)]}' { "log": [ { "commit": "abdb318b79d387a723af5db2aa79f812cefd0797", "tree": "7915b48a984df976ed639b7b0a99672e7dc049c6", "parents": [ "3534573b58fd7576d3dc8dd66a9973592ac08b2d" ], "author": { "name": "Jiro SEKIBA", "email": "jir@unicus.jp", "time": "Fri Nov 27 19:41:14 2009 +0900" }, "committer": { "name": "Ryusuke Konishi", "email": "konishi.ryusuke@lab.ntt.co.jp", "time": "Fri Nov 27 20:05:16 2009 +0900" }, "message": "nilfs2: replace mark_inode_dirty as nilfs_mark_inode_dirty\n\nReplace mark_inode_dirty() as nilfs_mark_inode_dirty()\nto reduce deep function calls.\n\nSigned-off-by: Jiro SEKIBA \u003cjir@unicus.jp\u003e\nSigned-off-by: Ryusuke Konishi \u003ckonishi.ryusuke@lab.ntt.co.jp\u003e\n" }, { "commit": "9ca941d4b62e72571948efe5a73c563b4cacc98d", "tree": "8c4a5cde6c2e626758c0aed85fe9cd7759747890", "parents": [ "0234576d041b9b2cc7043691ea61d2c2ca597aaa" ], "author": { "name": "Jiro SEKIBA", "email": "jir@unicus.jp", "time": "Fri Nov 27 19:41:06 2009 +0900" }, "committer": { "name": "Ryusuke Konishi", "email": "konishi.ryusuke@lab.ntt.co.jp", "time": "Fri Nov 27 20:05:15 2009 +0900" }, "message": "nilfs2: delete mark_inode_dirty in nilfs_new_inode\n\nIt is redundant to call mark_inode_dirty() in nilfs_new_inode() because\nall caller of nilfs_new_inode() will call mark_inode_dirty()\nafter calling nilfs_new_inode() directly or indirectly in transaction.\n\nSigned-off-by: Jiro SEKIBA \u003cjir@unicus.jp\u003e\nSigned-off-by: Ryusuke Konishi \u003ckonishi.ryusuke@lab.ntt.co.jp\u003e\n" }, { "commit": "9cb4e0d2b99e8b0e5e269d898ae6ab1967647c5a", "tree": "09ac1daa4985b4717afc818f89cfe2439664b44a", "parents": [ "09bf4aae0a3c471b721c43e7bdb6132200d907b2" ], "author": { "name": "Ryusuke Konishi", "email": "konishi.ryusuke@lab.ntt.co.jp", "time": "Fri Nov 06 01:00:48 2009 +0900" }, "committer": { "name": "Ryusuke Konishi", "email": "konishi.ryusuke@lab.ntt.co.jp", "time": "Fri Nov 20 10:05:47 2009 +0900" }, "message": "nilfs2: move out mark_inode_dirty calls from bmap routines\n\nPreviously, nilfs_bmap_add_blocks() and nilfs_bmap_sub_blocks() called\nmark_inode_dirty() after they changed the number of data blocks.\n\nThis moves these calls outside bmap outermost functions like\nnilfs_bmap_insert() or nilfs_bmap_truncate().\n\nThis will mitigate overhead for truncate or delete operation since\nthey repeatedly remove set of blocks. Nearly 10 percent improvement\nwas observed for removal of a large file:\n\n # dd if\u003d/dev/zero of\u003d/test/aaa bs\u003d1M count\u003d512\n # time rm /test/aaa\n\n real 2.968s -\u003e 2.705s\n\nFurther optimization may be possible by eliminating these\nmark_inode_dirty() uses though I avoid mixing separate changes here.\n\nSigned-off-by: Ryusuke Konishi \u003ckonishi.ryusuke@lab.ntt.co.jp\u003e\n" }, { "commit": "a49762fd119d191dcbb2f638a2dbc2ed53f4e2bb", "tree": "5cd999854d4b156c7bcd901694998e60b479f6e8", "parents": [ "e2073e78575e3690ea3cce67b11b7b1de8e85fd3" ], "author": { "name": "Ryusuke Konishi", "email": "konishi.ryusuke@lab.ntt.co.jp", "time": "Wed Nov 11 01:28:48 2009 +0900" }, "committer": { "name": "Ryusuke Konishi", "email": "konishi.ryusuke@lab.ntt.co.jp", "time": "Fri Nov 20 10:05:47 2009 +0900" }, "message": "nilfs2: remove buffer locking in nilfs_mark_inode_dirty\n\nThis lock is eliminable because inodes on the buffer can be updated\nindependently. Although a log writer also fills in bmap data on the\non-disk inodes, this update is exclusively done by a log writer lock.