Hi, I am resending this mail with just the title fixed to rc3. On Tue, Dec 01, 2009 at 06:41:37PM +0200, Eli Cohen wrote: > OFED-RDMAoE-1.5-rc3 is available > > The tarball is available on: > http://www.openfabrics.org/downloads/OFED/ofed-rdmaoe-1.5/OFED-RDMAoE-1.5-rc3.tgz > > > To get BUILD_ID run ofed_info > > Please report any issues in bugzilla https://bugs.openfabrics.org/ for > OFED 1.5, component RDMAoE > > > Release information: > -------------------- > Linux Operating Systems: > - RedHat EL4 up6 2.6.9-67.ELsmp > - RedHat EL4 up7 2.6.9-78.ELsmp > - RedHat EL4 up8 2.6.9-89.ELsmp > - RedHat EL5 up2 2.6.18-92.el5 > - RedHat EL5 up3 2.6.18-128.el5 > - RedHat EL5 up4 2.6.18-164.el5 > - SLES10 SP2 2.6.16.60-0.21-smp > - SLES10 SP3 2.6.16.60-0.54-smp > - SLES11 2.6.27.19-5-default > - OEL 4 up7 2.6.9-78.ELsmp > - OEL 5 up2 2.6.18-92.el5 > - CentOS5.2 2.6.18-92.el5 > - CentOS5.3 2.6.18-128.el5 > - Fedora Cor 2.6.29 * > - OpenSuSE 11 2.6.25.5-1.1 * > - kernel.org 2.6.29 and 2.6.30 > > > * Minimal QA for these versions > > Systems: > * x86_64 > * x86 > * ia64 > * ppc64 > > Main changes from 1.5 rc2: > ============================ > 1. Updated packages: > > libibverbs-1.1.2-0.8.g66ece2f.tar.gz > libipathverbs-1.2.tar.gz > dapl-2.0.25.tar.gz > compat-dapl-1.2.15.tar.gz > ib-bonding-0.9.0-41.src.rpm > rnfs-utils-1.1.5-10.OFED.src.rpm > > - Management packages: > libibumad-1.3.3_20091130_4ea1c4d.tar.gz > libibmad-1.3.3_20091130_4ea1c4d.tar.gz > opensm-3.3.3_20091130_4ea1c4d.tar.gz > infiniband-diags-1.5.3_20091130_4ea1c4d.tar.gz > ibutils-1.2-0.1.ge8e69b7.tar.gz > > - MPI packages: > mvapich2-1.4-2.src.rpm > mpitests-3.2-916.src.rpm (Updated IMB and OSU benchmark tests) > > 2. Bug fixes (attached log file) > > > Limitations: > ============ > - SLES10 SP3 on IA64 is not supported yet > > > Notes: > RDMAoE and mlx4_en capabilities are supported only for x86 and x86_64 > architectures. > >
> commit 437d9b4dfd4dbd58a8e52348c6f12b70ebf9bb1d > Merge: 5c0b86c... b9e3b12... > Author: Eli Cohen <[email protected]> > Date: Tue Dec 1 11:02:05 2009 +0200 > > Merge branch 'ofed_kernel_1_5' into rdmaoe > > commit b9e3b127bc904ef0bd9a3efb77bdf0c66a85f386 > Author: Yevgeny Petrilin <[email protected]> > Date: Tue Dec 1 10:27:50 2009 +0200 > > mlx4_core: Using selected vector number when setting EQ in cqc > > Signed-off-by: Yevgeny Petrilin <[email protected]> > > commit 56bfadef0b8fd97cfcd6b202f84d566ce6e9ccd3 > Author: Amir Vadai <[email protected]> > Date: Mon Nov 30 16:14:14 2009 +0200 > > sdp: cancel_work_sync on 2.6.22 didn't return a value > > Since the return value was used for debug message only, > gave up this message to make backports easier. > > Signed-off-by: Amir Vadai <[email protected]> > > commit 5c0b86cb9540abe2d130769d932e9b2c482218f0 > Merge: 3dc9160... 56bfade... > Author: Eli Cohen <[email protected]> > Date: Mon Nov 30 17:07:56 2009 +0200 > > Merge branch 'ofed_kernel_1_5' into rdmaoe > > commit 3dc9160cac856a60f3ea04bd253cf9f436572b5a > Merge: 4e401c5... c842ed7... > Author: Eli Cohen <[email protected]> > Date: Mon Nov 30 16:37:33 2009 +0200 > > Merge branch 'ofed_kernel_1_5' into rdmaoe > > commit 108d9abbab4251aa6806b57bf84a8ea63caa73d4 > Author: Amir Vadai <[email protected]> > Date: Mon Nov 30 15:04:05 2009 +0200 > > sdp: fix compilation error on kernel 2.6.20 > > Signed-off-by: Amir Vadai <[email protected]> > > commit c842ed79fe5811b2e42c3ea6fa5675125d5b4875 > Author: Yevgeny Petrilin <[email protected]> > Date: Mon Nov 30 09:47:45 2009 +0200 > > mlx4: Fixed a bug in dynamic completion vector allocation > > The cq allocation function would return -EINVAL in case that it > received MLX4_LEAST_ATTACHED_VECTOR as the vector number. > > Signed-off-by: Yevgeny Petrilin <[email protected]> > > commit 4e401c545d857ed52066c46a00c533b9f9a0facc > Author: Eli Cohen <[email protected]> > Date: Sun Nov 29 12:17:55 2009 +0200 > > From 0952e0395387e0cfcbeaa2232f92ee4545453014 Mon Sep 17 00:00:00 2001 > From: Eli Cohen <[email protected]> Date: Sun, 29 Nov 2009 11:51:48 +0200 > Subject: [PATCH 2/2] rdmaoe: Fix crash when using loopback > > When we use a loopback connection, we do not find the corresponding > netdev. > Failure to do so caused a crash when trying to retrieve the MTU. Use the > max > MTU for loopback devices. > Note: loopback still does not work in this version due to FW > configuration but > the system will not crash. > > Signed-off-by: Eli Cohen <[email protected]> > > commit f44485baeb44593e0faaa27e0a500c38add138e8 > Author: Eli Cohen <[email protected]> > Date: Sun Nov 29 12:11:46 2009 +0200 > > From c057fed4df8bccefdcba0d8411f2e007e131d968 Mon Sep 17 00:00:00 2001 > From: Eli Cohen <[email protected]> Date: Sun, 29 Nov 2009 10:12:25 +0200 > Subject: [PATCH 1/2] rmda_cm: avoid double free of path rec > > If cma_resolve_rdmaoe_route fails, set route->path_rec so that it will > not be > freed later in rdma_destroy_id(). > > Signed-off-by: Eli Cohen <[email protected]> > > commit 12e55fe428d0db1d2fee5b7cee90b350a6b439ea > Merge: d8a3012... ae05492... > Author: Eli Cohen <[email protected]> > Date: Mon Nov 30 10:09:09 2009 +0200 > > Merge branch 'ofed_kernel_1_5' into rdmaoe > > commit ae054922723792dcc041cd920f7d2266573ef9ce > Author: Amir Vadai <[email protected]> > Date: Sun Nov 29 14:23:58 2009 +0200 > > sdp: changed important prints in zcopy to warning instead of debug > > Signed-off-by: Amir Vadai <[email protected]> > > commit df44545b339ccee79de1e32c1262fc1fca28dfd3 > Author: Amir Vadai <[email protected]> > Date: Thu Nov 26 17:10:57 2009 +0200 > > sdp: flush rx_comp_work before destroying socket + make > dreq_wait_timeout_work flushing synced > > Signed-off-by: Amir Vadai <[email protected]> > > commit f017beef52008747bbddcb520f1d57fe51747d5d > Author: Amir Vadai <[email protected]> > Date: Thu Nov 26 17:28:58 2009 +0200 > > sdp: cleaned up debug prints > > Signed-off-by: Amir Vadai <[email protected]> > > commit d8a3012ae76db93798c3149edbbaf797838ce874 > Merge: 99606ee... 926a714... > Author: Eli Cohen <[email protected]> > Date: Thu Nov 26 15:06:55 2009 +0200 > > Merge branch 'ofed_kernel_1_5' into rdmaoe > > commit 926a7142d1edec675199066babb80c5c0775fb2e > Author: Andy Grover <[email protected]> > Date: Tue Nov 24 15:35:51 2009 -0800 > > RDS: Re-add pf/sol access via sysctl > > Although RDS has an official PF_RDS value now, existing software > expects to look for rds sysctls to determine it. We need to maintain > these for now, for backwards compatibility. > > Signed-off-by: Andy Grover <[email protected]> > > commit 773ff8caf5e1d24e5ada780ac18ac314994cc467 > Author: Amir Vadai <[email protected]> > Date: Wed Nov 25 14:59:46 2009 +0200 > > sdp: take a reference during zcopy_send > > Signed-off-by: Amir Vadai <[email protected]> > > commit bd02799a0716b9f4705376b854137ddaada660b3 > Author: Amir Vadai <[email protected]> > Date: Wed Nov 25 15:06:01 2009 +0200 > > sdp: Disable BZcopy + Enable ZCopy > > Signed-off-by: Amir Vadai <[email protected]> > > commit fda37e499eb8ba947a125a410441dcf3d89e73e1 > Author: Amir Vadai <[email protected]> > Date: Wed Nov 25 10:26:59 2009 +0200 > > sdp: changed some warnings into debug prints > > Signed-off-by: Amir Vadai <[email protected]> > > commit 4481a8ad8f83d9194b658f28b48c571f7aa4e921 > Author: Amir Vadai <[email protected]> > Date: Tue Nov 24 11:01:05 2009 +0200 > > sdp: fixed BUG1826 part 1 - schedule while atomic > > Allocate skb according to context > > Signed-off-by: Amir Vadai <[email protected]> > > commit 99606eeb30f2bd44988229c5f3a02df3c6638fc9 > Merge: f8cf54c... bb49143... > Author: Eli Cohen <[email protected]> > Date: Tue Nov 24 17:15:20 2009 +0200 > > Merge branch 'ofed_kernel_1_5' into rdmaoe > > commit bb49143753189b46c195635972e44ee512e5ee5c > Author: Eli Cohen <[email protected]> > Date: Tue Nov 24 17:14:23 2009 +0200 > > From 134e8ba45ac49b773ba912726b4faee6a2839fab Mon Sep 17 00:00:00 2001 > From: Eli Cohen <[email protected]> > Date: Tue, 24 Nov 2009 17:06:35 +0200 > Subject: [PATCH] mlx4: properly mask MGM entry members count > > The members_count field size is 24 bits so mask it properly when reading > it. > > Signed-off-by: Eli Cohen <[email protected]> > > commit f8cf54cb8c765d410f6f4cedcfb5e7a2714a5cdc > Author: Eli Cohen <[email protected]> > Date: Tue Nov 24 14:44:20 2009 +0200 > > From a08c9408580449f5a6912dc83993ba54225f69f2 Mon Sep 17 00:00:00 2001 > From: Eli Cohen <[email protected]> > Date: Tue, 24 Nov 2009 10:16:50 +0200 > Subject: [PATCH 5/6] rdmaoe/mlx4: Don't set the g bit in AH > > This not required to be done since it is implicit. > > Signed-off-by: Eli Cohen <[email protected]> > > commit edcd7662d74d99a16d59024b8ca2844bfd3bc5ae > Author: Eli Cohen <[email protected]> > Date: Tue Nov 24 14:40:09 2009 +0200 > > From cc3940c3cf16b53e4d0bf6f4a8c8a76d6d060ae0 Mon Sep 17 00:00:00 2001 > From: Eli Cohen <[email protected]> > Date: Mon, 23 Nov 2009 18:24:38 +0200 > Subject: [PATCH 4/6] mlx4: report grh for all AV's of an RDMAoE port > > This is preferred over using a reserved bit in the HW. > > Signed-off-by: Eli Cohen <[email protected]> > > commit ce392ca7f0790ab785a6cff86e4a37ea59767513 > Author: Eli Cohen <[email protected]> > Date: Tue Nov 24 14:33:28 2009 +0200 > > From 15aa5e57dadaba7c09ee4d0bb870526c735674ef Mon Sep 17 00:00:00 2001 > From: Eli Cohen <[email protected]> > Date: Mon, 23 Nov 2009 14:44:13 +0200 > Subject: [PATCH 3/6] rdmaoe/cma: fix bug in mapping ipv4 multicast > > The mapping function had a bug setting two bytes with 0xff where they > should > have been cleared. > > Signed-off-by: Eli Cohen <[email protected]> > > commit 8d25c7889c6e47d9f660224f3d48a173aef013e9 > Merge: cc015d3... f78229e... > Author: Eli Cohen <[email protected]> > Date: Tue Nov 24 14:19:05 2009 +0200 > > Merge branch 'ofed_kernel_1_5' into rdmaoe > > commit f78229e5764dd2c354adc36c766b166fb21dc24d > Merge: 9b7e7cf... cc39d9d... > Author: Vladimir Sokolovsky <[email protected]> > Date: Tue Nov 24 