+commit 29bf0a24f54c45d2742026756f31f1a1f26fb6f3
+Author: Arlin Davis <arlin.r.davis@intel.com>
+Date: Thu Apr 16 14:35:18 2009 -0700
+
+ dapltest: reset server listen ports to avoid collisions during long runs
+
+ If server is running continuously the port number increments
+ from base without reseting between tests. This will
+ eventually cause collisions in port space.
+
+ Signed-off-by: Arlin Davis <ardavis@ichips.intel.com>
+
+commit c27af8de0501d132b8152ec8546023cdba212de5
+Author: Sean Hefty <sean.hefty@intel.com>
+Date: Thu Apr 16 10:21:51 2009 -0700
+
+ To avoid duplicating port numbers between different tests, the next port
+ number to use must increment based on the number of endpoints per thread *
+ the number of threads.
+
+ Signed-off-by: Sean Hefty <sean.hefty@intel.com>
+
+commit 3084310197c20aaa50abe82260fc835786f591f5
+Author: Sean Hefty <sean.hefty@intel.com>
+Date: Thu Apr 16 10:21:45 2009 -0700
+
+ dapltest assumes that events across multiple endpoints occur in a specific
+ order. Since this is a false assumption, avoid this by directing events to
+ per endpoint EVDs, rather than using shared EVDs.
+
+ Signed-off-by: Sean Hefty <sean.hefty@intel.com>
+
+commit ef87a0a462f4fa07ac252e28d3aeb44af73cc202
+Author: Sean Hefty <sean.hefty@intel.com>
+Date: Thu Apr 16 10:21:41 2009 -0700
+
+ Synchronization is missing between removing items from an EVD and queuing
+ them. Since the removal thread is the user's, but the queuing thread is
+ not, the synchronization must be provided by DAPL. Hold the evd lock
+ around any calls to dapls_rbuf_*.
+
+ Signed-off-by: Sean Hefty <sean.hefty@intel.com>
+
+commit f5e86d28f803162ffdf94b41ec7435dec92f728d
+Author: Sean Hefty <sean.hefty@intel.com>
+Date: Thu Apr 16 10:21:26 2009 -0700
+
+ Communication to the CR thread is done using an internal socket. When a
+ new connection request is ready for processing, an object is placed on
+ the CR list, and data is written to the internal socket. The write causes
+ the CR thread to wake-up and process anything on its cr list.
+
+ If multiple objects are placed on the CR list around the same time, then
+ the CR thread will read in a single character, but process the entire list.
+ This results in additional data being left on the internal socket. When
+ the CR does a select(), it will find more data to read, read the data, but
+ not have any real work to do. The result is that the thread spins in a
+ loop checking for changes when none have occurred until all data on the
+ internal socket has been read.
+
+ Avoid this overhead by reading all data off the internal socket before
+ processing the CR list.
+
+ Signed-off-by: Sean Hefty <sean.hefty@intel.com>
+
+commit 2ab52e9b1ab37c6eb44206c135e0568a8c2d01fa
+Author: Sean Hefty <sean.hefty@intel.com>
+Date: Thu Apr 16 10:21:13 2009 -0700
+
+ The dapl connect call takes as input an address (sockaddr) and a port number
+ as separate input parameters. It modifies the sockaddr address to set the
+ port number before trying to connect. This leads to a situation in
+ dapltest with multiple threads that reference the same buffer for their
+ address, but specify different port numbers, where the different threads
+ end up trying to connect to the same remote port.
+
+ To solve this, do not modify the caller's address buffer and instead use
+ a local buffer. This fixes an issue seen running multithreaded tests with
+ dapltest.
+
+ Signed-off-by: Sean Hefty <sean.hefty@intel.com>
+
+commit 7947026ede478f08b4a7b8cb607f457765bf2afa
+Author: Sean Hefty <sean.hefty@intel.com>
+Date: Thu Apr 16 10:21:03 2009 -0700
+
+ Windows socket calls should check return values against SOCKET_ERROR to
+ determine if an error occurred.
+
+ Signed-off-by: Sean Hefty <sean.hefty@intel.com>
+
+commit e66e42fc44c50c8202f7c98f76d799a69aa3f1b6
+Author: Arlin Davis <arlin.r.davis@intel.com>
+Date: Fri Apr 10 08:33:41 2009 -0700
+
+ Build: add new file dapl/openib_cma/linux/openib_osd.h to EXTRA_DIST
+
+ Fix rpmbuild problem with new cma osd include file.
