunity to check what features doest ext4\nsupport in each particular copy. This adds easy interface by creating new\n\"features\" directory in sys/fs/ext4/. In that directory files\nadvertising feature names can be created.\n\nAdd lazy_itable_init to the feature list.\n\nSigned-off-by: Lukas Czerner \u003clczerner@redhat.com\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "bfff68738f1cb5c93dab1114634cea02aae9e7ba", "tree": "b6cdf3f26e86464c7088cab62d837eb32f559fb9", "parents": [ "e6fa0be699449d28a20e815bfe9ce26725ec4962" ], "author": { "name": "Lukas Czerner", "email": "lczerner@redhat.com", "time": "Wed Oct 27 21:30:05 2010 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Wed Oct 27 21:30:05 2010 -0400" }, "message": "ext4: add support for lazy inode table initialization\n\nWhen the lazy_itable_init extended option is passed to mke2fs, it\nconsiderably speeds up filesystem creation because inode tables are\nnot zeroed out. The fact that parts of the inode table are\nuninitialized is not a problem so long as the block group descriptors,\nwhich contain information regarding how much of the inode table has\nbeen initialized, has not been corrupted However, if the block group\nchecksums are not valid, e2fsck must scan the entire inode table, and\nthe the old, uninitialized data could potentially cause e2fsck to\nreport false problems.\n\nHence, it is important for the inode tables to be initialized as soon\nas possble. This commit adds this feature so that mke2fs can safely\nuse the lazy inode table initialization feature to speed up formatting\nfile systems.\n\nThis is done via a new new kernel thread called ext4lazyinit, which is\ncreated on demand and destroyed, when it is no longer needed. There\nis only one thread for all ext4 filesystems in the system. When the\nfirst filesystem with inititable mount option is mounted, ext4lazyinit\nthread is created, then the filesystem can register its request in the\nrequest list.\n\nThis thread then walks through the list of requests picking up\nscheduled requests and invoking ext4_init_inode_table(). Next schedule\ntime for the request is computed by multiplying the time it took to\nzero out last inode table with wait multiplier, which can be set with\nthe (init_itable\u003dn) mount option (default is 10). We are doing\nthis so we do not take the whole I/O bandwidth. When the thread is no\nlonger necessary (request list is empty) it frees the appropriate\nstructures and exits (and can be created later later by another\nfilesystem).\n\nWe do not disturb regular inode allocations in any way, it just do not\ncare whether the inode table is, or is not zeroed. But when zeroing, we\nhave to skip used inodes, obviously. Also we should prevent new inode\nallocations from the group, while zeroing is on the way. For that we\ntake write alloc_sem lock in ext4_init_inode_table() and read alloc_sem\nin the ext4_claim_inode, so when we are unlucky and allocator hits the\ngroup which is currently being zeroed, it just has to wait.\n\nThis can be suppresed using the mount option no_init_itable.\n\nSigned-off-by: Lukas Czerner \u003clczerner@redhat.com\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "fb1813f4a8a27bbd4735967e46931e61fc837a3e", "tree": "c9d7c9d851c81663a8e501ba5c14f2a4b332f893", "parents": [ "b853fd364810a241050778124842a8c415c72a69" ], "author": { "name": "Curt Wohlgemuth", "email": "curtw@google.com", "time": "Wed Oct 27 21:29:12 2010 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Wed Oct 27 21:29:12 2010 -0400" }, "message": "ext4: use dedicated slab caches for group_info structures\n\next4_group_info structures are currently allocated with kmalloc().\nWith a typical 4K block size, these are 136 bytes each -- meaning\nthey\u0027ll each consume a 256-byte slab object. On a system with many\next4 large partitions, that\u0027s a lot of wasted kernel slab space.\n(E.g., a single 1TB partition will have about 8000 block groups, using\nabout 2MB of slab, of which nearly 1MB is wasted.)\n\nThis patch creates an array of slab pointers created as needed --\ndepending on the superblock block size -- and uses these slabs to\nallocate the group info objects.\n\nGoogle-Bug-Id: 2980809\n\nSigned-off-by: Curt Wohlgemuth \u003ccurtw@google.com\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "5f248c9c251c60af3403902b26e08de43964ea0b", "tree": "6d3328e72a7e4015a64017eb30be18095c6a3c64", "parents": [ "f6cec0ae58c17522a7bc4e2f39dae19f199ab534", "dca332528bc69e05f67161e1ed59929633d5e63d" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Aug 10 11:26:52 2010 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Aug 10 11:26:52 2010 -0700" }, "message": "Merge branch \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs-2.6\n\n* \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs-2.6: (96 commits)\n no need for list_for_each_entry_safe()/resetting with superblock list\n Fix sget() race with failing mount\n vfs: don\u0027t hold s_umount over close_bdev_exclusive() call\n sysv: do not mark superblock dirty on remount\n sysv: do not mark superblock dirty on mount\n btrfs: remove junk sb_dirt change\n BFS: clean up the superblock usage\n AFFS: wait for sb synchronization when needed\n AFFS: clean up dirty flag usage\n cifs: truncate fallout\n mbcache: fix shrinker function return value\n mbcache: Remove unused features\n add f_flags to struct statfs(64)\n pass a struct path to vfs_statfs\n update VFS documentation for method changes.\n All filesystems that need invalidate_inode_buffers() are doing that explicitly\n convert remaining -\u003eclear_inode() to -\u003eevict_inode()\n Make -\u003edrop_inode() just return whether inode needs to be dropped\n fs/inode.c:clear_inode() is gone\n fs/inode.c:evict() doesn\u0027t care about delete vs. non-delete paths now\n ...\n\nFix up trivial conflicts in fs/nilfs2/super.c\n" }, { "commit": "0930fcc1ee2f0a810b938bc283a3a262d7adccbb", "tree": "5e5d10894f1e31a7aede75110b43d7a41347631d", "parents": [ "7da08fd17a6e42d80f0f3897a5cbd682e77bcdb4" ], "author": { "name": "Al Viro", "email": "viro@zeniv.linux.org.uk", "time": "Mon Jun 07 13:16:22 2010 -0400" }, "committer": { "name": "Al Viro", "email": "viro@zeniv.linux.org.uk", "time": "Mon Aug 09 16:48:30 2010 -0400" }, "message": "convert ext4 to -\u003eevict_inode()\n\npretty much brute-force...\n\nSigned-off-by: Al Viro \u003cviro@zeniv.linux.org.uk\u003e\n" }, { "commit": "0cfc9255a1efb0467de2162950197750570ecec0", "tree": "3bb2510d8b6d373d7db3a49bcda3cbbebd10330e", "parents": [ "8dd420466c7bfc459fa04680bd5690bfc41a4553" ], "author": { "name": "Eric Sandeen", "email": "sandeen@redhat.com", "time": "Thu Aug 05 01:46:37 2010 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Thu Aug 05 01:46:37 2010 -0400" }, "message": "ext4: re-inline ext4_rec_len_(to|from)_disk functions\n\ncommit 3d0518f4, \"ext4: New rec_len encoding for very\nlarge blocksizes\" made several changes to this path, but from\na perf perspective, un-inlining ext4_rec_len_from_disk() seems\nmost significant. This function is called from ext4_check_dir_entry(),\nwhich on a file-creation workload is called extremely often.\n\nI tested this with bonnie:\n\n# bonnie++ -u root -s 0 -f -x 200 -d /mnt/test -n 32\n\n(this does 200 iterations) and got this for the file creations:\n\next4 stock: Average \u003d 21206.8 files/s\next4 inlined: Average \u003d 22346.7 files/s (+5%)\n\nSigned-off-by: Eric Sandeen \u003csandeen@redhat.com\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "8b67f04ab9de5d8f3a71aef72bf02c995a506db5", "tree": "dd05968730762f5b18de4c6b0720843669e4e9db", "parents": [ "ca0e05e4b15193aeba72b995e90de990db7f8304" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Sun Aug 01 23:14:20 2010 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Sun Aug 01 23:14:20 2010 -0400" }, "message": "ext4: Add mount options in superblock\n\nAllow mount options to be stored in the superblock. Also add default\nmount option bits for nobarrier, block_validity, discard, and nodelalloc.