OFED 1.5.1 rc1 release is available
Vladimir Sokolovsky
Acked-by Steve WiseOFED 1.5.1-rc1 is available
Notes:
The tarball is available on:
http://www.openfabrics.org/builds/ofed-1.5.1/release/OFED-1.5.1-rc1.tgz
To get BUILD_ID run ofed_info
Please report any issues in bugzilla https://bugs.openfabrics.org/ for OFED 1.5.1
Vladimir & Tziporet
========================================================================
Supported Platforms and Operating Systems
---------------------------------------------
o CPU architectures:
- x86_64
- x86
- ppc64
- ia64
o Linux Operating Systems:
- RedHat EL4 up7 2.6.9-78.ELsmp
- RedHat EL4 up8 2.6.9-89.ELsmp
- 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 4 up8 2.6.9-89.ELsmp
- CentOS5.3 2.6.18-128.el5
- CentOS5.4 2.6.18-164.el5
- Fedora Core12 2.6.31.5-127.fc12 *
- OpenSuSE 11.2 2.6.31.5-0.1-default *
- kernel.org 2.6.29, 2.6.30,2.6.31 and 2.6.32 *
* Minimal QA for these versions
Main changes from 1.5:
=====================
1. Added RoCEE support
2. updated Open MPI to rev 1.4.1
3. Updated MVAPICH2 to rev 1.4-2
4. Updated DAPL to rev 2.0.26
5. Added extended atomic operations to ConnectX (kernel only)
6. Bug fixes
commit 95af56b8c3576e040e3bd06f5167b5cf2e16ad0a
Author: Vladimir Sokolovsky <vlad-VPRAkNaXOzVS1MOuV/[email protected]> Date: Thu Feb 11 13:00:30 2010 +0200
IB/core: Fix the ability to check for a receive by testing opcode & IB_WC_RECV
The previous patch that add support for masked atomic operation breaks the ability to check for a receive by testing opcode & IB_WC_RECV. This patch fixes this issue.
commit 96aded67b664b28939535845af00bf71aba85027
Author: Eli Cohen <eli-VPRAkNaXOzVS1MOuV/[email protected]> Date: Thu Feb 11 11:38:02 2010 +0200
rocee/cma: bug fix in cma_resolve_iboe_route()
make sure interface is bound and netdev found before proceeding with cma_resolve_iboe_route(). Failure to do so can result in kernel oops.
commit db05a7050eec3f9fa369b67774a2ab0875d2cf02
Author: Sean Hefty <[email protected]> Date: Wed Feb 10 10:04:27 2010 +0200
rdma/cm: revert associating an RDMA device when binding to loopback
Revert the following change from commit
6f8372b69c3198e06cecb1df2cb9682d0c55e657:
The defined behavior of rdma_bind_addr is to associate an RDMA device with an rdma_cm_id, as long as the user specified a non- zero address. (ie they weren't just trying to reserve a port) Currently, if the loopback address is passed to rdma_bind_addr, no device is associated with the rdma_cm_id. Fix this.
It turns out that openmpi depends on rdma_bind_addr NOT associating any RDMA device when binding to a loopback address. Openmpi is being updated to correct this, but until a new openmpi release is available, maintain the previous behavior: allow rdma_bind_addr to succeed, but do not bind to a device.
commit 3c43e70f78c2251cbf2eb54fe30f7cb13601c774
Merge: 45f08da baa5070
Author: Vladimir Sokolovsky <vlad-VPRAkNaXOzVS1MOuV/[email protected]>
Date: Wed Feb 10 09:40:05 2010 +0200
Merge branch 'ofed_1_5' of ssh://sofa.openfabrics.org/home/swise/scm/ofed_kernel into ofed_kernel_1_5
commit baa5070c056c52019a4027d10b600ae5df5ceece
Author: Tom Tucker <tom-/Yg/[email protected]> Date: Tue Feb 9 15:16:50 2010 -0600
svcrdma: Fixes to the NFSRDMA patch files.
