aboutsummaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAge
...
| * | | | ARM: i.MX5: Add nand oftree supportSascha Hauer2012-07-06
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This adds snippets to the i.MX51/53 devicetrees for the nand flash controller. Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de> Acked-by: Shawn Guo <shawn.guo@linaro.org> Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com> Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
| * | | | mtd: mxc_nand: add i.MX53 supportSascha Hauer2012-07-06
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The only relevant change between i.MX51 and i.MX53 is that a bitfield is shifted one bit to the left. Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de> Acked-by: Shawn Guo <shawn.guo@linaro.org> Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com> Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
| * | | | mtd: mxc_nand: swap iomem resource orderSascha Hauer2012-07-06
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The i.MX v3 nand controller (i.MX5) needs two memory resources. Traditionally we have the AXI resource first. For sorting in this driver into the devicetree it feels much more natural to have the IP resource first. This patch swaps the ordering of these two resources. Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de> Acked-by: Shawn Guo <shawn.guo@linaro.org> Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com> Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
| * | | | mtd: mxc_nand: Use managed resourcesSascha Hauer2012-07-06
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | To make the error path simpler and to make subsequent patches easier. Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de> Acked-by: Shawn Guo <shawn.guo@linaro.org> Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com> Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
| * | | | mtd: elbc nand: use drvdata to only remove the relevant chipScott Wood2012-07-06
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Previously the remove method was looping and removing all chips, which is obviously not the right thing to do — left over from when the driver was organized differently and that was the remove method for the entire controller. This would result in bad things happening if you have more than one NAND chip, and remove the module. This also fixes priv->dev to properly point to the chip's device rather than the controller's. Until now priv->dev was only used for error/debug prints (and it's an improvement there), so this shouldn't break anything. Signed-off-by: Scott Wood <scottwood@freescale.com> Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com> Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
| * | | | mtd: add JEDEC ID for w25q32dw to chip tableing. Federico Fuga2012-07-06
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Adds JEDEC ID for the 1.8V version of WinBond w25q32. Signed-off-by: Federico Fuga <fuga@studiofuga.com> Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com> Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
| * | | | nand: mxc_nand: Use clk_prepare_enable/clk_disable_unprepareFabio Estevam2012-07-06
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Prepare the clock before enabling it. Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com> Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com> Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
| * | | | mtd: nand: kill NAND_NO_READRDYBrian Norris2012-07-06
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | According to its documentation, the NAND_NO_READRDY option is always used when autoincrement is not supported. Autoincrement support was recently dropped, so we can drop this options as well (defaulting to "no read ready check"). Signed-off-by: Brian Norris <computersforpeace@gmail.com> Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com> Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
| * | | | mtd: nand: remove NAND_BBT_SEARCH optionBrian Norris2012-07-06
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This option was never used and isn't currently used. Signed-off-by: Brian Norris <computersforpeace@gmail.com> Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com> Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
| * | | | mtd: nand: add Eon Silicon Solutions manufacturer IDBrian Norris2012-07-06
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Eon's new NAND flash: EN27LN1G08. Signed-off-by: Brian Norris <computersforpeace@gmail.com> Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com> Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
| * | | | mtd: tests: use random32 instead of home-brewed generatorArtem Bityutskiy2012-07-06
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This is a clean-up patch which removes the own pseudo-random numbers generator from the speed- and stress-tests and makes them use the 'random32()' generator instead. [dwmw2: Merge later fix for negative offsets] Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com> Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
| * | | | mtd: sh_flctl: Use user oob data in hardware ECC modeBastian Hecht2012-07-06
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In hardware ecc mode, the flctl now writes and reads the oob data provided by the user. Additionally the ECC is now returned in normal page reads, not only when using the explicit NAND_CMD_READOOB command. Signed-off-by: Bastian Hecht <hechtb@gmail.com> Acked-by: Brian Norris <computersforpeace@gmail.com> Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com> Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
| * | | | mtd: sh_flctl: Restructure the hardware ECC handlingBastian Hecht2012-07-06
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | There are multiple reasons for a rewrite: - a race exists: when _4ECCEND is set, _4ECCFA may become true too meanwhile, which is lost and a non-correctable error is treated as correctable. - the ECC statistics don't get properly propagated to the base code. - empty pages would get marked as corrupted The rewrite resolves the issues and I hope it gives a more explicit code flow structure. Signed-off-by: Bastian Hecht <hechtb@gmail.com> Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com> Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