\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "79e8303677fc15f508b9877e0fea1925c4add6f3", "tree": "85891ab05b2a9e5dd47966fd16bc372df6d4bb88", "parents": [ "62d2b5f2dcd3707b070efb16bbfdf6947c38c194" ], "author": { "name": "Eric Sandeen", "email": "sandeen@redhat.com", "time": "Tue Jul 27 11:56:07 2010 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Tue Jul 27 11:56:07 2010 -0400" }, "message": "ext4: fix ext4_get_blocks references\n\next4_get_blocks got renamed to ext4_map_blocks, but left stale\ncomments and a prototype littered around.\n\nSigned-off-by: Eric Sandeen \u003csandeen@redhat.com\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "5b3ff237bef43b9e7fb7d1eb858e29b73fd664f9", "tree": "e3810d974a0e51a8c4b9046abdd36a5a4022abb3", "parents": [ "552ef8024f909d9b3a7442d0ab0d48a22de24e9e" ], "author": { "name": "jiayingz@google.com (Jiaying Zhang)", "email": "", "time": "Tue Jul 27 11:56:06 2010 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Tue Jul 27 11:56:06 2010 -0400" }, "message": "ext4: move aio completion after unwritten extent conversion\n\nThis patch is to be applied upon Christoph\u0027s \"direct-io: move aio_complete\ninto -\u003eend_io\" patch. It adds iocb and result fields to struct ext4_io_end_t,\nso that we can call aio_complete from ext4_end_io_nolock() after the extent\nconversion has finished.\n\nI have verified with Christoph\u0027s aio-dio test that used to fail after a few\nruns on an original kernel but now succeeds on the patched kernel.\n\nSee http://thread.gmane.org/gmane.comp.file-systems.ext4/19659 for details.\n\nSigned-off-by: Jiaying Zhang \u003cjiayingz@google.com\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "89eeddf03327e19cfcbb18efa98e5470e2f5c563", "tree": "2ca9f7ed13b348677a2d4b2fe8bf40e041a72f5f", "parents": [ "66e61a9e9504f61b9a928c9055368c81da613a50" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Tue Jul 27 11:56:04 2010 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Tue Jul 27 11:56:04 2010 -0400" }, "message": "ext4: Define s_jnl_backup_type in superblock\n\nThis has been in use by e2fsprogs for a while; define it to keep the\nsuper block fields in sync.\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "66e61a9e9504f61b9a928c9055368c81da613a50", "tree": "afcd8147b4e4771103f75f5a03b836950882dece", "parents": [ "1c13d5c0872870cca3e612aa045d492ead9ab004" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Tue Jul 27 11:56:04 2010 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Tue Jul 27 11:56:04 2010 -0400" }, "message": "ext4: Once a day, printk file system error information to dmesg\n\nThis allows us to grab any file system error messages by scraping\n/var/log/messages. This will make it easy for us to do error analysis\nacross the very large number of machines as we deploy ext4 across the\nfleet.\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "1c13d5c0872870cca3e612aa045d492ead9ab004", "tree": "6e3dd0d3f49ff56dda9fc6cd72c233759bc24e09", "parents": [ "c398eda0e43a791be0fca6f197a1e2bbb9f16070" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Tue Jul 27 11:56:03 2010 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Tue Jul 27 11:56:03 2010 -0400" }, "message": "ext4: Save error information to the superblock for analysis\n\nSave number of file system errors, and the time function name, line\nnumber, block number, and inode number of the first and most recent\nerrors reported on the file system in the superblock.\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "c398eda0e43a791be0fca6f197a1e2bbb9f16070", "tree": "d7b2b95490f96a75e116a3aa13c17767aa630342", "parents": [ "60fd4da34d55a9cc0d857fc76dc12cf8cab4ed02" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Tue Jul 27 11:56:40 2010 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Tue Jul 27 11:56:40 2010 -0400" }, "message": "ext4: Pass line numbers to ext4_error() and friends\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "60fd4da34d55a9cc0d857fc76dc12cf8cab4ed02", "tree": "72c6b3d9c7680070a751da56bd9068f71e7b106f", "parents": [ "90c7201b97bb7ac5a4e2605abc0efb5fdfb957f0" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Tue Jul 27 11:54:40 2010 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Tue Jul 27 11:54:40 2010 -0400" }, "message": "ext4: Cleanup ext4_check_dir_entry so __func__ is now implicit\n \nAlso start passing the line number to ext4_check_dir since we\u0027re going\nto need it in upcoming patch.\n \nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "e29136f80e775b0310273932b4297a62f5574a29", "tree": "d10d56b15b7968744ea0becdf61440aae7e609c5", "parents": [ "c67d859e39896e4286249da89c4ca0ef8bd949cb" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Tue Jun 29 12:54:28 2010 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Tue Jun 29 12:54:28 2010 -0400" }, "message": "ext4: Enhance ext4_grp_locked_error() to take block and function numbers\n\nAlso use a macro definition so that __func__ and __LINE__ is implicit.\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "c67d859e39896e4286249da89c4ca0ef8bd949cb", "tree": "e483a8e61fe876506d8ed819fbfe29eac6c493a7", "parents": [ "4a9cdec73f79b2858e9ecf0b6cfac7f6b200bf3a" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Tue Jun 29 11:07:07 2010 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Tue Jun 29 11:07:07 2010 -0400" }, "message": "ext4: clean up ext4_abort() so __func__ is now implicit\n\nUse a macro definition for ext4_abort() to clean up the .c files a wee\nbit.\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "4a9cdec73f79b2858e9ecf0b6cfac7f6b200bf3a", "tree": "0e06b201e009e5e2a661811e61ef3184138bb8d3", "parents": [ "c6ac12a6159c802ae8b757dd13563564e64333df" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Tue Jun 29 11:00:23 2010 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Tue Jun 29 11:00:23 2010 -0400" }, "message": "ext4: Add new superblock fields reserved for the Next3 snapshot feature\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "206f7ab4f49a2021fcb8687f25395be77711ddee", "tree": "19764d36a25ec1e088d32572a1c0936eeaf19636", "parents": [ "5a0790c2c4a18435759a70e1562450035d778339" ], "author": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Mon Jun 14 14:42:49 2010 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Mon Jun 14 14:42:49 2010 -0400" }, "message": "ext4: remove vestiges of nobh support\n\nThe nobh option was only supported for writeback mode, but given that all\nwrite paths actually create buffer heads it effectively was a no-op already.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "a0375156ca1041574b5d47cc7e32f10b891151b0", "tree": "b791961012b9348f289c3dda5dc071bdf9b736d2", "parents": [ "7e27d6e778cd87b6f2415515d7127eba53fe5d02" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Fri Jun 11 23:14:04 2010 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Fri Jun 11 23:14:04 2010 -0400" }, "message": "ext4: Clean up s_dirt handling\n\nWe don\u0027t need to set s_dirt in most of the ext4 code when journaling\nis enabled. In ext3/4 some of the summary statistics for # of free\ninodes, blocks, and directories are calculated from the per-block\ngroup statistics when the file system is mounted or unmounted. As a\nresult the superblock doesn\u0027t have to be updated, either via the\njournal or by setting s_dirt. There are a few exceptions, most\nnotably when resizing the file system, where the superblock needs to\nbe modified --- and in that case it should be done as a journalled\noperation if possible, and s_dirt set only in no-journal mode.\n\nThis patch will optimize out some unneeded disk writes when using ext4\nwith a journal.