\n\nSigned-off-by: Ryusuke Konishi \u003ckonishi.ryusuke@lab.ntt.co.jp\u003e\n" }, { "commit": "18dafac1a4c6c88867a50f9a82492976f20383d6", "tree": "27800ccab2a98cccc903ed68348de1abc9b18a7d", "parents": [ "1a7af63108f07637504300671a72432c34e10021" ], "author": { "name": "Jiro SEKIBA", "email": "jir@unicus.jp", "time": "Sun Nov 15 13:49:45 2009 +0900" }, "committer": { "name": "Ryusuke Konishi", "email": "konishi.ryusuke@lab.ntt.co.jp", "time": "Sun Nov 15 17:17:46 2009 +0900" }, "message": "nilfs2: deleted inconsistent comment in nilfs_load_inode_block()\n\nThe comment says, \"Caller of this function MUST lock s_inode_lock\",\nhowever just above the comment, it locks s_inode_lock in the function.\n\nSigned-off-by: Jiro SEKIBA \u003cjir@unicus.jp\u003e\nSigned-off-by: Ryusuke Konishi \u003ckonishi.ryusuke@lab.ntt.co.jp\u003e\n" }, { "commit": "3cc811bffdf35ebaf1467fbec71a49b57800fc74", "tree": "b00abe8ad3687faeacf9b543492f0b91c7abc600", "parents": [ "1f28fcd925b2b3157411bbd08f0024b55b70d8dd" ], "author": { "name": "Ryusuke Konishi", "email": "konishi.ryusuke@lab.ntt.co.jp", "time": "Mon Sep 28 13:02:46 2009 +0900" }, "committer": { "name": "Ryusuke Konishi", "email": "konishi.ryusuke@lab.ntt.co.jp", "time": "Tue Sep 29 20:32:13 2009 +0900" }, "message": "nilfs2: fix missing initialization of i_dir_start_lookup member\n\nThe i_dir_start_lookup field in nilfs_inode_info objects should be\ncleared when the objects are allocated, but the the initialization was\nmissing in case of reading from disk. This adds the initialization.\n\nSince the variable just gives a start page on directory lookups, the\nbug was nonfatal until now.\n\nSigned-off-by: Ryusuke Konishi \u003ckonishi.ryusuke@lab.ntt.co.jp\u003e\n" }, { "commit": "7f09410bbc4306f592cfb43812389ea1c7905a20", "tree": "18f179435f70c4ec9231883501062d5ea0357af5", "parents": [ "ac4cfdd6d141c319a7af8655f750ed504c187a74" ], "author": { "name": "Alexey Dobriyan", "email": "adobriyan@gmail.com", "time": "Mon Sep 21 17:01:10 2009 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Sep 22 07:17:24 2009 -0700" }, "message": "const: mark remaining address_space_operations const\n\nSigned-off-by: Alexey Dobriyan \u003cadobriyan@gmail.com\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "1b2f5a641bca91966fd91dc489238068a0ab4b69", "tree": "85ab9020e9628d43ffabc6ab354efcf6ea1ef3d2", "parents": [ "74fca6a42863ffacaf7ba6f1936a9f228950f657" ], "author": { "name": "Ryusuke Konishi", "email": "konishi.ryusuke@lab.ntt.co.jp", "time": "Sat Aug 22 19:10:07 2009 +0900" }, "committer": { "name": "Ryusuke Konishi", "email": "konishi.ryusuke@lab.ntt.co.jp", "time": "Mon Sep 14 18:27:12 2009 +0900" }, "message": "nilfs2: fix ignored error code in __nilfs_read_inode()\n\nThe __nilfs_read_inode function is ignoring the error code returned\nfrom nilfs_read_inode_common(), and wrongly delivers a success code\n(zero) when it escapes from the function in erroneous cases.\n\nThis adds the missing error handling.\n\nSigned-off-by: Ryusuke Konishi \u003ckonishi.ryusuke@lab.ntt.co.jp\u003e\n" }, { "commit": "d441b1c293149212045de00f346c8ea6cd41cce4", "tree": "cbc7aa520feeabaa90ff4b8ff83e6e8fade216c7", "parents": [ "5affd88a104af43f0063a12ad1ee4c7a587945dc" ], "author": { "name": "Al Viro", "email": "viro@zeniv.linux.org.uk", "time": "Mon Jun 08 19:56:34 2009 -0400" }, "committer": { "name": "Al Viro", "email": "viro@zeniv.linux.org.uk", "time": "Wed Jun 24 08:17:05 2009 -0400" }, "message": "switch nilfs2 to inode-\u003ei_acl\n\nActually, get rid of private analog, since nothing in there is\nusing ACLs at all so far.