]>
git.openfabrics.org - ~shefty/librdmacm.git/log
Sean Hefty [Wed, 8 Apr 2015 22:27:42 +0000 (15:27 -0700)]
refresh
Sean Hefty [Wed, 8 Apr 2015 22:27:42 +0000 (15:27 -0700)]
refresh
Sean Hefty [Wed, 8 Apr 2015 22:27:42 +0000 (15:27 -0700)]
refresh (create temporary patch)
Sean Hefty [Wed, 8 Apr 2015 22:27:42 +0000 (15:27 -0700)]
refresh (create temporary patch)
Sean Hefty [Wed, 8 Apr 2015 22:27:42 +0000 (15:27 -0700)]
Refresh of v1.0.20
Sean Hefty [Wed, 8 Apr 2015 22:27:38 +0000 (15:27 -0700)]
v1.0.20
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Sean Hefty [Wed, 8 Apr 2015 22:27:38 +0000 (15:27 -0700)]
new
Sean Hefty [Wed, 8 Apr 2015 22:27:38 +0000 (15:27 -0700)]
new
Sean Hefty [Wed, 8 Apr 2015 22:27:38 +0000 (15:27 -0700)]
v1.0.20
Sean Hefty [Wed, 8 Apr 2015 22:27:38 +0000 (15:27 -0700)]
external modifications
Modifications by tools other than StGit (e.g. git).
Sean Hefty [Wed, 8 Apr 2015 22:27:38 +0000 (15:27 -0700)]
external modifications
Modifications by tools other than StGit (e.g. git).
Ilya Nelkenbaum [Thu, 26 Mar 2015 19:41:11 +0000 (12:41 -0700)]
cma: Workaround for rdma_ucm kernel bug
For certain new kernels, IB_QP_SMAC bit is erroneously
set in QP attribute mask. This workaround turns
off that bit. It allows SSA connections (AF_IB)
to work. Without this workaround, there are issues
on the client side.
Kernel patch which caused issue is commit
dd5f03b
IB/core: Ethernet L2 attributes in verbs/cm structures
Kernel patch to fix this in upstream:
commit
c2be9dc0e0fa59cc43c2c7084fc42b430809a0fe
Author: Ilya Nelkenbaum <ilyan@mellanox.com>
Date: Thu Feb 5 13:53:48 2015 +0200
IB/core: When marshaling ucma path from user-space, clear unused fields
When marshaling a user path to the kernel struct ib_sa_path, we need
to zero smac and dmac and set the vlan id to the "no vlan" value.
This is to ensure that Ethernet attributes are not used with
InfiniBand QPs.
Fixes: dd5f03beb4f7 ("IB/core: Ethernet L2 attributes in verbs/cm structures")
Signed-off-by: Ilya Nelkenbaum <ilyan@mellanox.com>
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: Roland Dreier <roland@purestorage.com>
The fix was pushed to stable 3.14, 3.18 and 3.19 versions.
Signed-off-by: Ilya Nelkenbaum <ilyan@mellanox.com>
Signed-off-by: Hal Rosenstock <hal@mellanox.com>
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Sasha Kotchubievsky [Wed, 25 Mar 2015 22:27:01 +0000 (15:27 -0700)]
rsocket; Set POLLERR bit in revent in case of rs_connect_error
In non-blocking mode, on server side accept returns a new data
socket and select/poll should indicate when the socket is ready
for reading/writing.
This fix allows an application to receive an error in case of
connection failure.
Signed-off-by: Sasha Kotchubievsky <sashakot@mellanox.com>
Signed-off-by: Hal Rosenstock <hal@mellanox.com>
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Sean Hefty [Wed, 25 Mar 2015 22:46:28 +0000 (15:46 -0700)]
commit
Sean Hefty [Wed, 25 Mar 2015 22:46:28 +0000 (15:46 -0700)]
commit
Sean Hefty [Wed, 25 Mar 2015 22:45:38 +0000 (15:45 -0700)]
refresh
Sean Hefty [Wed, 25 Mar 2015 22:45:38 +0000 (15:45 -0700)]
refresh
Sean Hefty [Wed, 25 Mar 2015 22:45:38 +0000 (15:45 -0700)]
refresh (create temporary patch)
Sean Hefty [Wed, 25 Mar 2015 22:45:38 +0000 (15:45 -0700)]
refresh (create temporary patch)
Sean Hefty [Wed, 25 Mar 2015 22:45:37 +0000 (15:45 -0700)]
Refresh of librdmacm-rstream-c-add-missin
Sean Hefty [Wed, 25 Mar 2015 22:45:34 +0000 (15:45 -0700)]
import
Sean Hefty [Wed, 25 Mar 2015 22:45:34 +0000 (15:45 -0700)]
import
Hal Rosenstock [Fri, 6 Mar 2015 14:58:45 +0000 (09:58 -0500)]
[librdmacm] rstream.c: Add missing binding to source address in client_connect
This is needed for IPv6 connections.
