)]}' { "log": [ { "commit": "2318aa272072f6906de8e00a332da1485506b3c5", "tree": "a305711c90dab041e1c53a00403c90f06989b56b", "parents": [ "ed4e6a94d3053b9900b4a1338b8056d532a564c4", "9836b8b9499cb25ea32cad9fff640eef874c5431" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jan 03 11:41:43 2013 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jan 03 11:41:43 2013 -0800" }, "message": "Merge tag \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/jaegeuk/f2fs\n\nPull f2fs bug fixes from Jaegeuk Kim:\n \"This patch-set includes two major bug fixes:\n - incorrect IUsed provided by *df -i*, and\n - lookup failure of parent inodes in corner cases.\n\n [Other Bug Fixes]\n - Fix error handling routines\n - Trigger recovery process correctly\n - Resolve build failures due to missing header files\n\n [Etc]\n - Add a MAINTAINERS entry for f2fs\n - Fix and clean up variables, functions, and equations\n - Avoid warnings during compilation\"\n\n* tag \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/jaegeuk/f2fs:\n f2fs: unify string length declarations and usage\n f2fs: clean up unused variables and return values\n f2fs: clean up the start_bidx_of_node function\n f2fs: remove unneeded variable from f2fs_sync_fs\n f2fs: fix fsync_inode list addition logic and avoid invalid access to memory\n f2fs: remove unneeded initialization of nr_dirty in dirty_seglist_info\n f2fs: handle error from f2fs_iget_nowait\n f2fs: fix equation of has_not_enough_free_secs()\n f2fs: add MAINTAINERS entry\n f2fs: return a default value for non-void function\n f2fs: invalidate the node page if allocation is failed\n f2fs: add missing #include \u003clinux/prefetch.h\u003e\n f2fs: do f2fs_balance_fs in front of dir operations\n f2fs: should recover orphan and fsync data\n f2fs: fix handling errors got by f2fs_write_inode\n f2fs: fix up f2fs_get_parent issue to retrieve correct parent inode number\n f2fs: fix wrong calculation on f_files in statfs\n f2fs: remove set_page_dirty for atomic f2fs_end_io_write\n" }, { "commit": "ed4e6a94d3053b9900b4a1338b8056d532a564c4", "tree": "89b4183298e7ae08ec531652529bbd73076ceda7", "parents": [ "22f4f7b59afd02c3f5372e19a536fde339bfd1dc", "13d2eb012927b03ac1b80202af5aa9abc4003bd5" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jan 03 11:38:14 2013 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jan 03 11:38:14 2013 -0800" }, "message": "Merge git://git.kernel.org/pub/scm/linux/kernel/git/steve/gfs2-3.0-fixes\n\nPull GFS2 fixes from Steven Whitehouse:\n \"Here are four small bug fixes for GFS2. There is no common theme here\n really, just a few items that were fixed recently.\n\n The first fixes lock name generation when the glock number is 0. The\n second fixes a race allocating reservation structures and the final\n two fix a performance issue by making small changes in the allocation\n code.\"\n\n* git://git.kernel.org/pub/scm/linux/kernel/git/steve/gfs2-3.0-fixes:\n GFS2: Reset rd_last_alloc when it reaches the end of the rgrp\n GFS2: Stop looking for free blocks at end of rgrp\n GFS2: Fix race in gfs2_rs_alloc\n GFS2: Initialize hex string to \u00270\u0027\n" }, { "commit": "22f4f7b59afd02c3f5372e19a536fde339bfd1dc", "tree": "e75b8c8db82b6cebe7621be854bcf712faab97a7", "parents": [ "062fe95afe93a24b33d34491c54445ab5438d0ba", "e6449c9b2d90c1bd9a5985bf05ddebfd1631cd6b" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jan 03 11:37:49 2013 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jan 03 11:37:49 2013 -0800" }, "message": "Merge branch \u0027merge\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc\n\nPull powerpc fixes from Ben Herrenschmidt:\n \"Here are a couple of small powerpc fixes. They aren\u0027t new bugs (and\n they are both CCed to stable) but I didn\u0027t see the point of sitting on\n the fixes any longer.\"\n\n* \u0027merge\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc:\n powerpc: Add missing NULL terminator to avoid boot panic on PPC40x\n powerpc/vdso: Remove redundant locking in update_vsyscall_tz()\n" }, { "commit": "062fe95afe93a24b33d34491c54445ab5438d0ba", "tree": "f4c111d5f01e569c05e60a339cd9cd65be842203", "parents": [ "d1c3ed669a2d452cacfb48c2d171a1f364dae2ed" ], "author": { "name": "Luck, Tony", "email": "tony.luck@intel.com", "time": "Thu Jan 03 10:33:48 2013 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Jan 03 10:57:23 2013 -0800" }, "message": "Wire up finit_module syscall\n\nLinux was granted a new system call to load modules by file descriptor\nin commit 34e1169d996a (\"module: add syscall to load module from fd\").\n\nWire it up for ia64 (ready for the Chrome port :-)\n\nSigned-off-by: Tony Luck \u003ctony.luck@intel.com\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "e6449c9b2d90c1bd9a5985bf05ddebfd1631cd6b", "tree": "00539e68a8240643ace0a0dda9b705adc6cd47f7", "parents": [ "ce73ec6db47af84d1466402781ae0872a9e7873c" ], "author": { "name": "Gabor Juhos", "email": "juhosg@openwrt.org", "time": "Thu Dec 20 03:44:28 2012 +0000" }, "committer": { "name": "Benjamin Herrenschmidt", "email": "benh@kernel.crashing.org", "time": "Thu Jan 03 16:45:52 2013 +1100" }, "message": "powerpc: Add missing NULL terminator to avoid boot panic on PPC40x\n\nThe missing NULL terminator can cause a panic on\nPPC405 boards during boot:\n\n Linux/PowerPC load: console\u003dttyS0,115200 root\u003d/dev/mtdblock1 rootfstype\u003dsquashfs,jffs2 noinitrd init\u003d/etc/preinit\n Finalizing device tree... flat tree at 0x6a5160\n bootconsole [udbg0] enabled\n Page fault in user mode with in_atomic() \u003d 1 mm \u003d (null)\n NIP \u003d c0275f50 MSR \u003d fffffffe\n Oops: Weird page fault, sig: 11 [#1]\n PowerPC 40x Platform\n Modules linked in:\n NIP: c0275f50 LR: c0275f60 CTR: c0280000\n REGS: c0275eb0 TRAP: 636f7265 Not tainted (3.7.1)\n MSR: fffffffe \u003cVEC,VSX,EE,PR,FP,ME,SE,BE,IR,DR,PMM,RI\u003e CR: c06a6190 XER: 00000001\n TASK \u003d c02662a8[0] \u0027swapper\u0027 THREAD: c0274000\n GPR00: c0275ec0 c000c658 c027c4bf 00000000 c0275ee0 c000a0ec c020a1a8 c020a1f0\n GPR08: c020f631 c020f404 c025f078 c025f080 c0275f10\n Call Trace:\n ---[ end trace 31fd0ba7d8756001 ]---\n\n Kernel panic - not syncing: Attempted to kill the idle task!\n\nThe panic happens since commit 9597abe00c1bab2aedce6b49866bf6d1e81c9eed\n(sections: fix section conflicts in arch/powerpc), however the root\ncause of this is that the NULL terminator were not added in commit\na4f740cf33f7f6c164bbde3c0cdbcc77b0c4997c (of/flattree: Add of_flat_dt_match()\nhelper function).\n\nCc: Grant Likely \u003cgrant.likely@secretlab.ca\u003e\nCc: \u003cstable@vger.kernel.org\u003e\nSigned-off-by: Gabor Juhos \u003cjuhosg@openwrt.org\u003e\nSigned-off-by: Benjamin Herrenschmidt \u003cbenh@kernel.crashing.org\u003e\n" }, { "commit": "ce73ec6db47af84d1466402781ae0872a9e7873c", "tree": "05929e9c525a6a5ef7a46795499f5b34dbbddf8b", "parents": [ "d1c3ed669a2d452cacfb48c2d171a1f364dae2ed" ], "author": { "name": "Shan Hai", "email": "shan.hai@windriver.com", "time": "Thu Nov 08 15:57:49 2012 +0000" }, "committer": { "name": "Benjamin Herrenschmidt", "email": "benh@kernel.crashing.org", "time": "Thu Jan 03 16:45:51 2013 +1100" }, "message": "powerpc/vdso: Remove redundant locking in update_vsyscall_tz()\n\nThe locking in update_vsyscall_tz() is not only unnecessary because the vdso\ncode copies the data unproteced in __kernel_gettimeofday() but also\nintroduces a hard to reproduce race condition between update_vsyscall()\nand update_vsyscall_tz(), which causes user space process to loop\nforever in vdso code.\n\nThe following patch removes the locking from update_vsyscall_tz().\n\nLocking is not only unnecessary because the vdso code copies the data\nunprotected in __kernel_gettimeofday() but also erroneous because updating\nthe tb_update_count is not atomic and introduces a hard to reproduce race\ncondition between update_vsyscall() and update_vsyscall_tz(), which further\ncauses user space process to loop forever in vdso code.\n\nThe below scenario describes the race condition,\nx\u003d\u003d0\tBoot CPU\t\t\tother CPU\n\tproc_P: x\u003d\u003d0\n\t timer interrupt\n\t\tupdate_vsyscall\nx\u003d\u003d1\t\t x++;sync\t\tsettimeofday\n\t\t\t\t\t update_vsyscall_tz\nx\u003d\u003d2\t\t\t\t\t\tx++;sync\nx\u003d\u003d3\t\t sync;x++\n\t\t\t\t\t\tsync;x++\n\tproc_P: x\u003d\u003d3 (loops until x becomes even)\n\nBecause the ++ operator would be implemented as three instructions and not\natomic on powerpc.\n\nA similar change was made for x86 in commit 6c260d58634\n(\"x86: vdso: Remove bogus locking in update_vsyscall_tz\")\n\nSigned-off-by: Shan Hai \u003cshan.hai@windriver.com\u003e\nCC: \u003cstable@vger.kernel.org\u003e\nSigned-off-by: Benjamin Herrenschmidt \u003cbenh@kernel.crashing.org\u003e\n" }, { "commit": "d1c3ed669a2d452cacfb48c2d171a1f364dae2ed", "tree": "97ce168cf32ac88b9aa93408b0b681747416a504", "parents": [ "d50403dcc5092e9c7d6a2525dcfc5813f7c35194" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Jan 02 18:13:21 2013 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Jan 02 18:13:21 2013 -0800" }, "message": "Linux 3.8-rc2\n" }, { "commit": "d50403dcc5092e9c7d6a2525dcfc5813f7c35194", "tree": "7cc1113ae98da24f0c59ea1b1c126beca4971418", "parents": [ "ef05e9b9602660fd468e05f5fbe8f08ad7418b73", "2d7c22f67d5c342a6296127af4f224208449b779" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Jan 02 18:12:35 2013 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Jan 02 18:12:35 2013 -0800" }, "message": "Merge branch \u0027fixes-for-3.8\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/cooloney/linux-leds\n\nPull LED fix from Bryan Wu.\n\n* \u0027fixes-for-3.8\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/cooloney/linux-leds:\n leds: leds-gpio: set devm_gpio_request_one() flags param correctly\n" }, { "commit": "2d7c22f67d5c342a6296127af4f224208449b779", "tree": "e8ddaa212a4833bbb8c4ccf91a73167192955a5a", "parents": [ "a49f0d1ea3ec94fc7cf33a7c36a16343b74bd565" ], "author": { "name": "Javier Martinez Canillas", "email": "javier.martinez@collabora.co.uk", "time": "Thu Dec 20 04:56:59 2012 -0800" }, "committer": { "name": "Bryan Wu", "email": "cooloney@gmail.com", "time": "Wed Jan 02 17:58:41 2013 -0800" }, "message": "leds: leds-gpio: set devm_gpio_request_one() flags param correctly\n\ncommit a99d76f leds: leds-gpio: use gpio_request_one\n\nchanged the leds-gpio driver to use gpio_request_one() instead\nof gpio_request() + gpio_direction_output()\n\nUnfortunately, it also made a semantic change that breaks the\nleds-gpio driver.\n\nThe gpio_request_one() flags parameter was set to:\n\nGPIOF_DIR_OUT | (led_dat-\u003eactive_low ^ state)\n\nSince GPIOF_DIR_OUT is 0, the final flags value will just be the\nXOR\u0027ed value of led_dat-\u003eactive_low and state.\n\nThis value were used to distinguish between HIGH/LOW output initial\nlevel and call gpio_direction_output() accordingly.\n\nWith this new semantic gpio_request_one() will take the flags value\nof 1 as a configuration of input direction (GPIOF_DIR_IN) and will\ncall gpio_direction_input() instead of gpio_direction_output().\n\nint gpio_request_one(unsigned gpio, unsigned long flags, const char *label)\n{\n..\n\tif (flags \u0026 GPIOF_DIR_IN)\n\t\terr \u003d gpio_direction_input(gpio);\n\telse\n\t\terr \u003d gpio_direction_output(gpio,\n\t\t\t\t(flags \u0026 GPIOF_INIT_HIGH) ? 1 : 0);\n..\n}\n\nThe right semantic is to evaluate led_dat-\u003eactive_low ^ state and\nset the output initial level explicitly.\n\nSigned-off-by: Javier Martinez Canillas \u003cjavier.martinez@collabora.co.uk\u003e\nReported-by: Arnaud Patard \u003carnaud.patard@rtp-net.org\u003e\nTested-by: Ezequiel Garcia \u003cezequiel.garcia@free-electrons.com\u003e\nSigned-off-by: Bryan Wu \u003ccooloney@gmail.com\u003e\n" }, { "commit": "ef05e9b9602660fd468e05f5fbe8f08ad7418b73", "tree": "7de3ceb677e7bccb997e3ad80be09b99f528aa42", "parents": [ "080a62e2ceec28163194facb837d6388983aae5d", "8899b8d93ec64b7a8e54807a68a958e1206535e2" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Jan 02 17:46:14 2013 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Jan 02 17:46:14 2013 -0800" }, "message": "Merge git://www.linux-watchdog.org/linux-watchdog\n\nPull watchdog fixes from Wim Van Sebroeck:\n \"This fixes some small errors in the new da9055 driver, eliminates a\n compiler warning and adds DT support for the twl4030_wdt driver (so\n that we can have multiple watchdogs with DT on the omap platforms).\"\n\n* git://www.linux-watchdog.org/linux-watchdog:\n watchdog: twl4030_wdt: add DT support\n watchdog: omap_wdt: eliminate unused variable and a compiler warning\n watchdog: da9055: Don\u0027t update wdt_dev-\u003etimeout in da9055_wdt_set_timeout error path\n watchdog: da9055: Fix invalid free of devm_ allocated data\n" }, { "commit": "080a62e2ceec28163194facb837d6388983aae5d", "tree": "c39ea5bd760383e2300057adcc112ea3be6ce4a4", "parents": [ "8a7eab2b54b349d005181fd971cfa027b1976c7b", "812089e01b9f65f90fc8fc670d8cce72a0e01fbb" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Jan 02 17:44:29 2013 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Jan 02 17:44:29 2013 -0800" }, "message": "Merge tag \u00273.8-pci-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/helgaas/pci\n\nPull PCI updates from Bjorn Helgaas:\n \"Some fixes for v3.8. They include a fix for the new SR-IOV sysfs\n management support, an expanded quirk for Ricoh SD card readers, a\n Stratus DMI quirk fix, and a PME polling fix.\"\n\n* tag \u00273.8-pci-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/helgaas/pci:\n PCI: Reduce Ricoh 0xe822 SD card reader base clock frequency to 50MHz\n PCI/PM: Do not suspend port if any subordinate device needs PME polling\n PCI: Add PCIe Link Capability link speed and width names\n PCI: Work around Stratus ftServer broken PCIe hierarchy (fix DMI check)\n PCI: Remove spurious error for sriov_numvfs store and simplify flow\n" }, { "commit": "8a7eab2b54b349d005181fd971cfa027b1976c7b", "tree": "42ba4b4248c3907d0876097488a5c2b7de53fe92", "parents": [ "3d33fcc11bdd11b6949cf5c406726a094395dc4f" ], "author": { "name": "David Howells", "email": "dhowells@redhat.com", "time": "Wed Jan 02 15:13:02 2013 +0000" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Jan 02 17:36:10 2013 -0800" }, "message": "UAPI: Strip _UAPI prefix on header install no matter the whitespace\n\nCommit 56c176c9cac9 (\"UAPI: strip the _UAPI prefix from header guards\nduring header installation\") strips the _UAPI prefix from header guards,\nbut only if there\u0027s a single space between the cpp directive and the\nlabel.