\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "7ea8085910ef3dd4f3cad6845aaa2b580d39b115", "tree": "d9c1edb5906f943f7d70bfb4b65106e29772d379", "parents": [ "cc967be54710d97c05229b2e5ba2d00df84ddd64" ], "author": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Wed May 26 17:53:25 2010 +0200" }, "committer": { "name": "Al Viro", "email": "viro@zeniv.linux.org.uk", "time": "Thu May 27 22:05:02 2010 -0400" }, "message": "drop unused dentry argument to -\u003efsync\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Al Viro \u003cviro@zeniv.linux.org.uk\u003e\n" }, { "commit": "14ece1028b3ed53ffec1b1213ffc6acaf79ad77c", "tree": "8f5a13487139c68ef775c9d39b7be83386fdae06", "parents": [ "60e6679e28518ccd67169c4a539d8cc7490eb8a6" ], "author": { "name": "Frank Mayhar", "email": "fmayhar@google.com", "time": "Mon May 17 08:00:00 2010 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Mon May 17 08:00:00 2010 -0400" }, "message": "ext4: Make fsync sync new parent directories in no-journal mode\n\nAdd a new ext4 state to tell us when a file has been newly created; use\nthat state in ext4_sync_file in no-journal mode to tell us when we need\nto sync the parent directory as well as the inode and data itself. This\nfixes a problem in which a panic or power failure may lose the entire\nfile even when using fsync, since the parent directory entry is lost.\n\nAddresses-Google-Bug: #2480057\n\nSigned-off-by: Frank Mayhar \u003cfmayhar@google.com\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "60e6679e28518ccd67169c4a539d8cc7490eb8a6", "tree": "65f39d1e9da3006df5bc5532d0c8a7ec365599f0", "parents": [ "4d92dc0f00a775dc2e1267b0e00befb783902fe7" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Mon May 17 07:00:00 2010 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Mon May 17 07:00:00 2010 -0400" }, "message": "ext4: Drop whitespace at end of lines\n\nThis patch was generated using:\n\n#!/usr/bin/perl -i\nwhile (\u003c\u003e) {\n s/[ \t]+$//;\n print;\n}\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "4d92dc0f00a775dc2e1267b0e00befb783902fe7", "tree": "07bb835871dfcc290f5909fe6635ab420ff61420", "parents": [ "899ad0cea6ad7ff4ba24b16318edbc3cbbe03fad" ], "author": { "name": "Ben Hutchings", "email": "ben@decadent.org.uk", "time": "Mon May 17 06:00:00 2010 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Mon May 17 06:00:00 2010 -0400" }, "message": "ext4: Fix compat EXT4_IOC_ADD_GROUP\n\nstruct ext4_new_group_input needs to be converted because u64 has\nonly 32-bit alignment on some 32-bit architectures, notably i386.\n\nSigned-off-by: Ben Hutchings \u003cben@decadent.org.uk\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "899ad0cea6ad7ff4ba24b16318edbc3cbbe03fad", "tree": "5a9b5b84cf47321fbf1ac10bde64bbf3ff75a269", "parents": [ "f084db932e6fe877bf8362bc256fc850de196deb" ], "author": { "name": "Ben Hutchings", "email": "ben@decadent.org.uk", "time": "Mon May 17 05:00:00 2010 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Mon May 17 05:00:00 2010 -0400" }, "message": "ext4: Conditionally define compat ioctl numbers\n\nIt is unnecessary, and in general impossible, to define the compat\nioctl numbers except when building the filesystem with CONFIG_COMPAT\ndefined.\n\nSigned-off-by: Ben Hutchings \u003cben@decadent.org.uk\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "12e9b892002d9af057655d35b44db8ee9243b0dc", "tree": "c5831b4bcf98eebdd39158d08dab97c198f5c683", "parents": [ "24676da469f50f433baa347845639662c561d1f6" ], "author": { "name": "Dmitry Monakhov", "email": "dmonakhov@openvz.org", "time": "Sun May 16 22:00:00 2010 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Sun May 16 22:00:00 2010 -0400" }, "message": "ext4: Use bitops to read/modify i_flags in struct ext4_inode_info\n\nAt several places we modify EXT4_I(inode)-\u003ei_flags without holding\ni_mutex (ext4_do_update_inode, ...). These modifications are racy and\nwe can lose updates to i_flags. So convert handling of i_flags to use\nbitops which are atomic.\n\nhttps://bugzilla.kernel.org/show_bug.cgi?id\u003d15792\n\nSigned-off-by: Dmitry Monakhov \u003cdmonakhov@openvz.org\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "24676da469f50f433baa347845639662c561d1f6", "tree": "b4b8205f8b50376af286193d0dcbe76f2bc2d1e1", "parents": [ "2ed886852adfcb070bf350e66a0da0d98b2f3ab5" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Sun May 16 21:00:00 2010 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Sun May 16 21:00:00 2010 -0400" }, "message": "ext4: Convert calls of ext4_error() to EXT4_ERROR_INODE()\n\nEXT4_ERROR_INODE() tends to provide better error information and in a\nmore consistent format. Some errors were not even identifying the inode\nor directory which was corrupted, which made them not very useful.\n\nAddresses-Google-Bug: #2507977\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "e35fd6609b2fee54484d520deccb8f18bf7d38f3", "tree": "9b786445602819074f599c282b31bead166e8c03", "parents": [ "8e48dcfbd7c0892b4cfd064d682cc4c95a29df32" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Sun May 16 19:00:00 2010 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Sun May 16 19:00:00 2010 -0400" }, "message": "ext4: Add new abstraction ext4_map_blocks() underneath ext4_get_blocks()\n\nJack up ext4_get_blocks() and add a new function, ext4_map_blocks()\nwhich uses a much smaller structure, struct ext4_map_blocks which is\n20 bytes, as opposed to a struct buffer_head, which nearly 5 times\nbigger on an x86_64 machine. By switching things to use\next4_map_blocks(), we can save stack space by using ext4_map_blocks()\nsince we can avoid allocating a struct buffer_head on the stack.\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "8a57d9d61a6e361c7bb159dda797672c1df1a691", "tree": "39a01022ed2294f0acc94b45554c9a292db671dc", "parents": [ "6d19c42b7cf81c39632b6d4dbc514e8449bcd346" ], "author": { "name": "Curt Wohlgemuth", "email": "curtw@google.com", "time": "Sun May 16 15:00:00 2010 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Sun May 16 15:00:00 2010 -0400" }, "message": "ext4: check for a good block group before loading buddy pages\n\nThis adds a new field in ext4_group_info to cache the largest available\nblock range in a block group; and don\u0027t load the buddy pages until *after*\nwe\u0027ve done a sanity check on the block group.\n\nWith large allocation requests (e.g., fallocate(), 8MiB) and relatively full\npartitions, it\u0027s easy to have no block groups with a block extent large\nenough to satisfy the input request length. This currently causes the loop\nduring cr \u003d\u003d 0 in ext4_mb_regular_allocator() to load the buddy bitmap pages\nfor EVERY block group. That can be a lot of pages. The patch below allows\nus to call ext4_mb_good_group() BEFORE we load the buddy pages (although we\nhave check again after we lock the block group).\n\nAddresses-Google-Bug: #2578108\nAddresses-Google-Bug: #2704453\n\nSigned-off-by: Curt Wohlgemuth \u003ccurtw@google.com\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "9467c4fdd66f6810cecef0f1173330f3c6e67d45", "tree": "5fea180a10127c893b288dff2c8788b72d2eaea3", "parents": [ "35c2e967d067ff02dc944f2434f024419c2fe83a", "a9185b41a4f84971b930c519f0c63bd450c4810d" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 05 11:53:53 2010 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 05 11:53:53 2010 -0800" }, "message": "Merge branch \u0027write_inode2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs-2.6\n\n* \u0027write_inode2\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs-2.6:\n pass writeback_control to -\u003ewrite_inode\n make sure data is on disk before calling -\u003ewrite_inode\n" }, { "commit": "1f63b9c15b17d5af360c180f5c71537e954d5d3b", "tree": "4da17d6f2035ae093680fa2caa3e1c84b44bb237", "parents": [ "b24bc1e61cec2174faf5dfa632da16b6ca17144f", "64e290ec69be39f1887fa0b403c1e417b6b038e7" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 05 10:47:00 2010 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Mar 05 10:47:00 2010 -0800" }, "message": "Merge branch \u0027for_linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4\n\n* \u0027for_linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4: (36 commits)\n ext4: fix up rb_root initializations to use RB_ROOT\n ext4: Code cleanup for EXT4_IOC_MOVE_EXT ioctl\n ext4: Fix the NULL reference in double_down_write_data_sem()\n ext4: Fix insertion point of extent in mext_insert_across_blocks()\n ext4: consolidate in_range() definitions\n ext4: cleanup to use ext4_grp_offs_to_block()\n ext4: cleanup to use ext4_group_first_block_no()\n ext4: Release page references acquired in ext4_da_block_invalidatepages\n ext4: Fix ext4_quota_write cross block boundary behaviour\n ext4: Convert BUG_ON checks to use ext4_error() instead\n ext4: Use direct_IO_no_locking in ext4 dio read\n ext4: use ext4_get_block_write in buffer write\n ext4: mechanical rename some of the direct I/O get_block\u0027s identifiers\n ext4: make \"offset\" consistent in ext4_check_dir_entry()\n ext4: Handle non empty on-disk orphan link\n ext4: explicitly remove inode from orphan list after failed direct io\n ext4: fix error handling in migrate\n ext4: deprecate obsoleted mount options\n ext4: Fix fencepost error in chosing choosing group vs file preallocation.