commit 45f08da4451f76a7637f8bfa2d06607c9a7c66f1
Author: Vladimir Sokolovsky <vlad-VPRAkNaXOzVS1MOuV/[email protected]> Date: Tue Feb 9 10:43:02 2010 +0200
ofa_kernel.spec: Fix ibdev2netdev installation
ibdev2netdev will be installed if mlx4 kernel module selected for installation
commit a4d3aae90bf9c88af01b145dd9fbbc2382c57dcd
Author: Yevgeny Petrilin <yevgenyp-VPRAkNaXOzVS1MOuV/[email protected]> Date: Mon Feb 8 14:26:18 2010 +0200
mlx4_en: Ignoring missing vpd
commit 72eafe294c55c77ac3e3bc50a89dcd8dae7ba80a
Author: Vladimir Sokolovsky <vlad-VPRAkNaXOzVS1MOuV/[email protected]> Date: Mon Feb 8 11:30:48 2010 +0200
Fix backport patches for kernels <= 2.6.21
commit b50bbce79639329ca857f26fbf50120b2372e708
Author: Vladimir Sokolovsky <vlad-VPRAkNaXOzVS1MOuV/[email protected]> Date: Mon Feb 8 11:12:37 2010 +0200
mlx4/IB: Add support for enhanced atomic operations
Added support for masked atomic operations.
commit 8ae1b84f0d8901948085b952bc4a86d921ea05a6
Author: Vladimir Sokolovsky <vlad-VPRAkNaXOzVS1MOuV/[email protected]> Date: Mon Feb 8 11:11:32 2010 +0200
IB/core: Add support for enhanced atomic operations
- Add a new IB_WR_ATOMIC_MASKED_CMP_AND_SWP and IB_WR_ATOMIC_MASKED_FETCH_AND_ADD send opcodes that can be used to mark a "masked atomic compare and swap" and "masked atomic fetch and add" work request correspondingly.
- Add IB_DEVICE_MASKED_ATOMIC capability bit.
- Add mask fields to atomic struct of ib_send_wr
- Add new opcodes to ib_wc_opcode
commit 983e117ff68b071cb160a7a8ce02244fed45d2a4
Merge: 3e7a44f feba462
Author: Vladimir Sokolovsky <vlad-VPRAkNaXOzVS1MOuV/[email protected]>
Date: Sun Feb 7 12:07:51 2010 +0200
Merge branch 'ofed_kernel_1_5' of git://git.openfabrics.org/~amirv/ofed_1_5 into ofed_kernel_1_5
commit feba462d9055b2f3a0b77dbad8db0514d911442a
Author: Amir Vadai <amirv-VPRAkNaXOzVS1MOuV/[email protected]> Date: Thu Feb 4 17:31:52 2010 +0200
sdp: BUG1899 - fix warnings on RH4.8 by avoiding multiple deletions on the same timer.
commit 3e7a44f0ede1c1d0612a11d96a9b6954a455e314
Author: Vladimir Sokolovsky <vlad-VPRAkNaXOzVS1MOuV/[email protected]> Date: Thu Feb 4 16:36:03 2010 +0200
ofed_scripts: Renamed ibdev2eth -> ibdev2netdev
Script name changed after adding IPoIB interfaces support
commit 5a7496f73c4e207e67dc5d2b6687eb5964428d5e
Author: Vladimir Sokolovsky <vlad-VPRAkNaXOzVS1MOuV/[email protected]> Date: Thu Feb 4 16:02:23 2010 +0200
ofa_kernel.spec: Fix files installation
Create missing /usr/bin directory
commit 00e1c15d9bbeb4a23919defbba0fc1fc7553a939
Author: Eli Cohen <eli-VPRAkNaXOzVS1MOuV/[email protected]> Date: Thu Feb 4 12:53:51 2010 +0200
ib_core/rocee: fix failure to initilalize spinlock
ib_sa_add_one() skips calling ib_register_mad_agent() for rocee ports, but also failed to initialize the port spinlock which is still used.