Signed-off-by: Hal Rosenstock <hal@mellanox.com>
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Hal Rosenstock [Fri, 6 Mar 2015 14:58:45 +0000 (09:58 -0500)]
[librdmacm] rstream.c: Add missing binding to source address in client_connect
This is needed for IPv6 connections.
Signed-off-by: Hal Rosenstock <hal@mellanox.com>
Sean Hefty [Fri, 13 Feb 2015 17:44:49 +0000 (09:44 -0800)]
commit
Sean Hefty [Fri, 13 Feb 2015 17:44:49 +0000 (09:44 -0800)]
commit
Sean Hefty [Wed, 11 Feb 2015 01:01:38 +0000 (17:01 -0800)]
refresh
Sean Hefty [Wed, 11 Feb 2015 01:01:38 +0000 (17:01 -0800)]
refresh
Sean Hefty [Wed, 11 Feb 2015 01:01:35 +0000 (17:01 -0800)]
refresh (create temporary patch)
Sean Hefty [Wed, 11 Feb 2015 01:01:35 +0000 (17:01 -0800)]
refresh (create temporary patch)
Sean Hefty [Wed, 11 Feb 2015 01:01:33 +0000 (17:01 -0800)]
Refresh of rs-idm
Sean Hefty [Wed, 11 Feb 2015 00:52:48 +0000 (16:52 -0800)]
refresh
Sean Hefty [Wed, 11 Feb 2015 00:52:48 +0000 (16:52 -0800)]
refresh
Sean Hefty [Wed, 11 Feb 2015 00:51:35 +0000 (16:51 -0800)]
refresh (create temporary patch)
Sean Hefty [Wed, 11 Feb 2015 00:51:35 +0000 (16:51 -0800)]
refresh (create temporary patch)
Sean Hefty [Wed, 11 Feb 2015 00:51:35 +0000 (16:51 -0800)]
Refresh of rs-idm
Sean Hefty [Wed, 11 Feb 2015 00:51:30 +0000 (16:51 -0800)]
refresh
Sean Hefty [Wed, 11 Feb 2015 00:51:30 +0000 (16:51 -0800)]
refresh
Sean Hefty [Wed, 11 Feb 2015 00:51:20 +0000 (16:51 -0800)]
refresh (create temporary patch)
Sean Hefty [Wed, 11 Feb 2015 00:51:20 +0000 (16:51 -0800)]
refresh (create temporary patch)
Sean Hefty [Wed, 11 Feb 2015 00:51:19 +0000 (16:51 -0800)]
Refresh of rs-idm
Sean Hefty [Wed, 11 Feb 2015 00:50:08 +0000 (16:50 -0800)]
rsocket: Fix race in indexer map
Although insertions and removals of rsockets are protected
against accesses to the index map, when reading the map using
a non-rsocket (i.e. normal fd), the reading of the map may
overlap with the removal of an rsocket. This can result in
accessing freed memory.
We can avoid this by not freeing the memory when rsockets
no longer reference an index array. This ensures that the
memory is valid, and protects against reading the memory without
adding locking into the read path.
Problem reported by: Sasha Kotchubievsky <sashakot@mellanox.com>
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Sean Hefty [Wed, 11 Feb 2015 00:50:08 +0000 (16:50 -0800)]
rsocket: Fix race in indexer map
Although insertions and removals of rsockets are protected
against accesses to the index map, when reading the map using
a non-rsocket (i.e. normal fd), the reading of the map may
overlap with the removal of an rsocket. This can result in
accessing freed memory.
We can avoid this by not freeing the memory when rsockets
no longer reference an index array. This ensures that the
memory is valid, and protects against reading the memory without
adding locking into the read path.
Problem reported by: Sasha Kotchubievsky <sashakot@mellanox.com>
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Sean Hefty [Wed, 11 Feb 2015 00:50:08 +0000 (16:50 -0800)]
rsocket: Fix race in indexer map
Although insertions and removals of rsockets are protected
against accesses to the index map, when reading the map using
a non-rsocket (i.e. normal fd), the reading of the map may
overlap with the removal of an rsocket. This can result in
accessing freed memory.