\n jbd2: clean up an assertion in jbd2_journal_commit_transaction()\n ...\n" }, { "commit": "a9185b41a4f84971b930c519f0c63bd450c4810d", "tree": "268cf4e206cca12fb9e1dd68984e7c190e465b46", "parents": [ "26821ed40b4230259e770c9911180f38fcaa6f59" ], "author": { "name": "Christoph Hellwig", "email": "hch@lst.de", "time": "Fri Mar 05 09:21:37 2010 +0100" }, "committer": { "name": "Al Viro", "email": "viro@zeniv.linux.org.uk", "time": "Fri Mar 05 13:25:52 2010 -0500" }, "message": "pass writeback_control to -\u003ewrite_inode\n\nThis gives the filesystem more information about the writeback that\nis happening. Trond requested this for the NFS unstable write handling,\nand other filesystems might benefit from this too by beeing able to\ndistinguish between the different callers in more detail.\n\nSigned-off-by: Christoph Hellwig \u003chch@lst.de\u003e\nSigned-off-by: Al Viro \u003cviro@zeniv.linux.org.uk\u003e\n" }, { "commit": "731eb1a03a8445cde2cb23ecfb3580c6fa7bb690", "tree": "9978388978604207cc78da1c81b97625857256cb", "parents": [ "bda00de7e8569b1fcde27b68fa59e74e14c5f93a" ], "author": { "name": "Akinobu Mita", "email": "akinobu.mita@gmail.com", "time": "Wed Mar 03 23:55:01 2010 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Wed Mar 03 23:55:01 2010 -0500" }, "message": "ext4: consolidate in_range() definitions\n\nThere are duplicate macro definitions of in_range() in mballoc.h and\nballoc.c. This consolidates these two definitions into ext4.h, and\nchanges extents.c to use in_range() as well.\n\nSigned-off-by: Akinobu Mita \u003cakinobu.mita@gmail.com\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\nCc: Andreas Dilger \u003cadilger@sun.com\u003e\n" }, { "commit": "273df556b6ee2065bfe96edab5888d3dc9b108d8", "tree": "19c73685fce581e4ed85ff845e0b2fc485cedf9c", "parents": [ "b7adc1f363e72e9131a582cc2cb00eaf83f51a39" ], "author": { "name": "Frank Mayhar", "email": "fmayhar@google.com", "time": "Tue Mar 02 11:46:09 2010 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Tue Mar 02 11:46:09 2010 -0500" }, "message": "ext4: Convert BUG_ON checks to use ext4_error() instead\n\nConvert a bunch of BUG_ONs to emit a ext4_error() message and return\nEIO. This is a first pass and most notably does _not_ cover\nmballoc.c, which is a morass of void functions.\n\nSigned-off-by: Frank Mayhar \u003cfmayhar@google.com\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "744692dc059845b2a3022119871846e74d4f6e11", "tree": "ed246651aebcb8dae57de8c58dc20983064ee017", "parents": [ "c7064ef13b2181a489836349f9baf87df0dab28f" ], "author": { "name": "Jiaying Zhang", "email": "jiayingz@google.com", "time": "Thu Mar 04 16:14:02 2010 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Thu Mar 04 16:14:02 2010 -0500" }, "message": "ext4: use ext4_get_block_write in buffer write\n\nAllocate uninitialized extent before ext4 buffer write and\nconvert the extent to initialized after io completes.\nThe purpose is to make sure an extent can only be marked\ninitialized after it has been written with new data so\nwe can safely drop the i_mutex lock in ext4 DIO read without\nexposing stale data. This helps to improve multi-thread DIO\nread performance on high-speed disks.\n\nSkip the nobh and data\u003djournal mount cases to make things simple for now.\n\nSigned-off-by: Jiaying Zhang \u003cjiayingz@google.com\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "c7064ef13b2181a489836349f9baf87df0dab28f", "tree": "433d86d9ed4285e5c5e7f24fbbaa7f48fad09381", "parents": [ "b8b8afe236e97b6359d46d3a3f8c46455e192271" ], "author": { "name": "Jiaying Zhang", "email": "jiayingz@google.com", "time": "Tue Mar 02 13:28:44 2010 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Tue Mar 02 13:28:44 2010 -0500" }, "message": "ext4: mechanical rename some of the direct I/O get_block\u0027s identifiers\n\nThis commit renames some of the direct I/O\u0027s block allocation flags,\nvariables, and functions introduced in Mingming\u0027s \"Direct IO for holes\nand fallocate\" patches so that they can be used by ext4\u0027s buffered\nwrite path as well. Also changed the related function comments\naccordingly to cover both direct write and buffered write cases.\n\nSigned-off-by: Jiaying Zhang \u003cjiayingz@google.com\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "c8d46e41bc744c8fa0092112af3942fcd46c8b18", "tree": "fd8bad55c3c9903b6c798b838396bc832bbf7f4b", "parents": [ "73b50c1c92666d326b5fa2c945d46509f2f6d91f" ], "author": { "name": "Jiaying Zhang", "email": "jiayingz@google.com", "time": "Wed Feb 24 09:52:53 2010 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Wed Feb 24 09:52:53 2010 -0500" }, "message": "ext4: Add flag to files with blocks intentionally past EOF\n\nfallocate() may potentially instantiate blocks past EOF, depending\non the flags used when it is called.\n\ne2fsck currently has a test for blocks past i_size, and it\nsometimes trips up - noticeably on xfstests 013 which runs fsstress.\n\nThis patch from Jiayang does fix it up - it (along with\ne2fsprogs updates and other patches recently from Aneesh) has\nsurvived many fsstress runs in a row.\n\nSigned-off-by: Eric Sandeen \u003csandeen@redhat.com\u003e\nSigned-off-by: Jiaying Zhang \u003cjiayingz@google.com\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "003cb608a2533d0927a83bc4e07e46d7a622eda9", "tree": "19347ffcec1c09cda9c4fd01addd8797cf8e31c4", "parents": [ "43cf38eb5cea91245502df3fcee4dbfc1c74dd1c" ], "author": { "name": "Tejun Heo", "email": "tj@kernel.org", "time": "Tue Feb 02 14:39:01 2010 +0900" }, "committer": { "name": "Tejun Heo", "email": "tj@kernel.org", "time": "Wed Feb 17 11:17:38 2010 +0900" }, "message": "percpu: add __percpu sparse annotations to fs\n\nAdd __percpu sparse annotations to fs.\n\nThese annotations are to make sparse consider percpu variables to be\nin a different address space and warn if accessed without going\nthrough percpu accessors. This patch doesn\u0027t affect normal builds.\n\nSigned-off-by: Tejun Heo \u003ctj@kernel.org\u003e\nCc: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\nCc: Trond Myklebust \u003cTrond.Myklebust@netapp.com\u003e\nCc: Alex Elder \u003caelder@sgi.com\u003e\nCc: Christoph Hellwig \u003chch@infradead.org\u003e\nCc: Alexander Viro \u003cviro@zeniv.linux.org.uk\u003e\n" }, { "commit": "12062dddda450976b129dcb1bacd91acaf4d8030", "tree": "e64590b1147639cd3629f8a977b269410cd6bd13", "parents": [ "f710b4b96ba292dfed2153afc47e9063b0abfd89" ], "author": { "name": "Eric Sandeen", "email": "sandeen@redhat.com", "time": "Mon Feb 15 14:19:27 2010 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Mon Feb 15 14:19:27 2010 -0500" }, "message": "ext4: move __func__ into a macro for ext4_warning, ext4_error\n\nJust a pet peeve of mine; we had a mishash of calls with either __func__\nor \"function_name\" and the latter tends to get out of sync.\n\nI think it\u0027s easier to just hide the __func__ in a macro, and it\u0027ll\nbe consistent from then on.\n\nSigned-off-by: Eric Sandeen \u003csandeen@redhat.com\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "f710b4b96ba292dfed2153afc47e9063b0abfd89", "tree": "fdfdc979eb03c41c5785b4153e6337d61a4c89e2", "parents": [ "19f5fb7ad679bb361222c7916086435020c37cce" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Mon Jan 25 03:31:32 2010 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Mon Jan 25 03:31:32 2010 -0500" }, "message": "ext4: Reserve INCOMPAT_EA_INODE and INCOMPAT_DIRDATA feature codepoints\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "19f5fb7ad679bb361222c7916086435020c37cce", "tree": "9e301163075c4faaf340cf50afd51855c76acd8c", "parents": [ "d2eecb03936878ec574ade5532fa83df7d75dde7" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Sun Jan 24 14:34:07 2010 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Sun Jan 24 14:34:07 2010 -0500" }, "message": "ext4: Use bitops to read/modify EXT4_I(inode)-\u003ei_state\n\nAt several places we modify EXT4_I(inode)-\u003ei_state without holding\ni_mutex (ext4_release_file, ext4_bmap, ext4_journalled_writepage,\next4_do_update_inode, ...). These modifications are racy and we can\nlose updates to i_state. So convert handling of i_state to use bitops\nwhich are atomic.