+
+ Signed-off-by: Arlin Davis <ardavis@ichips.intel.com>
+
+commit acb213adb3268e9bf6999e2bf040d4a71212b701
+Author: Arlin Davis <arlin.r.davis@intel.com>
+Date: Fri Apr 10 08:32:24 2009 -0700
+
+ dapl scm: reduce wait time for thread startup.
+
+ thread startup wait reduce to 2ms to reduce open times.
+
+ Signed-off-by: Arlin Davis <ardavis@ichips.intel.com>
+
+commit 55459699fa9c0e5fb7e2b17822f0916412c64b35
+Author: Arlin Davis <arlin.r.davis@intel.com>
+Date: Fri Apr 10 08:31:22 2009 -0700
+
+ dapl-scm: getsockopt optlen needs initialized to size of optval
+
+ Signed-off-by: Arlin Davis <ardavis@ichips.intel.com>
+
+commit d710c5327e05a40796341d16b45a2b098b03f588
+Author: Sean Hefty <sean.hefty@intel.com>
+Date: Fri Apr 10 08:17:32 2009 -0700
+
+ The connection request thread adds sockets to a select list unless
+ the cr->socket is invalid and the cr request state is set to destroy. If the
+ cr->socket is invalid, but the cr->state is not destroy, then the cr->socket
+ is added to an FD set for select/poll. This results in select/poll
+ returning an error when select is called. As a result, the cr thread never
+ actually blocks during this state.
+
+ Fix this by only destroying a cr based on its state being set to destroy
+ and skip adding cr->sockets to the FD set when they are invalid.
+
+ Signed-off-by: Sean Hefty <sean.hefty@intel.com>
+
+commit 0be961c432f897d4f92d9a24dcb7c42ad30ea160
+Author: Sean Hefty <sean.hefty@intel.com>
+Date: Fri Apr 10 08:08:16 2009 -0700
+
+ Make sure all locks are initialized properly and don't zero their memory
+ once they are.
+
+ Signed-off-by: Sean Hefty <sean.hefty@intel.com>
+
+commit 9abdc26cd6154aa55588759ba54c9ca69e3fe2b5
+Author: Sean Hefty <sean.hefty@intel.com>
+Date: Fri Apr 10 08:08:13 2009 -0700
+
+ The lock functions are defined just a few lines beneath the prototypes
+ as inline. Remove the duplicate prototypes.
+
+ Signed-off-by: Sean Hefty <sean.hefty@intel.com>
+
+commit 9578c4aeb9878d98374e4b7abc02db182aef82c6
+Author: Sean Hefty <sean.hefty@intel.com>
+Date: Fri Apr 10 08:08:07 2009 -0700
+
+ Make sure all locks are initialized and don't zero out their memory once
+ they are.
+
+ Signed-off-by: Sean Hefty <sean.hefty@intel.com>
+
+commit 97edcbb662b489303ef68c0da02831efaddeed91
+Author: Sean Hefty <sean.hefty@intel.com>
+Date: Fri Apr 10 08:08:03 2009 -0700
+
+ The IBAL library allocates a small number of threads for callbacks to the
+ user. If the user blocks all of the callback threads, no additional
+ callbacks can be invoked. The DAPL IBAL provider cancels listen requests
+ from within an IBAL callback, then waits for a second callback to confirm
+ that the listen has been canceled. If there is a single IBAL callback
+ thread, or multiple listens are canceled simultaneously, then the provider
+ can deadlock waiting for a cancel callback that never occurs.
+
+ This problem is seen when running dapltest with multiple threads.
+
+ Signed-off-by: Sean Hefty <sean.hefty@intel.com>
+
+commit 3e56e63bcb68de352edadafdcfcc4cb222c08c7b
+Author: Sean Hefty <sean.hefty@intel.com>
+Date: Fri Apr 10 08:07:57 2009 -0700
+
+ We need to check the return value from select for errors before checking
+ the FD sets. An item may be in an FD set but select could have returned
+ an error.