13:25:48 2009 +0200 > > Merge branch 'ofed_1_5' of > ssh://[email protected]/~swise/scm/ofed_kernel into ofed_kernel_1_5 > > commit 9b7e7cf0b17afbc9aa0844e46338101aed0db032 > Merge: 9cf376f... ea2b694... > Author: Vladimir Sokolovsky <[email protected]> > Date: Tue Nov 24 10:43:47 2009 +0200 > > Merge branch 'ofed_kernel_1_5' of > git://git.openfabrics.org/~amirv/ofed_1_5 into ofed_kernel_1_5 > > commit ea2b694c1a744fb9ece2724dedd43b4b9efc046b > Author: Amir Vadai <[email protected]> > Date: Tue Nov 24 09:33:53 2009 +0200 > > sdp: Fixed annoying warning by memtrack > > kzalloc done in sdp_seq_open is freed by the system in a function > that memtrack can't see. Therefore it printed false warning. > > Signed-off-by: Amir Vadai <[email protected]> > > commit ab7b734e930d22a25f904c47e4f4898718cb150c > Author: Amir Vadai <[email protected]> > Date: Tue Nov 24 09:32:39 2009 +0200 > > sdp: fixed BUG1796 - running out of memory on rx > > rcv queue could grow endlessly because minimal RX buffers in QP > was set to SDP_MIN_TX_CREDITS + 1 - so there always were credits > available for the sender. > > Signed-off-by: Amir Vadai <[email protected]> > > commit ae23d8839e308cd28f0d75ce98fb7dc247c9459b > Author: Amir Vadai <[email protected]> > Date: Mon Nov 23 13:29:41 2009 +0200 > > sdp: fixed sparse warnings > > Signed-off-by: Amir Vadai <[email protected]> > > commit cc39d9d3858d37bf45ac58f880a36b9a2f2a8e8d > Author: Steve Wise <[email protected]> > Date: Mon Nov 23 11:46:02 2009 -0600 > > cxgb3: pull in page unmap fix. > > Signed-off-by: Steve Wise <[email protected]> > > commit 9cf376f7bad5bf5e52ca9ad0437c94e03115a30e > Author: Yevgeny Petrilin <[email protected]> > Date: Mon Nov 23 18:24:26 2009 +0200 > > mlx4_core: Added missing device ID's > > Signed-off-by: Yevgeny Petrilin <[email protected]> > > commit 29fbed919ea2c66a322d2c4c179b39c481292c93 > Author: Yevgeny Petrilin <[email protected]> > Date: Mon Nov 23 12:32:58 2009 +0200 > > Fixed 'round_jiffies_relative' definition in timer.h backports > > Signed-off-by: Yevgeny Petrilin <[email protected]> > > commit cc015d329c871892481d13fa42b98e9d0ab754c7 > Merge: 67362c5... 33518d2... > Author: Eli Cohen <[email protected]> > Date: Sun Nov 22 17:07:35 2009 +0200 > > Merge branch 'ofed_kernel_1_5' into rdmaoe > > commit 33518d251b6893262f7946243d5913511f29499e > Author: Yevgeny Petrilin <[email protected]> > Date: Sun Nov 22 15:42:20 2009 +0200 > > mlx4_core: Revert commit 2d455685bfb144e60adde8e76737fa820bea9c3c > > Signed-off-by: Yevgeny Petrilin <[email protected]> > > commit c442f6facdb64025a73c44452813f57bb4ef51ef > Author: Brian J. Murrell <[email protected]> > Date: Fri Nov 20 13:33:57 2009 -0600 > > Fix SLES10sp2 compile issue of ipv6_addr_loopback being redefined > > Signed-off-by: Brian J. Murrell <[email protected]> > > commit 750c71bed3b4d22532134958ce853e8808bc6c84 > Author: Jon Mason <[email protected]> > Date: Fri Nov 20 13:31:25 2009 -0600 > > Fix NFS compile warning on SLES11 > > Signed-off-by: Jon Mason <[email protected]> > > commit 67362c5af180f6449c95a04864f114b5fdd6b27e > Author: Eli Cohen <[email protected]> > Date: Fri Nov 20 07:54:04 2009 +0200 > > From 4bceabcbba92dd47592683241187e4d342f4cf60 Mon Sep 17 00:00:00 2001 > From: Eli Cohen <[email protected]> > Date: Fri, 20 Nov 2009 07:15:08 +0200 > Subject: [PATCH 3/3] rdmaoe: use correct union member > > There are many overlapping fields between the eth and ib members of the > union > but in the rdmaoe context use the ones of the eth member. > Note: I keep using the g_slid field of the ib member so that checking > later the > gid in query ah will use this field and be simpler. Although this goes > into a > reserved filed in WQE, the HW doesn't care about it. > > Signed-off-by: Eli Cohen <[email protected]> > > commit c09e71371dba5388b62934576391323e70a74f10 > Author: Eli Cohen <[email protected]> > Date: Fri Nov 20 07:34:14 2009 +0200 > > From 1013ee23724cd853175c4d4def9263aa8fefe30d Mon Sep 17 00:00:00 2001 > From: Eli Cohen <[email protected]> > Date: Thu, 19 Nov 2009 17:21:11 +0200 > Subject: [PATCH 2/3] mlx4_ib: Fix bug use before set > > mlx4_ib_query_ah referred to the port_num field before updating it. > This fixes bug #1832. > > Signed-off-by: Eli Cohen <[email protected]> > > commit fb7cc0eff85e2fb176819130f7d0965646500214 > Author: Eli Cohen <[email protected]> > Date: Fri Nov 20 07:26:51 2009 +0200 > > From 639b6c64666c36e24a68a609d9b304bdba7cb218 Mon Sep 17 00:00:00 2001 > From: Eli Cohen <[email protected]> > Date: Mon, 16 Nov 2009 17:11:25 +0200 > Subject: [PATCH 1/3] mlx4: use mac as an array of 6 bytes > > Instead of two arrays of 2 and 4 bytes. > > Signed-off-by: Eli Cohen <[email protected]> > > commit 7fc3a1c85eef027dc982b6257c5cef33e1d37070 > Author: Eli Cohen <[email protected]> > Date: Thu Nov 19 21:53:46 2009 +0200 > > Fix backports after merge > > Signed-off-by: Eli Cohen <[email protected]> > > commit 6cbfb22d2708ba400e6e8dd1edb477c2c365acf0 > Merge: cb24203... 9567e1b... > Author: Eli Cohen <[email protected]> > Date: Thu Nov 19 20:32:57 2009 +0200 > > Merge branch 'ofed_kernel_1_5' into rdmaoe > > commit 9567e1bfde37ecac308ae28cebd90af6e353761a > Author: Jon Mason <[email protected]> > Date: Thu Nov 19 10:07:06 2009 -0600 > > Fix build break caused by the 1671 patch not adding the new file > locations for SLES11_update and SLES10sp3. > > Signed-off-by: Jon Mason <[email protected]> > > commit 3e1f93df6696c7d07414c6e584a30b3459152061 > Author: Yevgeny Petrilin <[email protected]> > Date: Thu Nov 19 16:21:01 2009 +0200 > > Fixed workqueue backport for 2.6.18 and 2.6.19 systems > > Signed-off-by: Yevgeny Petrilin <[email protected]> > > commit 3097e6ac88b1ae0f2e655bca7e688891a5cf64e4 > Author: Yevgeny Petrilin <[email protected]> > Date: Thu Nov 19 10:53:26 2009 +0200 > > Fixed workqueue backport for Sles10 systems > > Signed-off-by: Yevgeny Petrilin <[email protected]> > > commit 3d459694b83b66b08e6722fbab873a7f78b66bcd > Author: Jon Mason <[email protected]> > Date: Wed Nov 18 17:46:30 2009 -0600 > > The EADDRNOTAVAIL patch causes more problems than it solves. Remove it. > > Signed-off-by: Jon Mason <[email protected]> > > commit e0d151154a46ce6d77ea39e3a072d621b7e6dabc > Author: Jon Mason <[email protected]> > Date: Thu Jul 30 17:41:24 2009 -0500 > > Lustre has a requirement to use either the Native OS version of NFS or > the OFED version of NFS. When disabling the OFED NFS to use the Native > NFS, there are headerfile conflicts between what Lustre expects to be > there and what actually is there. This patch works around this issue by > relocating the NFS specific headerfiles to a unique