\n\nCc: Jan Kara \u003cjack@suse.cz\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "1296cc85c26e94eb865d03f82140f27d598de467", "tree": "b8a2bc06ecd992a86179ebcd9c087994c676c291", "parents": [ "5f634d064c709ea02c3cdaa850a08323a4a4bf28" ], "author": { "name": "Aneesh Kumar K.V", "email": "aneesh.kumar@linux.vnet.ibm.com", "time": "Fri Jan 15 01:27:59 2010 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Fri Jan 15 01:27:59 2010 -0500" }, "message": "ext4: Drop EXT4_GET_BLOCKS_UPDATE_RESERVE_SPACE flag\n\nWe should update reserve space if it is delalloc buffer\nand that is indicated by EXT4_GET_BLOCKS_DELALLOC_RESERVE flag.\nSo use EXT4_GET_BLOCKS_DELALLOC_RESERVE in place of\nEXT4_GET_BLOCKS_UPDATE_RESERVE_SPACE\n\nSigned-off-by: Aneesh Kumar K.V \u003caneesh.kumar@linux.vnet.ibm.com\u003e\n" }, { "commit": "5f634d064c709ea02c3cdaa850a08323a4a4bf28", "tree": "3f81e9c56cd8348b7bb94f1d54efff696374c929", "parents": [ "1db913823c0f8360fccbd24ca67eb073966a5ffd" ], "author": { "name": "Aneesh Kumar K.V", "email": "aneesh.kumar@linux.vnet.ibm.com", "time": "Mon Jan 25 04:00:31 2010 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Mon Jan 25 04:00:31 2010 -0500" }, "message": "ext4: Fix quota accounting error with fallocate\n\nWhen we fallocate a region of the file which we had recently written,\nand which is still in the page cache marked as delayed allocated blocks\nwe need to make sure we don\u0027t do the quota update on writepage path.\nThis is because the needed quota updated would have already be done\nby fallocate.\n\nSigned-off-by: Aneesh Kumar K.V \u003caneesh.kumar@linux.vnet.ibm.com\u003e\n" }, { "commit": "9d0be50230b333005635967f7ecd4897dbfd181b", "tree": "59aefe29e33284e1d904b23eaf2cc98994431374", "parents": [ "ee5f4d9cdf32fd99172d11665c592a288c2b1ff4" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Fri Jan 01 02:41:30 2010 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Fri Jan 01 02:41:30 2010 -0500" }, "message": "ext4: Calculate metadata requirements more accurately\n\nIn the past, ext4_calc_metadata_amount(), and its sub-functions\next4_ext_calc_metadata_amount() and ext4_indirect_calc_metadata_amount()\nbadly over-estimated the number of metadata blocks that might be\nrequired for delayed allocation blocks. This didn\u0027t matter as much\nwhen functions which managed the reserved metadata blocks were more\naggressive about dropping reserved metadata blocks as delayed\nallocation blocks were written, but unfortunately they were too\naggressive. This was fixed in commit 0637c6f, but as a result the\nover-estimation by ext4_calc_metadata_amount() would lead to reserving\n2-3 times the number of pending delayed allocation blocks as\npotentially required metadata blocks. So if there are 1 megabytes of\nblocks which have been not yet been allocation, up to 3 megabytes of\nspace would get reserved out of the user\u0027s quota and from the file\nsystem free space pool until all of the inode\u0027s data blocks have been\nallocated.\n\nThis commit addresses this problem by much more accurately estimating\nthe number of metadata blocks that will be required. It will still\nsomewhat over-estimate the number of blocks needed, since it must make\na worst case estimate not knowing which physical blocks will be\nneeded, but it is much more accurate than before.\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "a9e7f4472075fb6937c545af3f6329e9946bbe66", "tree": "399e294982e1c9691332eca72942910a7f74e9f8", "parents": [ "fd8fbfc1709822bd94247c5b2ab15a5f5041e103" ], "author": { "name": "Dmitry Monakhov", "email": "dmonakhov@openvz.org", "time": "Mon Dec 14 15:21:14 2009 +0300" }, "committer": { "name": "Jan Kara", "email": "jack@suse.cz", "time": "Wed Dec 23 13:33:55 2009 +0100" }, "message": "ext4: Convert to generic reserved quota\u0027s space management.\n\nThis patch also fixes write vs chown race condition.\n\nAcked-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\nSigned-off-by: Dmitry Monakhov \u003cdmonakhov@openvz.org\u003e\nSigned-off-by: Jan Kara \u003cjack@suse.cz\u003e\n" }, { "commit": "1f2acb6017d8528135ec3b01ab7cd2be6ea0630b", "tree": "7263a5c1402237166cf425bc8f5336426159622c", "parents": [ "15121c18a22ae483279f76dc9e554334b800d0f7" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Fri Jan 22 17:40:42 2010 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Fri Jan 22 17:40:42 2010 -0500" }, "message": "ext4: Add block validity check when truncating indirect block mapped inodes\n\nAdd checks to ext4_free_branches() to make sure a block number found\nin an indirect block are valid before trying to free it. If a bad\nblock number is found, stop freeing the indirect block immediately,\nsince the file system is corrupt and we will need to run fsck anyway.\nThis also avoids spamming the logs, and specifically avoids\ndriver-level \"attempt to access beyond end of device\" errors obscure\nwhat is really going on.\n\nIf you get *really*, *really*, *really* unlucky, without this patch, a\nsupposed indirect block containing garbage might contain a reference\nto a primary block group descriptor, in which case\next4_free_branches() could end up zero\u0027ing out a block group\ndescriptor block, and if then one of the block bitmaps for a block\ngroup described by that bg descriptor block is not in memory, and is\nread in by ext4_read_block_bitmap(). This function calls\next4_valid_block_bitmap(), which assumes that bg_inode_table() was\nvalidated at mount time and hasn\u0027t been modified since. Since this\nassumption is no longer valid, it\u0027s possible for the value\n(ext4_inode_table(sb, desc) - group_first_block) to go negative, which\nwill cause ext4_find_next_zero_bit() to trigger a kernel GPF.\n\nAddresses-Google-Bug: #2220436\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "a1de02dccf906faba2ee2d99cac56799bda3b96a", "tree": "9871a11252e9a7d39882206c2057ec3080517365", "parents": [ "724e6d3fe8003c3f60bf404bf22e4e331327c596" ], "author": { "name": "Eric Sandeen", "email": "sandeen@redhat.com", "time": "Thu Feb 04 23:58:38 2010 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Thu Feb 04 23:58:38 2010 -0500" }, "message": "ext4: fix async i/o writes beyond 4GB to a sparse file\n\nThe \"offset\" member in ext4_io_end holds bytes, not blocks, so\next4_lblk_t is wrong - and too small (u32).\n\nThis caused the async i/o writes to sparse files beyond 4GB to fail\nwhen they wrapped around to 0.\n\nAlso fix up the type of arguments to ext4_convert_unwritten_extents(),\nit gets ssize_t from ext4_end_aio_dio_nolock() and\next4_ext_direct_IO().\n\nReported-by: Giel de Nijs \u003cgiel@vectorwise.com\u003e\nSigned-off-by: Eric Sandeen \u003csandeen@redhat.com\u003e\n" }, { "commit": "b436b9bef84de6893e86346d8fbf7104bc520645", "tree": "50fb9ae167bcd622e9adf47646bcf3b4c7dd111d", "parents": [ "194074acacebc169ded90a4657193f5180015051" ], "author": { "name": "Jan Kara", "email": "jack@suse.cz", "time": "Tue Dec 08 23:51:10 2009 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Tue Dec 08 23:51:10 2009 -0500" }, "message": "ext4: Wait for proper transaction commit on fsync\n\nWe cannot rely on buffer dirty bits during fsync because pdflush can come\nbefore fsync is called and clear dirty bits without forcing a transaction\ncommit. What we do is that we track which transaction has last changed\nthe inode and which transaction last changed allocation and force it to\ndisk on fsync.