\n\nMake it more flexible and able to handle tabs and multiple white space\ncharacters.\n\nSigned-off-by: David Howells \u003cdhowell@redhat.com\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "3d33fcc11bdd11b6949cf5c406726a094395dc4f", "tree": "c2c25e3d7b3c031d21048c5606cceb5fe09d37f8", "parents": [ "007f6c3a63e0831ff5a34bde072db08e917134c8" ], "author": { "name": "David Howells", "email": "dhowells@redhat.com", "time": "Wed Jan 02 15:12:55 2013 +0000" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Jan 02 17:36:10 2013 -0800" }, "message": "UAPI: Remove empty Kbuild files\n\nEmpty files can get deleted by the patch program, so remove empty Kbuild\nfiles and their links from the parent Kbuilds.\n\nSigned-off-by: David Howells \u003cdhowells@redhat.com\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "007f6c3a63e0831ff5a34bde072db08e917134c8", "tree": "f9c7b1cbd91f719165ce97463652c66ee81d3704", "parents": [ "58890c06691462ca29900d1116b28c7a3e131252", "37028758f92d0a3eb74bcfbecf6bc477072e9e28" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Jan 02 17:33:50 2013 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Jan 02 17:33:50 2013 -0800" }, "message": "Merge tag \u0027ecryptfs-3.8-rc2-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tyhicks/ecryptfs\n\nPull ecryptfs fixes from Tyler Hicks:\n \"Two self-explanatory fixes and a third patch which improves\n performance: when overwriting a full page in the eCryptfs page cache,\n skip reading in and decrypting the corresponding lower page.\"\n\n* tag \u0027ecryptfs-3.8-rc2-fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tyhicks/ecryptfs:\n fs/ecryptfs/crypto.c: make ecryptfs_encode_for_filename() static\n eCryptfs: fix to use list_for_each_entry_safe() when delete items\n eCryptfs: Avoid unnecessary disk read and data decryption during writing\n" }, { "commit": "58890c06691462ca29900d1116b28c7a3e131252", "tree": "2fe512c5a664176de8078f06264da850b1460fb5", "parents": [ "42288fe366c4f1ce7522bc9f27d0bc2a81c55264", "0fa6ebc600bc8e830551aee47a0e929e818a1868" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Jan 02 17:32:49 2013 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Jan 02 17:32:49 2013 -0800" }, "message": "Merge branch \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/sage/ceph-client\n\nPull Ceph fixes from Sage Weil:\n \"Two of Alex\u0027s patches deal with a race when reseting server\n connections for open RBD images, one demotes some non-fatal BUGs to\n WARNs, and my patch fixes a protocol feature bit failure path.\"\n\n* \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/sage/ceph-client:\n libceph: fix protocol feature mismatch failure path\n libceph: WARN, don\u0027t BUG on unexpected connection states\n libceph: always reset osds when kicking\n libceph: move linger requests sooner in kick_requests()\n" }, { "commit": "42288fe366c4f1ce7522bc9f27d0bc2a81c55264", "tree": "f881ebd4f4032e6cc8f5f848b2e61e9ff1d611af", "parents": [ "5439ca6b8ff8cf8d758c19eb28b617a5912904ee" ], "author": { "name": "Mel Gorman", "email": "mgorman@suse.de", "time": "Fri Dec 21 23:10:25 2012 +0000" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Jan 02 17:32:13 2013 -0800" }, "message": "mm: mempolicy: Convert shared_policy mutex to spinlock\n\nSasha was fuzzing with trinity and reported the following problem:\n\n BUG: sleeping function called from invalid context at kernel/mutex.c:269\n in_atomic(): 1, irqs_disabled(): 0, pid: 6361, name: trinity-main\n 2 locks held by trinity-main/6361:\n #0: (\u0026mm-\u003emmap_sem){++++++}, at: [\u003cffffffff810aa314\u003e] __do_page_fault+0x1e4/0x4f0\n #1: (\u0026(\u0026mm-\u003epage_table_lock)-\u003erlock){+.+...}, at: [\u003cffffffff8122f017\u003e] handle_pte_fault+0x3f7/0x6a0\n Pid: 6361, comm: trinity-main Tainted: G W\n 3.7.0-rc2-next-20121024-sasha-00001-gd95ef01-dirty #74\n Call Trace:\n __might_sleep+0x1c3/0x1e0\n mutex_lock_nested+0x29/0x50\n mpol_shared_policy_lookup+0x2e/0x90\n shmem_get_policy+0x2e/0x30\n get_vma_policy+0x5a/0xa0\n mpol_misplaced+0x41/0x1d0\n handle_pte_fault+0x465/0x6a0\n\nThis was triggered by a different version of automatic NUMA balancing\nbut in theory the current version is vunerable to the same problem.\n\ndo_numa_page\n -\u003e numa_migrate_prep\n -\u003e mpol_misplaced\n -\u003e get_vma_policy\n -\u003e shmem_get_policy\n\nIt\u0027s very unlikely this will happen as shared pages are not marked\npte_numa -- see the page_mapcount() check in change_pte_range() -- but\nit is possible.\n\nTo address this, this patch restores sp-\u003elock as originally implemented\nby Kosaki Motohiro. In the path where get_vma_policy() is called, it\nshould not be calling sp_alloc() so it is not necessary to treat the PTL\nspecially.\n\nSigned-off-by: KOSAKI Motohiro \u003ckosaki.motohiro@jp.fujitsu.com\u003e\nTested-by: KOSAKI Motohiro \u003ckosaki.motohiro@jp.fujitsu.com\u003e\nSigned-off-by: Mel Gorman \u003cmgorman@suse.de\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "5439ca6b8ff8cf8d758c19eb28b617a5912904ee", "tree": "676f9e0b25074d2d8c5ab29df30c962b3cb7311e", "parents": [ "a7a88b23737095e6c18a20c5d4eef9e25ec5b829", "0e9a9a1ad619e7e987815d20262d36a2f95717ca" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Jan 02 09:57:34 2013 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Jan 02 09:57:34 2013 -0800" }, "message": "Merge tag \u0027ext4_for_linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4\n\nPull ext4 bug fixes from Ted Ts\u0027o:\n \"Various bug fixes for ext4. Perhaps the most serious bug fixed is one\n which could cause file system corruptions when performing file punch\n operations.\"\n\n* tag \u0027ext4_for_linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4:\n ext4: avoid hang when mounting non-journal filesystems with orphan list\n ext4: lock i_mutex when truncating orphan inodes\n ext4: do not try to write superblock on ro remount w/o journal\n ext4: include journal blocks in df overhead calcs\n ext4: remove unaligned AIO warning printk\n ext4: fix an incorrect comment about i_mutex\n ext4: fix deadlock in journal_unmap_buffer()\n ext4: split off ext4_journalled_invalidatepage()\n jbd2: fix assertion failure in jbd2_journal_flush()\n ext4: check dioread_nolock on remount\n ext4: fix extent tree corruption caused by hole punch\n" }, { "commit": "a7a88b23737095e6c18a20c5d4eef9e25ec5b829", "tree": "da6000487d0a1e4694fd760ac55467bce091e854", "parents": [ "f2a07f40dbc603c15f8b06e6ec7f768af67b424f" ], "author": { "name": "Hugh Dickins", "email": "hughd@google.com", "time": "Wed Jan 02 02:04:23 2013 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Jan 02 09:27:10 2013 -0800" }, "message": "mempolicy: remove arg from mpol_parse_str, mpol_to_str\n\nRemove the unused argument (formerly no_context) from mpol_parse_str()\nand from mpol_to_str().\n\nSigned-off-by: Hugh Dickins \u003chughd@google.com\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "f2a07f40dbc603c15f8b06e6ec7f768af67b424f", "tree": "965f2db39cbb57c55da3ebfc31249e26eb746906", "parents": [ "128dd1759d96ad36c379240f8b9463e8acfd37a1" ], "author": { "name": "Hugh Dickins", "email": "hughd@google.com", "time": "Wed Jan 02 02:01:33 2013 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Jan 02 09:27:10 2013 -0800" }, "message": "tmpfs mempolicy: fix /proc/mounts corrupting memory\n\nRecently I suggested using \"mount -o remount,mpol\u003dlocal /tmp\" in NUMA\nmempolicy testing. Very nasty. Reading /proc/mounts, /proc/pid/mounts\nor /proc/pid/mountinfo may then corrupt one bit of kernel memory, often\nin a page table (causing \"Bad swap\" or \"Bad page map\" warning or \"Bad\npagetable\" oops), sometimes in a vm_area_struct or rbnode or somewhere\nworse. \"mpol\u003dprefer\" and \"mpol\u003dprefer:Node\" are equally toxic.\n\nRecent NUMA enhancements are not to blame: this dates back to 2.6.35,\nwhen commit e17f74af351c \"mempolicy: don\u0027t call mpol_set_nodemask() when\nno_context\" skipped mpol_parse_str()\u0027s call to mpol_set_nodemask(),\nwhich used to initialize v.preferred_node, or set MPOL_F_LOCAL in flags.\nWith slab poisoning, you can then rely on mpol_to_str() to set the bit\nfor node 0x6b6b, probably in the next page above the caller\u0027s stack.\n\nmpol_parse_str() is only called from shmem_parse_options(): no_context\nis always true, so call it unused for now, and remove !no_context code.\nSet v.nodes or v.preferred_node or MPOL_F_LOCAL as mpol_to_str() might\nexpect. Then mpol_to_str() can ignore its no_context argument also,\nthe mpol being appropriately initialized whether contextualized or not.\nRename its no_context unused too, and let subsequent patch remove them\n(that\u0027s not needed for stable backporting, which would involve rejects).\n\nI don\u0027t understand why MPOL_LOCAL is described as a pseudo-policy:\nit\u0027s a reasonable policy which suffers from a confusing implementation\nin terms of MPOL_PREFERRED with MPOL_F_LOCAL. I believe this would be\nmuch more robust if MPOL_LOCAL were recognized in switch statements\nthroughout, MPOL_F_LOCAL deleted, and MPOL_PREFERRED use the (possibly\nempty) nodes mask like everyone else, instead of its preferred_node\nvariant (I presume an optimization from the days before MPOL_LOCAL).\nBut that would take me too long to get right and fully tested.\n\nSigned-off-by: Hugh Dickins \u003chughd@google.com\u003e\nCc: stable@vger.kernel.org\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "128dd1759d96ad36c379240f8b9463e8acfd37a1", "tree": "ff5a716b9dcd319d203485e9aa86f3512d555fce", "parents": [ "4a490b78cb7e0e5efa44425df72a9fedc1c36366" ], "author": { "name": "Eric Wong", "email": "normalperson@yhbt.net", "time": "Tue Jan 01 21:20:27 2013 +0000" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Jan 02 09:16:43 2013 -0800" }, "message": "epoll: prevent missed events on EPOLL_CTL_MOD\n\nEPOLL_CTL_MOD sets the interest mask before calling f_op-\u003epoll() to\nensure events are not missed. Since the modifications to the interest\nmask are not protected by the same lock as ep_poll_callback, we need to\nensure the change is visible to other CPUs calling ep_poll_callback.\n\nWe also need to ensure f_op-\u003epoll() has an up-to-date view of past\nevents which occured before we modified the interest mask. So this\nbarrier also pairs with the barrier in wq_has_sleeper().\n\nThis should guarantee either ep_poll_callback or f_op-\u003epoll() (or both)\nwill notice the readiness of a recently-ready/modified item.\n\nThis issue was encountered by Andreas Voellmy and Junchang(Jason) Wang in:\nhttp://thread.gmane.org/gmane.linux.kernel/1408782/\n\nSigned-off-by: Eric Wong \u003cnormalperson@yhbt.net\u003e\nCc: Hans Verkuil \u003chans.verkuil@cisco.com\u003e\nCc: Jiri Olsa \u003cjolsa@redhat.com\u003e\nCc: Jonathan Corbet \u003ccorbet@lwn.net\u003e\nCc: Al Viro \u003cviro@zeniv.linux.org.uk\u003e\nCc: Davide Libenzi \u003cdavidel@xmailserver.org\u003e\nCc: Hans de Goede \u003chdegoede@redhat.com\u003e\nCc: Mauro Carvalho Chehab \u003cmchehab@infradead.org\u003e\nCc: David Miller \u003cdavem@davemloft.net\u003e\nCc: Eric Dumazet \u003ceric.dumazet@gmail.com\u003e\nCc: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nCc: Andreas Voellmy \u003candreas.voellmy@yale.edu\u003e\nTested-by: \"Junchang(Jason) Wang\" \u003cjunchang.wang@yale.edu\u003e\nCc: netdev@vger.kernel.org\nCc: linux-fsdevel@vger.kernel.org\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "8899b8d93ec64b7a8e54807a68a958e1206535e2", "tree": "22bc88d77fdf1fd2489c8be62f51bf95b9a6db8c", "parents": [ "412b3729dd0234771c67452b8999191f1e8d8630" ], "author": { "name": "Aaro Koskinen", "email": "aaro.koskinen@iki.fi", "time": "Sun Dec 23 22:03:37 2012 +0200" }, "committer": { "name": "Wim Van Sebroeck", "email": "wim@iguana.be", "time": "Wed Jan 02 12:07:05 2013 +0100" }, "message": "watchdog: twl4030_wdt: add DT support\n\nAdd DT support for twl4030_wdt. This is needed to get twl4030_wdt to\nprobe when booting with DT.\n\nSigned-off-by: Aaro Koskinen \u003caaro.koskinen@iki.fi\u003e\nSigned-off-by: Wim Van Sebroeck \u003cwim@iguana.be\u003e\n" }, { "commit": "412b3729dd0234771c67452b8999191f1e8d8630", "tree": "2acd40d2daf8ebbda60aaa14ffcc0f58b4870fd4", "parents": [ "98e4a293895dda2b74476ac3a9f79c58b5d0155a" ], "author": { "name": "Aaro Koskinen", "email": "aaro.koskinen@iki.fi", "time": "Sun Dec 23 22:03:36 2012 +0200" }, "committer": { "name": "Wim Van Sebroeck", "email": "wim@iguana.be", "time": "Wed Jan 02 12:06:58 2013 +0100" }, "message": "watchdog: omap_wdt: eliminate unused variable and a compiler warning\n\nWe forgot to delete this in the commit 4f4753d9 (watchdog: omap_wdt:\nconvert to devm_ functions), and as a result the following compilation\nwarning was introduced:\n\ndrivers/watchdog/omap_wdt.c: In function \u0027omap_wdt_remove\u0027:\ndrivers/watchdog/omap_wdt.c:299:19: warning: unused variable \u0027res\u0027 [-Wunused-variable]\n\nSigned-off-by: Aaro Koskinen \u003caaro.koskinen@iki.fi\u003e\nReviewed-by: Paul Walmsley \u003cpaul@pwsan.com\u003e\nSigned-off-by: Wim Van Sebroeck \u003cwim@iguana.be\u003e\n" }, { "commit": "98e4a293895dda2b74476ac3a9f79c58b5d0155a", "tree": "d79ceba10dcf42c5c416b685aa80b2f047054d6b", "parents": [ "ee8c94adff9bd8609e70fb2ecdfaa71f561ed40d" ], "author": { "name": "Axel Lin", "email": "axel.lin@ingics.com", "time": "Sat Dec 22 11:07:01 2012 +0800" }, "committer": { "name": "Wim Van Sebroeck", "email": "wim@iguana.be", "time": "Wed Jan 02 12:06:49 2013 +0100" }, "message": "watchdog: da9055: Don\u0027t update wdt_dev-\u003etimeout in da9055_wdt_set_timeout error path\n\nOtherwise, WDIOC_GETTIMEOUT returns wrong value if set_timeout fails.\nThis patch also removes unnecessary ret variable in da9055_wdt_ping function.