Michael Brown
|
251cc84ed6
|
Started implementing create_qp() and destroy_qp().
|
17 years ago |
Michael Brown
|
83a6cc8c9b
|
Don't get stuck in an infinite loop on negative integers!
|
17 years ago |
Michael Brown
|
6d15a193aa
|
Add fls() for non-constant values.
|
17 years ago |
Michael Brown
|
b21d4ca21e
|
Revert to dev_priv/owner_priv scheme, rather than container_of; it
makes it easier to put the generic allocation code into infiniband.c
|
17 years ago |
Michael Brown
|
e238bb1e43
|
destroy_cq() now implemented (not tested).
|
17 years ago |
Michael Brown
|
18edcf6634
|
create_cq() implemented (but not tested).
|
17 years ago |
Michael Brown
|
725a574042
|
Match doorbell layout to expected usage.
|
17 years ago |
Michael Brown
|
5a43293c38
|
Started to add code for CQ creation
|
17 years ago |
Michael Brown
|
156b409ccc
|
Rearrange data structures to maximise embedding (and hence minimise
the number of separate allocations that need to be done).
|
17 years ago |
Michael Brown
|
791f992657
|
Command interface now reasonably friendly.
|
17 years ago |
Michael Brown
|
bf9bd93856
|
First (working) draft of command interface.
|
17 years ago |
Michael Brown
|
baa885ee8e
|
Kill off more dead code.
|
17 years ago |
Michael Brown
|
37fc40bc8c
|
post_recv() now works, and we can pass data on the IPoIB queue pair
using entirely our own code.
|
17 years ago |
Michael Brown
|
838b972cd3
|
Kill off some dead code
|
17 years ago |
Michael Brown
|
31b82ea1dd
|
Hack up IB structures at start-of-day, instead of on each use.
|
17 years ago |
Michael Brown
|
8deef093d9
|
Direct polling of TX completion queue now works.
|
17 years ago |
Michael Brown
|
687afdcdd9
|
Add const attribute to byte-swapping functions
|
17 years ago |
Michael Brown
|
6a791649f0
|
Updated MLX_* accessor macros to use implicit type information.
|
17 years ago |
Michael Brown
|
a3a91fedc1
|
Started added poll_cq() verb.
Started reworking MLX_EXTRACT(), MLX_POPULATE() etc. to automatically
determine type information.
|
17 years ago |
Michael Brown
|
37a036bd48
|
Map the whole of physical memory
|
17 years ago |
Michael Brown
|
970951666f
|
arbel_post_send() has been observed to transmit a packet!
|
17 years ago |
Michael Brown
|
8b27da9de1
|
Gets a response out of the hardware. (An error completion, to be precise.)
|
17 years ago |
Michael Brown
|
21d4ab3ce2
|
Prefix arbel-specific functions etc. with arbel_
|
17 years ago |
Michael Brown
|
38a73b55c4
|
Now at least compiles
|
17 years ago |
Michael Brown
|
9d08b7c692
|
Starting to introduce an Infiniband device abstraction
|
17 years ago |
Michael Brown
|
75fbc96f75
|
Remove some dead code
|
17 years ago |
Michael Brown
|
08e8dfd801
|
Now handling TX completions in our poll loop.
|
17 years ago |
Michael Brown
|
e69863b5fb
|
Proof-of-concept to manually parse completion event
|
17 years ago |
Michael Brown
|
e9df4f691e
|
Add EX_FLD_BE()
|
17 years ago |
Michael Brown
|
da23e8d287
|
Start constructing a generic poll() routine.
|
17 years ago |
Michael Brown
|
30a19c3f1c
|
Can now both send and receive packets. LL header format not yet
fixed; still using a quick hack-up just to be able to pass through
data.
|
17 years ago |
Michael Brown
|
99411e36da
|
Added netdev_priv() to ease Linux driver compatibility.
|
17 years ago |
Michael Brown
|
eecf71e4a0
|
Removed file checked in accidentally
|
17 years ago |
Michael Brown
|
03c90e183f
|
Dump received packet, including GRH.
|
17 years ago |
Michael Brown
|
7e4e5af462
|
Use RFC4390 whenever hardware address exceeds 16 bytes; this allows us
to construct DHCP packets suitable for Infiniband.
|
17 years ago |
Michael Brown
|
771455653b
|
Bugfix: DHCP message type should be a one-byte option... (d'oh)
|
17 years ago |
Michael Brown
|
9e32e8e513
|
Now sends IP packets correctly.
|
17 years ago |
Michael Brown
|
c0d99245a9
|
Bugfix: DHCP message type should be a one-byte option... (d'oh)
|
17 years ago |
Michael Brown
|
7b6d11e713
|
Started IB driver rewrite
|
17 years ago |
Michael Brown
|
6f799db5bd
|
Use "/bin/echo -e" instead of shell builtin echo; some shells don't
implement the -e option.
(Thanks to Jim McQuillan for this suggestion.)
|
17 years ago |
Michael Brown
|
972f293e46
|
Check for correct block number in tftp_rx_data().
(Problem observed by Clay McClure in VMware Fusion.)
|
17 years ago |
Michael Brown
|
fbec308f41
|
Re-added the kpxe prefix. I have no idea when this disappeared.
|
17 years ago |
Michael Brown
|
66207bb912
|
Add the possibility to create .o files from arbitrary binary blobs.
|
17 years ago |
Michael Brown
|
a82e6ec2de
|
Added missing .kpxe prefix.
|
17 years ago |
Michael Brown
|
00d93c6dd9
|
Minor style fix: structure fields are generally initialised in the
order in which they occur.
|
17 years ago |
Marty Connor
|
14e1d3cc4d
|
TFTP default blocksize fix from Vampyre
|
17 years ago |
Marty Connor
|
63f0e4e796
|
allpxes allroms alldsks fixes from Vampyre
|
17 years ago |
Marty Connor
|
eaca053174
|
Add sis900 variant (patch from Vampyre)
|
17 years ago |
Michael Brown
|
74a49af4ab
|
Force MAC address for testing purposes
|
17 years ago |
Michael Brown
|
01b755704a
|
Auto-assign drive number only if set to 0xff; almost all applications
require that we use drive 0x80, so it should probably be the default.
|
17 years ago |