\n\nSigned-off-by: Jan Kara \u003cjack@suse.cz\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "e6362609b6c71c5b802026be9cf263bbdd67a50e", "tree": "be908b6b0566f70d31378bf11b548a0d59ae0218", "parents": [ "4433871130f36585fde38e7dd817433296648945" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Mon Nov 23 07:17:05 2009 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Mon Nov 23 07:17:05 2009 -0500" }, "message": "ext4: call ext4_forget() from ext4_free_blocks()\n\nAdd the facility for ext4_forget() to be called from\next4_free_blocks(). This simplifies the code in a large number of\nplaces, and centralizes most of the work of calling ext4_forget() into\na single place.\n\nAlso fix a bug in the extents migration code; it wasn\u0027t calling\next4_forget() when releasing the indirect blocks during the\nconversion. As a result, if the system cashed during or shortly after\nthe extents migration, and the released indirect blocks get reused as\ndata blocks, the journal replay would corrupt the data blocks. With\nthis new patch, fixing this bug was as simple as adding the\nEXT4_FREE_BLOCKS_FORGET flags to the call to ext4_free_blocks().\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\nCc: \"Aneesh Kumar K.V\" \u003caneesh.kumar@linux.vnet.ibm.com\u003e\n" }, { "commit": "4433871130f36585fde38e7dd817433296648945", "tree": "5dd91fbb345f5a4240f4d2df271da49ca1a6037b", "parents": [ "b7e57e7c2a41826e51fe060fae5158bfc7a04e81" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Sun Nov 22 07:44:56 2009 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Sun Nov 22 07:44:56 2009 -0500" }, "message": "ext4: fold ext4_free_blocks() and ext4_mb_free_blocks()\n\next4_mb_free_blocks() is only called by ext4_free_blocks(), and the\nlatter function doesn\u0027t really do much. So merge the two functions\ntogether, such that ext4_free_blocks() is now found in\nfs/ext4/mballoc.c. This saves about 200 bytes of compiled text space.\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "d6797d14b1640d088652c72508b529a3aea479e3", "tree": "2e608c3e362b94439a2e7503b7d97cb3891bb101", "parents": [ "e3bb52ae2bb9573e84c17b8e3560378d13a5c798" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Sun Nov 22 20:52:12 2009 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Sun Nov 22 20:52:12 2009 -0500" }, "message": "ext4: move ext4_forget() to ext4_jbd2.c\n\nThe ext4_forget() function better belongs in ext4_jbd2.c. This will\nallow us to do some cleanup of the ext4_journal_revoke() and\next4_journal_forget() functions, as well as giving us better error\nreporting since we can report the caller of ext4_forget() when things\ngo wrong.\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "5328e635315734d42080de9a5a1ee87bf4cae0a4", "tree": "4952c04c9b1faec9a3c82a3743805efad05ed96e", "parents": [ "2bba702d4f88d7b010ec37e2527b552588404ae7" ], "author": { "name": "Eric Sandeen", "email": "sandeen@redhat.com", "time": "Thu Nov 19 14:25:42 2009 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Thu Nov 19 14:25:42 2009 -0500" }, "message": "ext4: make trim/discard optional (and off by default)\n\nIt is anticipated that when sb_issue_discard starts doing\nreal work on trim-capable devices, we may see issues. Make\nthis mount-time optional, and default it to off until we know\nthat things are working out OK.\n\nSigned-off-by: Eric Sandeen \u003csandeen@redhat.com\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "5f5249507e4b5c4fc0f9c93f33d133d8c95f47e1", "tree": "0b0a790f568c07298c3a4122572e84392c787648", "parents": [ "109f55651954def97fa41ee71c464d268c512ab0" ], "author": { "name": "Mingming", "email": "cmm@us.ibm.com", "time": "Tue Nov 10 10:48:04 2009 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Tue Nov 10 10:48:04 2009 -0500" }, "message": "ext4: skip conversion of uninit extents after direct IO if there isn\u0027t any\n\nAt the end of direct I/O operation, ext4_ext_direct_IO() always called\next4_convert_unwritten_extents(), regardless of whether there were any\nunwritten extents involved in the I/O or not.\n\nThis commit adds a state flag so that ext4_ext_direct_IO() only calls\next4_convert_unwritten_extents() when necessary.\n\nSigned-off-by: Mingming Cao \u003ccmm@us.ibm.com\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "d4da6c9ccf648f3f1cb5bf9d981a62c253d30e28", "tree": "709f8bdc50a3a1d47632047eb3670b4a4a6ff045", "parents": [ "c35102c3e15f90fe604523a2fbffd9dc158b455a" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Mon Nov 02 10:15:27 2009 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Mon Nov 02 10:15:27 2009 -0800" }, "message": "Revert \"ext4: Remove journal_checksum mount option and enable it by default\"\n\nThis reverts commit d0646f7b636d067d715fab52a2ba9c6f0f46b0d7, as\nrequested by Eric Sandeen.\n\nIt can basically cause an ext4 filesystem to miss recovery (and thus get\nmounted with errors) if the journal checksum does not match.\n\nQuoth Eric:\n\n \"My hand-wavy hunch about what is happening is that we\u0027re finding a\n bad checksum on the last partially-written transaction, which is\n not surprising, but if we have a wrapped log and we\u0027re doing the\n initial scan for head/tail, and we abort scanning on that bad\n checksum, then we are essentially running an unrecovered filesystem.\n\n But that\u0027s hand-wavy and I need to go look at the code.\n\n We lived without journal checksums on by default until now, and at\n this point they\u0027re doing more harm than good, so we should revert\n the default-changing commit until we can fix it and do some good\n power-fail testing with the fixes in place.\"\n\nSee\n\n\thttp://bugzilla.kernel.org/show_bug.cgi?id\u003d14354\n\nfor all the gory details.\n\nRequested-by: Eric Sandeen \u003csandeen@redhat.com\u003e\nCc: Theodore Tso \u003ctytso@mit.edu\u003e\nCc: Alexey Fisher \u003cbug-track@fisher-privat.net\u003e\nCc: Maxim Levitsky \u003cmaximlevitsky@gmail.com\u003e\nCc: Aneesh Kumar K.V \u003caneesh.kumar@linux.vnet.ibm.com\u003e\nCc: Mathias Burén \u003cmathias.buren@gmail.com\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "c1fccc0696bcaff6008c11865091f5ec4b0937ab", "tree": "0e8269e07d40278c136b7f93aba7ab118f75f532", "parents": [ "0ef122494020521309be855bfdeeb41f34bf8c94" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Wed Sep 30 01:13:55 2009 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Wed Sep 30 01:13:55 2009 -0400" }, "message": "ext4: Fix time encoding with extra epoch bits\n\n\"Looking at ext4.h, I think the setting of extra time fields forgets to\nmask the epoch bits so the epoch part overwrites nsec part. The second\nchange is only for coherency (2 -\u003e EXT4_EPOCH_BITS).\"\n\nThanks to Damien Guibouret for pointing out this problem.\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "296c355cd6443d89fa251885a8d78778fe111dc4", "tree": "5cf7c8b115617dc3829a16a5969894d37b73173c", "parents": [ "90576c0b9a0b5323fc4bd7f23f49be0d234f36d1" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Wed Sep 30 00:32:42 2009 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Wed Sep 30 00:32:42 2009 -0400" }, "message": "ext4: Use tracepoints for mb_history trace file\n\nThe /proc/fs/ext4/\u003cdev\u003e/mb_history was maintained manually, and had a\nnumber of problems: it required a largish amount of memory to be\nallocated for each ext4 filesystem, and the s_mb_history_lock\nintroduced a CPU contention problem. \n\nBy ripping out the mb_history code and replacing it with ftrace\ntracepoints, and we get more functionality: timestamps, event\nfiltering, the ability to correlate mballoc history with other ext4\ntracepoints, etc.\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "8d5d02e6b176565c77ff03604908b1453a22044d", "tree": "0d29e4f28233f24960c7921c1c0a7608077bf713", "parents": [ "4c0425ff68b1b87b802ffeda7b6a46ff7da7241c" ], "author": { "name": "Mingming Cao", "email": "cmm@us.ibm.com", "time": "Mon Sep 28 15:48:29 2009 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Mon Sep 28 15:48:29 2009 -0400" }, "message": "ext4: async direct IO for holes and fallocate support\n\nFor async direct IO that covers holes or fallocate, the end_io\ncallback function now queued the convertion work on workqueue but\ndon\u0027t flush the work rightaway as it might take too long to afford.