\n\nSigned-off-by: Axel Lin \u003caxel.lin@ingics.com\u003e\nSigned-off-by: Wim Van Sebroeck \u003cwim@iguana.be\u003e\n" }, { "commit": "ee8c94adff9bd8609e70fb2ecdfaa71f561ed40d", "tree": "ec8b1cd9f8c525db48cf5720e5412066def83fc2", "parents": [ "4a490b78cb7e0e5efa44425df72a9fedc1c36366" ], "author": { "name": "Axel Lin", "email": "axel.lin@ingics.com", "time": "Fri Dec 21 21:09:06 2012 +0800" }, "committer": { "name": "Wim Van Sebroeck", "email": "wim@iguana.be", "time": "Wed Jan 02 12:06:43 2013 +0100" }, "message": "watchdog: da9055: Fix invalid free of devm_ allocated data\n\nIt is not required to free devm_ allocated data. Since kref_put\nneeds a valid release function, da9055_wdt_release_resources()\nis not deleted.\n\nSigned-off-by: Axel Lin \u003caxel.lin@ingics.com\u003e\nSigned-off-by: Wim Van Sebroeck \u003cwim@iguana.be\u003e\n" }, { "commit": "13d2eb012927b03ac1b80202af5aa9abc4003bd5", "tree": "e2145f0cdd69a53edeb915efd9712e2ebba19f45", "parents": [ "15bd50ad82a6d3421af1abe82e2554898abc4141" ], "author": { "name": "Bob Peterson", "email": "rpeterso@redhat.com", "time": "Thu Dec 20 13:23:04 2012 -0500" }, "committer": { "name": "Steven Whitehouse", "email": "swhiteho@redhat.com", "time": "Wed Jan 02 10:05:27 2013 +0000" }, "message": "GFS2: Reset rd_last_alloc when it reaches the end of the rgrp\n\nIn function rg_mblk_search, it\u0027s searching for multiple blocks in\na given state (e.g. \"free\"). If there\u0027s an active block reservation\nits goal is the next free block of that. If the resource group\ncontains the dinode\u0027s goal block, that\u0027s used for the search. But\nif neither is the case, it uses the rgrp\u0027s last allocated block.\nThat way, consecutive allocations appear after one another on media.\nThe problem comes in when you hit the end of the rgrp; it would never\nstart over and search from the beginning. This became a problem,\nsince if you deleted all the files and data from the rgrp, it would\nnever start over and find free blocks. So it had to keep searching\nfurther out on the media to allocate blocks. This patch resets the\nrd_last_alloc after it does an unsuccessful search at the end of\nthe rgrp.\n\nSigned-off-by: Bob Peterson \u003crpeterso@redhat.com\u003e\nSigned-off-by: Steven Whitehouse \u003cswhiteho@redhat.com\u003e\n" }, { "commit": "15bd50ad82a6d3421af1abe82e2554898abc4141", "tree": "f23d595b1306ad4cddf64c181ac831af8cb3042d", "parents": [ "f1213cacc7ffc7d4cdef3692f22b28a2df3216f5" ], "author": { "name": "Bob Peterson", "email": "rpeterso@redhat.com", "time": "Thu Dec 20 13:21:07 2012 -0500" }, "committer": { "name": "Steven Whitehouse", "email": "swhiteho@redhat.com", "time": "Wed Jan 02 10:05:10 2013 +0000" }, "message": "GFS2: Stop looking for free blocks at end of rgrp\n\nThis patch adds a return code check after calling function\ngfs2_rbm_from_block while determining the free extent size.\nThat way, when the end of an rgrp is reached, it won\u0027t try\nto process unaligned blocks after the end.\n\nSigned-off-by: Bob Peterson \u003crpeterso@redhat.com\u003e\nSigned-off-by: Steven Whitehouse \u003cswhiteho@redhat.com\u003e\n" }, { "commit": "f1213cacc7ffc7d4cdef3692f22b28a2df3216f5", "tree": "323a502b74dc45a726b05c5b0bddccb77ba516df", "parents": [ "ec1487528bed94c4aaff3687834fe94203880fd6" ], "author": { "name": "Abhijith Das", "email": "adas@redhat.com", "time": "Wed Dec 19 10:48:01 2012 -0500" }, "committer": { "name": "Steven Whitehouse", "email": "swhiteho@redhat.com", "time": "Wed Jan 02 10:04:53 2013 +0000" }, "message": "GFS2: Fix race in gfs2_rs_alloc\n\nQE aio tests uncovered a race condition in gfs2_rs_alloc where it\u0027s possible\nto come out of the function with a valid ip-\u003ei_res allocation but it gets\nfreed before use resulting in a NULL ptr dereference.\n\nThis patch envelopes the initial short-circuit check for non-NULL ip-\u003ei_res\ninto the mutex lock. With this patch, I was able to successfully run the\nreproducer test multiple times.\n\nResolves: rhbz#878476\nSigned-off-by: Abhi Das \u003cadas@redhat.com\u003e\nSigned-off-by: Steven Whitehouse \u003cswhiteho@redhat.com\u003e\n" }, { "commit": "ec1487528bed94c4aaff3687834fe94203880fd6", "tree": "596a9a256e1465b3baf3ee882a6a06c1d2311f34", "parents": [ "4a490b78cb7e0e5efa44425df72a9fedc1c36366" ], "author": { "name": "Nathan Straz", "email": "nstraz@redhat.com", "time": "Tue Dec 11 17:01:24 2012 -0500" }, "committer": { "name": "Steven Whitehouse", "email": "swhiteho@redhat.com", "time": "Wed Jan 02 10:04:00 2013 +0000" }, "message": "GFS2: Initialize hex string to \u00270\u0027\n\nWhen generating the DLM lock name, a value of 0 would skip\nthe loop and leave the string unchanged. This left locks with\na value of 0 unlabeled. Initializing the string to \u00270\u0027 fixes this.\n\nSigned-off-by: Nathan Straz \u003cnstraz@redhat.com\u003e\nSigned-off-by: Steven Whitehouse \u003cswhiteho@redhat.com\u003e\n" }, { "commit": "4a490b78cb7e0e5efa44425df72a9fedc1c36366", "tree": "8a867e39c4e555e4ba10772748b0bde8fe789e20", "parents": [ "8d91a42e54eebc43f4d8f6064751ccba73528275", "d5757dbe79870d825d0dec30074d48683e1d7e9a" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Dec 30 10:00:37 2012 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Dec 30 10:00:37 2012 -0800" }, "message": "Merge branch \u0027drm-next\u0027 of git://people.freedesktop.org/~airlied/linux\n\nPull DRM update from Dave Airlie:\n \"This is a bit larger due to me not bothering to do anything since\n before Xmas, and other people working too hard after I had clearly\n given up.\n\n It\u0027s got the 3 main x86 driver fixes pulls, and a bunch of tegra\n fixes, doesn\u0027t fix the Ironlake bug yet, but that does seem to be\n getting closer.\n\n - radeon: gpu reset fixes and userspace packet support\n - i915: watermark fixes, workarounds, i830/845 fix,\n - nouveau: nvd9/kepler microcode fixes, accel is now enabled and\n working, gk106 support\n - tegra: misc fixes.\"\n\n* \u0027drm-next\u0027 of git://people.freedesktop.org/~airlied/linux: (34 commits)\n Revert \"drm: tegra: protect DC register access with mutex\"\n drm: tegra: program only one window during modeset\n drm: tegra: clean out old gem prototypes\n drm: tegra: remove redundant tegra2_tmds_config entry\n drm: tegra: protect DC register access with mutex\n drm: tegra: don\u0027t leave clients host1x member uninitialized\n drm: tegra: fix front_porch \u003c-\u003e back_porch mixup\n drm/nve0/graph: fix fuc, and enable acceleration on all known chipsets\n drm/nvc0/graph: fix fuc, and enable acceleration on GF119\n drm/nouveau/bios: cache ramcfg strap on later chipsets\n drm/nouveau/mxm: silence output if no bios data\n drm/nouveau/bios: parse/display extra version component\n drm/nouveau/bios: implement opcode 0xa9\n drm/nouveau/bios: update gpio parsing apis to match current design\n drm/nouveau: initial support for GK106\n drm/radeon: add WAIT_UNTIL to evergreen VM safe reg list\n drm/i915: disable shrinker lock stealing for create_mmap_offset\n drm/i915: optionally disable shrinker lock stealing\n drm/i915: fix flags in dma buf exporting\n drm/radeon: add support for MEM_WRITE packet\n ...\n" }, { "commit": "8d91a42e54eebc43f4d8f6064751ccba73528275", "tree": "198199e546cd355971f26f7a0d0073cf1af6e4ab", "parents": [ "4fe2dfabe48edb799065d05fe96bfb7745cd5688", "f64d204b5cf21b282445ba2fa357d970028c9f52" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Dec 30 09:59:21 2012 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Dec 30 09:59:21 2012 -0800" }, "message": "Merge tag \u0027omap-late-cleanups\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc\n\nPull late ARM cleanups for omap from Olof Johansson:\n \"From Tony Lindgren:\n\n Here are few more patches to finish the omap changes for multiplatform\n conversion that are not strictly fixes, but were too complex to do\n with the dependencies during the merge window. Those are to move of\n serial-omap.h to platform_data, and the removal of remaining\n cpu_is_omap macro usage outside mach-omap2.\n\n Then there are several trivial fixes for typos and few minimal\n omap2plus_defconfig updates.\"\n\n* tag \u0027omap-late-cleanups\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc:\n arch/arm/mach-omap2/dpll3xxx.c: drop if around WARN_ON\n OMAP2: Fix a typo - replace regist with register.\n ARM/omap: use module_platform_driver macro\n ARM: OMAP2+: PMU: Remove unused header\n ARM: OMAP4: remove duplicated include from omap_hwmod_44xx_data.c\n ARM: OMAP2+: omap2plus_defconfig: enable twl4030 SoC audio\n ARM: OMAP2+: omap2plus_defconfig: Add tps65217 support\n ARM: OMAP2+: enable devtmpfs and devtmpfs automount\n ARM: OMAP2+: omap_twl: Change TWL4030_MODULE_PM_RECEIVER to TWL_MODULE_PM_RECEIVER\n ARM: OMAP2+: Drop plat/cpu.h for omap2plus\n ARM: OMAP: Split fb.c to remove last remaining cpu_is_omap usage\n MAINTAINERS: Add an entry for omap related .dts files\n" }, { "commit": "4fe2dfabe48edb799065d05fe96bfb7745cd5688", "tree": "a257d72db9d39616376633fe0d410fc8c1f4ef7b", "parents": [ "ecccd1248d6e6986130ffcc3b0d003cb46a485c0", "2e376799b2f6af2e8baeb407b6437a4d7b650ac0" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Dec 30 09:58:36 2012 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Dec 30 09:58:36 2012 -0800" }, "message": "Merge tag \u0027fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc\n\nPull ARM SoC fixes from Olof Johansson:\n \"It\u0027s been quiet over the holidays, but we have had a couple of trivial\n fixes coming in for the newly introduced sunxi platform; one to add it\n to the multiplatform defconfig for build coverage, and one fixup for\n device tree strings.\"\n\n* tag \u0027fixes\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc:\n sunxi: Change the machine compatible string.\n ARM: multi_v7_defconfig: Add ARCH_SUNXI\n" }, { "commit": "d5757dbe79870d825d0dec30074d48683e1d7e9a", "tree": "f2edbd03bea099694bd3ce9effb6165cc104c55d", "parents": [ "500df2e5d870ef5be3d37f0798f770028b69fd47" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Sun Dec 30 21:58:20 2012 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Sun Dec 30 21:58:20 2012 +1000" }, "message": "Revert \"drm: tegra: protect DC register access with mutex\"\n\nThis reverts commit 83c0bcb694be31dcd6c04bdd935b96a95a0af548.\n\nLucas pointed out this was a mistake, and I missed the discussion,\nso just revert it out to save a rebase.\n\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\n" }, { "commit": "500df2e5d870ef5be3d37f0798f770028b69fd47", "tree": "7680e6b82fd5576762c7a5b8aea22183a0e3d03b", "parents": [ "e39250aa5246b43aae882274e345620f665ab5ef" ], "author": { "name": "Lucas Stach", "email": "dev@lynxeye.de", "time": "Wed Dec 19 21:38:57 2012 +0000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Sun Dec 30 14:01:35 2012 +1000" }, "message": "drm: tegra: program only one window during modeset\n\nThe intention is to program exactly WIN_A, not WIN_A and possibly\nothers.\n\nSigned-off-by: Lucas Stach \u003cdev@lynxeye.de\u003e\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\n" }, { "commit": "e39250aa5246b43aae882274e345620f665ab5ef", "tree": "43249f55eefeab8c5099c5a6e26c393377d774be", "parents": [ "fa416ddc0ae5996d894b10f7f49efc2a494b048b" ], "author": { "name": "Lucas Stach", "email": "dev@lynxeye.de", "time": "Wed Dec 19 21:38:56 2012 +0000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Sun Dec 30 14:01:34 2012 +1000" }, "message": "drm: tegra: clean out old gem prototypes\n\nThere is no gem.c anymore, those functions are implemented by the\ndrm_cma_helpers now.\n\nSigned-off-by: Lucas Stach \u003cdev@lynxeye.de\u003e\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\n" }, { "commit": "fa416ddc0ae5996d894b10f7f49efc2a494b048b", "tree": "8567565e5d970bc754f175e85694b37ab8debd5f", "parents": [ "83c0bcb694be31dcd6c04bdd935b96a95a0af548" ], "author": { "name": "Lucas Stach", "email": "dev@lynxeye.de", "time": "Wed Dec 19 21:38:55 2012 +0000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Sun Dec 30 14:01:33 2012 +1000" }, "message": "drm: tegra: remove redundant tegra2_tmds_config entry\n\nThe 720p and 1080p entries are completely redundant, as we are matching\nthe table entries against \u003c\u003dpclk.\nAlso generalize the comment, as we are using those table entries even\nwhen driving other modes than the standard TV ones.\n\nSigned-off-by: Lucas Stach \u003cdev@lynxeye.de\u003e\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\n" }, { "commit": "83c0bcb694be31dcd6c04bdd935b96a95a0af548", "tree": "f6b6448f3080240513a40208098e99625010a31d", "parents": [ "4026bfb39a3e63e32b3c4a648bb1ac1fd8c6b162" ], "author": { "name": "Lucas Stach", "email": "dev@lynxeye.de", "time": "Wed Dec 19 21:38:54 2012 +0000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Sun Dec 30 14:01:33 2012 +1000" }, "message": "drm: tegra: protect DC register access with mutex\n\nWindow properties are programmed through a shared aperture and have to\nhappen atomically. Also we do the read-update-write dance on some of the\nshared regs.\nTo make sure that different functions don\u0027t stumble over each other\nprotect the register access with a mutex.\n\nSigned-off-by: Lucas Stach \u003cdev@lynxeye.de\u003e\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\n" }, { "commit": "4026bfb39a3e63e32b3c4a648bb1ac1fd8c6b162", "tree": "bff6d6da10564ecfbc3c132d0598fce1ee419f54", "parents": [ "4049508988a6ad03f51c33bf035bd9b603454830" ], "author": { "name": "Lucas Stach", "email": "dev@lynxeye.de", "time": "Wed Dec 19 21:38:53 2012 +0000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Sun Dec 30 14:01:32 2012 +1000" }, "message": "drm: tegra: don\u0027t leave clients host1x member uninitialized\n\nNo real problem for now, as nothing is using this, but leaving it\nunitialized is asking for trouble later on.\n\nSigned-off-by: Lucas Stach \u003cdev@lynxeye.de\u003e\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\n" }, { "commit": "4049508988a6ad03f51c33bf035bd9b603454830", "tree": "65e2f7405c513b180819ab4545faaf8c97c040e2", "parents": [ "8be0e5c427c18a59ce261c496ae2193cbcbafffd" ], "author": { "name": "Lucas Stach", "email": "dev@lynxeye.de", "time": "Wed Dec 19 21:38:52 2012 +0000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Sun Dec 30 14:01:31 2012 +1000" }, "message": "drm: tegra: fix front_porch \u003c-\u003e back_porch mixup\n\nFixes wrong picture offset observed when using HDMI output with a\nTechnisat HD TV.