commit ba1f35e562e7d4dfdc46e015fc7427127ff43512
Author: Eli Cohen <eli-VPRAkNaXOzVS1MOuV/[email protected]> Date: Wed Feb 3 14:00:48 2010 +0200
CMA: Fix iWarp failures to bind to a device
rdma_addr_get_sgid() relies on dev_addr->transport to retrieve the correct GID based on the hardware address. However, when called from cma_acquire_dev(), the transport field is not yet valid. The solution is to avoid calling rdma_addr_get_sgid() from cma_acquire_dev() and find the device based on it's
GID: for ethernet, assume first it is rocee and search the GID table, if not found generate the GID by copying it from the hardware address.
commit 3467dec0e02e33c4666b92fb7193bc308fbfcfcb
Author: Eli Cohen <eli-VPRAkNaXOzVS1MOuV/[email protected]> Date: Thu Feb 4 12:18:28 2010 +0200
ofed_scripts: Added ibdev2eth script
When working with RoCEE, you may need to know how to associate (ib device, port) with eth device. The ibdev2eth script does just that.
This is especially useful when there are a number of HCAs in the system.
commit 774e45b826d81d1c6ad2c3e2f1d993f3125d2466
Merge: 6fff638 f8ab933
Author: Vladimir Sokolovsky <vlad-VPRAkNaXOzVS1MOuV/[email protected]>
Date: Wed Feb 3 19:58:12 2010 +0200
Merge branch 'ofed_1_5' of ssh://[email protected]/~swise/scm/ofed_kernel into ofed_kernel_1_5
commit 6fff63846eb4de7b301d6417640cb65965cb1bb5
Author: Sebastien Dugue <sebastien.dugue-6ktuUTfB/[email protected]> Date: Tue Feb 2 11:11:04 2010 +0200
Fix 2.6.3[12] user_mad backports
The ib_umad_port's sm_sem semaphore is initialized as a mutex which although it builds (with a warning) badly breaks at runtime.
Fix this by using the proper initialization function sema_init().
Also fix those DOSish end of lines.
commit ce27ef63091d76784bc09e0e494c22a61468a2d3
Author: Andy Grover <[email protected]> Date: Mon Feb 1 15:32:13 2010 -0800
qlgc_vnic: Fix typo in header guard
A typo in the header guard may cause compilation errors against Xen-enabled kernels.
commit 40fbb82255896a5c39b35035e2fd5202246b6508
Author: Yevgeny Petrilin <yevgenyp-VPRAkNaXOzVS1MOuV/[email protected]> Date: Sun Jan 31 11:17:20 2010 +0200
mlx4_en: No need to set smaller tx queue len
commit a2b7387dd83a4ca41ac4367cc8092cef59051893
Author: Yevgeny Petrilin <yevgenyp-VPRAkNaXOzVS1MOuV/[email protected]> Date: Sun Jan 31 10:57:47 2010 +0200
mlx4_en: removed redundant backport for vlan_features
commit f89895b1b3f1ae2af850fa98a6c8ccd81f3938e8
Author: Yevgeny Petrilin <yevgenyp-VPRAkNaXOzVS1MOuV/[email protected]> Date: Sun Jan 24 11:49:50 2010 +0200
mlx4_en: Setting offloads for vlan interfaces
commit f8ab9334621cc7b109ce2e692fe5bbb39af766c2
Author: Steve Wise <[email protected]> Date: Fri Jan 29 09:24:53 2010 -0600
cxgb3: advertise 256K max CQ depth.
commit 7323ca442eae728964419a63b1101ff344cb0637
Author: Tina Yang <[email protected]> Date: Thu Jan 28 16:19:41 2010 -0800
RDS: Fix send locking issue
Fix a deadlock between rds_rdma_send_complete() and rds_send_remove_from_sock() when rds socket lock and rds message lock are acquired out-of-order.
commit 26d8079f3f31a08569d6fe4fac726617eaf5d2d1
Author: Andy Grover <[email protected]> Date: Wed Jan 20 15:01:05 2010 -0800
RDS: Fix congestion issues for loopback
We have two kinds of loopback: software (via loop transport) and hardware (via IB). sw is used for 127.0.0.1, and doesn't support rdma ops. hw is used for sends to local device IPs, and supports rdma. Both are used in different cases.
For both of these, when there is a congestion map update, we want to call rds_cong_map_updated() but not actually send anything -- since loopback local and foreign congestion maps point to the same spot, they're already in sync.