location and only > references them there if OFED NFS is selected. > > This patch resolves bug 1671. > > Signed-Off-By: Jon Mason <[email protected]> > > commit f6d607f1d80de11ef9a01fc0517d2ad6e1936a06 > Author: Jon Mason <[email protected]> > Date: Mon Nov 16 14:08:23 2009 -0600 > > Not all of the calls to send_write are properly handling the error case > and unmapping the outstanding ib_dma_maps when an error is encountered. > Also, send_write is not returning en error when the connection is closed > while it is sleeping waiting for free slots on the send queue. This > patch corrects the calls to send_write that did not cleanup after > themselves and now returns an error if the connection is closed in all > cases. > > This fixes bug 1824. > > Signed-off-by: Jon Mason <[email protected]> > > commit 1d9c4ee11209429a009aa10f6a35b85984a80c26 > Author: Jon Mason <[email protected]> > Date: Mon Nov 16 13:54:16 2009 -0600 > > Undo commit e692689e3b77c0cdd7594bb225c09a0767899693 > > The change was unneeded and provides no fix for NFSRDMA running over > ConnectX. > > Signed-off-by: Jon Mason <[email protected]> > > commit 064b54f7d9923da072eae74f97e1f695474b7f01 > Author: Moni Shoua <[email protected]> > Date: Wed Nov 18 13:19:15 2009 +0200 > > ofed_scripts: Do not remove slaves before stopping stack > > There is no need to unenslave IPoIB interfaces before stopping the stack. > This can cause a kernel crash as described in > https://bugs.openfabrics.org/show_bug.cgi?id=1821 > > Signed-off-by: Moni Shoua <[email protected]> > > commit 2d455685bfb144e60adde8e76737fa820bea9c3c > Author: Yevgeny Petrilin <[email protected]> > Date: Wed Nov 18 10:49:18 2009 +0200 > > mlx4_core: Using QUERY_STAT_CFG for port enablement > > Signed-off-by: Yevgeny Petrilin <[email protected]> > > commit 6f4b4859c0735de0e2aab4fcdc4efd677b34bf2f > Author: Yevgeny Petrilin <[email protected]> > Date: Wed Nov 18 10:48:44 2009 +0200 > > mlx4_core: Added transceiver type to QUERY_PORT > > Signed-off-by: Yevgeny Petrilin <[email protected]> > > commit 4e0958d8db84ec233a75d289f7d636f974741d9d > Author: Yevgeny Petrilin <[email protected]> > Date: Tue Nov 17 18:20:22 2009 +0200 > > mlx4_core: Updated 'cancel_delayed_work_synch' backport > > Signed-off-by: Yevgeny Petrilin <[email protected]> > > commit cb2420324e5acbe609d26764bb0b0fef09578498 > Merge: a26f683... 69907e1... > Author: Eli Cohen <[email protected]> > Date: Thu Nov 12 15:15:15 2009 +0200 > > Merge branch 'ofed_kernel_1_5' into rdmaoe > > commit 69907e1c6d6d8b5751ec34d4f162bac60f563a2d > Merge: 9fac87d... a9b3185... > Author: Vladimir Sokolovsky <[email protected]> > Date: Thu Nov 12 10:00:50 2009 +0200 > > Merge branch '1822' of > ssh://[email protected]/home/jon/scm/ofed_kernel-1.5 into > ofed_kernel_1_5 > > commit 9fac87ded0d34d79b82a932a46c63e0c91c33c0c > Author: Yevgeny Petrilin <[email protected]> > Date: Wed Nov 11 10:12:43 2009 +0200 > > mlx4: Saving port configuration > > Added a list of port configurations per PCI device, > now when the driver is recovering from catastrophic event > it will be able to restore its previous configuration. > > Signed-off-by: Yevgeny Petrilin <[email protected]> > > commit c0ac820ae29daf121271b0b3a2c9f042ff651b4c > Author: Yevgeny Petrilin <[email protected]> > Date: Tue Nov 10 17:09:32 2009 +0200 > > mlx4: Default value for automatic completion vector selection > > When the vector number passed to mlx4_cq_alloc is > MLX4_LEAST_ATTACHED_VECTOR (0xffffffff), > the driver selects the completion vector that has the least CQ's attached > to it and attaches the CQ to the chosen vector. > IB_CQ_VECTOR_LEAST_ATTACHED is defined in rdma/ib_verbs.h, when mlx4_ib > driver, > recieves this cq vector number, it uses MLX4_LEAST_ATTACHED_VECTOR an CQ > creation. > > Signed-off-by: Yevgeny Petrilin <[email protected]> > > commit a9b3185fc5fed02ef73219239ba6226bfa6451f0 > Author: Jon Mason <[email protected]> > Date: Tue Nov 10 14:44:37 2009 -0600 > > SLES11 kernels linux-2.6.27.25-0.1 and greater have redefined the > vm_operations_struct to be different than the one that shipped in SLES11 > GA (linux-2.6.27.19-5). This breaks compilation on the newer kernels. > This patch modified NFS to refer to the new definition of that struct. > > This patch fixes bug 1822 > > Signed-off-by: Jon Mason <[email protected]> > > commit a26f683337d672aedc885623b6e0992275200969 > Merge: 766e316... 97ea31f... > Author: Eli Cohen <[email protected]> > Date: Wed Nov 11 13:25:35 2009 +0200 > > Merge branch 'ofed_kernel_1_5' into rdmaoe > > commit 9574afc6c3b78ff20e2041810aa02fedb19c4216 > Author: Jon Mason <[email protected]> > Date: Tue Nov 10 14:19:03 2009 -0600 > > The SLES11 kernel updates have redefined some structs needed by NFS, > thus causing build breaks on the updated kernels. Since there is no > elegant way to differentiate between the versions, a separate backport > must be made for the updated kernels. > > Needed for bug 1822 > > Signed-off-by: Jon Mason <[email protected]> > > commit 97ea31f3aad9b31d4bf1e2845fe33505b6d08998 > Author: Vladimir Sokolovsky <[email protected]> > Date: Mon Nov 9 02:25:25 2009 +0200 > > mlx4/IB: Add set_4k_mtu module parameter. > > It control Infiniband link MTU for all IB ports in a host. > > Signed-off-by: Vladimir Sokolovsky <[email protected]> > > commit 766e31663a08421026f379661bd3a08f407957a4 > Merge: c35c3bd... 51df451... > Author: Eli Cohen <[email protected]> > Date: Sun Nov 8 17:05:14 2009 +0200 > > Merge branch 'ofed_kernel_1_5' into rdmaoe > > commit 51df451753f115f271198e14c7df51ada82e2d93 > Author: Eli Cohen <[email protected]> > Date: Sun Nov 8 16:33:35 2009 +0200 > > IPoIB: fix crashes when creatig a child interface > > In kernels < 2.6.17, netdev_register_sysfs() is called only when > rtnl_unlock() > gets called so we can't add files to sysfs from within ipoib_vlan_add() > since > it hold rtnl lock. The solution is to free it the lock and acquire it > again so > netdev_register_sysfs() gets called. This fixes bug #1770. > > Signed-off-by: Eli Cohen <[email protected]> > > commit 2d4234db198f0b2e1f625a400ec98ebab2c2cbf2 > Author: Vladimir Sokolovsky <[email protected]> > Date: Sun Nov 8 12:32:37 2009 +0200 > > IA64/backport: Fix SLES10 SP2/3 backports > > Signed-Off-By: Jon Mason <[email protected]> > Signed-off-by: Vladimir Sokolovsky <[email protected]> > _______________________________________________ > ewg mailing list > [email protected] > http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg _______________________________________________ ewg mailing list [email protected] http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