\n\nSigned-off-by: Lucas Stach \u003cdev@lynxeye.de\u003e\nAcked-by: Mark Zhang \u003cmarkz@nvidia.com\u003e\nTested-by: Mark Zhang \u003cmarkz@nvidia.com\u003e\nSigned-off-by: Dave Airlie \u003cairlied@redhat.com\u003e\n" }, { "commit": "8be0e5c427c18a59ce261c496ae2193cbcbafffd", "tree": "26dac80685ba46246642b901460a5042ac4e187c", "parents": [ "b1d778b970ce52e02ca6a7f34ba167fe95bc1cc4", "da494d7ca5e0a1afca3480826b5060e15c951e80" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Sun Dec 30 13:54:12 2012 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Sun Dec 30 13:54:12 2012 +1000" }, "message": "Merge branch \u0027drm-intel-fixes\u0027 of git://people.freedesktop.org/~danvet/drm-intel into drm-next\n\nSome fixes for 3.8:\n- Watermark fixups from Chris Wilson (4 pieces).\n- 2 snb workarounds, seem to be recently added to our internal DB.\n- workaround for the infamous i830/i845 hang, seems now finally solid!\n Based on Chris\u0027 fix for SNA, now also for UXA/mesa\u0026old SNA.\n- Some more fixlets for shrinker-pulls-the-rug issues (Chris\u0026me).\n- Fix dma-buf flags when exporting (you).\n- Disable the VGA plane if it\u0027s enabled on lid open - similar fix in\n spirit to the one I\u0027ve sent you last weeek, BIOS\u0027 really like to mess\n with the display when closing the lid (awesome debug work from Krzysztof\n Mazur).\n\n* \u0027drm-intel-fixes\u0027 of git://people.freedesktop.org/~danvet/drm-intel:\n drm/i915: disable shrinker lock stealing for create_mmap_offset\n drm/i915: optionally disable shrinker lock stealing\n drm/i915: fix flags in dma buf exporting\n i915: ensure that VGA plane is disabled\n drm/i915: Preallocate the drm_mm_node prior to manipulating the GTT drm_mm manager\n drm: Export routines for inserting preallocated nodes into the mm manager\n drm/i915: don\u0027t disable disconnected outputs\n drm/i915: Implement workaround for broken CS tlb on i830/845\n drm/i915: Implement WaSetupGtModeTdRowDispatch\n drm/i915: Implement WaDisableHiZPlanesWhenMSAAEnabled\n drm/i915: Prefer CRTC \u0027active\u0027 rather than \u0027enabled\u0027 during WM computations\n drm/i915: Clear self-refresh watermarks when disabled\n drm/i915: Double the cursor self-refresh latency on Valleyview\n drm/i915: Fixup cursor latency used for IVB lp3 watermarks\n" }, { "commit": "b1d778b970ce52e02ca6a7f34ba167fe95bc1cc4", "tree": "c57d08b4807c9a18db7f3a67975854d196533a3c", "parents": [ "344f9067d59a429093fc3e6ed596d1bef2f77c9d", "668bbc81baf0f34df832d8aca5c7d5e19a493c68" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Sun Dec 30 13:02:48 2012 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Sun Dec 30 13:02:48 2012 +1000" }, "message": "Merge branch \u0027drm-fixes-3.8\u0027 of git://people.freedesktop.org/~agd5f/linux into drm-next\n\nMisc fixes for reset and new packets for userspace usage.\n\n* \u0027drm-fixes-3.8\u0027 of git://people.freedesktop.org/~agd5f/linux:\n drm/radeon: add WAIT_UNTIL to evergreen VM safe reg list\n drm/radeon: add support for MEM_WRITE packet\n drm/radeon: restore modeset late in GPU reset path\n drm/radeon: avoid deadlock in pm path when waiting for fence\n drm/radeon: don\u0027t leave fence blocked process on failed GPU reset\n" }, { "commit": "344f9067d59a429093fc3e6ed596d1bef2f77c9d", "tree": "098f1d5c63ca234c1946b5ad28571ebddca446b1", "parents": [ "a49f0d1ea3ec94fc7cf33a7c36a16343b74bd565", "eca15296a9c2a5d5d7d8281a710ba4bd0c2e7cd3" ], "author": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Sun Dec 30 13:01:52 2012 +1000" }, "committer": { "name": "Dave Airlie", "email": "airlied@redhat.com", "time": "Sun Dec 30 13:01:52 2012 +1000" }, "message": "Merge branch \u0027drm-nouveau-fixes-3.8\u0027 of git://anongit.freedesktop.org/git/nouveau/linux-2.6 into drm-next\n\nFixes the accel support for nvd9 + kepler chipsets, also fixes GK106 support.\n\n* \u0027drm-nouveau-fixes-3.8\u0027 of git://anongit.freedesktop.org/git/nouveau/linux-2.6:\n drm/nve0/graph: fix fuc, and enable acceleration on all known chipsets\n drm/nvc0/graph: fix fuc, and enable acceleration on GF119\n drm/nouveau/bios: cache ramcfg strap on later chipsets\n drm/nouveau/mxm: silence output if no bios data\n drm/nouveau/bios: parse/display extra version component\n drm/nouveau/bios: implement opcode 0xa9\n drm/nouveau/bios: update gpio parsing apis to match current design\n drm/nouveau: initial support for GK106\n" }, { "commit": "ecccd1248d6e6986130ffcc3b0d003cb46a485c0", "tree": "bc9fdf3caf32e846776f13985f837c89455cbb39", "parents": [ "101e5c7470eb7f77ae87f966b9155f0dbb5b4698" ], "author": { "name": "Zlatko Calusic", "email": "zlatko.calusic@iskon.hr", "time": "Fri Dec 28 03:16:38 2012 +0100" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Fri Dec 28 08:42:39 2012 -0800" }, "message": "mm: fix null pointer dereference in wait_iff_congested()\n\nAn unintended consequence of commit 4ae0a48b5efc (\"mm: modify\npgdat_balanced() so that it also handles order-0\") is that\nwait_iff_congested() can now be called with NULL \u0027struct zone *\u0027\nproducing kernel oops like this:\n\n BUG: unable to handle kernel NULL pointer dereference\n IP: [\u003cffffffff811542d9\u003e] wait_iff_congested+0x59/0x140\n\nThis trivial patch fixes it.\n\nReported-by: Zhouping Liu \u003czliu@redhat.com\u003e\nReported-and-tested-by: Sedat Dilek \u003csedat.dilek@gmail.com\u003e\nCc: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nCc: Mel Gorman \u003cmgorman@suse.de\u003e\nCc: Hugh Dickins \u003chughd@google.com\u003e\nSigned-off-by: Zlatko Calusic \u003czlatko.calusic@iskon.hr\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "2e376799b2f6af2e8baeb407b6437a4d7b650ac0", "tree": "414e4b844580de1d6489cb6e5d308be65cc1b22b", "parents": [ "a49f0d1ea3ec94fc7cf33a7c36a16343b74bd565", "43880f709dc59840849e31b01735ac587195ef8a" ], "author": { "name": "Olof Johansson", "email": "olof@lixom.net", "time": "Fri Dec 28 08:53:01 2012 +0100" }, "committer": { "name": "Olof Johansson", "email": "olof@lixom.net", "time": "Fri Dec 28 08:53:01 2012 +0100" }, "message": "Merge tag \u0027sunxi-fixes-for-3.8-rc2\u0027 of git://github.com/mripard/linux into fixes\n\nFrom Maxime Ripard:\nFixes for the sunxi core to be merged in 3.8-rc2\n\n* tag \u0027sunxi-fixes-for-3.8-rc2\u0027 of git://github.com/mripard/linux:\n sunxi: Change the machine compatible string.\n ARM: multi_v7_defconfig: Add ARCH_SUNXI\n" }, { "commit": "9836b8b9499cb25ea32cad9fff640eef874c5431", "tree": "b7d05ce46e6b1f86aea3794ba7b0eddf9db8c9bd", "parents": [ "2b50638decdb9a8585654a5acf1c8ce5962f1951" ], "author": { "name": "Leon Romanovsky", "email": "leon@leon.nu", "time": "Thu Dec 27 19:55:46 2012 +0200" }, "committer": { "name": "Jaegeuk Kim", "email": "jaegeuk.kim@samsung.com", "time": "Fri Dec 28 11:27:53 2012 +0900" }, "message": "f2fs: unify string length declarations and usage\n\nThis patch is intended to unify string length declarations and usage.\nThere are number of calls to strlen which return size_t object.\nThe size of this object depends on compiler if it will be bigger,\nequal or even smaller than an unsigned int\n\nSigned-off-by: Leon Romanovsky \u003cleon@leon.nu\u003e\nSigned-off-by: Jaegeuk Kim \u003cjaegeuk.kim@samsung.com\u003e\n" }, { "commit": "2b50638decdb9a8585654a5acf1c8ce5962f1951", "tree": "71887dfa0124dc314003af82cc23ad510ff880c3", "parents": [ "ce19a5d4321911f98d42e4d724630ae48f413719" ], "author": { "name": "Jaegeuk Kim", "email": "jaegeuk.kim@samsung.com", "time": "Wed Dec 26 14:39:50 2012 +0900" }, "committer": { "name": "Jaegeuk Kim", "email": "jaegeuk.kim@samsung.com", "time": "Fri Dec 28 11:27:52 2012 +0900" }, "message": "f2fs: clean up unused variables and return values\n\nThis patch cleans up a couple of unnecessary codes related to unused variables\nand return values.\n\nSigned-off-by: Jaegeuk Kim \u003cjaegeuk.kim@samsung.com\u003e\n" }, { "commit": "ce19a5d4321911f98d42e4d724630ae48f413719", "tree": "4b97b609a1d544da234ac53cd369bd5b80fc60c8", "parents": [ "64c576fe51bc6b19e99340d2d0e1bda89f66db25" ], "author": { "name": "Jaegeuk Kim", "email": "jaegeuk.kim@samsung.com", "time": "Wed Dec 26 12:03:22 2012 +0900" }, "committer": { "name": "Jaegeuk Kim", "email": "jaegeuk.kim@samsung.com", "time": "Fri Dec 28 11:27:52 2012 +0900" }, "message": "f2fs: clean up the start_bidx_of_node function\n\nThis patch also resolves the following warning reported by kbuild test robot.\n\nfs/f2fs/gc.c: In function \u0027start_bidx_of_node\u0027:\nfs/f2fs/gc.c:453:21: warning: \u0027bidx\u0027 may be used uninitialized in this function\n\nSigned-off-by: Jaegeuk Kim \u003cjaegeuk.kim@samsung.com\u003e\n" }, { "commit": "64c576fe51bc6b19e99340d2d0e1bda89f66db25", "tree": "dc8e96c3df9c31d7a6d890fb0b8fff853007ca9c", "parents": [ "fd8bb65f796f041ee6ba400255ca9021bc45a992" ], "author": { "name": "Namjae Jeon", "email": "namjae.jeon@samsung.com", "time": "Sat Dec 22 12:10:27 2012 +0900" }, "committer": { "name": "Jaegeuk Kim", "email": "jaegeuk.kim@samsung.com", "time": "Fri Dec 28 11:27:45 2012 +0900" }, "message": "f2fs: remove unneeded variable from f2fs_sync_fs\n\nWe can directly return \u00270\u0027 from the function, instead of introducing a\n\u0027ret\u0027 variable.\n\nSigned-off-by: Namjae Jeon \u003cnamjae.jeon@samsung.com\u003e\nSigned-off-by: Amit Sahrawat \u003ca.sahrawat@samsung.com\u003e\nSigned-off-by: Jaegeuk Kim \u003cjaegeuk.kim@samsung.com\u003e\n" }, { "commit": "fd8bb65f796f041ee6ba400255ca9021bc45a992", "tree": "db675ff9033743c4d98f62422e04c9103fea777f", "parents": [ "344324f10fad05e40b1047c5e09ebbc77e43c24f" ], "author": { "name": "Namjae Jeon", "email": "namjae.jeon@samsung.com", "time": "Sat Dec 22 12:10:12 2012 +0900" }, "committer": { "name": "Jaegeuk Kim", "email": "jaegeuk.kim@samsung.com", "time": "Fri Dec 28 11:27:36 2012 +0900" }, "message": "f2fs: fix fsync_inode list addition logic and avoid invalid access to memory\n\nIn function find_fsync_dnodes() - the fsync inodes gets added to the list, but\nin one path suppose f2fs_iget results in error, in such case - error gets added\nto the fsync inode list.\nIn next call to recover_data()-\u003eget_fsync_inode()\nentry \u003d list_entry(this, struct fsync_inode_entry, list);\n if (entry-\u003einode-\u003ei_ino \u003d\u003d ino)\nThis can result in \"invalid access to memory\" when it encounters \u0027error\u0027 as\nentry in the fsync inode list.\nSo, add the fsync inode entry to the list only in case of no errors.\nAnd, free the object at that point itself in case of issue.\n\nSigned-off-by: Namjae Jeon \u003cnamjae.jeon@samsung.com\u003e\nSigned-off-by: Amit Sahrawat \u003ca.sahrawat@samsung.com\u003e\nSigned-off-by: Jaegeuk Kim \u003cjaegeuk.kim@samsung.com\u003e\n" }, { "commit": "344324f10fad05e40b1047c5e09ebbc77e43c24f", "tree": "e5ea654ddb10af7aa22c734c317ac844ec0ffdb2", "parents": [ "06025f4df88e9e41f4ebcf6b4c3df30661332bc9" ], "author": { "name": "Namjae Jeon", "email": "namjae.jeon@samsung.com", "time": "Sat Dec 22 12:09:58 2012 +0900" }, "committer": { "name": "Jaegeuk Kim", "email": "jaegeuk.kim@samsung.com", "time": "Fri Dec 28 11:27:05 2012 +0900" }, "message": "f2fs: remove unneeded initialization of nr_dirty in dirty_seglist_info\n\nSince, the memory for the object of dirty_seglist_info is allocated\nusing kzalloc - which returns zeroed out memory. So, there is no need\nto initialize the nr_dirty values with zeroes.\n\nSigned-off-by: Namjae Jeon \u003cnamjae.jeon@samsung.com\u003e\nSigned-off-by: Amit Sahrawat \u003ca.sahrawat@samsung.com\u003e\nSigned-off-by: Jaegeuk Kim \u003cjaegeuk.kim@samsung.com\u003e\n" }, { "commit": "0fa6ebc600bc8e830551aee47a0e929e818a1868", "tree": "e9768481dfd72099c1615a357c4e2f489d596d68", "parents": [ "122070a2ffc91f87fe8e8493eb0ac61986c5557c" ], "author": { "name": "Sage Weil", "email": "sage@inktank.com", "time": "Thu Dec 27 20:27:04 2012 -0600" }, "committer": { "name": "Alex Elder", "email": "elder@inktank.com", "time": "Thu Dec 27 20:27:04 2012 -0600" }, "message": "libceph: fix protocol feature mismatch failure path\n\nWe should not set con-\u003estate to CLOSED here; that happens in\nceph_fault() in the caller, where it first asserts that the state\nis not yet CLOSED. Avoids a BUG when the features don\u0027t match.\n\nSince the fail_protocol() has become a trivial wrapper, replace\ncalls to it with direct calls to reset_connection().\n\nSigned-off-by: Sage Weil \u003csage@inktank.com\u003e\nReviewed-by: Alex Elder \u003celder@inktank.com\u003e\n" }, { "commit": "122070a2ffc91f87fe8e8493eb0ac61986c5557c", "tree": "b19b8bced463ffdb54f3b5d3b2c287bd81084582", "parents": [ "e6d50f67a6b1a6252a616e6e629473b5c4277218" ], "author": { "name": "Alex Elder", "email": "elder@inktank.com", "time": "Wed Dec 26 10:43:57 2012 -0600" }, "committer": { "name": "Alex Elder", "email": "elder@inktank.com", "time": "Thu Dec 27 20:27:04 2012 -0600" }, "message": "libceph: WARN, don\u0027t BUG on unexpected connection states\n\nA number of assertions in the ceph messenger are implemented with\nBUG_ON(), killing the system if connection\u0027s state doesn\u0027t match\nwhat\u0027s expected. At this point our state model is (evidently) not\nwell understood enough for these assertions to trigger a BUG().\nConvert all BUG_ON(con-\u003estate...) calls to be WARN_ON(con-\u003estate...)\nso we learn about these issues without killing the machine.\n\nWe now recognize that a connection fault can occur due to a socket\nclosure at any time, regardless of the state of the connection. So\nthere is really nothing we can assert about the state of the\nconnection at that point so eliminate that assertion.\n\nReported-by: Ugis \u003cugis22@gmail.com\u003e\nTested-by: Ugis \u003cugis22@gmail.com\u003e\nSigned-off-by: Alex Elder \u003celder@inktank.com\u003e\nReviewed-by: Sage Weil \u003csage@inktank.com\u003e\n" }, { "commit": "e6d50f67a6b1a6252a616e6e629473b5c4277218", "tree": "ff0ca019ba50a5f65d12d628af0aee2695d5219b", "parents": [ "ab60b16d3c31b9bd9fd5b39f97dc42c52a50b67d" ], "author": { "name": "Alex Elder", "email": "elder@inktank.com", "time": "Wed Dec 26 14:31:40 2012 -0600" }, "committer": { "name": "Alex Elder", "email": "elder@inktank.com", "time": "Thu Dec 27 20:27:04 2012 -0600" }, "message": "libceph: always reset osds when kicking\n\nWhen ceph_osdc_handle_map() is called to process a new osd map,\nkick_requests() is called to ensure all affected requests are\nupdated if necessary to reflect changes in the osd map. This\nhappens in two cases: whenever an incremental map update is\nprocessed; and when a full map update (or the last one if there is\nmore than one) gets processed.