summaryrefslogtreecommitdiff
path: root/sys/dev/raidframe
AgeCommit message (Expand)Author
2004-03-02Another spot where we're willing to wait for pool_get to give us somethingoster
2004-03-02A few more cases where RF_DEBUG_PSS can be used.oster
2004-03-02DO_WAIT and DO_SIGNAL macros are a bit much, as they are only each usedoster
2004-03-02We're willing to wait, so we should always get something valid fromoster
2004-03-02Fix more cases where mutex locks were being held longer than necessary.oster
2004-03-01Use RF_ACC_TRACE to #if out more chunks of code related onlyoster
2004-03-01We always wait until we get memory, so the NULL check doesn't buy us anything.oster
2004-02-29Bah. RF_ACC_TRACE bits need a bit more work before they can be turned off.oster
2004-02-29As a LOCK_DEBUG kernel complains, we should really initialzeoster
2004-02-29Unlock the mcpair mutex before we call rf_DispatchDAG(), and re-lock it onceoster
2004-02-29Oops.. might still need this mutex in a few places.oster
2004-02-29Minor shuffling of variable declarations to clean RF_ACC_TRACE #definesoster
2004-02-29Add "RF_ACC_TRACE" as a new #define to rf_archs.h.oster
2004-02-29Minor comment reformatting.oster
2004-02-29When the 'if' condition is true, the function will return, so thereoster
2004-02-29Adjust _rf_ShutdownCreate() so that it is willing to wait for moreoster
2004-02-29InitHdrNode() might as well return 'void'. Nothing ever pays attentionoster
2004-02-29rf_MakeAllocList() will always get memory. No point in checking foroster
2004-02-29We'd better have gotten a dag header from the pool. In any event, callersoster
2004-02-29TransferDagMemory() doesn't exist, so these lines are just wasting space.oster
2004-02-29Stripe functions are now handled by a linked-list instead of aoster
2004-02-27Add forgotten pool_destroy().oster
2004-02-27Use a dynamically allocated linked list of dagLists instead of using aoster
2004-02-25Poolify alloclist stuff. Saves a whopping 32 bytes while decreasing codeoster
2004-02-24occured -> occurred. From Peter Postma.wiz
2004-02-21unlockQNodes, unlockDataNodes, and unlockParityNodes are not used. Turf.oster
2004-02-08Relegate some printf's to DEBUG land.oster
2004-01-25Make VOP_STRATEGY(bp) a real VOP as discussed on tech-kern.hannken
2004-01-23Always ask for a new RF_StripeLockDesc_t "just in case", and thenoster
2004-01-15Add missing simple_lock_init(). Noted by skrueger-at-europe-dot-com.oster
2004-01-14Add 2 missing RF_LOCK_MUTEX()es. Problem pointed out byoster
2004-01-10Since the LOCK and UNLOCK flags are never used, no need in extracting them.oster
2004-01-10store a i/o priority hint in struct buf for buffer queue discipline.yamt
2004-01-10iCleanup the RF_CREATE_PARAM3(). Middle two "arguments" were nothingoster
2004-01-10More lu_flags disappear into oblivion.oster
2004-01-09Nuke lu_flag and code associated with it being non-zero.oster
2004-01-09[This file should have been committed with the otheroster
2004-01-09Cleanup rf_enableAtomicRMW and its use. According to the comments, weoster
2004-01-06Clean up a few comments, and improve code formatting somewhat.oster
2004-01-05remove terminate_disk_queues from RF_Raid_s. The hist_diskreq[]oster
2004-01-04raidPtr->reconControl->percentCompleted only gets used in oneoster
2004-01-04As noted by Thierry Deval in a posting to misc/at/openbsd.org,oster
2004-01-02Fix the "We panic if we can't create a DAG" problem that's existedoster
2004-01-02Remove "row" from DEBUG statements.hannken
2004-01-01Cleanup some unused desc->flags:oster
2004-01-01Finally cleanup RF_BZERO. Turns out that 'bp' was only used to obtainoster
2004-01-01Tidy up the #includes a bit.oster
2004-01-01Nuke a few more 'row'-related bits that were missed in the row-removal.oster
2004-01-01Nuke a bunch of unused variables:oster
2003-12-31Remove 3 unused debugging options:oster