summaryrefslogtreecommitdiffstats
path: root/drivers/block/drbd/drbd_actlog.c
Commit message (Collapse)AuthorAgeFilesLines
...
* drbd: wait for meta data IO completion even with failed disk, unless ↵Lars Ellenberg2012-11-091-10/+7
| | | | | | | | | | | | | force-detached The intention of force-detach is to be able to deal with a completely unresponsive lower level IO stack, which does not even deliver error completions anymore, but no completion at all. In all other cases, we must still wait for the meta data IO completion. Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: announce FLUSH/FUA capability to upper layersLars Ellenberg2012-11-091-3/+9
| | | | | | | | In 8.4, we may have bios spanning two activity log extents. Fixup drbd_al_begin_io() and drbd_al_complete_io() to deal with zero sized bios. Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: differentiate between normal and forced detachLars Ellenberg2012-11-081-1/+1
| | | | | | | | | | | | | | | Aborting local requests (not waiting for completion from the lower level disk) is dangerous: if the master bio has been completed to upper layers, data pages may be re-used for other things already. If local IO is still pending and later completes, this may cause crashes or corrupt unrelated data. Only abort local IO if explicitly requested. Intended use case is a lower level device that turned into a tarpit, not completing io requests, not even doing error completion. Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: move the drbd_work_queue from drbd_socket to drbd_connectionLars Ellenberg2012-11-081-2/+2
| | | | | | | | | | | | | | cherry-picked and adapted from drbd 9 devel branch In 8.4, we don't distinguish between "resource work" and "connection work" yet, we have one worker for both, as we still have only one connection. We only ever used the "data.work", no need to keep the "meta.work" around. Move tconn->data.work to tconn->sender_work. Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: New disk option al-updatesPhilipp Reisner2012-11-081-2/+10
| | | | | | | | | | By disabling al-updates one might increase performace. The price for that is that in case a crashed primary (that had al-updates disabled) is reintegraded, it will receive a full-resync instead of a bitmap based resync. Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: missing wakeup after drbd_rs_del_allLars Ellenberg2012-11-081-0/+1
| | | | | Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: Consider the disk-timeout also for meta-data IO operationsPhilipp Reisner2012-11-081-3/+15
| | | | | | | | | | | | | | | | | | If the backing device is already frozen during attach, we failed to recognize that. The current disk-timeout code works on top of the drbd_request objects. During attach we do not allow IO and therefore never generate a drbd_request object but block before that in drbd_make_request(). This patch adds the timeout to all drbd_md_sync_page_io(). Before this patch we used to go from D_ATTACHING directly to D_DISKLESS if IO failed during attach. We can no longer do this since we have to stay in D_FAILED until all IO ops issued to the backing device returned. Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: fix potential spinlock deadlockPhilipp Reisner2012-11-081-8/+12
| | | | | | | | | | | | | | | | drbd_try_clear_on_disk_bm() has a sanity check for the number of blocks left to be resynced (rs_left) in the current resync extent. If it detects a mismatch, it complains, and forces a disconnect using drbd_force_state(mdev, NS(conn, C_DISCONNECTING)); Unfortunately, this may be called while holding the req_lock, and drbd_force_state() want's to aquire that lock itself. Deadlock. Don't force a disconnect, but fix up rs_left by recounting and reassigning the number of dirty blocks in that extent. Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: Silenced compiler warningsPhilipp Reisner2012-11-081-8/+3
| | | | | | | | Since version 4.6.1 gcc warns about variables that get a value assigned, but which are never read later on. Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: Improve error reporting in drbd_md_sync_page_io()Andreas Gruenbacher2012-11-081-2/+2
| | | | | Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: Remove dead codePhilipp Reisner2012-11-081-8/+0
| | | | | Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: Fixed w_restart_disk_io() to handle non active AL-extentsPhilipp Reisner2012-11-081-25/+45
| | | | | | | | | | | Since we now apply the AL in user space onto the bitmap, the AL is not active for the requests we want to reply. For that a al_write_transaction() that might be called from worker context became necessary. Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: on attach, enforce clean meta dataLars Ellenberg2012-11-081-259/+0
| | | | | | | | | | | | | | | | | | | | | Detection of unclean shutdown has moved into user space. The kernel code will, whenever it updates the meta data, mark it as "unclean", and will refuse to attach to such unclean meta data. "drbdadm up" now schedules "drbdmeta apply-al", which will apply the activity log to the bitmap, and/or reinitialize it, if necessary, as well as set a "clean" indicator flag. This moves a bit code out of kernel space. As a side effect, it also prevents some 8.3 module from accidentally ignoring the 8.4 style activity log, if someone should downgrade, whether on purpose, or accidentally because he changed kernel versions without providing an 8.4 for the new kernel, and the new kernel comes with in-tree 8.3. Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: detach from frozen backing devicePhilipp Reisner2012-11-081-17/+58