The old code never called sw loop's xmit_cong_map(),so rds_cong_map_updated() wasn't being called for it. sw loop ports would not work right with the congestion monitor.
Fixing that meant that hw loopback now would send congestion maps to itself. This is also undesirable (racy), so we check for this case in the ib-specific xmit code.
commit 64120f74851b94b7a49f32e0b1202dd6b944e0d5
Author: Steve Wise <[email protected]> Date: Tue Jan 26 15:28:42 2010 -0600
cxgb3: bump fw to 7.8
commit 5d6a8491f0075d1b964a7bff460535b56692ea7f
Author: Steve Wise <[email protected]> Date: Tue Jan 26 15:28:42 2010 -0600
iw_cxgb3: Add big cqs patch.
commit b57100b0c9654859359fdee005c2d8b756e61913
Author: Steve Wise <[email protected]> Date: Tue Jan 26 15:28:41 2010 -0600
cxgb3: Doorbell overflow avoidance/recovery.
commit 1ed36496633af886879b97b478c2b0833ecc10d4
Author: Steve Wise <[email protected]> Date: Tue Jan 26 15:28:35 2010 -0600
Core: Add idr_for_each() to backports.
commit 6d1ad4a5df4589a280b5d6cbb21a7e91b2df2089
Author: Steve Wise <[email protected]> Date: Tue Jan 26 13:41:47 2010 -0600
cxgb3: Allow read/write of t3 regs.
commit e2ccd34b1d3d8af083bd922be46b351ac6a0057e
Author: Jack Morgenstein <[email protected]> Date: Tue Jan 26 12:33:16 2010 +0200
mlx4_core: adjust T11 fix so that it does not overwrite other flag bits.
The T11 patch wrote an entire byte in the INIT_HCA flags dword. Correct this so that only a single bit gets set.
In addition, change T11 logic so that the T11 bit gets set only if the ConnectX supports T11 format.
Adjust fix mlx4_1310 RoCEE so that it does not collide.
commit b6d857394b9fb76c628088e5207e2137eb937499
Author: Amir Vadai <amirv-VPRAkNaXOzVS1MOuV/[email protected]> Date: Sun Jan 24 17:12:34 2010 +0200
sdp: must use ib_sg_dma_*, not sg_dma_* for mapping
This fixes OFED bug 1895, althoug some warnings are still generated, when running qperf sdp_bw with large sizes (using zcopy), on the truescale adapters.
commit 985bfdd2a130c05b275d21e27c2b0f1fde4191bd
Author: Ella Podvalny <ella-VPRAkNaXOzVS1MOuV/[email protected]> Date: Thu Jan 21 11:56:42 2010 +0200
kernel_addons: for fcoe RH5.2
commit 1cf680464b85e1eff7c8a935d03ac3c456c51cae
Author: Ella Podvalny <ella-VPRAkNaXOzVS1MOuV/[email protected]> Date: Thu Jan 21 09:43:57 2010 +0200
kernel_fixes: adding t11 module parameter
commit c2e6887a72d6db0ad6d7f19c531d533ffe39de75
Author: Or Gerlitz <ogerlitz-smomgflXvOZWk0Htik3J/[email protected]> Date: Sun Jan 17 11:34:35 2010 +0200
iser: Add backport for RHEL 5.4
patch the ofed provided iser code to match the RHEL 5.4 provided iser code, mainly by doing iscsi2_xxx instead of iscsi_xxx calls. This trivial transformation works since the 5.4 iser/iscsi stack was upgraded to 2.6.30 or so.
commit a2d4b687aa050f21dffcbe922e82db54cc3d9fd9
Author: Or Gerlitz <ogerlitz-smomgflXvOZWk0Htik3J/[email protected]> Date: Sun Jan 17 11:31:51 2010 +0200
IB/iser: Rewrite SG handling for RDMA logic
After dma-mapping an SG list provided by the SCSI midlayer, iser has to make sure the mapped SG is "aligned for RDMA" in the sense that its possible to produce one mapping in the HCA IOMMU which represents the whole SG. Next, the mapped SG is formatted for registration with the HCA.