| * | | | mtd: sh_flctl: Group sector accesses into a single transferBastian Hecht2012-07-06
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When we use hardware ecc, the flctl is run in so-called "sector access mode". We can bundle 4 sector accesses when using 2KiB page sizes to read a whole page at once and speed up things. Signed-off-by: Bastian Hecht <hechtb@gmail.com> Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com> Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
| * | | | mtd: sh_flctl: Simplify the hardware ecc page read/writeBastian Hecht2012-07-06
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | As the equation mtd->writesize == eccsteps * eccsize holds, we can simplify the code. The second loop of the 1st hunk is never entered, so we delete it. Signed-off-by: Bastian Hecht <hechtb@gmail.com> Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com> Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
| * | | | mtd: sh_flctl: Fix hardware ECC behaviourBastian Hecht2012-07-06
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The flctl uses 10 bytes ECC data for every 512 bytes sector. This patch makes the controller write all 40 bytes instead of 10 bytes only. Signed-off-by: Bastian Hecht <hechtb@gmail.com> Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com> Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
| * | | | mtd: sh_flctl: Use different OOB layoutBastian Hecht2012-07-06
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The flctl hardware has changed and a new OOB layout must be adapted for 2KiB page size NAND chips when using hardware ECC. The related bit fields ECCPOS[0-2] are gone — the bits are marked as reserved now in the datasheet. As there are no official users of the hardware ECC so far, they are completely removed. Signed-off-by: Bastian Hecht <hechtb@gmail.com> Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com> Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
| * | | | mtd: sh_flctl: Add support for error IRQBastian Hecht2012-07-06
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When the data transfer between the controller and the NAND chip fails, we now get notified. Signed-off-by: Bastian Hecht <hechtb@gmail.com> Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com> Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
| * | | | mtd: sh_flctl: Add missing iounmap()Bastian Hecht2012-07-06
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Add the unmapping for the error case and for the driver removal. Signed-off-by: Bastian Hecht <hechtb@gmail.com> Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com> Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com> Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
| * | | | mtd: nand: gpmi: need to use {read,write}_oob_rawBrian Norris2012-07-06
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch is simply an added warning in the comments. Ideally, this patch need not be merged, but rather, a developer will write a proper solution that can use the ecc.read_oob_raw and ecc.write_oob_raw interfaces. Signed-off-by: Brian Norris <computersforpeace@gmail.com> Cc: Huang Shijie <b32955@freescale.com> Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com> Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
| * | | | mtd: cafe_nand: spelling mistakeBrian Norris2012-07-06
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Signed-off-by: Brian Norris <computersforpeace@gmail.com> Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com> Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
| * | | | mtd: mtdoops: refactor loopBrian Norris2012-07-06
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | We can clean up the loop logic a bit, here. This refactoring was enabled in part by: Commit bb4a09866 [mtdoops: clean-up new MTD API usage] Signed-off-by: Brian Norris <computersforpeace@gmail.com> Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com> Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
| * | | | mtd: move mtd_read_oob() definition out of mtd.hBrian Norris2012-07-06
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | mtd_read_oob() will be expanded a little, so don't leave it in the header as a static inline function. Signed-off-by: Brian Norris <computersforpeace@gmail.com> Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com> Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
* | | | | Merge branch 'for-linus' of ↵Linus Torvalds2012-10-09
|\ \ \ \ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs Pull btrfs update from Chris Mason: "This is a large pull, with the bulk of the updates coming from: - Hole punching - send/receive fixes - fsync performance - Disk format extension allowing more hardlinks inside a single directory (btrfs-progs patch required to enable the compat bit for this one) I'm cooking more unrelated RAID code, but I wanted to make sure this original batch makes it in. The largest updates here are relatively old and have been in testing for some time." * 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs: (121 commits) btrfs: init ref_index to zero in add_inode_ref Btrfs: remove repeated eb->pages check in, disk-io.c/csum_dirty_buffer Btrfs: fix page leakage Btrfs: do not warn_on when we cannot alloc a page for an extent buffer Btrfs: don't bug on enomem in readpage Btrfs: cleanup pages properly when ENOMEM in compression Btrfs: make filesystem read-only when submitting barrier fails Btrfs: detect corrupted filesystem after write I/O errors Btrfs: make compress and nodatacow mount options mutually exclusive btrfs: fix message printing Btrfs: don't bother committing delayed inode updates when fsyncing btrfs: move inline function code to header file Btrfs: remove unnecessary IS_ERR in bio_readpage_error() btrfs: remove unused function btrfs_insert_some_items() Btrfs: don't commit instead of overcommitting Btrfs: confirmation of value is added before trace_btrfs_get_extent() is called Btrfs: be smarter about dropping things from the tree log Btrfs: don't lookup csums for prealloc extents Btrfs: cache extent state when writing out dirty metadata pages Btrfs: do not hold the file extent leaf locked when adding extent item ...