| | | | | | | | | | | | | | | | | | | | * drbd-8.3: documentation: Documented detach's --force and disk's --disk-timeout drbd: Implemented the disk-timeout option drbd: Force flag for the detach operation drbd: Allow new IOs while the local disk in in FAILED state drbd: Bitmap IO functions can not return prematurely if the disk breaks drbd: Added a kref to bm_aio_ctx drbd: Hold a reference to ldev while doing meta-data IO drbd: Keep a reference to the bio until the completion handler finished drbd: Implemented wait_until_done_or_disk_failure() drbd: Replaced md_io_mutex by an atomic: md_io_in_use drbd: moved md_io into mdev drbd: Immediately allow completion of IOs, that wait for IO completions on a failed disk drbd: Keep a reference to barrier acked requests Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: skip spurious wait_event in drbd_al_begin_ioLars Ellenberg2012-11-081-8/+10
| | | | | | | | | | | | | | | | | | | Activity log transaction writes are serialized on a bit lock. If several CPUs race to write an AL transaction, those that did not get the lock the first time may continue as soon as there are no more pending transactions. The do not need to all grab the lock in turn, just to realize that the AL is clean already, and they have nothing to do. This also closes a potential deadlock with drbd_adm_disk_opts. Once it got the AL bit lock, it knows there are no pending transactions, the AL is clean, and it should be safe to wait for all element references to drop to zero. Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: fix copy/paste error in commentLars Ellenberg2012-11-081-1/+1
| | | | | Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: improvements to activate/deactivate multiple activity log extentsLars Ellenberg2012-11-081-23/+23
| | | | | | | | | | | | | Recent commit drbd: get rid of bio_split, allow bios of "arbitrary" size had a reference count leak: it only deactivated the first of several activity log extents for intervals crossing extent boundaries. This commit generalizes on bios spanning multiple activity log extents in drbd_al_begin_io, and adds the necessary loop around lc_put in drbd_al_complete_io as well. Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: prepare to activate two activity log extents at onceLars Ellenberg2012-11-081-5/+13
| | | | | Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: preparation commit, pass drbd_interval to drbd_al_begin/complete_ioLars Ellenberg2012-11-081-4/+4
| | | | | | | | | We want to avoid bio_split for bios crossing activity log boundaries. So we may need to activate two activity log extents "atomically". drbd_al_begin_io() needs to know more than just the start sector. Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: introduce the "initialized" activity log transaction typeLars Ellenberg2012-11-081-2/+14
| | | | | | | | | So we can initialize a clean on disk activity log area, without the module complaining with loud assert messages because of checksum or magic value mismatches. Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: drbd_md_sync_page_io(): Return 0 upon success and an error code otherwiseAndreas Gruenbacher2012-11-081-8/+6
| | | | | Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: _drbd_md_sync_page_io(): Return 0 upon success and an error code otherwiseAndreas Gruenbacher2012-11-081-6/+7
| | | | | Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: Make all worker callbacks return 0 upon success and an error code ↵Andreas Gruenbacher2012-11-081-8/+8
| | | | | | | otherwise Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: Temporarily change the return type of all worker callbacksAndreas Gruenbacher2012-11-081-3/+3
| | | | | | | | This helps to ensure that we don't miss one of them when changing their return value semantics. Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: Do not modify the connection state with something else that ↵Philipp Reisner2012-11-041-1/+1
| | | | | | | conn_request_state() Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: switch configuration interface from connector to genetlinkLars Ellenberg2012-11-041-1/+2
| | | | | Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: introduce a bio_set to allocate housekeeping bios fromLars Ellenberg2011-10-141-1/+1
| | | | | | | | Don't rely on availability of bios from the global fs_bio_set, we should use our own bio_set for meta data IO. Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: new on-disk activity log transaction formatLars Ellenberg2011-10-141-156/+253
| | | | | | | | | | | Use a new on-disk transaction format for the activity log, which allows for multiple changes to the active set per transaction. Using 4k transaction blocks, we can now get rid of the work-around code to deal with devices not supporting 512 byte logical block size. Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* lru_cache: allow multiple changes per transactionLars Ellenberg2011-10-141-38/+12
| | | | | | | | Allow multiple changes to the active set of elements in lru_cache. The only current user of lru_cache, drbd, is driving this generalisation. Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: fix typo in commentLars Ellenberg2011-10-141-1/+1
| | | | | Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: Use the IS_ALIGNED() macro in some more placesAndreas Gruenbacher2011-10-141-3/+3
| | | | | Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: Generalized the work callbacksPhilipp Reisner2011-09-281-3/+5