\n\nBut when fsync is called after all the data is completed, user expects\nthe metadata also being updated before fsync returns.\n\nThus we need to flush the conversion work when fsync() is called.\nThis patch keep track of a listed of completed async direct io that\nhas a work queued on workqueue. When fsync() is called, it will go\nthrough the list and do the conversion.\n\nSigned-off-by: Mingming Cao \u003ccmm@us.ibm.com\u003e\n" }, { "commit": "4c0425ff68b1b87b802ffeda7b6a46ff7da7241c", "tree": "a8718f5f4574af8e15fd876b24f4aec88d62451b", "parents": [ "0031462b5b392f90d17f1d75abb795883c44e969" ], "author": { "name": "Mingming Cao", "email": "cmm@us.ibm.com", "time": "Mon Sep 28 15:48:41 2009 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Mon Sep 28 15:48:41 2009 -0400" }, "message": "ext4: Use end_io callback to avoid direct I/O fallback to buffered I/O\n\nCurrently the DIO VFS code passes create \u003d 0 when writing to the\nmiddle of file. It does this to avoid block allocation for holes, so\nas not to expose stale data out when there is a parallel buffered read\n(which does not hold the i_mutex lock). Direct I/O writes into holes\nfalls back to buffered IO for this reason.\n\nSince preallocated extents are treated as holes when doing a\nget_block() look up (buffer is not mapped), direct IO over fallocate\nalso falls back to buffered IO. Thus ext4 actually silently falls\nback to buffered IO in above two cases, which is undesirable.\n\nTo fix this, this patch creates unitialized extents when a direct I/O\nwrite into holes in sparse files, and registering an end_io callback which\nconverts the uninitialized extent to an initialized extent after the\nI/O is completed.\n\nSinged-Off-By: Mingming Cao \u003ccmm@us.ibm.com\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "0031462b5b392f90d17f1d75abb795883c44e969", "tree": "e8323861b8dede0f3ddbfc8324d650bf1f4fd74b", "parents": [ "9f0ccfd8e07d61b413e6536ffa02fbf60d2e20d8" ], "author": { "name": "Mingming Cao", "email": "cmm@us.ibm.com", "time": "Mon Sep 28 15:49:08 2009 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Mon Sep 28 15:49:08 2009 -0400" }, "message": "ext4: Split uninitialized extents for direct I/O\n\nWhen writing into an unitialized extent via direct I/O, and the direct\nI/O doesn\u0027t exactly cover the unitialized extent, split the extent\ninto uninitialized and initialized extents before submitting the I/O.\nThis avoids needing to deal with an ENOSPC error in the end_io\ncallback that gets used for direct I/O.\n\nWhen the IO is complete, the written extent will be marked as initialized.\n\nSinged-Off-By: Mingming Cao \u003ccmm@us.ibm.com\u003e \nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "55138e0bc29c0751e2152df9ad35deea542f29b3", "tree": "40dbc77de3e7a426030bd740ba8e2fd0d2171523", "parents": [ "71780577306fd1e76c7a92e3b308db624d03adb9" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Tue Sep 29 13:31:31 2009 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Tue Sep 29 13:31:31 2009 -0400" }, "message": "ext4: Adjust ext4_da_writepages() to write out larger contiguous chunks\n\nWork around problems in the writeback code to force out writebacks in\nlarger chunks than just 4mb, which is just too small. This also works\naround limitations in the ext4 block allocator, which can\u0027t allocate\nmore than 2048 blocks at a time. So we need to defeat the round-robin\ncharacteristics of the writeback code and try to write out as many\nblocks in one inode before allowing the writeback code to move on to\nanother inode. We add a a new per-filesystem tunable,\nmax_writeback_mb_bump, which caps this to a default of 128mb per\ninode.\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "0a80e9867db154966b2a771042e10452ac110e1e", "tree": "51f43830b3c0dd733b0d446399eb3ea2b6bc84ed", "parents": [ "5534fb5bb35a62a94e0bd1fa2421f7fb6e894f10" ], "author": { "name": "Eric Sandeen", "email": "sandeen@redhat.com", "time": "Thu Sep 17 11:55:58 2009 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Thu Sep 17 11:55:58 2009 -0400" }, "message": "ext4: replace MAX_DEFRAG_SIZE with EXT_MAX_BLOCK\n\nThere\u0027s no reason to redefine the maximum allowable offset\nin an extent-based file just for defrag; \nEXT_MAX_BLOCK already does this.\n\nSigned-off-by: Eric Sandeen \u003csandeen@redhat.com\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "1b9c12f44c1eb614fd3b8822bfe8f1f5d8e53737", "tree": "5de0737dd71557dadaa5e3f98d6e62bd3b9d7c51", "parents": [ "fb0a387dcdcd21aab1b09ee7fd80b7c979bdbbfd" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Thu Sep 17 08:32:22 2009 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Thu Sep 17 08:32:22 2009 -0400" }, "message": "ext4: store EXT4_EXT_MIGRATE in i_state instead of i_flags\n\nEXT4_EXT_MIGRATE is only intended to be used for an in-memory flag,\nand the hex value assigned to it collides with FS_DIRECTIO_FL (which\nis also stored in i_flags). There\u0027s no reason for the\nEXT4_EXT_MIGRATE bit to be stored in i_flags, so we switch it to use\ni_state instead.\n\nCc: \"Aneesh Kumar K.V\" \u003caneesh.kumar@linux.vnet.ibm.com\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "fb0a387dcdcd21aab1b09ee7fd80b7c979bdbbfd", "tree": "dcb12c5fce8f7ccb8b183936ea71a29aba3f3846", "parents": [ "c40ce3c9ea97425a12d7e44031a98fe50add6fc1" ], "author": { "name": "Eric Sandeen", "email": "sandeen@redhat.com", "time": "Wed Sep 16 14:45:10 2009 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Wed Sep 16 14:45:10 2009 -0400" }, "message": "ext4: limit block allocations for indirect-block files to \u003c 2^32\n\nToday, the ext4 allocator will happily allocate blocks past\n2^32 for indirect-block files, which results in the block\nnumbers getting truncated, and corruption ensues.\n\nThis patch limits such allocations to \u003c 2^32, and adds\nBUG_ONs if we do get blocks larger than that.\n\nThis should address RH Bug 519471, ext4 bitmap allocator \nmust limit blocks to \u003c 2^32\n\n* ext4_find_goal() is modified to choose a goal \u003c UINT_MAX,\n so that our starting point is in an acceptable range.\n\n* ext4_xattr_block_set() is modified such that the goal block\n is \u003c UINT_MAX, as above.\n\n* ext4_mb_regular_allocator() is modified so that the group\n search does not continue into groups which are too high\n\n* ext4_mb_use_preallocated() has a check that we don\u0027t use\n preallocated space which is too far out\n\n* ext4_alloc_blocks() and ext4_xattr_block_set() add some BUG_ONs\n\nNo attempt has been made to limit inode locations to \u003c 2^32,\nso we may wind up with blocks far from their inodes. Doing\nthis much already will lead to some odd ENOSPC issues when the\n\"lower 32\" gets full, and further restricting inodes could\nmake that even weirder.\n\nFor high inodes, choosing a goal of the original, % UINT_MAX,\nmay be a bit odd, but then we\u0027re in an odd situation anyway,\nand I don\u0027t know of a better heuristic.\n\nSigned-off-by: Eric Sandeen \u003csandeen@redhat.com\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "d0646f7b636d067d715fab52a2ba9c6f0f46b0d7", "tree": "f7b3ae9510e8c789651d99fee0c22867e6ccba94", "parents": [ "a3710fd1ee8cd542c5de63cf2c39f8912031f867" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Sat Sep 05 12:50:43 2009 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Sat Sep 05 12:50:43 2009 -0400" }, "message": "ext4: Remove journal_checksum mount option and enable it by default\n\nThere\u0027s no real cost for the journal checksum feature, and we should\nmake sure it is enabled all the time.\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n\n\n" }, { "commit": "b3a3ca8ca0c3c29abc5b2bfe94bb14f3f4590df9", "tree": "1006f780d43ec8ff90ed11b1eb0d5ed961c9f613", "parents": [ "de89de6e0cf4b1eb13f27137cf2aa40d287aabdf" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Mon Aug 31 23:13:11 2009 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Mon Aug 31 23:13:11 2009 -0400" }, "message": "ext4: Add new tracepoint: trace_ext4_da_write_pages()\n\nAdd a new tracepoint which shows the pages that will be written using\nwrite_cache_pages() by ext4_da_writepages().