Problem reported by: Sasha Kotchubievsky <sashakot@mellanox.com>
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Sean Hefty [Wed, 11 Feb 2015 00:50:08 +0000 (16:50 -0800)]
new
Sean Hefty [Wed, 11 Feb 2015 00:50:08 +0000 (16:50 -0800)]
new
Sean Hefty [Wed, 11 Feb 2015 00:50:08 +0000 (16:50 -0800)]
rsocket: Fix race in indexer map
Although insertions and removals of rsockets are protected
against accesses to the index map, when reading the map using
a non-rsocket (i.e. normal fd), the reading of the map may
overlap with the removal of an rsocket. This can result in
accessing freed memory.
Problem reported by:
Sean Hefty [Tue, 10 Feb 2015 20:27:35 +0000 (12:27 -0800)]
rename
Sean Hefty [Tue, 10 Feb 2015 20:27:35 +0000 (12:27 -0800)]
rename
Sean Hefty [Tue, 10 Feb 2015 20:27:20 +0000 (12:27 -0800)]
pop
Sean Hefty [Tue, 10 Feb 2015 20:27:20 +0000 (12:27 -0800)]
pop
Sean Hefty [Tue, 10 Feb 2015 20:25:35 +0000 (12:25 -0800)]
refresh
Sean Hefty [Tue, 10 Feb 2015 20:25:35 +0000 (12:25 -0800)]
refresh
Sean Hefty [Tue, 10 Feb 2015 20:25:35 +0000 (12:25 -0800)]
refresh (create temporary patch)
Sean Hefty [Tue, 10 Feb 2015 20:25:35 +0000 (12:25 -0800)]
refresh (create temporary patch)
Sean Hefty [Tue, 10 Feb 2015 20:25:35 +0000 (12:25 -0800)]
Refresh of raccept
Sean Hefty [Tue, 10 Feb 2015 20:24:23 +0000 (12:24 -0800)]
refresh
Sean Hefty [Tue, 10 Feb 2015 20:24:23 +0000 (12:24 -0800)]
refresh
Sean Hefty [Tue, 10 Feb 2015 20:24:23 +0000 (12:24 -0800)]
refresh (create temporary patch)
Sean Hefty [Tue, 10 Feb 2015 20:24:23 +0000 (12:24 -0800)]
refresh (create temporary patch)
Sean Hefty [Tue, 10 Feb 2015 20:24:23 +0000 (12:24 -0800)]
Refresh of raccept
Sean Hefty [Tue, 10 Feb 2015 20:24:01 +0000 (12:24 -0800)]
refresh
Sean Hefty [Tue, 10 Feb 2015 20:24:01 +0000 (12:24 -0800)]
refresh
Sean Hefty [Tue, 10 Feb 2015 20:24:01 +0000 (12:24 -0800)]
refresh (create temporary patch)
Sean Hefty [Tue, 10 Feb 2015 20:24:01 +0000 (12:24 -0800)]
refresh (create temporary patch)
Sean Hefty [Tue, 10 Feb 2015 20:24:01 +0000 (12:24 -0800)]
Refresh of raccept
Sasha Kotchubievsky [Tue, 10 Feb 2015 20:23:52 +0000 (12:23 -0800)]
rsocket: Return -1 from raccept in case of EAGAIN error
If no pending connections are present for the rsocket
marked as nonblocknig, raccept should fail with the error
EAGAIN or EWOULDBLOCK.
Signed-off-by: Sasha Kotchubievsky <sashakot@mellanox.com>
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Sasha Kotchubievsky [Tue, 10 Feb 2015 20:23:52 +0000 (12:23 -0800)]
rsocket: Return -1 from raccept in case of EAGAIN error
If no pending connections are present for the rsocket
marked as nonblocknig, raccept should fail with the error
EAGAIN or EWOULDBLOCK.
Signed-off-by: Sasha Kotchubievsky <sashakot@mellanox.com>
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Sean Hefty [Tue, 10 Feb 2015 20:23:52 +0000 (12:23 -0800)]
rsocket: Return -1 from raccept in case of EAGAIN error
If no pending connections are present for the rsocket
marked as nonblocknig, raccept should fail with the error
EAGAIN or EWOULDBLOCK.