This patch re-writes the logic that does the above, to make it clearer and simpler. It also fixes a bug in the being aligned for RDMA checks, where a "start" check wasn't done but rather only "end" check.
commit 921bfa2708441a6b30ac2cd5bf0372f269551a76
Author: Vladimir Sokolovsky <vlad-VPRAkNaXOzVS1MOuV/[email protected]> Date: Wed Jan 13 13:30:20 2010 +0200
Fix RHEL4U[6-8] backports
commit 070b34d4a72dcbff6ac39b408bfabcf1bdce7ebc
Author: Vladimir Sokolovsky <vlad-VPRAkNaXOzVS1MOuV/[email protected]> Date: Tue Jan 12 19:02:01 2010 +0200
rocee: Fix SLES10 SP3 backport
commit a2bf1fdf807bfbdc136c919a225097c6fbaf9aea
Author: Eli Cohen <eli-VPRAkNaXOzVS1MOuV/[email protected]> Date: Tue Jan 12 14:15:46 2010 +0200
Fix backport for netif_oper_up
For kernels that do not provide it, suffice with netif_running()
commit e42e22b7942e6bf40cd7b9b926679c54aba61116
Author: Eli Cohen <eli-VPRAkNaXOzVS1MOuV/[email protected]> Date: Tue Jan 12 11:19:57 2010 +0200
Update backports for dev_get_by_index
commit 93f7c9d1e7b16476b8ab134193be86efb74b1e5f
Author: Eli Cohen <eli-VPRAkNaXOzVS1MOuV/[email protected]> Date: Mon Jan 11 22:29:28 2010 +0200
Fix backports
Remove net_eq calls for kernels that do not support it.
commit 94e4afe453846babd2e6fd641bc0f9a454b3c480
Author: Eli Cohen <eli-VPRAkNaXOzVS1MOuV/[email protected]> Date: Mon Jan 11 18:51:46 2010 +0200
Fix compilation errors after adding RoCEE
commit 4297fe4c8642cba55642265863e5d359b8cb79b2
Author: Eli Cohen <eli-VPRAkNaXOzVS1MOuV/[email protected]> Date: Thu Jan 7 16:11:13 2010 +0200
Fix backports after adding RoCEE
commit 44fc4a53e979b5d95b0362e63be64bef6cbb1fe4
Author: Eli Cohen <eli-VPRAkNaXOzVS1MOuV/[email protected]> Date: Thu Jan 7 11:04:43 2010 +0200
Add support for RoCEE
Add support for RoCEE based on v7 of the patches sent to the rmda mailing list. Also add support for counters in mlx4.
commit 3a8f3b86da8cfb9b19048cda7073874cfb795d58
Author: Ella Podvalny <ella-VPRAkNaXOzVS1MOuV/[email protected]> Date: Wed Jan 6 13:30:02 2010 +0200
commit 7c3e183793968ad9d2353215d5a6f7df63a82b2f
Author: David J. Wilder <dwilder-r/[email protected]> Date: Tue Dec 15 09:43:08 2009 -0800
bacmports: create an if_inet6.h kernel_addon for the 2.6.9-[67,78,89]
kernels.
Signed-off-by: David Wilder <dwilder-r/[email protected]>
------------------------------------------------------
.../backport/2.6.9_U6/include/net/if_inet6.h | 20 ++++++++++++++++++++
.../backport/2.6.9_U7/include/net/if_inet6.h | 20 ++++++++++++++++++++
.../backport/2.6.9_U8/include/net/if_inet6.h | 20 ++++++++++++++++++++
3 files changed, 60 insertions(+), 0 deletions(-)commit 8d3b1a25d2f883515c8dfb4d40d8765b7919029a
Author: David Wilder <dwilder-r/[email protected]> Date: Sun Dec 13 17:29:07 2009 +0200
Backport patches for rdma cm
commit 2c0ff2f658441a52d30864f865499535fea6990b
Author: Vladimir Sokolovsky <vlad-VPRAkNaXOzVS1MOuV/[email protected]> Date: Wed Dec 2 12:25:04 2009 +0200
rdma/cm: Fixup IPv6 support and IPv4 routing corner cases for RDMA CM
commit 5d96b90156bd87f1578f6497eda6c3150a16736d
Author: Andy Grover <[email protected]> Date: Mon Jan 4 21:25:53 2010 -0800
RDS/TCP: Wait to wake thread when write space available
Instead of waking the send thread whenever any send space is available, wait until it is at least half empty. This is modeled on how sock_def_write_space() does it, and may help to minimize context switches.