\n\nSigned-off-by: Al Viro \u003cviro@zeniv.linux.org.uk\u003e\n" }, { "commit": "c3a7abf06ce719a51139e62a034590be99abbc2c", "tree": "14d61bbd8c34d1b1c7997c9afad79158b7af2914", "parents": [ "fa032744ad41de1b0a1807e7c379c6196e72ad80" ], "author": { "name": "Ryusuke Konishi", "email": "konishi.ryusuke@lab.ntt.co.jp", "time": "Mon May 25 02:47:14 2009 +0900" }, "committer": { "name": "Ryusuke Konishi", "email": "konishi.ryusuke@lab.ntt.co.jp", "time": "Wed Jun 10 23:41:12 2009 +0900" }, "message": "nilfs2: support contiguous lookup of blocks\n\nAlthough get_block() callback function can return extent of contiguous\nblocks with bh-\u003eb_size, nilfs_get_block() function did not support\nthis feature.\n\nThis adds contiguous lookup feature to the block mapping codes of\nnilfs, and allows the nilfs_get_blocks() function to return the extent\ninformation by applying the feature.\n\nSigned-off-by: Ryusuke Konishi \u003ckonishi.ryusuke@lab.ntt.co.jp\u003e\n" }, { "commit": "e85dc1d5295779d2e6d973b7ed524d6bd3825a3f", "tree": "8f13c049839deb9f6cde4deb440b2a3758f10f64", "parents": [ "30bda0b8ae470997dc681d6fd82108bac5caebf8" ], "author": { "name": "Ryusuke Konishi", "email": "konishi.ryusuke@lab.ntt.co.jp", "time": "Wed May 27 07:27:12 2009 +0900" }, "committer": { "name": "Ryusuke Konishi", "email": "konishi.ryusuke@lab.ntt.co.jp", "time": "Wed Jun 10 23:41:11 2009 +0900" }, "message": "nilfs2: enable sync_page method\n\nThis adds a missing sync_page method which unplugs bio requests when\nwaiting for page locks. This will improve read performance of nilfs.\n\nHere is a measurement result using dd command.\n\nWithout this patch:\n\n # mount -t nilfs2 /dev/sde1 /test\n # dd if\u003d/test/aaa of\u003d/dev/null bs\u003d512k\n 1024+0 records in\n 1024+0 records out\n 536870912 bytes (537 MB) copied, 6.00688 seconds, 89.4 MB/s\n\nWith this patch:\n\n # mount -t nilfs2 /dev/sde1 /test\n # dd if\u003d/test/aaa of\u003d/dev/null bs\u003d512k\n 1024+0 records in\n 1024+0 records out\n 536870912 bytes (537 MB) copied, 3.54998 seconds, 151 MB/s\n\nSigned-off-by: Ryusuke Konishi \u003ckonishi.ryusuke@lab.ntt.co.jp\u003e\n" }, { "commit": "258ef67e246fd548e7ad91c23004ad157c03cce5", "tree": "a956d6acdd0c5e8c29f1a0993ef8a426ca834523", "parents": [ "7cde31d7d6959b2c608aa6b200eb68892d3a6063" ], "author": { "name": "Hisashi Hifumi", "email": "hifumi.hisashi@oss.ntt.co.jp", "time": "Wed May 13 11:19:40 2009 +0900" }, "committer": { "name": "Ryusuke Konishi", "email": "konishi.ryusuke@lab.ntt.co.jp", "time": "Wed Jun 10 23:41:11 2009 +0900" }, "message": "NILFS2: Pagecache usage optimization on NILFS2\n\nHi,\n\nI introduced \"is_partially_uptodate\" aops for NILFS2.\n\nA page can have multiple buffers and even if a page is not uptodate, some buffers\ncan be uptodate on pagesize !\u003d blocksize environment.\nThis aops checks that all buffers which correspond to a part of a file\nthat we want to read are uptodate. If so, we do not have to issue actual\nread IO to HDD even if a page is not uptodate because the portion we\nwant to read are uptodate.\n\"block_is_partially_uptodate\" function is already used by ext2/3/4.\nWith the following patch random read/write mixed workloads or random read after\nrandom write workloads can be optimized and we can get performance improvement.\n\nI did a performance test using the sysbench.\n\n1 --file-block-size\u003d8K --file-total-size\u003d2G --file-test-mode\u003drndrw --file-fsync-freq\u003d0 --fil\ne-rw-ratio\u003d1 run\n\n-2.6.30-rc5\n\nTest execution summary:\n total time: 151.2907s\n total number of events: 200000\n total time taken by event execution: 2409.8387\n per-request statistics:\n min: 0.0000s\n avg: 0.0120s\n max: 0.9306s\n approx. 