\n\nIn the former case, the kick_requests() call is followed immediately\nby a call to reset_changed_osds() to ensure any connections to osds\naffected by the map change are reset. But for full map updates\nthis isn\u0027t done.\n\nBoth cases should be doing this osd reset.\n\nRather than duplicating the reset_changed_osds() call, move it into\nthe end of kick_requests().\n\nSigned-off-by: Alex Elder \u003celder@inktank.com\u003e\nReviewed-by: Sage Weil \u003csage@inktank.com\u003e\n" }, { "commit": "ab60b16d3c31b9bd9fd5b39f97dc42c52a50b67d", "tree": "f5ac3a40c46f6587ddfc3c98bc4a12d299fd6e36", "parents": [ "c3e946ce7276faf0b302acd25c7b874edbeba661" ], "author": { "name": "Alex Elder", "email": "elder@inktank.com", "time": "Wed Dec 19 15:52:36 2012 -0600" }, "committer": { "name": "Alex Elder", "email": "elder@inktank.com", "time": "Thu Dec 27 20:27:04 2012 -0600" }, "message": "libceph: move linger requests sooner in kick_requests()\n\nThe kick_requests() function is called by ceph_osdc_handle_map()\nwhen an osd map change has been indicated. Its purpose is to\nre-queue any request whose target osd is different from what it\nwas when it was originally sent.\n\nIt is structured as two loops, one for incomplete but registered\nrequests, and a second for handling completed linger requests.\nAs a special case, in the first loop if a request marked to linger\nhas not yet completed, it is moved from the request list to the\nlinger list. This is as a quick and dirty way to have the second\nloop handle sending the request along with all the other linger\nrequests.\n\nBecause of the way it\u0027s done now, however, this quick and dirty\nsolution can result in these incomplete linger requests never\ngetting re-sent as desired. The problem lies in the fact that\nthe second loop only arranges for a linger request to be sent\nif it appears its target osd has changed. This is the proper\nhandling for *completed* linger requests (it avoids issuing\nthe same linger request twice to the same osd).\n\nBut although the linger requests added to the list in the first loop\nmay have been sent, they have not yet completed, so they need to be\nre-sent regardless of whether their target osd has changed.\n\nThe first required fix is we need to avoid calling __map_request()\non any incomplete linger request. Otherwise the subsequent\n__map_request() call in the second loop will find the target osd\nhas not changed and will therefore not re-send the request.\n\nSecond, we need to be sure that a sent but incomplete linger request\ngets re-sent. If the target osd is the same with the new osd map as\nit was when the request was originally sent, this won\u0027t happen.\nThis can be fixed through careful handling when we move these\nrequests from the request list to the linger list, by unregistering\nthe request *before* it is registered as a linger request. This\nworks because a side-effect of unregistering the request is to make\nthe request\u0027s r_osd pointer be NULL, and *that* will ensure the\nsecond loop actually re-sends the linger request.\n\nProcessing of such a request is done at that point, so continue with\nthe next one once it\u0027s been moved.\n\nSigned-off-by: Alex Elder \u003celder@inktank.com\u003e\nReviewed-by: Sage Weil \u003csage@inktank.com\u003e\n" }, { "commit": "06025f4df88e9e41f4ebcf6b4c3df30661332bc9", "tree": "9d5fab4bb021444db75538a70095675af227dded", "parents": [ "029cd28c1f739bbfc5105035696d5f1f4e45d161" ], "author": { "name": "Namjae Jeon", "email": "namjae.jeon@samsung.com", "time": "Sat Dec 22 12:09:43 2012 +0900" }, "committer": { "name": "Jaegeuk Kim", "email": "jaegeuk.kim@samsung.com", "time": "Fri Dec 28 11:26:13 2012 +0900" }, "message": "f2fs: handle error from f2fs_iget_nowait\n\nIn case f2fs_iget_nowait returns error, it results in truncate_hole being\ncalled with \u0027error\u0027 value as inode pointer. There is no check in truncate_hole\nfor valid inode, so it could result in crash due \"invalid access to memory\".\nAvoid this by handling error condition properly.\n\nSigned-off-by: Namjae Jeon \u003cnamjae.jeon@samsung.com\u003e\nSigned-off-by: Amit Sahrawat \u003ca.sahrawat@samsung.com\u003e\nSigned-off-by: Jaegeuk Kim \u003cjaegeuk.kim@samsung.com\u003e\n" }, { "commit": "029cd28c1f739bbfc5105035696d5f1f4e45d161", "tree": "738b26494ab4b9c959398e729caefe16956bee79", "parents": [ "f58ad8f51a5a0f34c8d41ffc065e0f1c896c991c" ], "author": { "name": "Jaegeuk Kim", "email": "jaegeuk.kim@samsung.com", "time": "Fri Dec 21 17:20:21 2012 +0900" }, "committer": { "name": "Jaegeuk Kim", "email": "jaegeuk.kim@samsung.com", "time": "Fri Dec 28 11:24:10 2012 +0900" }, "message": "f2fs: fix equation of has_not_enough_free_secs()\n\nPractically, has_not_enough_free_secs() should calculate with the numbers of\ncurrent node and directory data blocks together.\nActually the equation was implemented in need_to_flush().\n\nSo, this patch removes need_flush() and moves the equation into\nhas_not_enough_free_secs().\n\nSigned-off-by: Jaegeuk Kim \u003cjaegeuk.kim@samsung.com\u003e\n" }, { "commit": "f58ad8f51a5a0f34c8d41ffc065e0f1c896c991c", "tree": "d0c6e7f9216deabf4760b8d200856a34f5a08b9e", "parents": [ "12a67146e35ba1d04ac4a5430eaaa8790158d60e" ], "author": { "name": "Jaegeuk Kim", "email": "jaegeuk.kim@samsung.com", "time": "Fri Dec 21 12:12:27 2012 +0900" }, "committer": { "name": "Jaegeuk Kim", "email": "jaegeuk.kim@samsung.com", "time": "Fri Dec 28 11:24:09 2012 +0900" }, "message": "f2fs: add MAINTAINERS entry\n\nThis patch adds myself to MAINTAINERS entry for the f2fs file system.\n\nSigned-off-by: Jaegeuk Kim \u003cjaegeuk.kim@samsung.com\u003e\n" }, { "commit": "12a67146e35ba1d04ac4a5430eaaa8790158d60e", "tree": "9a3a2c0768d04fba20e0be8cc07f9cb6ae4a71c5", "parents": [ "71e9fec548a95b2a4cf378646addd5d3098684a2" ], "author": { "name": "Jaegeuk Kim", "email": "jaegeuk.kim@samsung.com", "time": "Fri Dec 21 11:47:05 2012 +0900" }, "committer": { "name": "Jaegeuk Kim", "email": "jaegeuk.kim@samsung.com", "time": "Fri Dec 28 11:24:09 2012 +0900" }, "message": "f2fs: return a default value for non-void function\n\nThis patch resolves a build warning reported by kbuild test robot.\n\n\"\nfs/f2fs/segment.c: In function \u0027__get_segment_type\u0027:\nfs/f2fs/segment.c:806:1: warning: control reaches end of non-void\nfunction [-Wreturn-type]\n\"\n\nSigned-off-by: Jaegeuk Kim \u003cjaegeuk.kim@samsung.com\u003e\n" }, { "commit": "71e9fec548a95b2a4cf378646addd5d3098684a2", "tree": "129aef9aacfb87ec29d67a8ca95422a82ab9db12", "parents": [ "690e4a3ead5f88fc95f7650816d1376aa2e79db5" ], "author": { "name": "Jaegeuk Kim", "email": "jaegeuk.kim@samsung.com", "time": "Thu Dec 20 15:10:06 2012 +0900" }, "committer": { "name": "Jaegeuk Kim", "email": "jaegeuk.kim@samsung.com", "time": "Fri Dec 28 11:24:09 2012 +0900" }, "message": "f2fs: invalidate the node page if allocation is failed\n\nThe new_node_page() is processed as the following procedure.\n\n1. A new node page is allocated.\n2. Set PageUptodate with proper footer information.\n3. Check if there is a free space for allocation\n 4.a. If there is no space, f2fs returns with -ENOSPC.\n 4.b. Otherwise, go next.\n\nIn the case of step #4.a, f2fs remains a wrong node page in the page cache\nwith the uptodate flag.\n\nAlso, even though a new node page is allocated successfully, an error can be\noccurred afterwards due to allocation failure of the other data structures.\nIn such a case, remove_inode_page() would be triggered, so that we have to\nclear uptodate flag in truncate_node() too.\n\nSo, we should remove the uptodate flag, if allocation is failed.\n\nSigned-off-by: Jaegeuk Kim \u003cjaegeuk.kim@samsung.com\u003e\n" }, { "commit": "690e4a3ead5f88fc95f7650816d1376aa2e79db5", "tree": "2d52d94ff4e52191391e313e0f88af49013a3a53", "parents": [ "1efef832020ef392deb2cd3d74e0c316711245be" ], "author": { "name": "Geert Uytterhoeven", "email": "geert@linux-m68k.org", "time": "Wed Dec 19 22:19:30 2012 +0100" }, "committer": { "name": "Jaegeuk Kim", "email": "jaegeuk.kim@samsung.com", "time": "Fri Dec 28 11:22:43 2012 +0900" }, "message": "f2fs: add missing #include \u003clinux/prefetch.h\u003e\n\nm68k allmodconfig:\n\nfs/f2fs/data.c: In function ‘read_end_io’:\nfs/f2fs/data.c:311: error: implicit declaration of function ‘prefetchw’\n\nfs/f2fs/segment.c: In function ‘f2fs_end_io_write’:\nfs/f2fs/segment.c:628: error: implicit declaration of function ‘prefetchw’\n\nSigned-off-by: Geert Uytterhoeven \u003cgeert@linux-m68k.org\u003e\nSigned-off-by: Jaegeuk Kim \u003cjaegeuk.kim@samsung.com\u003e\n" }, { "commit": "101e5c7470eb7f77ae87f966b9155f0dbb5b4698", "tree": "151e031713daa8cfe0d06586d30c5c3699d45dd0", "parents": [ "ddf75ae34e61bc8472f8b54281ad29bc87274be1", "86266ca0e23b2b8e7e38edda1675cd085a9abc14" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Dec 27 10:46:47 2012 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Dec 27 10:46:47 2012 -0800" }, "message": "Merge tag \u0027hwmon-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging\n\nPull hwmon fixes from Guenter Roeck:\n\n - Report i2c errors to userspace in lm73 driver\n\n - Fix problem with DIV_ROUND_CLOSEST and unsigned divisors in emc6w201\n driver\n\n* tag \u0027hwmon-for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging:\n hwmon: (emc6w201) Fix DIV_ROUND_CLOSEST problem with unsigned divisors\n hwmon: (lm73} Detect and report i2c bus errors\n" }, { "commit": "ddf75ae34e61bc8472f8b54281ad29bc87274be1", "tree": "cd93e82253bc35d1bd97859f7b6353223e0c2f84", "parents": [ "7fd83b47cebb9e4fafab0ff9a058d2bebf29b8f5", "48c6d1217e3dc743e7d3ad9b9def8d4810d13a85" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Dec 27 10:42:46 2012 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Dec 27 10:42:46 2012 -0800" }, "message": "Merge branch \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/ebiederm/user-namespace\n\nPull namespace fixes from Eric Biederman:\n \"This tree includes two bug fixes for problems Oleg spotted on his\n review of the recent pid namespace work. A small fix to not enable\n bottom halves with irqs disabled, and a trivial build fix for f2fs\n with user namespaces enabled.\"\n\n* \u0027for-linus\u0027 of git://git.kernel.org/pub/scm/linux/kernel/git/ebiederm/user-namespace:\n f2fs: Don\u0027t assign e_id in f2fs_acl_from_disk\n proc: Allow proc_free_inum to be called from any context\n pidns: Stop pid allocation when init dies\n pidns: Outlaw thread creation after unshare(CLONE_NEWPID)\n" }, { "commit": "7fd83b47cebb9e4fafab0ff9a058d2bebf29b8f5", "tree": "eb669f55b4e144afd49349f4922b78009f938ccd", "parents": [ "ad4b3fb7ff9940bcdb1e4cd62bd189d10fa636ba", "ae782bb16c35ce27512beeda9be6024c88f85b08" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Dec 27 10:40:30 2012 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Thu Dec 27 10:40:30 2012 -0800" }, "message": "Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net\n\nPull networking fixes from David Miller:\n\n1) GRE tunnel drivers don\u0027t set the transport header properly, they also\n blindly deref the inner protocol ipv4 and needs some checks. Fixes\n from Isaku Yamahata.\n\n2) Fix sleeps while atomic in netdevice rename code, from Eric Dumazet.\n\n3) Fix double-spinlock in solos-pci driver, from Dan Carpenter.\n\n4) More ARP bug fixes. Fix lockdep splat in arp_solicit() and then the\n bug accidentally added by that fix. From Eric Dumazet and Cong Wang.\n\n5) Remove some __dev* annotations that slipped back in, as well as all\n HOTPLUG references. From Greg KH\n\n6) RDS protocol uses wrong interfaces to access scatter-gather elements,\n causing a regression. From Mike Marciniszyn.\n\n7) Fix build error in cpts driver, from Richard Cochran.\n\n8) Fix arithmetic in packet scheduler, from Stefan Hasko.\n\n9) Similarly, fix association during calculation of random backoff in\n batman-adv. From Akinobu Mita.\n\n* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: (21 commits)\n ipv6/ip6_gre: set transport header correctly\n ipv4/ip_gre: set transport header correctly to gre header\n IB/rds: suppress incompatible protocol when version is known\n IB/rds: Correct ib_api use with gs_dma_address/sg_dma_len\n net/vxlan: Use the underlying device index when joining/leaving multicast groups\n tcp: should drop incoming frames without ACK flag set\n netprio_cgroup: define sk_cgrp_prioidx only if NETPRIO_CGROUP is enabled\n cpts: fix a run time warn_on.\n cpts: fix build error by removing useless code.\n batman-adv: fix random jitter calculation\n arp: fix a regression in arp_solicit()\n net: sched: integer overflow fix\n CONFIG_HOTPLUG removal from networking core\n Drivers: network: more __dev* removal\n bridge: call br_netpoll_disable in br_add_if\n ipv4: arp: fix a lockdep splat in arp_solicit()\n tuntap: dont use a private kmem_cache\n net: devnet_rename_seq should be a seqcount\n ip_gre: fix possible use after free\n ip_gre: make ipgre_tunnel_xmit() not parse network header as IP unconditionally\n ...\n" }, { "commit": "0e9a9a1ad619e7e987815d20262d36a2f95717ca", "tree": "f8bfb87bf543ccd931bc18bbdb050d1725a81573", "parents": [ "721e3eba21e43532e438652dd8f1fcdfce3187e7" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Thu Dec 27 01:42:50 2012 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Thu Dec 27 01:42:50 2012 -0500" }, "message": "ext4: avoid hang when mounting non-journal filesystems with orphan list\n\nWhen trying to mount a file system which does not contain a journal,\nbut which does have a orphan list containing an inode which needs to\nbe truncated, the mount call with hang forever in\next4_orphan_cleanup() because ext4_orphan_del() will return\nimmediately without removing the inode from the orphan list, leading\nto an uninterruptible loop in kernel code which will busy out one of\nthe CPU\u0027s on the system.