| * | | | | btrfs: init ref_index to zero in add_inode_refChris Mason2012-10-09
| | | | | | | | | | | | | | | | | | | | | | | | Signed-off-by: Chris Mason <chris.mason@fusionio.com>
| * | | | | Btrfs: remove repeated eb->pages check in, disk-io.c/csum_dirty_bufferWang Sheng-Hui2012-10-09
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In csum_dirty_buffer, we first get eb from page->private. Then we check if the page is the first page of eb. Later we check it again. Remove the repeated check here. Signed-off-by: Wang Sheng-Hui <shhuiw@gmail.com>
| * | | | | Btrfs: fix page leakageJosef Bacik2012-10-09
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Alloc_dummy_extent_buffer will not free the first page in the eb array if we fail to allocate a page, fix this. Thanks, Reported-by: David Sterba <dave@jikos.cz> Signed-off-by: Josef Bacik <jbacik@fusionio.com>
| * | | | | Btrfs: do not warn_on when we cannot alloc a page for an extent bufferJosef Bacik2012-10-09
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | It's just annoying and the user will have gotten a nice OOM killer message so they are already fully aware they are screwed :). Thanks, Reported-by: Jérôme Poulin <jeromepoulin@gmail.com> Signed-off-by: Josef Bacik <jbacik@fusionio.com>
| * | | | | Btrfs: don't bug on enomem in readpageJosef Bacik2012-10-09
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Get rid of the BUG_ON(ret == -ENOMEM) in __extent_read_full_page. Thanks, Reported-by: Jérôme Poulin <jeromepoulin@gmail.com> Signed-off-by: Josef Bacik <jbacik@fusionio.com>
| * | | | | Btrfs: cleanup pages properly when ENOMEM in compressionJosef Bacik2012-10-09
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | We were freeing non-existent pages which was causing a panic for a user who was suffering from ENOMEM. This patch fixes the problem. Thanks, Reported-by: Jérôme Poulin <jeromepoulin@gmail.com> Signed-off-by: Josef Bacik <jbacik@fusionio.com>
| * | | | | Btrfs: make filesystem read-only when submitting barrier failsStefan Behrens2012-10-09
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | So far the return code of barrier_all_devices() is ignored, which means that errors are ignored. The result can be a corrupt filesystem which is not consistent. This commit adds code to evaluate the return code of barrier_all_devices(). The normal btrfs_error() mechanism is used to switch the filesystem into read-only mode when errors are detected. In order to decide whether barrier_all_devices() should return error or success, the number of disks that are allowed to fail the barrier submission is calculated. This calculation accounts for the worst RAID level of metadata, system and data. If single, dup or RAID0 is in use, a single disk error is already considered to be fatal. Otherwise a single disk error is tolerated. The calculation of the number of disks that are tolerated to fail the barrier operation is performed when the filesystem gets mounted, when a balance operation is started and finished, and when devices are added or removed. Signed-off-by: Stefan Behrens <sbehrens@giantdisaster.de>
| * | | | | Btrfs: detect corrupted filesystem after write I/O errorsStefan Behrens2012-10-09
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In check-integrity, detect when a superblock is written that points to blocks that have not been written to disk due to I/O write errors. Signed-off-by: Stefan Behrens <sbehrens@giantdisaster.de>
| * | | | | Btrfs: make compress and nodatacow mount options mutually exclusiveAndrei Popa2012-10-09
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | If a filesystem is mounted with compression and then remounted by adding nodatacow, the compression is disabled but the compress flag is still visible. Also, if a filesystem is mounted with nodatacow and then remounted with compression, nodatacow flag is still present but it's not active. This patch: - removes compress flags and notifies that the compression has been disabled if the filesystem is mounted with nodatacow - removes nodatacow and nodatasum flags if mounted with compress. Signed-off-by: Andrei Popa <andrei.popa@i-neo.ro>
| * | | | | btrfs: fix message printingDaniel J Blueman2012-10-09
