2007-10-29 |
Michael Brown | Merge branch '3leaf' |
tree | commitdiff |
2007-10-10 |
Michael Brown | Change ROM names to lower case. |
tree | commitdiff |
2007-10-10 |
Michael Brown | Ensure that empty e820 regions are skipped even at... |
tree | commitdiff |
2007-10-10 |
Michael Brown | Ensure that empty e820 regions are skipped even at... |
tree | commitdiff |
2007-10-10 |
Michael Brown | Add a missing PCI ID that is tested as working. |
tree | commitdiff |
2007-10-04 |
Michael Brown | Force a netdevice poll in net_tx() before attempting... |
tree | commitdiff |
2007-10-04 |
Michael Brown | Force a netdevice poll in net_tx() before attempting... |
tree | commitdiff |
2007-09-28 |
Michael Brown | Redefine bzimage_exec_context::mem_limit to be the... |
tree | commitdiff |
2007-09-26 |
Michael Brown | Quick and very dirty hack to get r8169 driver working... |
tree | commitdiff |
2007-09-26 |
Michael Brown | Standardise DBG_PRINTF()->DBG(), and remove custom... |
tree | commitdiff |
2007-09-25 |
Michael Brown | Switch rm_ss and rm_sp back to being words; it'll make... |
tree | commitdiff |
2007-09-25 |
Michael Brown | Don't use the "rep ss movsb" trick to copy the RM stack... |
tree | commitdiff |
2007-09-25 |
Michael Brown | rm_ss, rm_sp, and pm_esp don't need to be accessed... |
tree | commitdiff |
2007-09-25 |
Michael Brown | Minor refactoring to eliminate duplication. |
tree | commitdiff |
2007-09-25 |
Michael Brown | Use full protected mode for access to high memory withi... |
tree | commitdiff |
2007-09-21 |
Michael Brown | Added more verbose memory-map debugging. |
tree | commitdiff |
2007-09-21 |
Michael Brown | Added more verbose memory-map debugging. |
tree | commitdiff |
2007-09-21 |
Michael Brown | Merge branch 'master' into 3leaf-delivery |
tree | commitdiff |
2007-09-21 |
Michael Brown | Merge branch 'master' of rom.etherboot.org:/gpxe |
tree | commitdiff |
2007-09-21 |
Michael Brown | Infinband requires longer link-layer addresses and... |
tree | commitdiff |
2007-09-21 |
Michael Brown | Added defines for Infiniband files (which can't be... |
tree | commitdiff |
2007-09-21 |
Michael Brown | Minor debug message improvement. |
tree | commitdiff |
2007-09-21 |
Michael Brown | Use RFC4390 whenever hardware address exceeds 16 bytes... |
tree | commitdiff |
2007-09-21 |
Michael Brown | Add const attribute to byte-swapping functions |
tree | commitdiff |
2007-09-21 |
Michael Brown | Don't get stuck in an infinite loop on negative integers! |
tree | commitdiff |
2007-09-21 |
Michael Brown | Add fls() for non-constant values. |
tree | commitdiff |
2007-09-21 |
Michael Brown | Moved iobuf.h assertions outside the static inline... |
tree | commitdiff |
2007-09-18 |
Michael Brown | Reverted driver/net/mlx_ipoib to clean master state |
tree | commitdiff |
2007-09-18 |
Michael Brown | Merge branch 'master' into 3leaf-rewrite |
tree | commitdiff |
2007-09-18 |
Michael Brown | IB can't afford many RX buffers, because the MTU is... |
tree | commitdiff |
2007-09-18 |
Michael Brown | Added arbel.c |
tree | commitdiff |
2007-09-18 |
Michael Brown | Remove the last remaining visible hack. |
tree | commitdiff |
2007-09-18 |
Michael Brown | Separated out to a clean new drivers/infiniband directory. |
tree | commitdiff |
2007-09-18 |
Michael Brown | First version that works with SELF_INIT! |
tree | commitdiff |
2007-09-18 |
Michael Brown | Gets most of the way through initialisation. |
tree | commitdiff |
2007-09-18 |
Michael Brown | Added sw2hw_mpt |
tree | commitdiff |
2007-09-18 |
Michael Brown | INIT_HCA block now matches. |
tree | commitdiff |
2007-09-18 |
Michael Brown | Seems to get through the ICM mapping process |
tree | commitdiff |
2007-09-18 |
Michael Brown | ICM allocation code in place; about to start test-and... |
tree | commitdiff |
2007-09-18 |
Michael Brown | ICM allocation code *should* now be complete. |
tree | commitdiff |
2007-09-18 |
Michael Brown | Started with the ICM partitioning code. |
tree | commitdiff |
2007-09-17 |
Michael Brown | Restructured bits of initialisation (not testable at... |
tree | commitdiff |
2007-09-17 |
Michael Brown | Starting the firmware directly now works. |
tree | commitdiff |
2007-09-17 |
Michael Brown | Start migrating the remaining initialisation steps. |
tree | commitdiff |
2007-09-17 |
Michael Brown | Don't use the mailboxless version of the HW2SW_CQ comma... |
tree | commitdiff |
2007-09-17 |
Michael Brown | Removed more hacks, and adjusted number of queue entrie... |
tree | commitdiff |
2007-09-17 |
Michael Brown | Remove hacks, and fix leaving the multicast GID. |
tree | commitdiff |
2007-09-17 |
Michael Brown | Added an almost obscene amount of debugging and asserti... |
tree | commitdiff |
2007-09-17 |
Michael Brown | Minor debug message improvement. |
tree | commitdiff |
2007-09-17 |