commit 7f4f82ec58c10dab756f3dc78f536d1893239806
Author: Andy Grover <[email protected]> Date: Mon Jan 4 21:11:40 2010 -0800
RDS: use IB_CQ_VECTOR_LEAST_ATTACHED for cq's
commit f3ac1b5ad844c0b5ffc736b46915e33f1205ff72
Author: Andy Grover <[email protected]> Date: Thu Dec 17 11:02:27 2009 -0800
RDS: update copy_to_user state in tcp transport
Other transports use rds_page_copy_user, which updates our s_copy_to_user counter. TCP doesn't, so it needs to explicity call rds_stats_add().
commit 129a51ce8dd7407985c96936f723e33e180342e5
Author: Amir Vadai <amirv-VPRAkNaXOzVS1MOuV/[email protected]> Date: Thu Dec 31 11:25:39 2009 +0200
sdp: use IB_CQ_VECTOR_LEAST_ATTACHED for cq's
commit 63714a73c94ccb758b51d7d901b412113b2df511
Author: Amir Vadai <amirv-VPRAkNaXOzVS1MOuV/[email protected]> Date: Wed Dec 30 15:43:10 2009 +0200
sdp: make statistics per cpu
- Caused lots of cache misses during multi stream traffic
commit 1f3f22e79c24ac733319fbe6ee63a59779e72ac0
Author: Amir Vadai <amirv-VPRAkNaXOzVS1MOuV/[email protected]> Date: Wed Dec 30 10:10:09 2009 +0200
sdp: added statistics instead of prints
bug_id,"bug_severity","priority","op_sys","assigned_to","bug_status","resolution","short_short_desc"
1899,"major","P3","RHEL 4","[email protected]","RESOLVED","FIXED","Getting timer related warnings when running sdp tests on RHEL 4.8" 1907,"major","P3","RHEL 5","[email protected]","RESOLVED","INVALID","Running ‘rds-tools/rds-gen –l 4096’ over RoCEE results in kernel panic" 1908,"major","P3","RHEL 5","[email protected]","RESOLVED","FIXED","Running ‘rds-tools/rds-gen –l 4096’ over RoCEE results in kernel panic" 1910,"major","P3","All","[email protected]","RESOLVED","FIXED","Running ucmatose over RoCEE results in Kernel panic" 1920,"critical","P3","All","[email protected]","RESOLVED","FIXED","OFED_1_5_1 doesn't support Drexler's device IDs" 1900,"normal","P3","All","[email protected]","RESOLVED","FIXED","The example test Ibv_srq_pingpong always fail when run over IB link" 1901,"normal","P3","All","[email protected]","RESOLVED","INVALID","The example test Ibv_srq_pingpong always fail when run over IB link" 1902,"normal","P3","All","[email protected]","RESOLVED","INVALID","The example test Ibv_srq_pingpong always fail when run over IB link" 1903,"normal","P3","All","[email protected]","RESOLVED","INVALID","The example test Ibv_srq_pingpong always fail when run over IB link" 1887,"normal","P3","All","[email protected]","RESOLVED","FIXED","qperf doesn't support operation between DDR and QDR servers" 1821,"major","P3","RHEL 5","[email protected]","RESOLVED","FIXED","Crash in bonding"
1921,"minor","P3","RHEL 5","[email protected]","RESOLVED","INVALID","Installation patch has broken hunk"
1774,"critical","P3","SLES 11","[email protected]","VERIFIED","FIXED","Installing OFED1.5 RC1 on SLES11 on machine with OFED-distro does not remove ofed-kmp" 1859,"normal","P2","All","[email protected]","CLOSED","FIXED","IPOIB Can miss a change in dgid"
_______________________________________________
ewg mailing list [email protected]
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg