linux.git/drivers/infiniband, branch v2.6.15 Linux kernel source tree IB/uverbs: track multicast group membership for userspace QPs 2005-11-30T00:57:01+00:00 Jack Morgenstein jackm@mellanox.co.il 2005-11-30T00:57:01+00:00 f4e401562c11c7ca65592ebd749353cf0b19af7b uverbs needs to track which multicast groups is each qp attached to, in order to properly detach when cleanup is performed on device file close. Signed-off-by: Jack Morgenstein <jackm@mellanox.co.il> Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il> Signed-off-by: Roland Dreier <rolandd@cisco.com> 
uverbs needs to track which multicast groups is each qp
attached to, in order to properly detach when cleanup
is performed on device file close.

Signed-off-by: Jack Morgenstein <jackm@mellanox.co.il>
Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
IB/mthca: fix posting of send lists of length >= 255 on mem-free HCAs 2005-11-29T19:33:46+00:00 Michael S. Tsirkin mst@mellanox.co.il 2005-11-29T19:33:46+00:00 e0ae9ecf469fdd3c1ad999efbf4fe6b782f49900 On mem-free HCAs, when posting a long list of send requests, a doorbell must be rung every 255 requests. Add code to handle this. Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il> Signed-off-by: Roland Dreier <rolandd@cisco.com> 
On mem-free HCAs, when posting a long list of send requests, a
doorbell must be rung every 255 requests.  Add code to handle this.

Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
IPoIB: fix error handling in ipoib_open 2005-11-29T18:55:58+00:00 Roland Dreier rolandd@cisco.com 2005-11-29T18:55:58+00:00 267ee88ed34c76dc527eeb3d95f9f9558ac99973 If ipoib_ib_dev_up() fails after ipoib_ib_dev_open() is called, then ipoib_ib_dev_stop() needs to be called to clean up. Signed-off-by: Roland Dreier <rolandd@cisco.com> 
If ipoib_ib_dev_up() fails after ipoib_ib_dev_open() is called, then
ipoib_ib_dev_stop() needs to be called to clean up.

Signed-off-by: Roland Dreier <rolandd@cisco.com>
IPoIB: protect child list in ipoib_ib_dev_flush 2005-11-29T18:53:30+00:00 Michael S. Tsirkin mst@mellanox.co.il 2005-11-29T18:53:30+00:00 4f71055a45a503273c039d80db8ba9b13cb17549 race condition: ipoib_ib_dev_flush is accessing child list without locks. Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il> Signed-off-by: Roland Dreier <rolandd@cisco.com> 
race condition: ipoib_ib_dev_flush is accessing child list without locks.

Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
IPoIB: don't zero members after we allocate with kzalloc 2005-11-29T18:25:23+00:00 Roland Dreier rolandd@cisco.com 2005-11-29T18:25:23+00:00 2e86541ec878de9ec5771600a77f451a80bebfc4 ipoib_mcast_alloc() uses kzalloc(), so there's no need to zero out members of the mcast struct after it's allocated. Signed-off-by: Roland Dreier <rolandd@cisco.com> 
ipoib_mcast_alloc() uses kzalloc(), so there's no need to zero out
members of the mcast struct after it's allocated.

Signed-off-by: Roland Dreier <rolandd@cisco.com>
IPoIB: reinitialize mcast structs' completions for every query 2005-11-29T18:18:45+00:00 Michael S. Tsirkin mst@mellanox.co.il 2005-11-29T18:18:45+00:00 de922487890936470660e89f9095aee980637989 Make sure mcast->done is initialized to uncompleted value before we submit a new query, so that it's safe to wait on. Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il> Signed-off-by: Roland Dreier <rolandd@cisco.com> 
Make sure mcast->done is initialized to uncompleted value before we
submit a new query, so that it's safe to wait on.

Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
IPoIB: always set path->query to NULL when query finishes 2005-11-29T18:13:54+00:00 Roland Dreier rolandd@cisco.com 2005-11-29T18:13:54+00:00 5872a9fc28e6cd3a4e51479a50970d19a01573b3 Always set path->query to NULL when the SA path record query completes, rather than only when we don't have an address handle. Signed-off-by: Roland Dreier <rolandd@cisco.com> 
Always set path->query to NULL when the SA path record query
completes, rather than only when we don't have an address handle.

Signed-off-by: Roland Dreier <rolandd@cisco.com>
IPoIB: reinitialize path struct's completion for every query 2005-11-29T05:20:34+00:00 Roland Dreier rolandd@cisco.com 2005-11-29T05:20:34+00:00 65c7eddaba33995e013ef3c04718f6dc8fdf2335 It's possible that IPoIB will issue multiple SA queries for the same path struct. Therefore the struct's completion needs to be initialized for each query rather than only once when the struct is allocated, or else we might not wait long enough for later queries to finish and free the path struct too soon. Signed-off-by: Roland Dreier <rolandd@cisco.com> 
It's possible that IPoIB will issue multiple SA queries for the same
path struct.  Therefore the struct's completion needs to be
initialized for each query rather than only once when the struct is
allocated, or else we might not wait long enough for later queries to
finish and free the path struct too soon.

Signed-off-by: Roland Dreier <rolandd@cisco.com>
IB/umad: fix RMPP handling 2005-11-28T21:07:20+00:00 Michael S. Tsirkin mst@mellanox.co.il 2005-11-28T21:07:20+00:00 bf6d9e23a36c8a01bf6fbb945387d8ca3870ff71 ib_umad_write in user_mad.c is looking at rmpp_hdr field in MAD before checking that the MAD actually has the RMPP header. So for a MAD without RMPP header it looks like we are actually checking a bit inside M_Key, or something. Signed-off-by: Jack Morgenstein <jackm@mellanox.co.il> Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il> Signed-off-by: Roland Dreier <rolandd@cisco.com> 
ib_umad_write in user_mad.c is looking at rmpp_hdr field in MAD before
checking that the MAD actually has the RMPP header.  So for a MAD
without RMPP header it looks like we are actually checking a bit
inside M_Key, or something.

Signed-off-by: Jack Morgenstein <jackm@mellanox.co.il>
Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
IB/mthca: reset QP's last pointers when transitioning to reset state 2005-11-28T19:19:43+00:00 Michael S. Tsirkin mst@mellanox.co.il 2005-11-28T19:19:43+00:00 187a25863fe014486ee834164776b2a587d6934d last pointer is not updated when QP is modified to reset state. This causes data corruption if WQEs are already posted on the queue. Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il> Signed-off-by: Roland Dreier <rolandd@cisco.com> 
last pointer is not updated when QP is modified to reset state.  This
causes data corruption if WQEs are already posted on the queue.

Signed-off-by: Michael S. Tsirkin <mst@mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>