summaryrefslogtreecommitdiffstats
path: root/src/drivers
Commit message (Collapse)AuthorAgeFilesLines
...
| * ICM allocation code in place; about to start test-and-compare runs.Michael Brown2007-09-182-31/+187
| |
| * ICM allocation code *should* now be complete.Michael Brown2007-09-182-10/+54
| |
| * Started with the ICM partitioning code.Michael Brown2007-09-182-4/+116
| |
| * Restructured bits of initialisation (not testable at present)Michael Brown2007-09-172-42/+116
| |
| * Starting the firmware directly now works.Michael Brown2007-09-172-24/+159
| |
| * Start migrating the remaining initialisation steps.Michael Brown2007-09-172-4/+70
| |
| * Don't use the mailboxless version of the HW2SW_CQ command; it seems toMichael Brown2007-09-171-4/+7
| | | | | | | | crash the machine.
| * Removed more hacks, and adjusted number of queue entries for likely usage.Michael Brown2007-09-172-22/+3Star
| |
| * Remove hacks, and fix leaving the multicast GID.Michael Brown2007-09-171-25/+1Star
| |
| * Added an almost obscene amount of debugging and assertion code whileMichael Brown2007-09-172-9/+49
| | | | | | | | | | | | | | | | tracking down a bug that turned out to be a free_iob() used where I needed a netdev_tx_complete(). This left the freed I/O buffer on the net device's TX list, with bad, bad consequences later. Also fixed the bug in question.
| * Some interesting packet corruption happening now.Michael Brown2007-09-172-43/+181
| |
| * Broadcast GID is now calculated by IPoIB layer.Michael Brown2007-09-172-23/+107
| |
| * Cleaned up some debug messages.Michael Brown2007-09-172-34/+15Star
| |
| * Unicasts seem to be working. :)Michael Brown2007-09-172-37/+118
| |
| * Dead code disablingMichael Brown2007-09-172-18/+4Star
| |
| * Obtains a response to the get path record!Michael Brown2007-09-174-16/+214
| |
| * Use pkey table access to determine broadcast GID directly.Michael Brown2007-09-171-9/+64
| |
| * Prepare for adding a metadata queue to IPoIBMichael Brown2007-09-171-51/+120
| |
| * Dead code removalMichael Brown2007-09-171-470/+5Star
| |
| * IPoIB code separated out to ipoib.c.Michael Brown2007-09-176-6/+559
| |
| * Read port GID directly using MAD IFC.Michael Brown2007-09-172-15/+127
| |
| * Multicast join now works.Michael Brown2007-09-172-4/+157
| |
| * Now transmits packets on our own allocated IPoIB queue pair. :)Michael Brown2007-09-161-1/+1
| |
| * Almost working with own-queue allocation.Michael Brown2007-09-166-11/+166
| |
| * create_qp() and destroy_qp() now written (but not tested).Michael Brown2007-09-164-16/+128
| |
| * Started implementing create_qp() and destroy_qp().Michael Brown2007-09-162-41/+205
| |
| * Revert to dev_priv/owner_priv scheme, rather than container_of; itMichael Brown2007-09-162-101/+158
| | | | | | | | makes it easier to put the generic allocation code into infiniband.c
| * destroy_cq() now implemented (not tested).Michael Brown2007-09-162-2/+67
| |
| * create_cq() implemented (but not tested).Michael Brown2007-09-162-30/+78
| |
| * Match doorbell layout to expected usage.Michael Brown2007-09-162-17/+17
| |
| * Started to add code for CQ creationMichael Brown2007-09-163-21/+212
| |
| * Rearrange data structures to maximise embedding (and hence minimiseMichael Brown2007-09-162-75/+131
| | | | | | | | the number of separate allocations that need to be done).
| * Command interface now reasonably friendly.Michael Brown2007-09-162-50/+94
| |
| * First (working) draft of command interface.Michael Brown2007-09-162-4/+159
| |
| * Kill off more dead code.Michael Brown2007-09-161-123/+0Star
| |
| * post_recv() now works, and we can pass data on the IPoIB queue pairMichael Brown2007-09-163-59/+194
| | | | | | | | using entirely our own code.
| * Kill off some dead codeMichael Brown2007-09-151-83/+0Star
| |
| * Hack up IB structures at start-of-day, instead of on each use.Michael Brown2007-09-151-0/+54
| |
| * Direct polling of TX completion queue now works.Michael Brown2007-09-152-61/+269
| |
| * Updated MLX_* accessor macros to use implicit type information.Michael Brown2007-09-153-100/+130
| |
| * Started added poll_cq() verb.Michael Brown2007-09-152-79/+189
| | | | | | | | | | Started reworking MLX_EXTRACT(), MLX_POPULATE() etc. to automatically determine type information.
| * Map the whole of physical memoryMichael Brown2007-09-152-7/+9
| |
| * arbel_post_send() has been observed to transmit a packet!Michael Brown2007-09-152-20/+52
| |
| * Gets a response out of the hardware. (An error completion, to be precise.)Michael Brown2007-09-152-12/+96
| |
| * Prefix arbel-specific functions etc. with arbel_Michael Brown2007-09-151-41/+44
| |
| * Now at least compilesMichael Brown2007-09-153-58/+158
| |
| * Starting to introduce an Infiniband device abstractionMichael Brown2007-09-141-0/+138
| |
| * Remove some dead codeMichael Brown2007-09-141-245/+10Star
| |
| * Now handling TX completions in our poll loop.Michael Brown2007-09-141-49/+64
| |
| * Proof-of-concept to manually parse completion eventMichael Brown2007-09-141-2/+21
| |
generated by cgit v1.2.3-55-g7522 (git 2.39.0) at 2026-02-27 12:48:32 +0100