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Fix various messages to include newline and module prefix. Signed-off-by: Daniel J Blueman <daniel@quora.org>
| * | | | | Btrfs: don't bother committing delayed inode updates when fsyncingJosef Bacik2012-10-09
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | We can just copy the in memory inode into the tree log directly, no sense in updating the fs tree so we can copy it into the tree log tree. Thanks, Signed-off-by: Josef Bacik <jbacik@fusionio.com>
| * | | | | btrfs: move inline function code to header fileRobin Dong2012-10-09
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When building btrfs from kernel code, it will report: fs/btrfs/extent_io.h:281: warning: 'extent_buffer_page' declared inline after being called fs/btrfs/extent_io.h:281: warning: previous declaration of 'extent_buffer_page' was here fs/btrfs/extent_io.h:280: warning: 'num_extent_pages' declared inline after being called fs/btrfs/extent_io.h:280: warning: previous declaration of 'num_extent_pages' was here because of the wrong declaration of inline functions. Signed-off-by: Robin Dong <sanbai@taobao.com>
| * | | | | Btrfs: remove unnecessary IS_ERR in bio_readpage_error()Tsutomu Itoh2012-10-09
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Because the value of extent_map is only a correct value or NULL, so IS_ERR is unnecessary. Signed-off-by: Tsutomu Itoh <t-itoh@jp.fujitsu.com>
| * | | | | btrfs: remove unused function btrfs_insert_some_items()Robin Dong2012-10-09
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The function btrfs_insert_some_items() would not be called by any other functions, so remove it. Signed-off-by: Robin Dong <sanbai@taobao.com>
| * | | | | Btrfs: don't commit instead of overcommittingJosef Bacik2012-10-09
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | I don't think we have the same problem that this was supposed to fix originally since we can allocate chunks in the enospc path now. This code is causing us to constantly commit the transaction as we get close to using all of our available space in our currently allocated chunks, instead of allocating another chunk and carrying on with life, which is not nice for performance. Thanks, Signed-off-by: Josef Bacik <jbacik@fusionio.com>
| * | | | | Btrfs: confirmation of value is added before trace_btrfs_get_extent() is calledTsutomu Itoh2012-10-09
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | We should confirm the value of extent_map before calling trace_btrfs_get_extent() because the value of extent_map has the possibility of NULL. Signed-off-by: Tsutomu Itoh <t-itoh@jp.fujitsu.com>
| * | | | | Btrfs: be smarter about dropping things from the tree logJosef Bacik2012-10-09
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When we truncate existing items in the tree log we've been searching for each individual item and removing them. This is unnecessary churn and searching, just keep track of the slot we are on and how many items we need to delete and delete them all at once. Thanks, Signed-off-by: Josef Bacik <jbacik@fusionio.com>
| * | | | | Btrfs: don't lookup csums for prealloc extentsJosef Bacik2012-10-09
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The tree logging stuff was looking up csums to copy over for prealloc extents which is just work we don't need to be doing. Thanks, Signed-off-by: Josef Bacik <jbacik@fusionio.com>
| * | | | | Btrfs: cache extent state when writing out dirty metadata pagesJosef Bacik2012-10-09
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Everytime we write out dirty pages we search for an offset in the tree, convert the bits in the state, and then when we wait we search for the offset again and clear the bits. So for every dirty range in the io tree we are doing 4 rb searches, which is suboptimal. With this patch we are only doing 2 searches for every cycle (modulo weird things happening). Thanks, Signed-off-by: Josef Bacik <jbacik@fusionio.com>
| * | | | | Btrfs: do not hold the file extent leaf locked when adding extent itemJosef Bacik2012-10-09
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | For some reason we unlock everything except the leaf we are on, set the path blocking and then add the extent item for the extent we just finished writing. I can't for the life of me figure out why we would want to do this, and the history doesn't really indicate that there was a real reason for it, so just remove it. This will reduce our tree lock contention on heavy writes. Thanks, Signed-off-by: Josef Bacik <jbacik@fusionio.com>
| * | | | | Btrfs: do not async metadata csumming in certain situationsJosef Bacik2012-10-09
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | There are a coule scenarios where farming metadata csumming off to an async thread doesn't help. The first is if our processor supports crc32c, in which case the csumming will be fast and so the overhead of the async model is not worth the cost. The other case is for our tree log. We will be making that stuff dirty and writing it out and waiting for it immediately. Even with software crc32c this gives me a ~15% increase in speed with O_SYNC workloads. Thanks, Signed-off-by: Josef Bacik <jbacik@fusionio.com>
| * | | | | btrfs: fix min csum item size warnings in 32bitZach Brown2012-10-09
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | commit 7ca4be45a0255ac8f08c05491c6add2dd87dd4f8 limited csum items to PAGE_CACHE_SIZE. It used min() with incompatible types in 32bit which generates warnings: fs/btrfs/file-item.c: In function ‘btrfs_csum_file_blocks’: fs/btrfs/file-item.c:717: warning: comparison of distinct pointer types lacks a cast This uses min_t(u32,) to fix the warnings. u32 seemed reasonable because btrfs_root->leafsize is u32 and PAGE_CACHE_SIZE is unsigned long. Signed-off-by: Zach Brown <zab@zabbo.net>
| * | | | | Btrfs: run delayed refs first when out of spaceJosef Bacik2012-10-09
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Running delayed refs is faster than running delalloc, so lets do that first to try and reclaim space. This makes my fs_mark test about 20% faster. Thanks, Signed-off-by: Josef Bacik <jbacik@fusionio.com>
| * | | | | Btrfs: fix orphan transaction on the freezed filesystemMiao Xie2012-10-09
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | With the following debug patch: static int btrfs_freeze(struct super_block *sb) { + struct btrfs_fs_info *fs_info = btrfs_sb(sb); + struct btrfs_transaction *trans; + + spin_lock(&fs_info->trans_lock); + trans = fs_info->running_transaction; + if (trans) { + printk("Transid %llu, use_count %d, num_writer %d\n", + trans->transid, atomic_read(&trans->use_count), + atomic_read(&trans->num_writers)); + } + spin_unlock(&fs_info->trans_lock); return 0; } I found there was a orphan transaction after the freeze operation was done. It is because the transaction may not be committed when the transaction handle end even though it is the last handle of the current transaction. This design avoid committing the transaction frequently, but also introduce the above problem. So I add btrfs_attach_transaction() which can catch the current transaction and commit it. If there is no transaction, it will return ENOENT, and do not anything. This function also can be used to instead of btrfs_join_transaction_freeze() because it don't increase the writer counter and don't start a new transaction, so it also can fix the deadlock between sync and freeze. Besides that, it is used to instead of btrfs_join_transaction() in transaction_kthread(), because if there is no transaction, the transaction kthread needn't anything. Signed-off-by: Miao Xie <miaox@cn.fujitsu.com>
| * | | | | Btrfs: add a type field for the transaction handleMiao Xie2012-10-09
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch add a type field into the transaction handle structure, in this way, we needn't implement various end-transaction functions and can make the code more simple and readable. Signed-off-by: Miao Xie <miaox@cn.fujitsu.com>
| * | | | | Btrfs: fix memory leak in start_transaction()Miao Xie2012-10-09
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch fixes memory leak of the transaction handle which happened when starting transaction failed on a freezed fs. Signed-off-by: Miao Xie <miaox@cn.fujitsu.com>
generated by cgit v1.2.2 (git 2.25.0) at 2025-08-24 06:16:41 -0400