| Age | Commit message (Collapse) | Author | |
|---|---|---|---|
| 2014-07-12 | add a size argument to free. will be used soon, but for now default to 0. | Ted Unangst | |
| after discussions with beck deraadt kettenis. | |||
| 2014-07-12 | essentially mechanical conversion of softraid rebuild from | Bret Lambert | |
| workq to taskq ok jsing@ | |||
| 2014-07-10 | Stop using a shutdown hook for softraid(4) and explicitly shutdown | Martin Pieuchot | |
| the disciplines right after vfs_shutdown(). This change is required in order to be able to set `cold' to 1 before traversing the device (mainbus) tree for DVACT_POWERDOWN when halting a machine. Yes, this is ugly because sr_shutdown() needs to sleep. But at least it is obvious and hopefully somebody will be ofended and fix it. In order to properly flush the cache of the disks under softraid0, sr_shutdown() now propagates DVACT_POWERDOWN for this particular subtree of devices which are not under mainbus. As a side effect sd(4) shutdown hook should no longer be necessary. Tested by stsp@ and Jean-Philippe Ouellet. ok deraadt@, stsp@, jsing@ | |||
| 2014-01-22 | Add some debug messages to the rebuild process and fix a bogus comment. | Joel Sing | |
| 2014-01-22 | Add a debug flag for rebuild. | Joel Sing | |
| 2014-01-22 | Make rebuild a discipline specific function pointer. For now, this defaults | Joel Sing | |
| to the existing rebuild code. ok krw@ | |||
| 2014-01-22 | The sr_rebuild function does not actually do the rebuild, it only starts | Joel Sing | |
| the rebuild thread. Rename it to sr_rebuild_start to make this clearer. ok krw@ | |||
| 2014-01-22 | Move sr_dump from the RAID5 code into shared code. Rename it to | Joel Sing | |
| sr_dump_block and place it under the debug define in the process. | |||
| 2014-01-22 | Switch metadata saves from the system workq to the system taskq. | Joel Sing | |
| ok dlg@ | |||
| 2014-01-21 | Actually use the correct TAILQ field. | Joel Sing | |
| 2014-01-21 | Allow for variable sized work units. | Joel Sing | |
| ok krw@ | |||
| 2014-01-21 | Use a TAILQ for tracking work units, rather than an array. This will allow | Joel Sing | |
| for variable size work units. While here, remove some pointless checks. ok krw@ | |||
| 2014-01-21 | Stop bzeroing the work unit when it is initialised. Rather, only initialise | Joel Sing | |
| the specific fields that are either not persistent for the life of the workunit and/or not reinitialised in sr_wu_release_ccbs. This is also necessary for upcoming changes. ok krw@ | |||
| 2014-01-20 | Convert the softraid work unit workqs to taskqs. | Joel Sing | |
| ok dlg@ | |||
| 2014-01-20 | Replace dedicated swu_fake variable with a work unit flag. | Joel Sing | |
| ok krw@ | |||
| 2014-01-19 | Assert that a workunit has ccbs if it is being scheduled. Otherwise we | Joel Sing | |
| deadlock since nothing will ever come back to tell us that the workunit is completed. ok krw@ | |||
| 2014-01-18 | Move the block get/put routines into the common code, instead of having | Joel Sing | |
| RAID 6 borrow them from RAID 5. | |||
| 2014-01-18 | Rename softraid RAIDP to softraid RAID5. | Joel Sing | |
| Discussed with krw@ | |||
| 2014-01-18 | Remove the RAID 4 discipline from softraid. Anyone sensible would use RAID | Joel Sing | |
| 5 instead of RAID 4, assuming both were functional. Discussed with krw@ | |||
| 2014-01-05 | On some architectures you can get away with using an uninitialised mutex; | Joel Sing | |
| this is not the case on hppa (practically the only architecture where unlocked has a non-zero value). This one has been hiding since r1.223... Found the hard way by juanfra@ | |||
| 2013-11-19 | Print daddr_t's with %lld, size_t's with %zu, int's with %d and | Kenneth R Westerback | |
| u_int32_t's with %u. Translate DL_GETPSIZE() results to 512-byte-block values with DL_SECTOBLK(). ok jsing@ | |||
| 2013-11-04 | substantial namespace cleanup. Might go a little bit too far, but we | Theo de Raadt | |
| can expose some of the kernel structures with split .h files if need be. Discussed with various, including jsing. | |||
| 2013-11-01 | Sprinkle (long long) casts where %lld is being used to print daddr_t | Kenneth R Westerback | |
| variables. Some random whitespace/knf repairs encountered on the way. ok miod@ on inspection, feedback & more suggestions from millert@ | |||
| 2013-07-19 | Don't use uninitialized variable when a simple 0 will do. | Kenneth R Westerback | |
| Found by Maxime Villard. ok jsing@ | |||
| 2013-07-01 | When an I/O error occurs on a softraid chunk, only take it offline if the | Joel Sing | |