Michael Brown | Moved iobuf.h assertions outside the static inline... |
tree | commitdiff |
2007-09-17 |
Michael Brown | Some interesting packet corruption happening now. |
tree | commitdiff |
2007-09-17 |
Michael Brown | Broadcast GID is now calculated by IPoIB layer. |
tree | commitdiff |
2007-09-17 |
Michael Brown | Cleaned up some debug messages. |
tree | commitdiff |
2007-09-17 |
Michael Brown | Unicasts seem to be working. :) |
tree | commitdiff |
2007-09-17 |
Michael Brown | Dead code disabling |
tree | commitdiff |
2007-09-17 |
Michael Brown | Obtains a response to the get path record! |
tree | commitdiff |
2007-09-17 |
Michael Brown | Use pkey table access to determine broadcast GID directly. |
tree | commitdiff |
2007-09-17 |
Michael Brown | Prepare for adding a metadata queue to IPoIB |
tree | commitdiff |
2007-09-17 |
Michael Brown | Dead code removal |
tree | commitdiff |
2007-09-17 |
Michael Brown | IPoIB code separated out to ipoib.c. |
tree | commitdiff |
2007-09-17 |
Michael Brown | Read port GID directly using MAD IFC. |
tree | commitdiff |
2007-09-16 |
Michael Brown | Multicast join now works. |
tree | commitdiff |
2007-09-16 |
Michael Brown | Now transmits packets on our own allocated IPoIB queue... |
tree | commitdiff |
2007-09-16 |
Michael Brown | Almost working with own-queue allocation. |
tree | commitdiff |
2007-09-16 |
Michael Brown | create_qp() and destroy_qp() now written (but not tested). |
tree | commitdiff |
2007-09-16 |
Michael Brown | Started implementing create_qp() and destroy_qp(). |
tree | commitdiff |
2007-09-16 |
Michael Brown | Don't get stuck in an infinite loop on negative integers! |
tree | commitdiff |
2007-09-16 |
Michael Brown | Add fls() for non-constant values. |
tree | commitdiff |
2007-09-16 |
Michael Brown | Revert to dev_priv/owner_priv scheme, rather than conta... |
tree | commitdiff |
2007-09-16 |
Michael Brown | destroy_cq() now implemented (not tested). |
tree | commitdiff |
2007-09-16 |
Michael Brown | create_cq() implemented (but not tested). |
tree | commitdiff |
2007-09-16 |
Michael Brown | Match doorbell layout to expected usage. |
tree | commitdiff |
2007-09-16 |
Michael Brown | Started to add code for CQ creation |
tree | commitdiff |
2007-09-16 |
Michael Brown | Rearrange data structures to maximise embedding (and... |
tree | commitdiff |
2007-09-16 |
Michael Brown | Command interface now reasonably friendly. |
tree | commitdiff |
2007-09-16 |
Michael Brown | First (working) draft of command interface. |
tree | commitdiff |
2007-09-15 |
Michael Brown | Kill off more dead code. |
tree | commitdiff |
2007-09-15 |
Michael Brown | post_recv() now works, and we can pass data on the... |
tree | commitdiff |
2007-09-15 |
Michael Brown | Kill off some dead code |
tree | commitdiff |
2007-09-15 |
Michael Brown | Hack up IB structures at start-of-day, instead of on... |
tree | commitdiff |
2007-09-15 |
Michael Brown | Direct polling of TX completion queue now works. |
tree | commitdiff |
2007-09-15 |
Michael Brown | Add const attribute to byte-swapping functions |
tree | commitdiff |
2007-09-15 |
Michael Brown | Updated MLX_* accessor macros to use implicit type... |
tree | commitdiff |
2007-09-15 |
Michael Brown | Started added poll_cq() verb. |
tree | commitdiff |
2007-09-15 |
Michael Brown | Map the whole of physical memory |
tree | commitdiff |
2007-09-15 |
Michael Brown | arbel_post_send() has been observed to transmit a packet! |
tree | commitdiff |
2007-09-15 |
Michael Brown | Gets a response out of the hardware. (An error complet... |
tree | commitdiff |
2007-09-14 |
Michael Brown | Prefix arbel-specific functions etc. with arbel_ |
tree | commitdiff |
2007-09-14 |
Michael Brown | Now at least compiles |
tree | commitdiff |
2007-09-14 |
Michael Brown | Starting to introduce an Infiniband device abstraction |
tree | commitdiff |
2007-09-14 |
Michael Brown | Remove some dead code |
tree | commitdiff |
2007-09-14 |
Michael Brown | Now handling TX completions in our poll loop. |
tree | commitdiff |
2007-09-14 |
Michael Brown | Proof-of-concept to manually parse completion event |
tree | commitdiff |
2007-09-14 |
Michael Brown | Add EX_FLD_BE() |
tree | commitdiff |
2007-09-13 |
Michael Brown | Start constructing a generic poll() routine. |
tree | commitdiff |
2007-09-13 |
Michael Brown | Can now both send and receive packets. LL header forma... 3leaf-rewrite |
tree | commitdiff |
2007-09-13 |
Michael Brown | Added netdev_priv() to ease Linux driver compatibility. |
tree | commitdiff |
2007-09-13 |
Michael Brown | Removed file checked in accidentally |
tree | commitdiff |
2007-09-13 |
Michael Brown | Dump received packet, including GRH. |
tree | commitdiff |
2007-09-13 |
Michael Brown | Use RFC4390 whenever hardware address exceeds 16 bytes... |
tree | commitdiff |
next |