+
+ Signed-off-by: Sean Hefty <sean.hefty@intel.com>
+
+commit a8a977becaeefe0d7f8e01e01631a11988d2d54e
+Author: Sean Hefty <sean.hefty@intel.com>
+Date: Fri Apr 10 08:07:53 2009 -0700
+
+ Signed-off-by: Sean Hefty <sean.hefty@intel.com>
+
+commit ecc79cc0a1ae2bdbb3dfd19e15b3b562ac9a2957
+Author: Sean Hefty <sean.hefty@intel.com>
+Date: Fri Apr 10 08:07:49 2009 -0700
+
+ Enable building with CQ_WAIT_OBJECTS support to directly wait on CQ
+ completion channels in the Windows version of the openib_scm provider.
+ Also minor fixup to use DAPL_DBG_TYPE_UTIL for debug log messages
+ instead of DAPL_DBG_TYPE_CM.
+
+ Signed-off-by: Sean Hefty <sean.hefty@intel.com>
+
+commit 73728763666a46df5789af93b50db53cdf64afd6
+Author: Sean Hefty <sean.hefty@intel.com>
+Date: Fri Apr 10 08:07:44 2009 -0700
+
+ The IBAL-SCM provider will run into an inifinite loop if the check for
+ cr->socket > SCM_MAX_CONN - 1 fails. The code continues back to the start
+ of the while loop without moving to the next connection request entry
+ in the list.
+
+ Signed-off-by: Sean Hefty <sean.hefty@intel.com>
+
+commit 9b1b396539926d36ffacfff04fbe7c081e436b45
+Author: Sean Hefty <sean.hefty@intel.com>
+Date: Fri Apr 10 08:07:40 2009 -0700
+
+ next_cr is set just before and inside the check
+ if ((cr->socket == DAPL_INVALID_SOCKET && cr->state == SCM_DESTROY)
+ Remove setting it inside the if statement.
+
+ Signed-off-by: Sean Hefty <sean.hefty@intel.com>
+
+commit 7b49a9850f62276bb7bfccb2d85a1e94e311813c
+Author: Sean Hefty <sean.hefty@intel.com>
+Date: Fri Apr 10 08:07:35 2009 -0700
+
+ Some errors on windows are more easily interpretted in hex than decimal.
+
+ Signed-off-by: Sean Hefty <sean.hefty@intel.com>
+
+commit 08ee072a1396ac2c28983878dbc6b02feb035787
+Author: Sean Hefty <sean.hefty@intel.com>
+Date: Fri Apr 10 08:07:32 2009 -0700
+
+ The WinOF HCA driver cannot handle transitioning from RTS -> RESET ->
+ INIT -> ERROR. Simply delete the QP and re-create it to reinitialize
+ the endpoint until the bug is fixed.
+
+ Signed-off-by: Sean Hefty <sean.hefty@intel.com>
+
+commit 51ef5d96ce67d6141ec02b2a318f1b6e12be1bcf
+Author: Sean Hefty <sean.hefty@intel.com>
+Date: Fri Apr 10 08:07:23 2009 -0700
+
+ Signed-off-by: Sean Hefty <sean.hefty@intel.com>
+
+commit a8582be0e1fc89e856f1d0b43a3c1b271295a352
+Author: Sean Hefty <sean.hefty@intel.com>
+Date: Fri Apr 10 08:07:18 2009 -0700
+
+ Convert the openib_cma provider to common code between linux and windows.
+
+ Signed-off-by: Sean Hefty <sean.hefty@intel.com>
+
+commit cc2a71dfe0c35a70f6b1ba66070a3a06059a8bb5
+Author: Sean Hefty <sean.hefty@intel.com>
+Date: Fri Apr 10 08:06:53 2009 -0700
+
+ Move from using pipes to sockets for internal communication. This
+ avoids issues with windows only supporting select() on sockets.
+
+ Remove windows specific definition of dapl_dbg_log.
+
+ Update to latest windows libibverbs implementation using completion
+ channel abstraction to improve windows scalability and simplify
+ porting where FD's are accessed directly in Linux.
+
+ Signed-off-by: Sean Hefty <sean.hefty@intel.com>
+
+commit b3ad2ed97399a24a869841e17d1314e11c379aae
+Author: Arlin Davis <arlin.r.davis@intel.com>
+Date: Tue Mar 31 05:41:50 2009 -0800
+
+ Release 2.0.17
+
+ Signed-off-by: Arlin Davis <ardavis@ichips.intel.com>
+
commit 5d732929f8a90a490994e8e35a3666c3647ad4fe
Author: Arlin Davis <arlin.r.davis@intel.com>
Date: Tue Mar 31 05:22:11 2009 -0800
projects / ~ardavis / dapl.git / commitdiff