| | | | | | | | No longer work callbacks must operate on a mdev. From now on they can also operate on a tconn. Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: Moved the mdev member into drbd_work (from drbd_request and ↵Philipp Reisner2011-09-281-0/+2
| | | | | | | drbd_peer_request) Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: moved data and meta from mdev to tconnPhilipp Reisner2011-08-291-2/+2
| | | | | | | | | | | Patch mostly: sed -i -e 's/mdev->data/mdev->tconn->data/g' \ -e 's/mdev->meta/mdev->tconn->meta/g' \ *.[ch] Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: Replace the ERR_IF macro with an assert-like macroAndreas Gruenbacher2011-08-291-8/+13
| | | | | | | | | Remove the file name and line number from the syslog messages generated: we have no duplicate function names, and no function contains the same assertion more than once. Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: Endianness convert the constants instead of the variablesAndreas Gruenbacher2011-08-251-1/+1
| | | | | | | Converting the constants happens at compile time. Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: we should write meta data updates with FLUSH FUALars Ellenberg2011-06-301-1/+1
| | | | | | | | | | | | We used to write these with BIO_RW_BARRIER aka REQ_HARDBARRIER (unless disabled in the configuration). The correct semantic now would be to write with FLUSH/FUA. For example, with activity log transactions, FUA alone is not enough, we need the corresponding bitmap update (and all related application updates) on stable storage as well. Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: Fix spellingBart Van Assche2011-05-241-1/+1
| | | | | | | | Found these with the help of ispell -l. Signed-off-by: Bart Van Assche <bvanassche@acm.org> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com> Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com>
* Fix common misspellingsLucas De Marchi2011-03-311-1/+1
| | | | | | Fixes generated by 'codespell' and manually reviewed. Signed-off-by: Lucas De Marchi <lucas.demarchi@profusion.mobi>
* drbd: Remove unused function atodb_endio()Andreas Gruenbacher2011-03-101-27/+0
| | | | | Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: kill dead codeLars Ellenberg2011-03-101-93/+0
| | | | | | | | This code became obsolete and unused last December with drbd: bitmap keep track of changes vs on-disk bitmap Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: bitmap keep track of changes vs on-disk bitmapLars Ellenberg2011-03-101-103/+29
| | | | | | | | | | When we set or clear bits in a bitmap page, also set a flag in the page->private pointer. This allows us to skip writes of unchanged pages. Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: Use the standard bool, true, and false keywordsAndreas Gruenbacher2011-03-101-4/+4
| | | | | Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: Get rid of unnecessary macros (2)Andreas Gruenbacher2011-03-101-2/+2
| | | | | | | | The FAULT_ACTIVE macro just wraps the drbd_insert_fault macro for no apparent reason. Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: Removed 20 seconds upper bound for side-steppingPhilipp Reisner2011-03-101-3/+4
| | | | | | | | | | | | Given low-enough network bandwidth combined with a IO pattern that hammers onto a single RS-extent, side-stepping might be necessary for much longer times. Changed the code to print a single informal message after 20 seconds, but it keeps on stepping aside forever. Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: Implemented side-stepping in drbd_res_begin_io()Philipp Reisner2011-03-101-5/+19
| | | | | | | | | | | | | | | | Before: drbd_rs_begin_io() locked app-IO out of an RS extent, and waited then until all previous app-IO in that area finished. (But not only until the disk-IO was finished but until the barrier/epoch ack came in for that == round trip time latency ++) After: As soon as a new app-IO waits wants to start new IO on that RS extent, drbd_rs_begin_io() steps aside (clearing the BME_NO_WRITES flag again). It retries after 100ms. Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: Implemented priority inheritance for resync requestsPhilipp Reisner2011-03-101-2/+2
| | | | | | | | | | | | We only issue resync requests if there is no significant application IO going on. = Application IO has higher priority than resnyc IO. If application IO can not be started because the resync process locked an resync_lru entry, start the IO operations necessary to release the lock ASAP. Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: New packet for Ahead/Behind mode: P_OUT_OF_SYNCPhilipp Reisner2011-03-101-4/+6
| | | | | Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: properly use max_hw_sectors to limit the our bio sizeLars Ellenberg2011-03-101-4/+4
| | | | | | | | | | | | | To ease tracking of bios in some hash tables, we want it to not cross certain boundaries (128k, used to be 32k). We limit the maximum bio size using queue parameters. Historically some defines and variables we use there have been named max_segment_size, which was misguided. Rename them to max_bio_size, and use [blk_]queue_max_hw_sectors where appropriate. Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
* drbd: factor out advancement of resync marks for progress reportingLars Ellenberg2011-03-101-13/+17
| | | | | | | | This is in preparation to unify progress reporting of online-verify and resync requests. Signed-off-by: Philipp Reisner <philipp.reisner@linbit.com> Signed-off-by: Lars Ellenberg <lars.ellenberg@linbit.com>
generated by cgit v1.2.3 (git 2.25.1) at 2024-10-11 01:09:00 +0000