95 percentile: 0.0439s\n\nThreads fairness:\n events (avg/stddev): 12500.0000/238.52\n execution time (avg/stddev): 150.6149/0.01\n\n-2.6.30-rc5-patched\n\nTest execution summary:\n total time: 140.8828s\n total number of events: 200000\n total time taken by event execution: 2240.8577\n per-request statistics:\n min: 0.0000s\n avg: 0.0112s\n max: 0.8750s\n approx. 95 percentile: 0.0418s\n\nThreads fairness:\n events (avg/stddev): 12500.0000/218.43\n execution time (avg/stddev): 140.0536/0.01\n\narch: ia64\npagesize: 16k\n\nThanks.\n\nSigned-off-by: Hisashi Hifumi \u003chifumi.hisashi@oss.ntt.co.jp\u003e\nSigned-off-by: Ryusuke Konishi \u003ckonishi.ryusuke@lab.ntt.co.jp\u003e\n" }, { "commit": "612392307cb09e49051225092cbbd7049bd8db93", "tree": "401a227d0fae219aae7b682bb613bb388b4a7682", "parents": [ "e339ad31f59925b48a92ee3947692fdf9758b8c7" ], "author": { "name": "Ryusuke Konishi", "email": "konishi.ryusuke@lab.ntt.co.jp", "time": "Mon Apr 06 19:02:00 2009 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Apr 07 08:31:20 2009 -0700" }, "message": "nilfs2: support nanosecond timestamp\n\nAfter a review of user\u0027s feedback for finding out other compatibility\nissues, I found nilfs improperly initializes timestamps in inode;\nCURRENT_TIME was used there instead of CURRENT_TIME_SEC even though nilfs\ndidn\u0027t have nanosecond timestamps on disk. A few users gave us the report\nthat the tar program sometimes failed to expand symbolic links on nilfs,\nand it turned out to be the cause.\n\nInstead of applying the above displacement, I\u0027ve decided to support\nnanosecond timestamps on this occation. Fortunetaly, a needless 64-bit\nfield was in the nilfs_inode struct, and I found it\u0027s available for this\npurpose without impact for the users.\n\nSo, this will do the enhancement and resolve the tar problem.\n\nSigned-off-by: Ryusuke Konishi \u003ckonishi.ryusuke@lab.ntt.co.jp\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "458c5b0822a669d170fdb7bb16c9145f652ebe06", "tree": "f2c53998492306a8aaf6d20ded229b69cba17348", "parents": [ "e62687468588f47c32256e3c8c36157c40111b6e" ], "author": { "name": "Ryusuke Konishi", "email": "konishi.ryusuke@lab.ntt.co.jp", "time": "Mon Apr 06 19:01:56 2009 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Apr 07 08:31:19 2009 -0700" }, "message": "nilfs2: clean up sketch file\n\nThe sketch file is a file to mark checkpoints with user data. It was\nexperimentally introduced in the original implementation, and now\nobsolete. The file was handled differently with regular files; the file\nsize got truncated when a checkpoint was created.\n\nThis stops the special treatment and will treat it as a regular file.\nMost users are not affected because mkfs.nilfs2 no longer makes this file.\n\nSigned-off-by: Ryusuke Konishi \u003ckonishi.ryusuke@lab.ntt.co.jp\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "1f5abe7e7dbcd83e73212c6cb135a6106cea6a0b", "tree": "f80e97297d5badebd31bbb17003d76a4ea30453a", "parents": [ "2c2e52fc4fca251e68f90821c9ff5cb18be4df58" ], "author": { "name": "Ryusuke Konishi", "email": "konishi.ryusuke@lab.ntt.co.jp", "time": "Mon Apr 06 19:01:55 2009 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Apr 07 08:31:19 2009 -0700" }, "message": "nilfs2: replace BUG_ON and BUG calls triggerable from ioctl\n\nPekka Enberg advised me:\n\u003e It would be nice if BUG(), BUG_ON(), and panic() calls would be\n\u003e converted to proper error handling using WARN_ON() calls. The BUG()\n\u003e call in nilfs_cpfile_delete_checkpoints(), for example, looks to be\n\u003e triggerable from user-space via the ioctl() system call.