\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "a36b44988cef1fc007535107013571fa691a2d7f", "tree": "b9d659b8ac2a1068f6fbc08ac093902333512c70", "parents": [ "1927805e6599d8602d2c0af6a0155c85acc0b214" ], "author": { "name": "Eric Sandeen", "email": "sandeen@redhat.com", "time": "Tue Aug 25 22:36:45 2009 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Tue Aug 25 22:36:45 2009 -0400" }, "message": "ext4: use ext4_grpblk_t more extensively\n\nunsigned short is potentially too small to track blocks within\na group; today it is safe due to restrictions in e2fsprogs but\nwe have _lo / _hi bits for group blocks with the intent to go\nup to 32 bits, so clean this up now.\n\nThere are many more places where we use unsigned/int/unsigned int\nto contain a group block but this should at least fix all the\nshort types.\n\nI added a few comments to the struct ext4_group_info definition\nas well.\n\nSigned-off-by: Eric Sandeen \u003csandeen@redhat.com\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "0373130d5bc783751c1fbad948886916a21d4559", "tree": "a5b4d21b84345354ff965714bd6a6c473bae4b0f", "parents": [ "bf43d84b185e2ff54598f8c58a5a8e63148b6e90" ], "author": { "name": "Eric Sandeen", "email": "sandeen@redhat.com", "time": "Mon Aug 17 23:51:29 2009 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Mon Aug 17 23:51:29 2009 -0400" }, "message": "ext4: open-code ext4_mb_update_group_info\n\next4_mb_update_group_info is only called in one place, and it\u0027s\nextremely simple. There\u0027s no reason to have it in a separate function\nin a separate file as far as I can tell, it just obfuscates what\u0027s\nreally going on.\n\nPerhaps it was intended to keep the grp-\u003ebb_* manipulation local to\nmballoc.c but we\u0027re already accessing other grp-\u003e fields in balloc.c\ndirectly so this seems ok.\n\nSigned-off-by: Eric Sandeen \u003csandeen@redhat.com\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "487caeef9fc08c0565e082c40a8aaf58dad92bbb", "tree": "69920293cfe3a50bdbbf845be785350e7c203a2b", "parents": [ "9599b0e597d810be9b8f759ea6e9619c4f983c5e" ], "author": { "name": "Jan Kara", "email": "jack@suse.cz", "time": "Mon Aug 17 22:17:20 2009 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Mon Aug 17 22:17:20 2009 -0400" }, "message": "ext4: Fix possible deadlock between ext4_truncate() and ext4_get_blocks()\n\nDuring truncate we are sometimes forced to start a new transaction as\nthe amount of blocks to be journaled is both quite large and hard to\npredict. So far we restarted a transaction while holding i_data_sem\nand that violates lock ordering because i_data_sem ranks below a\ntransaction start (and it can lead to a real deadlock with\next4_get_blocks() mapping blocks in some page while having a\ntransaction open).\n\nWe fix the problem by dropping the i_data_sem before restarting the\ntransaction and acquire it afterwards. It\u0027s slightly subtle that this\nworks:\n\n1) By the time ext4_truncate() is called, all the page cache for the\ntruncated part of the file is dropped so get_block() should not be\ncalled on it (we only have to invalidate extent cache after we\nreacquire i_data_sem because some extent from not-truncated part could\nextend also into the part we are going to truncate).\n\n2) Writes, migrate or defrag hold i_mutex so they are stopped for all\nthe time of the truncate.\n\nThis bug has been found and analyzed by Theodore Tso \u003ctytso@mit.edu\u003e.\n\nSigned-off-by: Jan Kara \u003cjack@suse.cz\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "50797481a7bdee548589506d7d7b48b08bc14dcd", "tree": "19989d27e3a69c6c2c507f798f55a2d9a47a5d27", "parents": [ "4ba74d00a20256e22f159cb288ff34b587608917" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Fri Sep 18 13:34:02 2009 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Fri Sep 18 13:34:02 2009 -0400" }, "message": "ext4: Avoid group preallocation for closed files\n\nCurrently the group preallocation code tries to find a large (512)\nfree block from which to do per-cpu group allocation for small files.\nThe problem with this scheme is that it leaves the filesystem horribly\nfragmented. In the worst case, if the filesystem is unmounted and\nremounted (after a system shutdown, for example) we forget the fact\nthat wee were using a particular (now-partially filled) 512 block\nextent. So the next time we try to allocate space for a small file,\nwe will find *another* completely free 512 block chunk to allocate\nsmall files. Given that there are 32,768 blocks in a block group,\nafter 64 iterations of \"mount, write one 4k file in a directory,\nunmount\", the block group will have 64 files, each separated by 511\nblocks, and the block group will no longer have any free 512\ncompletely free chunks of blocks for group preallocation space.\n\nSo if we try to allocate blocks for a file that has been closed, such\nthat we know the final size of the file, and the filesystem is not\nbusy, avoid using group preallocation.\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "4ba74d00a20256e22f159cb288ff34b587608917", "tree": "6859cf49f8043ec11846e2c8ebc836499e1048c4", "parents": [ "0ef90db93a4ddfc300af288c2a1bfc1e6c79da64" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Sun Aug 09 22:01:13 2009 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Sun Aug 09 22:01:13 2009 -0400" }, "message": "ext4: Fix bugs in mballoc\u0027s stream allocation mode\n\nThe logic around sbi-\u003es_mb_last_group and sbi-\u003es_mb_last_start was all\nscrewed up. These fields were getting unconditionally all the time,\nset even when stream allocation had not taken place, and if they were\nbeing used when the file was smaller than s_mb_stream_request, which\nis when the allocation should _not_ be doing stream allocation.\n\nFix this by determining whether or not we stream allocation should\ntake place once, in ext4_mb_group_or_file(), and setting a flag which\ngets used in ext4_mb_regular_allocator() and ext4_mb_use_best_found().\nThis simplifies the code and assures that we are consistently using\n(or not using) the stream allocation logic.\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "0ef90db93a4ddfc300af288c2a1bfc1e6c79da64", "tree": "9a16d71efd6458777391c993f3b964cb5cebfb7d", "parents": [ "6ba495e9259cd9a0b40ebd6c315143535c92542f" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Sun Aug 09 16:46:13 2009 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Sun Aug 09 16:46:13 2009 -0400" }, "message": "ext4: Display the mballoc flags in mb_history in hex instead of decimal\n\nDisplaying the flags in base 16 makes it easier to see which flags\nhave been set.\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "726447d803802cd0be8f62d17c4a34421781b938", "tree": "ef1d7d4b9afb0f81b9eb6f09980c152d35ceb3cf", "parents": [ "089ceecc1ea4a69ed8bcc5c7c7b96ce487e26b33" ], "author": { "name": "Eric Sandeen", "email": "sandeen@redhat.com", "time": "Mon Jul 13 10:24:17 2009 -0400" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Mon Jul 13 10:24:17 2009 -0400" }, "message": "ext4: naturally align struct ext4_allocation_request\n\nAs Ted noted, the ext4_allocation_request isn\u0027t well aligned. Looking\nat it with pahole we\u0027re wasting space on 64-bit arches:\n\nstruct ext4_allocation_request {\n struct inode * inode; /* 0 8 */\n ext4_lblk_t logical; /* 8 4 */\n\n /* XXX 4 bytes hole, try to pack */\n\n ext4_fsblk_t goal; /* 16 8 */\n ext4_lblk_t lleft; /* 24 4 */\n\n /* XXX 4 bytes hole, try to pack */\n\n ext4_fsblk_t pleft; /* 32 8 */\n ext4_lblk_t lright; /* 40 4 */\n\n /* XXX 4 bytes hole, try to pack */\n\n ext4_fsblk_t pright; /* 48 8 */\n unsigned int len; /* 56 4 */\n unsigned int flags; /* 60 4 */\n /* --- cacheline 1 boundary (64 bytes) --- */\n\n /* size: 64, cachelines: 1, members: 9 */\n /* sum members: 52, holes: 3, sum holes: 12 */\n};\n\nGrouping 32-bit members together closes these holes and shrinks the\nstructure by 12 bytes. which is important since ext4 can get on the\nhairy edge of stack overruns.\n\nSigned-off-by: Eric Sandeen \u003csandeen@redhat.com\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" } ], "next": "d4bfe2f76d785cc77611a4bda8cedaff358d8c7d" }