| discipline supports redundancy. In the non-redundant case, there is little to gain my failing the chunk, in fact it just makes any form of data recovery significantly harder. ok krw@ todd@ | |||
| 2013-06-11 | final removal of daddr64_t. daddr_t has been 64 bit for a long enough | Theo de Raadt | |
| test period; i think 3 years ago the last bugs fell out. ok otto beck others | |||
| 2013-05-21 | Ensure that sc_lock is held when calling sr_meta_native_bootprobe() - the | Joel Sing | |
| softraid status functions will complain otherwise. Found the hard way by ajacoutot@ | |||
| 2013-05-21 | sr_raid_startwu() is no longer called from interrupt context, hence does | Joel Sing | |
| not need to invoke a workq-based callback. | |||
| 2013-05-21 | Provide a function that handles the scheduling of work units. This | Joel Sing | |
| simplifies the discipline code, avoids code duplication and moves the scheduling logic into a single location. ok krw@ | |||
| 2013-05-21 | Make the sr_raid_startwu() function handle deferred work units, instead of | Joel Sing | |
| the caller having to handle the state changes. | |||
| 2013-05-21 | Fix missing work unit state initialisations. | Joel Sing | |
| 2013-05-21 | Use a state to indicate that a work unit should only be constructed and not | Joel Sing | |
| scheduled, rather than trying to imply this from the rebuild flag. | |||
| 2013-04-26 | Add a SR_WUF_DISCIPLINE flag that identifies work units that have resulted | Joel Sing | |
| from discipline specific I/O. Such work units are not associated with a SCSI xfer and are returned via sr_wu_put() on completion. | |||
| 2013-04-23 | Rename ccb_flag to ccb_flags. | Joel Sing | |
| 2013-04-21 | Convert RAID1 to the new work unit completion functions and generic | Joel Sing | |
| interrupt handler. Disciplines such as RAID1/4/5/6 need a way to intercept I/O when the work unit is complete, but before the SCSI xfer is complete. This is provided via a sd_scsi_wu_done hook, which enables work units to be restarted or otherwise modified before the SCSI xfer completion occurs. ok krw@ | |||
| 2013-03-31 | Use the discipline name instead of the product string when reporting the | Joel Sing | |
| attachment. | |||
| 2013-03-31 | The return value of sd_free_resources() is never checked and all of the | Joel Sing | |
| implementations only ever returned zero. Change the return value from int to void instead of pretending it might mean something. | |||
| 2013-03-31 | Fix compilation with debug. | Joel Sing | |
| 2013-03-31 | Provide default resource allocation and free functions. Convert all | Joel Sing | |
| disciplines except for AOE and CRYPTO, which require custom handlers. | |||
| 2013-03-30 | Provide wrappers for scsi_io_get() and scsi_io_put(), that also include | Joel Sing | |
| the sd_sync check/wakeup. Remove some unnecessary NULL checks whilst here. ok krw@ | |||
| 2013-03-30 | Provide a default discipline interrupt handling function and migrate all of | Joel Sing | |
| the disciplines that now have the same interrupt code. ok krw@ | |||
| 2013-03-29 | Decouple wakeups on work unit completion from the type of I/O being | Joel Sing | |
| performed. ok krw@ | |||
| 2013-03-29 | sd_wu_pending is only decremented when scsi_done() or scsi_io_put() are | Joel Sing | |
| called. As a result, factor out the the sd_sync check/wakeup code and move it to after the scsi_done() call in sr_scsi_done(). Perform the same sd_sync check/wakeup after scsi_io_put() calls (including the addition of some that were previously missed). ok krw@ | |||
| 2013-03-25 | Factor out the code that is used to recreate work units - one copy of the | Joel Sing | |
| code is sufficient. ok krw@ | |||
| 2013-03-10 | Rename sc_dis to sc_targets, since it is an array of targets that also | Joel Sing | |
| happen to be softraid disciplines. ok krw@ | |||
| 2013-03-05 | Track attached softraid disciplines via a queue. This prevents the need to | Joel Sing | |
| scan the sparse SCSI targets array and simplifies code. Disciplines are now also shutdown in reverse attach order, making manually stacked softraid volumes somewhat more practical. ok krw@ | |||
| 2013-01-18 | The workunit completion needs to run at IPL_BIO, at least for now. | Joel Sing | |
| 2013-01-18 | Tweak workunit completion. | Joel Sing | |
| 2013-01-18 | Implement common workunit completion functions and handle the completion | Joel Sing | |
| via a workq callback. Also provide a discipline specific hook that allows it to optionally handle the final part of the workunit completion. ok krw@ | |||
| 2013-01-18 | Reset I/O counters when releasing ccbs. | Joel Sing | |
| ok krw@ | |||
 |
index : src | |
| OpenBSD base system |
| summaryrefslogtreecommitdiff |