\n\nThis can be trivially reproduced by trying to mount the file system\nfound in tests/f_orphan_extents_inode/image.gz from the e2fsprogs\nsource tree. If a malicious user were to put this on a USB stick, and\nmount it on a Linux desktop which has automatic mounts enabled, this\ncould be considered a potential denial of service attack. (Not a big\ndeal in practice, but professional paranoids worry about such things,\nand have even been known to allocate CVE numbers for such problems.)\n\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\nReviewed-by: Zheng Liu \u003cwenqing.lz@taobao.com\u003e\nCc: stable@vger.kernel.org\n" }, { "commit": "721e3eba21e43532e438652dd8f1fcdfce3187e7", "tree": "5d1d118d6b671b98edbeabee15e15869305d129f", "parents": [ "d096ad0f79a782935d2e06ae8fb235e8c5397775" ], "author": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Thu Dec 27 01:42:48 2012 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Thu Dec 27 01:42:48 2012 -0500" }, "message": "ext4: lock i_mutex when truncating orphan inodes\n\nCommit c278531d39 added a warning when ext4_flush_unwritten_io() is\ncalled without i_mutex being taken. It had previously not been taken\nduring orphan cleanup since races weren\u0027t possible at that point in\nthe mount process, but as a result of this c278531d39, we will now see\na kernel WARN_ON in this case. Take the i_mutex in\next4_orphan_cleanup() to suppress this warning.\n\nReported-by: Alexander Beregalov \u003ca.beregalov@gmail.com\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\nReviewed-by: Zheng Liu \u003cwenqing.lz@taobao.com\u003e\nCc: stable@vger.kernel.org\n" }, { "commit": "ae782bb16c35ce27512beeda9be6024c88f85b08", "tree": "10d52422a5e8b0cf7380cf5807eb7036855c3989", "parents": [ "861aa6d56de8147249c8a04a4e87fb7c3600cca1" ], "author": { "name": "Isaku Yamahata", "email": "yamahata@valinux.co.jp", "time": "Mon Dec 24 16:51:04 2012 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Dec 26 15:19:56 2012 -0800" }, "message": "ipv6/ip6_gre: set transport header correctly\n\nip6gre_xmit2() incorrectly sets transport header to inner payload\ninstead of GRE header. It seems copy-and-pasted from ipip.c.\nSet transport header to gre header.\n(In ipip case the transport header is the inner ip header, so that\u0027s\ncorrect.)\n\nFound by inspection. In practice the incorrect transport header\ndoesn\u0027t matter because the skb usually is sent to another net_device\nor socket, so the transport header isn\u0027t referenced.\n\nSigned-off-by: Isaku Yamahata \u003cyamahata@valinux.co.jp\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "861aa6d56de8147249c8a04a4e87fb7c3600cca1", "tree": "61554559d5bdb5a76b03f821b36d1f33fdf1fc22", "parents": [ "a49675988c127b5b5876c252e5db2ee0410a10c2" ], "author": { "name": "Isaku Yamahata", "email": "yamahata@valinux.co.jp", "time": "Mon Dec 24 16:51:03 2012 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Dec 26 15:19:56 2012 -0800" }, "message": "ipv4/ip_gre: set transport header correctly to gre header\n\nipgre_tunnel_xmit() incorrectly sets transport header to inner payload\ninstead of GRE header. It seems copy-and-pasted from ipip.c.\nSo set transport header to gre header.\n(In ipip case the transport header is the inner ip header, so that\u0027s\ncorrect.)\n\nFound by inspection. In practice the incorrect transport header\ndoesn\u0027t matter because the skb usually is sent to another net_device\nor socket, so the transport header isn\u0027t referenced.\n\nSigned-off-by: Isaku Yamahata \u003cyamahata@valinux.co.jp\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "a49675988c127b5b5876c252e5db2ee0410a10c2", "tree": "06775f343edbf3d15a32e7779d3767a023628b6e", "parents": [ "f2e9bd70327d788011cf787a51ceba5925bbc63a" ], "author": { "name": "Marciniszyn, Mike", "email": "mike.marciniszyn@intel.com", "time": "Fri Dec 21 08:01:54 2012 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Dec 26 15:17:37 2012 -0800" }, "message": "IB/rds: suppress incompatible protocol when version is known\n\nAdd an else to only print the incompatible protocol message\nwhen version hasn\u0027t been established.\n\nSigned-off-by: Mike Marciniszyn \u003cmike.marciniszyn@intel.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "f2e9bd70327d788011cf787a51ceba5925bbc63a", "tree": "af2b58950283f0bf49b9f725dd30ab306d28a610", "parents": [ "af9b078e354614fad1e2e0cef04e216ae37cc389" ], "author": { "name": "Marciniszyn, Mike", "email": "mike.marciniszyn@intel.com", "time": "Fri Dec 21 08:01:49 2012 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Dec 26 15:17:37 2012 -0800" }, "message": "IB/rds: Correct ib_api use with gs_dma_address/sg_dma_len\n\n0b088e00 (\"RDS: Use page_remainder_alloc() for recv bufs\")\nadded uses of sg_dma_len() and sg_dma_address(). This makes\nRDS DOA with the qib driver.\n\nIB ulps should use ib_sg_dma_len() and ib_sg_dma_address\nrespectively since some HCAs overload ib_sg_dma* operations.\n\nSigned-off-by: Mike Marciniszyn \u003cmike.marciniszyn@intel.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "af9b078e354614fad1e2e0cef04e216ae37cc389", "tree": "b8f4b9d5912bb71130be2b2ed6429134b70f561a", "parents": [ "c3ae62af8e755ea68380fb5ce682e60079a4c388" ], "author": { "name": "Yan Burman", "email": "yanb@mellanox.com", "time": "Thu Dec 20 03:36:08 2012 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Dec 26 15:09:55 2012 -0800" }, "message": "net/vxlan: Use the underlying device index when joining/leaving multicast groups\n\nThe socket calls from vxlan to join/leave multicast group aren\u0027t\nusing the index of the underlying device, as a result the stack uses\nthe first interface that is up. This results in vxlan being non functional\nover a device which isn\u0027t the 1st to be up.\nFix this by providing the iflink field to the vxlan instance\nto the multicast calls.\n\nSigned-off-by: Yan Burman \u003cyanb@mellanox.com\u003e\nAcked-by: Stephen Hemminger \u003cshemminger@vyatta.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "c3ae62af8e755ea68380fb5ce682e60079a4c388", "tree": "ef786563e00bac6b58b96cc7f0ebbaf487982fa6", "parents": [ "3d0dcfbd8fa2a1e63fabb5f8edac8b8a27860d98" ], "author": { "name": "Eric Dumazet", "email": "edumazet@google.com", "time": "Wed Dec 26 12:44:34 2012 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Dec 26 15:08:55 2012 -0800" }, "message": "tcp: should drop incoming frames without ACK flag set\n\nIn commit 96e0bf4b5193d (tcp: Discard segments that ack data not yet\nsent) John Dykstra enforced a check against ack sequences.\n\nIn commit 354e4aa391ed5 (tcp: RFC 5961 5.2 Blind Data Injection Attack\nMitigation) I added more safety tests.\n\nBut we missed fact that these tests are not performed if ACK bit is\nnot set.\n\nRFC 793 3.9 mandates TCP should drop a frame without ACK flag set.\n\n\" fifth check the ACK field,\n if the ACK bit is off drop the segment and return\"\n\nNot doing so permits an attacker to only guess an acceptable sequence\nnumber, evading stronger checks.\n\nMany thanks to Zhiyun Qian for bringing this issue to our attention.\n\nSee :\nhttp://web.eecs.umich.edu/~zhiyunq/pub/ccs12_TCP_sequence_number_inference.pdf\n\nReported-by: Zhiyun Qian \u003czhiyunq@umich.edu\u003e\nSigned-off-by: Eric Dumazet \u003cedumazet@google.com\u003e\nCc: Nandita Dukkipati \u003cnanditad@google.com\u003e\nCc: Neal Cardwell \u003cncardwell@google.com\u003e\nCc: John Dykstra \u003cjohn.dykstra1@gmail.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "ad4b3fb7ff9940bcdb1e4cd62bd189d10fa636ba", "tree": "aa9d7bc0e606df0104b47716c90a38c6689fc713", "parents": [ "637704cbc95c02d18741b4a6e7a5d2397f8b28ce" ], "author": { "name": "Christoffer Dall", "email": "cdall@cs.columbia.edu", "time": "Fri Dec 21 13:03:50 2012 -0500" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Wed Dec 26 14:57:34 2012 -0800" }, "message": "mm: Fix PageHead when !CONFIG_PAGEFLAGS_EXTENDED\n\nUnfortunately with !CONFIG_PAGEFLAGS_EXTENDED, (!PageHead) is false, and\n(PageHead) is true, for tail pages. If this is indeed the intended\nbehavior, which I doubt because it breaks cache cleaning on some ARM\nsystems, then the nomenclature is highly problematic.\n\nThis patch makes sure PageHead is only true for head pages and PageTail\nis only true for tail pages, and neither is true for non-compound pages.\n\n[ This buglet seems ancient - seems to have been introduced back in Apr\n 2008 in commit 6a1e7f777f61: \"pageflags: convert to the use of new\n macros\". And the reason nobody noticed is because the PageHead()\n tests are almost all about just sanity-checking, and only used on\n pages that are actual page heads. The fact that the old code returned\n true for tail pages too was thus not really noticeable. - Linus ]\n\nSigned-off-by: Christoffer Dall \u003ccdall@cs.columbia.edu\u003e\nAcked-by: Andrea Arcangeli \u003caarcange@redhat.com\u003e\nCc: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nCc: Will Deacon \u003cWill.Deacon@arm.com\u003e\nCc: Steve Capper \u003cSteve.Capper@arm.com\u003e\nCc: Christoph Lameter \u003ccl@linux.com\u003e\nCc: stable@kernel.org # 2.6.26+\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "3d0dcfbd8fa2a1e63fabb5f8edac8b8a27860d98", "tree": "2f318c33150ddbf86580e7128c1905b232aef6ac", "parents": [ "ccb6e984a1f0ee03c3e2607809fe3809fa872a2e" ], "author": { "name": "Li Zefan", "email": "lizefan@huawei.com", "time": "Tue Dec 25 20:48:24 2012 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Dec 26 14:16:23 2012 -0800" }, "message": "netprio_cgroup: define sk_cgrp_prioidx only if NETPRIO_CGROUP is enabled\n\nsock-\u003esk_cgrp_prioidx won\u0027t be used at all if CONFIG_NETPRIO_CGROUP\u003dn.\n\nSigned-off-by: Li Zefan \u003clizefan@huawei.com\u003e\nAcked-by: Neil Horman \u003cnhorman@tuxdriver.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "ccb6e984a1f0ee03c3e2607809fe3809fa872a2e", "tree": "5243c1ece37dec05c079e7aace60b07de1317d99", "parents": [ "cbc44dbe1fc5df0959bf075405dcef07989fed37" ], "author": { "name": "Richard Cochran", "email": "richardcochran@gmail.com", "time": "Sun Dec 23 21:19:10 2012 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Dec 26 14:15:09 2012 -0800" }, "message": "cpts: fix a run time warn_on.\n\nThis patch fixes a warning in clk_enable by calling clk_prepare_enable\ninstead.\n\nSigned-off-by: Richard Cochran \u003crichardcochran@gmail.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "cbc44dbe1fc5df0959bf075405dcef07989fed37", "tree": "b56613a3b5eb22ad90961c53c7cc5ba9addff6c3", "parents": [ "143cdd8f33909ff5a153e3f02048738c5964ba26" ], "author": { "name": "Richard Cochran", "email": "richardcochran@gmail.com", "time": "Sun Dec 23 21:19:09 2012 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Dec 26 14:15:09 2012 -0800" }, "message": "cpts: fix build error by removing useless code.\n\nThe cpts driver tries to obtain the input clock frequency by calling the\nclock\u0027s internal \u0027recalc\u0027 method. Since \u003cplat/clock.h\u003e has been removed,\nthis code can no longer compile.\n\nHowever, the driver never makes use of the frequency value, so this patch\nfixes the issue by removing the offending code altogether.\n\nSigned-off-by: Richard Cochran \u003crichardcochran@gmail.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "143cdd8f33909ff5a153e3f02048738c5964ba26", "tree": "f702045bb91845840e7d27c69dead53d2bea1793", "parents": [ "cf0be88057baceae033a82d669128b282308c742" ], "author": { "name": "Akinobu Mita", "email": "akinobu.mita@gmail.com", "time": "Wed Dec 26 02:32:10 2012 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Wed Dec 26 14:13:23 2012 -0800" }, "message": "batman-adv: fix random jitter calculation\n\nbatadv_iv_ogm_emit_send_time() attempts to calculates a random integer\nin the range of \u0027orig_interval +- BATADV_JITTER\u0027 by the below lines.\n\n msecs \u003d atomic_read(\u0026bat_priv-\u003eorig_interval) - BATADV_JITTER;\n msecs +\u003d (random32() % 2 * BATADV_JITTER);\n\nBut it actually gets \u0027orig_interval\u0027 or \u0027orig_interval - BATADV_JITTER\u0027\nbecause \u0027%\u0027 and \u0027*\u0027 have same precedence and associativity is\nleft-to-right.\n\nThis adds the parentheses at the appropriate position so that it matches\noriginal intension.\n\nSigned-off-by: Akinobu Mita \u003cakinobu.mita@gmail.com\u003e\nAcked-by: Antonio Quartulli \u003cordex@autistici.org\u003e\nCc: Marek Lindner \u003clindner_marek@yahoo.de\u003e\nCc: Simon Wunderlich \u003csiwu@hrz.tu-chemnitz.de\u003e\nCc: Antonio Quartulli \u003cordex@autistici.org\u003e\nCc: b.a.t.m.a.n@lists.open-mesh.org\nCc: \"David S. Miller\" \u003cdavem@davemloft.net\u003e\nCc: netdev@vger.kernel.org\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "812089e01b9f65f90fc8fc670d8cce72a0e01fbb", "tree": "cc187999ed4f669cd5bed0b9de661e9d99ec3753", "parents": [ "c733b77475707cc3980542c86ee0ad5c841d544c" ], "author": { "name": "Andy Lutomirski", "email": "luto@amacapital.net", "time": "Sat Dec 01 12:37:20 2012 -0800" }, "committer": { "name": "Bjorn Helgaas", "email": "bhelgaas@google.com", "time": "Wed Dec 26 10:43:06 2012 -0700" }, "message": "PCI: Reduce Ricoh 0xe822 SD card reader base clock frequency to 50MHz\n\nOtherwise it fails like this on cards like the Transcend 16GB SDHC card:\n\n mmc0: new SDHC card at address b368\n mmcblk0: mmc0:b368 SDC 15.0 GiB\n mmcblk0: error -110 sending status command, retrying\n mmcblk0: error -84 transferring data, sector 0, nr 8, cmd response 0x900, card status 0xb0\n\nTested on my Lenovo x200 laptop.\n\n[bhelgaas: changelog]\nSigned-off-by: Andy Lutomirski \u003cluto@amacapital.net\u003e\nSigned-off-by: Bjorn Helgaas \u003cbhelgaas@google.com\u003e\nAcked-by: Chris Ball \u003ccjb@laptop.org\u003e\nCC: Manoj Iyer \u003cmanoj.iyer@canonical.com\u003e\nCC: stable@vger.kernel.org" }, { "commit": "c733b77475707cc3980542c86ee0ad5c841d544c", "tree": "5a2fed17d994e0e88737283b3af5f0fa20375cb1", "parents": [ "130f1b8f35f14d27c43da755f3c9226318c17f57" ], "author": { "name": "Huang Ying", "email": "ying.huang@intel.com", "time": "Wed Dec 26 10:39:23 2012 -0700" }, "committer": { "name": "Bjorn Helgaas", "email": "bhelgaas@google.com", "time": "Wed Dec 26 10:39:23 2012 -0700" }, "message": "PCI/PM: Do not suspend port if any subordinate device needs PME polling\n\nUlrich reported that his USB3 cardreader does not work reliably when\nconnected to the USB3 port. It turns out that USB3 controller failed to\nawaken when plugging in the USB3 cardreader. Further experiments found\nthat the USB3 host controller can only be awakened via polling, not via PME\ninterrupt. But if the PCIe port to which the USB3 host controller is\nconnected is suspended, we cannot poll the controller because its config\nspace is not accessible when the PCIe port is in a low power state.