Signed-off-by: Sasha Kotchubievsky <sashakot@mellanox.com>
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Sean Hefty [Tue, 10 Feb 2015 20:23:52 +0000 (12:23 -0800)]
new
Sean Hefty [Tue, 10 Feb 2015 20:23:52 +0000 (12:23 -0800)]
new
Sean Hefty [Tue, 10 Feb 2015 20:23:52 +0000 (12:23 -0800)]
rsocket: Return -1 from raccept in case of EAGAIN error
If no pending connections are present for the rsocket
marked as nonblocknig, raccept should fail with the error
EAGAIN or EWOULDBLOCK.
Signed-off-by: Sasha Kotchubievsky <sashakot@mellanox.com>
Sean Hefty [Tue, 10 Feb 2015 20:08:58 +0000 (12:08 -0800)]
commit
Sean Hefty [Tue, 10 Feb 2015 20:08:58 +0000 (12:08 -0800)]
commit
Sean Hefty [Tue, 10 Feb 2015 20:02:04 +0000 (12:02 -0800)]
refresh
Sean Hefty [Tue, 10 Feb 2015 20:02:04 +0000 (12:02 -0800)]
refresh
Sean Hefty [Tue, 10 Feb 2015 20:02:04 +0000 (12:02 -0800)]
refresh (create temporary patch)
Sean Hefty [Tue, 10 Feb 2015 20:02:04 +0000 (12:02 -0800)]
refresh (create temporary patch)
Sean Hefty [Tue, 10 Feb 2015 20:02:04 +0000 (12:02 -0800)]
Refresh of setfl
Sean Hefty [Tue, 10 Feb 2015 19:55:16 +0000 (11:55 -0800)]
refresh
Sean Hefty [Tue, 10 Feb 2015 19:55:16 +0000 (11:55 -0800)]
refresh
Sean Hefty [Tue, 10 Feb 2015 19:55:16 +0000 (11:55 -0800)]
refresh (create temporary patch)
Sean Hefty [Tue, 10 Feb 2015 19:55:16 +0000 (11:55 -0800)]
refresh (create temporary patch)
Sean Hefty [Tue, 10 Feb 2015 19:55:15 +0000 (11:55 -0800)]
Refresh of setfl
Sean Hefty [Tue, 10 Feb 2015 19:55:12 +0000 (11:55 -0800)]
rsockets: Fix F_SETFL to use an int rather than long
The flags are specified as an int, not a long. The
rsocket code handles the flag incorrectly.
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Sean Hefty [Tue, 10 Feb 2015 19:55:12 +0000 (11:55 -0800)]
rsockets: Fix F_SETFL to use an int rather than long
The flags are specified as an int, not a long. The
rsocket code handles the flag incorrectly.
Signed-off-by: Sean Hefty <sean.hefty@intel.com>
Sean Hefty [Tue, 10 Feb 2015 19:55:12 +0000 (11:55 -0800)]
new
Sean Hefty [Tue, 10 Feb 2015 19:55:12 +0000 (11:55 -0800)]
new
Sean Hefty [Tue, 10 Feb 2015 19:55:12 +0000 (11:55 -0800)]
rsockets: Fix F_SETFL to use an int rather than long
The flags are specified as an int, not a long. The
rsocket code handles the flag incorrectly.
Sean Hefty [Fri, 6 Feb 2015 05:52:46 +0000 (21:52 -0800)]
commit
Sean Hefty [Fri, 6 Feb 2015 05:52:46 +0000 (21:52 -0800)]
commit
Sean Hefty [Fri, 6 Feb 2015 05:23:14 +0000 (21:23 -0800)]
refresh
Sean Hefty [Fri, 6 Feb 2015 05:23:14 +0000 (21:23 -0800)]
refresh
Sean Hefty [Fri, 6 Feb 2015 05:23:13 +0000 (21:23 -0800)]
refresh (create temporary patch)
Sean Hefty [Fri, 6 Feb 2015 05:23:13 +0000 (21:23 -0800)]
refresh (create temporary patch)
Sean Hefty [Fri, 6 Feb 2015 05:23:13 +0000 (21:23 -0800)]
Refresh of nonblock
Sean Hefty [Fri, 6 Feb 2015 05:17:07 +0000 (21:17 -0800)]
refresh
Sean Hefty [Fri, 6 Feb 2015 05:17:07 +0000 (21:17 -0800)]
refresh
Sean Hefty [Fri, 6 Feb 2015 05:17:07 +0000 (21:17 -0800)]
refresh (create temporary patch)
Sean Hefty [Fri, 6 Feb 2015 05:17:07 +0000 (21:17 -0800)]
refresh (create temporary patch)