\n\nThis will follow the comment and keep them to a minimum.\n\nAcked-by: Pekka Enberg \u003cpenberg@cs.helsinki.fi\u003e\nSigned-off-by: Ryusuke Konishi \u003ckonishi.ryusuke@lab.ntt.co.jp\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "47420c799830d4676e544dbec56b2a7f787528f5", "tree": "dd61f6c96942b07f762129c893d9cbbbeff60735", "parents": [ "a2e7d2df82cafb76f76809ddf6e2caa8afe4f75e" ], "author": { "name": "Ryusuke Konishi", "email": "konishi.ryusuke@lab.ntt.co.jp", "time": "Mon Apr 06 19:01:45 2009 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Apr 07 08:31:17 2009 -0700" }, "message": "nilfs2: avoid double error caused by nilfs_transaction_end\n\nPekka Enberg pointed out that double error handlings found after\nnilfs_transaction_end() can be avoided by separating abort operation:\n\n OK, I don\u0027t understand this. The only way nilfs_transaction_end() can\n fail is if we have NILFS_TI_SYNC set and we fail to construct the\n segment. But why do we want to construct a segment if we don\u0027t commit?\n\n I guess what I\u0027m asking is why don\u0027t we have a separate\n nilfs_transaction_abort() function that can\u0027t fail for the erroneous\n case to avoid this double error value tracking thing?\n\nThis does the separation and renames nilfs_transaction_end() to\nnilfs_transaction_commit() for clarification.\n\nSince, some calls of these functions were used just for exclusion control\nagainst the segment constructor, they are replaced with semaphore\noperations.\n\nAcked-by: Pekka Enberg \u003cpenberg@cs.helsinki.fi\u003e\nSigned-off-by: Ryusuke Konishi \u003ckonishi.ryusuke@lab.ntt.co.jp\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "f30bf3e40f80ef50c17f55271deae3abc03e793e", "tree": "2e6d69611d1443688c1a79c5e5aed60cf38672e7", "parents": [ "9ff05123e3bfbb1d2b68ba1d9bf1f7d1dffc1453" ], "author": { "name": "Ryusuke Konishi", "email": "konishi.ryusuke@lab.ntt.co.jp", "time": "Mon Apr 06 19:01:38 2009 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Apr 07 08:31:15 2009 -0700" }, "message": "nilfs2: fix missed-sync issue for do_sync_mapping_range()\n\nChris Mason pointed out that there is a missed sync issue in\nnilfs_writepages():\n\nOn Wed, 17 Dec 2008 21:52:55 -0500, Chris Mason wrote:\n\u003e It looks like nilfs_writepage ignores WB_SYNC_NONE, which is used by\n\u003e do_sync_mapping_range().\n\nwhere WB_SYNC_NONE in do_sync_mapping_range() was replaced with\nWB_SYNC_ALL by Nick\u0027s patch (commit:\nee53a891f47444c53318b98dac947ede963db400).\n\nThis fixes the problem by letting nilfs_writepages() write out the log of\nfile data within the range if sync_mode is WB_SYNC_ALL.\n\nThis involves removal of nilfs_file_aio_write() which was previously\nneeded to ensure O_SYNC sync writes.\n\nCc: Chris Mason \u003cchris.mason@oracle.com\u003e\nSigned-off-by: Ryusuke Konishi \u003ckonishi.ryusuke@lab.ntt.co.jp\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "05fe58fdc10df9ebea04c0eaed57adc47af5c184", "tree": "04fbeb7b48e6362ec11e14631b1f42eb74b9c865", "parents": [ "6c98cd4ecb0ae35f0368c5c2190712689c2064aa" ], "author": { "name": "Ryusuke Konishi", "email": "konishi.ryusuke@lab.ntt.co.jp", "time": "Mon Apr 06 19:01:32 2009 -0700" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Apr 07 08:31:14 2009 -0700" }, "message": "nilfs2: inode operations\n\nThis adds inode level operations of the nilfs2 file system.\n\nSigned-off-by: Ryusuke Konishi \u003ckonishi.ryusuke@lab.ntt.co.jp\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" } ] }