\n\nTo solve the issue, the PCIe port will not be suspended if any subordinate\ndevice needs PME polling.\n\n[bhelgaas: use bool consistently rather than mixing int/bool]\nReference: http://lkml.kernel.org/r/50841CCC.9030809@uli-eckhardt.de\nReported-by: Ulrich Eckhardt \u003cusb@uli-eckhardt.de\u003e\nTested-by: Sarah Sharp \u003csarah.a.sharp@linux.intel.com\u003e\nSigned-off-by: Huang Ying \u003cying.huang@intel.com\u003e\nSigned-off-by: Bjorn Helgaas \u003cbhelgaas@google.com\u003e\nAcked-by: Rafael J. Wysocki \u003crafael.j.wysocki@intel.com\u003e\nCC: stable@vger.kernel.org\t# v3.6+\n" }, { "commit": "130f1b8f35f14d27c43da755f3c9226318c17f57", "tree": "40429fe6acb8aed6729db8e3016bf036c0078e1e", "parents": [ "1278998f8ff6d66044ed00b581bbf14aacaba215" ], "author": { "name": "Bjorn Helgaas", "email": "bhelgaas@google.com", "time": "Wed Dec 26 10:39:23 2012 -0700" }, "committer": { "name": "Bjorn Helgaas", "email": "bhelgaas@google.com", "time": "Wed Dec 26 10:39:23 2012 -0700" }, "message": "PCI: Add PCIe Link Capability link speed and width names\n\nAdd standard #defines for the Supported Link Speeds field in the PCIe\nLink Capabilities register.\n\nNote that prior to PCIe spec r3.0, these encodings were defined:\n\n 0001b 2.5GT/s Link speed supported\n 0010b 5.0GT/s and 2.5GT/s Link speed supported\n\nStarting with spec r3.0, these encodings refer to bits 0 and 1 in the\nSupported Link Speeds Vector in the Link Capabilities 2 register, and bits\n0 and 1 there mean 2.5 GT/s and 5.0 GT/s, respectively. Therefore, code\nthat followed r2.0 and interpreted 0x1 as 2.5GT/s and 0x2 as 5.0GT/s will\ncontinue to work, and we can identify a device using the new encodings\nbecause it will have a non-zero Link Capabilities 2 register.\n\nSigned-off-by: Bjorn Helgaas \u003cbhelgaas@google.com\u003e\n" }, { "commit": "1278998f8ff6d66044ed00b581bbf14aacaba215", "tree": "624a50cde8213ea2659d803e6df397696f267dd9", "parents": [ "faa48a507fd328013886426f9437fd7e2e7b820b" ], "author": { "name": "Myron Stowe", "email": "myron.stowe@redhat.com", "time": "Wed Dec 26 10:39:23 2012 -0700" }, "committer": { "name": "Bjorn Helgaas", "email": "bhelgaas@google.com", "time": "Wed Dec 26 10:39:23 2012 -0700" }, "message": "PCI: Work around Stratus ftServer broken PCIe hierarchy (fix DMI check)\n\nCommit 284f5f9 was intended to disable the \"only_one_child()\" optimization\non Stratus ftServer systems, but its DMI check is wrong. It looks for\nDMI_SYS_VENDOR that contains \"ftServer\", when it should look for\nDMI_SYS_VENDOR containing \"Stratus\" and DMI_PRODUCT_NAME containing\n\"ftServer\".\n\nTested on Stratus ftServer 6400.\n\nReported-by: Fadeeva Marina \u003castarta@rat.ru\u003e\nReference: https://bugzilla.kernel.org/show_bug.cgi?id\u003d51331\nSigned-off-by: Myron Stowe \u003cmyron.stowe@redhat.com\u003e\nSigned-off-by: Bjorn Helgaas \u003cbhelgaas@google.com\u003e\nCC: stable@vger.kernel.org\t# v3.5+\n" }, { "commit": "faa48a507fd328013886426f9437fd7e2e7b820b", "tree": "61c299361e749c7886d79ec6f0daad40ef947e66", "parents": [ "a49f0d1ea3ec94fc7cf33a7c36a16343b74bd565" ], "author": { "name": "Bjorn Helgaas", "email": "bhelgaas@google.com", "time": "Wed Dec 26 10:39:22 2012 -0700" }, "committer": { "name": "Bjorn Helgaas", "email": "bhelgaas@google.com", "time": "Wed Dec 26 10:39:22 2012 -0700" }, "message": "PCI: Remove spurious error for sriov_numvfs store and simplify flow\n\nIf we request \"num_vfs\" and the driver\u0027s sriov_configure() method enables\nexactly that number (\"num_vfs_enabled\"), we complain \"Invalid value for\nnumber of VFs to enable\" and return an error. We should silently return\nsuccess instead.\n\nAlso, use kstrtou16() since numVFs is defined to be a 16-bit field and\nrework to simplify control flow.\n\nReported-by: Greg Rose \u003cgregory.v.rose@intel.com\u003e\nReference: http://lkml.kernel.org/r/20121214101911.00002f59@unknown\nSigned-off-by: Bjorn Helgaas \u003cbhelgaas@google.com\u003e\nTested-by: Donald Dutile \u003cddutile@redhat.com\u003e\n" }, { "commit": "48c6d1217e3dc743e7d3ad9b9def8d4810d13a85", "tree": "b8c5539ab8c13674dce2ff5a860982ae28c9c21d", "parents": [ "dfb2ea45becb198beeb75350d0b7b7ad9076a38f" ], "author": { "name": "Eric W. Biederman", "email": "ebiederm@xmission.com", "time": "Sat Dec 22 01:52:39 2012 -0800" }, "committer": { "name": "Eric W. Biederman", "email": "ebiederm@xmission.com", "time": "Tue Dec 25 20:05:15 2012 -0800" }, "message": "f2fs: Don\u0027t assign e_id in f2fs_acl_from_disk\n\nWith user namespaces enabled building f2fs fails with:\n\n CC fs/f2fs/acl.o\nfs/f2fs/acl.c: In function ‘f2fs_acl_from_disk’:\nfs/f2fs/acl.c:85:21: error: ‘struct posix_acl_entry’ has no member named ‘e_id’\nmake[2]: *** [fs/f2fs/acl.o] Error 1\nmake[2]: Target `__build\u0027 not remade because of errors.\n\ne_id is a backwards compatibility field only used for file systems\nthat haven\u0027t been converted to use kuids and kgids. When the posix\nacl tag field is neither ACL_USER nor ACL_GROUP assigning e_id is\nunnecessary. Remove the assignment so f2fs will build with user\nnamespaces enabled.\n\nCc: Namjae Jeon \u003cnamjae.jeon@samsung.com\u003e\nCc: Amit Sahrawat \u003ca.sahrawat@samsung.com\u003e\nAcked-by: Jaegeuk Kim \u003cjaegeuk.kim@samsung.com\u003e\nSigned-off-by: \"Eric W. Biederman\" \u003cebiederm@xmission.com\u003e\n" }, { "commit": "1efef832020ef392deb2cd3d74e0c316711245be", "tree": "f9b777202cb0cca945c42aa4aec0ffe0743776ab", "parents": [ "30f0c75858c46a0273ccb838de401b1f5fdebe6f" ], "author": { "name": "Jaegeuk Kim", "email": "jaegeuk.kim@samsung.com", "time": "Wed Dec 19 16:25:21 2012 +0900" }, "committer": { "name": "Jaegeuk Kim", "email": "jaegeuk.kim@samsung.com", "time": "Wed Dec 26 10:39:52 2012 +0900" }, "message": "f2fs: do f2fs_balance_fs in front of dir operations\n\nIn order to conserve free sections to deal with the worst-case scenarios, f2fs\nshould be able to freeze all the directory operations especially when there are\nnot enough free sections. The f2fs_balance_fs() is for this use.\n\nWhen FS utilization becomes almost 100%, directory operations can be failed due\nto -ENOSPC frequently, which produces some dirty node pages occasionally.\n\nPreviously, in such a case, f2fs_balance_fs() is not able to be triggered since\nit is triggered only if the directory operation ends up with success.\n\nSo, this patch triggers f2fs_balance_fs() at first before handling directory\noperations.\n\nSigned-off-by: Jaegeuk Kim \u003cjaegeuk.kim@samsung.com\u003e\n" }, { "commit": "30f0c75858c46a0273ccb838de401b1f5fdebe6f", "tree": "11a1a02a34d8cb4604cffdb30d1b365d0243aae7", "parents": [ "398b1ac5a57219823f942a8d3665b27ab99354de" ], "author": { "name": "Jaegeuk Kim", "email": "jaegeuk.kim@samsung.com", "time": "Wed Dec 19 16:09:19 2012 +0900" }, "committer": { "name": "Jaegeuk Kim", "email": "jaegeuk.kim@samsung.com", "time": "Wed Dec 26 10:39:52 2012 +0900" }, "message": "f2fs: should recover orphan and fsync data\n\nThe recovery routine should do all the time regardless of normal umount action.\n\nSigned-off-by: Jaegeuk Kim \u003cjaegeuk.kim@samsung.com\u003e\n" }, { "commit": "398b1ac5a57219823f942a8d3665b27ab99354de", "tree": "7e4806bf658cdffdd977f83a1365f673b9f8e487", "parents": [ "38e0abdcfb5e69aa61a1e9b474d434afc1c177a9" ], "author": { "name": "Jaegeuk Kim", "email": "jaegeuk.kim@samsung.com", "time": "Wed Dec 19 15:28:39 2012 +0900" }, "committer": { "name": "Jaegeuk Kim", "email": "jaegeuk.kim@samsung.com", "time": "Wed Dec 26 10:39:52 2012 +0900" }, "message": "f2fs: fix handling errors got by f2fs_write_inode\n\nRuslan reported that f2fs hangs with an infinite loop in f2fs_sync_file():\n\n\twhile (sync_node_pages(sbi, inode-\u003ei_ino, \u0026wbc) \u003d\u003d 0)\n\t\tf2fs_write_inode(inode, NULL);\n\nThe reason was revealed that the cold flag is not set even thought this inode is\na normal file. Therefore, sync_node_pages() skips to write node blocks since it\nonly writes cold node blocks.\n\nThe cold flag is stored to the node_footer in node block, and whenever a new\nnode page is allocated, it is set according to its file type, file or directory.\n\nBut, after sudden-power-off, when recovering the inode page, f2fs doesn\u0027t recover\nits cold flag.\n\nSo, let\u0027s assign the cold flag in more right places.\n\nOne more thing:\nIf f2fs_write_inode() returns an error due to whatever situations, there would\nbe no dirty node pages so that sync_node_pages() returns zero.\n(i.e., zero means nothing was written.)\n\nReported-by: Ruslan N. Marchenko \u003cme@ruff.mobi\u003e\nSigned-off-by: Jaegeuk Kim \u003cjaegeuk.kim@samsung.com\u003e\n" }, { "commit": "38e0abdcfb5e69aa61a1e9b474d434afc1c177a9", "tree": "3e221e1dda048aa5c087a7d88e9ba93e987be748", "parents": [ "1362b5e347e27102ea0fa99c9932bca1ecde330f" ], "author": { "name": "Namjae Jeon", "email": "namjae.jeon@samsung.com", "time": "Thu Dec 13 23:44:11 2012 +0900" }, "committer": { "name": "Jaegeuk Kim", "email": "jaegeuk.kim@samsung.com", "time": "Wed Dec 26 10:39:52 2012 +0900" }, "message": "f2fs: fix up f2fs_get_parent issue to retrieve correct parent inode number\n\nTest Case:\n[NFS Client]\nls -lR .\n\n[NFS Server]\nwhile [ 1 ]\ndo\necho 3 \u003e /proc/sys/vm/drop_caches\ndone\n\nError on NFS Client: \"No such file or directory\"\n\nWhen cache is dropped at the server, it results in lookup failure at the\nNFS client due to non-connection with the parent. The default path is it\ninitiates a lookup by calculating the hash value for the name, even though\nthe hash values stored on the disk for \".\" and \"..\" is maintained as zero,\nwhich results in failure from find_in_block due to not matching HASH values.\nFix up, by using the correct hashing values for these entries.\n\nSigned-off-by: Namjae Jeon \u003cnamjae.jeon@samsung.com\u003e\nSigned-off-by: Amit Sahrawat \u003ca.sahrawat@samsung.com\u003e\nSigned-off-by: Jaegeuk Kim \u003cjaegeuk.kim@samsung.com\u003e\n" }, { "commit": "1362b5e347e27102ea0fa99c9932bca1ecde330f", "tree": "c7166c18990972677de0049a83c532c3802d3f60", "parents": [ "dfb7c0ceab57fee7618f4c9c31c5a89254e8530a" ], "author": { "name": "Jaegeuk Kim", "email": "jaegeuk.kim@samsung.com", "time": "Wed Dec 12 19:45:49 2012 +0900" }, "committer": { "name": "Jaegeuk Kim", "email": "jaegeuk.kim@samsung.com", "time": "Wed Dec 26 10:39:51 2012 +0900" }, "message": "f2fs: fix wrong calculation on f_files in statfs\n\nIn f2fs_statfs(), f_files should be the total number of available inodes\ninstead of the currently allocated inodes.\nSo, this patch should resolve the reported bug below.\n\nNote that, showing 10% usage is not a bug, since f2fs reveals whole volume size\nas much as possible and shows the space overhead as *used*.\nThis policy is fair enough with respect to other file systems.\n\n\u003cReported Bug\u003e\n(loop0 is backed by 1GiB file)\n\n$ mkfs.f2fs /dev/loop0\n\nF2FS-tools: Ver: 1.1.0 (2012-12-11)\nInfo: sector size \u003d 512\nInfo: total sectors \u003d 2097152 (in 512bytes)\nInfo: zone aligned segment0 blkaddr: 512\nInfo: format successful\n\n$ mount /dev/loop0 mnt/\n\n$ df mnt/\nFilesystem 1K-blocks Used Available Use% Mounted on\n/dev/loop0 1046528 98312 929784 10%\n/home/zeta/linux-devel/mtd-bench/mnt\n\n$ df mnt/ -i\nFilesystem Inodes IUsed IFree IUse% Mounted on\n/dev/loop0 1 -465918 465919 - /home/zeta/linux-devel/mtd-bench/mnt\n\nNotice IUsed is negative. Also, 10% usage on a fresh f2fs seems too\nmuch to be correct.\n\nReported-and-Tested-by: Ezequiel Garcia \u003celezegarcia@gmail.com\u003e\nSigned-off-by: Jaegeuk Kim \u003cjaegeuk.kim@samsung.com\u003e\n" }, { "commit": "dfb7c0ceab57fee7618f4c9c31c5a89254e8530a", "tree": "562e1bd2ee29b43b1b1c2bd0f5ce30f3c09064d9", "parents": [ "637704cbc95c02d18741b4a6e7a5d2397f8b28ce" ], "author": { "name": "Jaegeuk Kim", "email": "jaegeuk.kim@samsung.com", "time": "Wed Dec 12 09:47:39 2012 +0900" }, "committer": { "name": "Jaegeuk Kim", "email": "jaegeuk.kim@samsung.com", "time": "Wed Dec 26 10:39:51 2012 +0900" }, "message": "f2fs: remove set_page_dirty for atomic f2fs_end_io_write\n\nWe should guarantee not to do *scheduling while atomic*.\nI found, in atomic f2fs_end_io_write(), there is a set_page_dirty() call\nto deal with IO errors.\n\nBut, set_page_dirty() calls:\n -\u003e f2fs_set_data_page_dirty()\n -\u003e set_dirty_dir_page()\n -\u003e cond_resched() which results in scheduling.\n\nIn order to avoid this, I\u0027d like to remove simply set_page_dirty(),\nsince the page is already marked as ERROR and f2fs will be operated\nas the read-only mode as well.\nSo, there is no recovery issue with this.\n\nSigned-off-by: Jaegeuk Kim \u003cjaegeuk.kim@samsung.com\u003e\n" }, { "commit": "dfb2ea45becb198beeb75350d0b7b7ad9076a38f", "tree": "7efa7f2073cf0e0ce085dd78adea855df3033d50", "parents": [ "c876ad7682155958d0c9c27afe9017925c230d64" ], "author": { "name": "Eric W. Biederman", "email": "ebiederm@xmission.com", "time": "Fri Dec 21 20:38:00 2012 -0800" }, "committer": { "name": "Eric W. Biederman", "email": "ebiederm@xmission.com", "time": "Tue Dec 25 16:23:12 2012 -0800" }, "message": "proc: Allow proc_free_inum to be called from any context\n\nWhile testing the pid namespace code I hit this nasty warning.\n\n[ 176.262617] ------------[ cut here ]------------\n[ 176.263388] WARNING: at /home/eric/projects/linux/linux-userns-devel/kernel/softirq.c:160 local_bh_enable_ip+0x7a/0xa0()\n[ 176.265145] Hardware name: Bochs\n[ 176.265677] Modules linked in:\n[ 176.266341] Pid: 742, comm: bash Not tainted 3.7.0userns+ #18\n[ 176.266564] Call Trace:\n[ 176.266564] [\u003cffffffff810a539f\u003e] warn_slowpath_common+0x7f/0xc0\n[ 176.266564] [\u003cffffffff810a53fa\u003e] warn_slowpath_null+0x1a/0x20\n[ 176.266564] [\u003cffffffff810ad9ea\u003e] local_bh_enable_ip+0x7a/0xa0\n[ 176.266564] [\u003cffffffff819308c9\u003e] _raw_spin_unlock_bh+0x19/0x20\n[ 176.266564] [\u003cffffffff8123dbda\u003e] proc_free_inum+0x3a/0x50\n[ 176.266564] [\u003cffffffff8111d0dc\u003e] free_pid_ns+0x1c/0x80\n[ 176.266564] [\u003cffffffff8111d195\u003e] put_pid_ns+0x35/0x50\n[ 176.266564] [\u003cffffffff810c608a\u003e] put_pid+0x4a/0x60\n[ 176.266564] [\u003cffffffff8146b177\u003e] tty_ioctl+0x717/0xc10\n[ 176.266564] [\u003cffffffff810aa4d5\u003e] ? wait_consider_task+0x855/0xb90\n[ 176.266564] [\u003cffffffff81086bf9\u003e] ? default_spin_lock_flags+0x9/0x10\n[ 176.266564] [\u003cffffffff810cab0a\u003e] ? remove_wait_queue+0x5a/0x70\n[ 176.266564] [\u003cffffffff811e37e8\u003e] do_vfs_ioctl+0x98/0x550\n[ 176.266564] [\u003cffffffff810b8a0f\u003e] ? recalc_sigpending+0x1f/0x60\n[ 176.266564] [\u003cffffffff810b9127\u003e] ? __set_task_blocked+0x37/0x80\n[ 176.266564] [\u003cffffffff810ab95b\u003e] ? sys_wait4+0xab/0xf0\n[ 176.266564] [\u003cffffffff811e3d31\u003e] sys_ioctl+0x91/0xb0\n[ 176.266564] [\u003cffffffff810a95f0\u003e] ? task_stopped_code+0x50/0x50\n[ 176.266564] [\u003cffffffff81939199\u003e] system_call_fastpath+0x16/0x1b\n[ 176.266564] ---[ end trace 387af88219ad6143 ]---\n\nIt turns out that spin_unlock_bh(proc_inum_lock) is not safe when\nput_pid is called with another spinlock held and irqs disabled.\n\nFor now take the easy path and use spin_lock_irqsave(proc_inum_lock)\nin proc_free_inum and spin_loc_irq in proc_alloc_inum(proc_inum_lock).\n\nSigned-off-by: \"Eric W. Biederman\" \u003cebiederm@xmission.com\u003e\n" }, { "commit": "c876ad7682155958d0c9c27afe9017925c230d64", "tree": "926064bd7909f60daed3b6b963555e57cab7b520", "parents": [ "8382fcac1b813ad0a4e68a838fc7ae93fa39eda0" ], "author": { "name": "Eric W. Biederman", "email": "ebiederm@xmission.com", "time": "Fri Dec 21 20:27:12 2012 -0800" }, "committer": { "name": "Eric W. Biederman", "email": "ebiederm@xmission.com", "time": "Tue Dec 25 16:10:05 2012 -0800" }, "message": "pidns: Stop pid allocation when init dies\n\nOleg pointed out that in a pid namespace the sequence.\n- pid 1 becomes a zombie\n- setns(thepidns), fork,...\n- reaping pid 1.\n- The injected processes exiting.\n\nCan lead to processes attempting access their child reaper and\ninstead following a stale pointer.\n\nThat waitpid for init can return before all of the processes in\nthe pid namespace have exited is also unfortunate.\n\nAvoid these problems by disabling the allocation of new pids in a pid\nnamespace when init dies, instead of when the last process in a pid\nnamespace is reaped.\n\nPointed-out-by: Oleg Nesterov \u003coleg@redhat.com\u003e\nReviewed-by: Oleg Nesterov \u003coleg@redhat.com\u003e\nSigned-off-by: \"Eric W. Biederman\" \u003cebiederm@xmission.com\u003e\n" }, { "commit": "d096ad0f79a782935d2e06ae8fb235e8c5397775", "tree": "b6590d5236e831a6b4564fb88d3fc791952ee1f8", "parents": [ "0875a2b448fcaba67010850cf9649293a5ef653d" ], "author": { "name": "Michael Tokarev", "email": "mjt@tls.msk.ru", "time": "Tue Dec 25 14:08:16 2012 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Tue Dec 25 14:08:16 2012 -0500" }, "message": "ext4: do not try to write superblock on ro remount w/o journal\n\nWhen a journal-less ext4 filesystem is mounted on a read-only block\ndevice (blockdev --setro will do), each remount (for other, unrelated,\nflags, like suid\u003d\u003enosuid etc) results in a series of scary messages\nfrom kernel telling about I/O errors on the device.\n\nThis is becauese of the following code ext4_remount():\n\n if (sbi-\u003es_journal \u003d\u003d NULL)\n ext4_commit_super(sb, 1);\n\nat the end of remount procedure, which forces writing (flushing) of\na superblock regardless whenever it is dirty or not, if the filesystem\nis readonly or not, and whenever the device itself is readonly or not.\n\nWe only need call ext4_commit_super when the file system had been\npreviously mounted read/write.\n\nThanks to Eric Sandeen for help in diagnosing this issue.\n\nSigned-off-By: Michael Tokarev \u003cmjt@tls.msk.ru\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\nCc: stable@vger.kernel.org\n" }, { "commit": "0875a2b448fcaba67010850cf9649293a5ef653d", "tree": "520abd528cf70057950c4d3f77c0af4e59da93cd", "parents": [ "a28a9178e8fcd9b94f7333184ce78e816c8cb2af" ], "author": { "name": "Eric Sandeen", "email": "sandeen@redhat.com", "time": "Tue Dec 25 13:56:01 2012 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Tue Dec 25 13:56:01 2012 -0500" }, "message": "ext4: include journal blocks in df overhead calcs\n\nTo more accurately calculate overhead for \"bsd\" style\ndf reporting, we should count the journal blocks as\noverhead as well.\n\nSigned-off-by: Eric Sandeen \u003csandeen@redhat.com\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\nTested-by: Eric Whitney \u003cenwlinux@gmail.com\u003e\n" }, { "commit": "a28a9178e8fcd9b94f7333184ce78e816c8cb2af", "tree": "cf841622e5d0eaaa717ee6f10dafcfc185a40132", "parents": [ "ad96f7115593e962dd22a0519021eafaba56f5e3" ], "author": { "name": "Eric Sandeen", "email": "sandeen@redhat.com", "time": "Tue Dec 25 13:33:13 2012 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Tue Dec 25 13:33:13 2012 -0500" }, "message": "ext4: remove unaligned AIO warning printk\n\nAlthough I put this in, I now think it was a bad decision. For most\nusers, there is very little to be done in this case. They get the\nmessage, once per day, with no real context or proposed action. TBH,\nit generates support calls when it probably does not need to; the\nmessage sounds more dire than the situation really is.\n\nJust nuke it. Normal investigation via blktrace or whatnot can\nreveal poor IO patterns if bad performance is encountered.\n\nSigned-off-by: Eric Sandeen \u003csandeen@redhat.com\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "ad96f7115593e962dd22a0519021eafaba56f5e3", "tree": "bb022d40e7bf5808e0a2cb4cacb71e8d6bf828ef", "parents": [ "53e872681fed6a43047e71bf927f77d06f467988" ], "author": { "name": "Andy Lutomirski", "email": "luto@amacapital.net", "time": "Tue Dec 25 13:31:52 2012 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Tue Dec 25 13:31:52 2012 -0500" }, "message": "ext4: fix an incorrect comment about i_mutex\n\ni_mutex is not held when -\u003esync_file is called.\n\nReviewed-by: Jan Kara \u003cjack@suse.cz\u003e\nSigned-off-by: Andy Lutomirski \u003cluto@amacapital.net\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "53e872681fed6a43047e71bf927f77d06f467988", "tree": "8b5061acbaf222b3f25df54ddbcaa0b1123c471a", "parents": [ "4520fb3c3690f2643006d85f09ecb74554c10e95" ], "author": { "name": "Jan Kara", "email": "jack@suse.cz", "time": "Tue Dec 25 13:29:52 2012 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Tue Dec 25 13:29:52 2012 -0500" }, "message": "ext4: fix deadlock in journal_unmap_buffer()\n\nWe cannot wait for transaction commit in journal_unmap_buffer()\nbecause we hold page lock which ranks below transaction start. We\nsolve the issue by bailing out of journal_unmap_buffer() and\njbd2_journal_invalidatepage() with -EBUSY. Caller is then responsible\nfor waiting for transaction commit to finish and try invalidation\nagain. Since the issue can happen only for page stradding i_size, it\nis simple enough to manually call jbd2_journal_invalidatepage() for\nsuch page from ext4_setattr(), check the return value and wait if\nnecessary.\n\nSigned-off-by: Jan Kara \u003cjack@suse.cz\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "4520fb3c3690f2643006d85f09ecb74554c10e95", "tree": "2355adc52f5d861073e37ed9b4cb8ccc165cd519", "parents": [ "d7961c7fa4d2e3c3f12be67e21ba8799b5a7238a" ], "author": { "name": "Jan Kara", "email": "jack@suse.cz", "time": "Tue Dec 25 13:28:54 2012 -0500" }, "committer": { "name": "Theodore Ts\u0027o", "email": "tytso@mit.edu", "time": "Tue Dec 25 13:28:54 2012 -0500" }, "message": "ext4: split off ext4_journalled_invalidatepage()\n\nIn data\u003djournal mode we don\u0027t need delalloc or DIO handling in invalidatepage\nand similarly in other modes we don\u0027t need the journal handling. So split\ninvalidatepage implementations.\n\nSigned-off-by: Jan Kara \u003cjack@suse.cz\u003e\nSigned-off-by: \"Theodore Ts\u0027o\" \u003ctytso@mit.edu\u003e\n" }, { "commit": "8382fcac1b813ad0a4e68a838fc7ae93fa39eda0", "tree": "a484c6a712a7361a07d9514ea86f835b70ad3414", "parents": [ "a49f0d1ea3ec94fc7cf33a7c36a16343b74bd565" ], "author": { "name": "Eric W. Biederman", "email": "ebiederm@xmission.com", "time": "Thu Dec 20 19:26:06 2012 -0800" }, "committer": { "name": "Eric W. Biederman", "email": "ebiederm@xmission.com", "time": "Mon Dec 24 22:53:14 2012 -0800" }, "message": "pidns: Outlaw thread creation after unshare(CLONE_NEWPID)\n\nThe sequence:\nunshare(CLONE_NEWPID)\nclone(CLONE_THREAD|CLONE_SIGHAND|CLONE_VM)\n\nCreates a new process in the new pid namespace without setting\npid_ns-\u003echild_reaper. After forking this results in a NULL\npointer dereference.\n\nAvoid this and other nonsense scenarios that can show up after\ncreating a new pid namespace with unshare by adding a new\ncheck in copy_prodcess.\n\nPointed-out-by: Oleg Nesterov \u003coleg@redhat.com\u003e\nAcked-by: Oleg Nesterov \u003coleg@redhat.com\u003e\nSigned-off-by: \"Eric W. Biederman\" \u003cebiederm@xmission.com\u003e\n" }, { "commit": "cf0be88057baceae033a82d669128b282308c742", "tree": "2d152ce4faa18fe3618c9a3ce280162d4ab4e2de", "parents": [ "d2fe85da52e89b8012ffad010ef352a964725d5f" ], "author": { "name": "Cong Wang", "email": "xiyou.wangcong@gmail.com", "time": "Sun Dec 23 15:23:16 2012 +0000" }, "committer": { "name": "David S. Miller", "email": "davem@davemloft.net", "time": "Mon Dec 24 18:42:58 2012 -0800" }, "message": "arp: fix a regression in arp_solicit()\n\nSedat reported the following commit caused a regression:\n\ncommit 9650388b5c56578fdccc79c57a8c82fb92b8e7f1\nAuthor: Eric Dumazet \u003cedumazet@google.com\u003e\nDate: Fri Dec 21 07:32:10 2012 +0000\n\n ipv4: arp: fix a lockdep splat in arp_solicit\n\nThis is due to the 6th parameter of arp_send() needs to be NULL\nfor the broadcast case, the above commit changed it to an all-zero\narray by mistake.\n\nReported-by: Sedat Dilek \u003csedat.dilek@gmail.com\u003e\nTested-by: Sedat Dilek \u003csedat.dilek@gmail.com\u003e\nCc: Sedat Dilek \u003csedat.dilek@gmail.com\u003e\nCc: Eric Dumazet \u003cedumazet@google.com\u003e\nCc: David S. Miller \u003cdavem@davemloft.net\u003e\nCc: Julian Anastasov \u003cja@ssi.bg\u003e\nSigned-off-by: Cong Wang \u003cxiyou.wangcong@gmail.com\u003e\nAcked-by: Eric Dumazet \u003cedumazet@google.com\u003e\nSigned-off-by: David S. Miller \u003cdavem@davemloft.net\u003e\n" }, { "commit": "637704cbc95c02d18741b4a6e7a5d2397f8b28ce", "tree": "af99b09a5bc5c31a102df4b2fb4fe6cf10c518f4", "parents": [ "4ae0a48b5efc44a95f5e7bb578f9de71fd35bfd0", "0b255e927d47b550620dfd3475ee74b0f52e09c8" ], "author": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Dec 23 09:48:33 2012 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Dec 23 09:48:33 2012 -0800" }, "message": "Merge branch \u0027i2c-embedded/for-next\u0027 of git://git.pengutronix.de/git/wsa/linux\n\nPull i2c __dev* attribute removal from Wolfram Sang:\n \"The squashed patches from Bill to get rid of the __dev* annotations in\n the i2c subsystem. I couldn\u0027t include it in my previous pull request\n due to some dependency with the mfd subsystem. I had this patch in\n linux-next for two days before rc1 and nothing popped up.\"\n\n* \u0027i2c-embedded/for-next\u0027 of git://git.pengutronix.de/git/wsa/linux:\n i2c: remove __dev* attributes from subsystem\n" }, { "commit": "4ae0a48b5efc44a95f5e7bb578f9de71fd35bfd0", "tree": "eb71914a38a71928cb6b528985eabafab3def51b", "parents": [ "9c016d61097cc39427a2f5025bdd97ac633d26a6" ], "author": { "name": "Zlatko Calusic", "email": "zlatko.calusic@iskon.hr", "time": "Sun Dec 23 15:12:54 2012 +0100" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Dec 23 09:46:36 2012 -0800" }, "message": "mm: modify pgdat_balanced() so that it also handles order-0\n\nTeach pgdat_balanced() about order-0 allocations so that we can simplify\ncode in a few places in vmstat.c.\n\nSuggested-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Zlatko Calusic \u003czlatko.calusic@iskon.hr\u003e\nCc: Mel Gorman \u003cmgorman@suse.de\u003e\nCc: Hugh Dickins \u003chughd@google.com\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "9c016d61097cc39427a2f5025bdd97ac633d26a6", "tree": "3a694f797800c0f3c34fb4f6148c824f64b5ca68", "parents": [ "a49f0d1ea3ec94fc7cf33a7c36a16343b74bd565" ], "author": { "name": "Rafael J. Wysocki", "email": "rafael.j.wysocki@intel.com", "time": "Sun Dec 23 14:39:32 2012 +0100" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Sun Dec 23 09:39:14 2012 -0800" }, "message": "Partly revert \"[media] uvcvideo: Set error_idx properly for extended controls API failures\"\n\nCommit f0ed2ce840b3 (\"[media] uvcvideo: Set error_idx properly for\nextended controls API failures\") causes user space to behave incorrectly\non one of my test machines (there is no sound under KDE 4.9.4 using\npulseaudio and there is a knotify4 process occupying one of the CPU\ncores 100% of the time). Reverting that commit entirely fixes the\nproblem for me.\n\nHowever, commit f0ed2ce840b3 appears to do more than it follows from its\nchangelog, because the changelog only says about the changes related to\nctrls-\u003eerror_idx, while the commit additionally changes error codes\nreturned by various functions in uvc_ctrl.c and uvc_v4l2.c. It turns\nout that the changes of the returned error codes confuse the user spce,\nso it is sufficient to revert the part of commit f0ed2ce840b3 not\nmentioned in its changelog to fix the problem.\n\n[ \u0027ENOENT\u0027 is not a valid error return from an ioctl to begin with, and\n I don\u0027t understand how anybody ever even thought it would be. - Linus ]\n\nSigned-off-by: Rafael J. Wysocki \u003crafael.j.wysocki@intel.com\u003e\nCc: Laurent Pinchart \u003claurent.pinchart@ideasonboard.com\u003e\nCc: Mauro Carvalho Chehab \u003cmchehab@redhat.com\u003e\nCc: Hans Verkuil \u003chans.verkuil@cisco.com\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" } ], "next": "43880f709dc59840849e31b01735ac587195ef8a" }