Hi,
OFED-1.4-rc5 release is available on
http://www.openfabrics.org/downloads/OFED/ofed-1.4/OFED-1.4-rc5.tgz
To get BUILD_ID run ofed_info
Please report any issues in bugzilla https://bugs.openfabrics.org/ for
OFED 1.4
Vladimir & Tziporet
========================================================================
Release information:
------------------------------
Linux Operating Systems:
- RedHat EL4 up4: 2.6.9-42.ELsmp *
- RedHat EL4 up5: 2.6.9-55.ELsmp
- RedHat EL4 up6: 2.6.9-67.ELsmp
- RedHat EL4 up7: 2.6.9-78.ELsmp
- RedHat EL5: 2.6.18-8.el5
- RedHat EL5 up1: 2.6.18-53.el5
- RedHat EL5 up2: 2.6.18-92.el5
- CentOS 5.2: 2.6.18-92.el5
- Fedora C9: 2.6.25-14.fc9 *
- SLES10: 2.6.16.21-0.8-smp
- SLES10 SP1: 2.6.16.46-0.12-smp
- SLES10 SP1 up1: 2.6.16.53-0.16-smp
- SLES10 SP2: 2.6.16.60-0.21-smp
- OpenSuSE 10.3: 2.6.22.5-31 *
- kernel.org: 2.6.26 and 2.6.27
* Minimal QA for these versions
Systems:
* x86_64
* x86
* ia64
* ppc64
Main Changes from OFED-1.4-rc4
==============================
- Updated MPI packages: mvapich-1.1.0-3141, mvapich2-1.2p1-1
- Updated bonding package: ib-bonding-0.9.0-34
- Updated qperf: qperf-0.4.2-1
- 8 bugs fixed (see attached for details)
- Attached kernel git tree changes:
Tasks that should be completed for the release:
================================
1. High priority bug fixes
2. Documentation update
commit 43fd69cda1fdb4451a7b0ffaf24354aba3326721
Author: Yevgeny Petrilin <[EMAIL PROTECTED]>
Date: Sun Nov 16 15:52:51 2008 +0200
mlx4_core: Ensure sense task has finished before removing device
Signed-off-by: Yevgeny Petrilin <[EMAIL PROTECTED]>
commit 09dc51b1934ffca6d7fab8fbe5eeae66e879f4bd
Author: Joachim Fenkes <[EMAIL PROTECTED]>
Date: Fri Nov 14 16:15:58 2008 +0100
IB/ehca: Remove suppress-DMEM code for 2.6.2[456]
Our "suppress DMEM" code works in these kernels only if
CONFIG_HOTPLUG_SPARSE_MEMORY is disabled, so patch it out again.
Signed-off-by: Joachim Fenkes <[EMAIL PROTECTED]>
commit c0257d6a2cee4b3e20838fad390a57f744935339
Author: Vladimir Sokolovsky <[EMAIL PROTECTED]>
Date: Sun Nov 16 10:14:15 2008 +0200
IPoIB/backport: This patch reverts 'd0de13622d5ac658efe7c51521dbdbe0752aa3dd'
commit for RHEL4 U[4-7]. Which cause IPoIB failure after ifconfig down
and then up on IPoIB interface.
https://bugs.openfabrics.org/show_bug.cgi?id=1370
Signed-off-by: Vladimir Sokolovsky <[EMAIL PROTECTED]>
commit eb4287e25ca7943a886cba26ab7fd0354aad73a9
Author: Dave Olson <[EMAIL PROTECTED]>
Date: Sat Nov 15 08:27:42 2008 -0800
IB/ipath: backport patch creating /dev/ipath device node.
RHEL4 U5 and U7 need that patch also.
Signed-off-by: Yannick Cote <[EMAIL PROTECTED]>
commit 602e3687b049671e70c57f51fb42648ef54a1234
Merge: 4eea91e... 3feae52...
Author: Vladimir Sokolovsky <[EMAIL PROTECTED]>
Date: Sat Nov 15 11:43:46 2008 +0200
Merge branch 'ofed_kernel' of git://git.openfabrics.org/~swise/ofed-1.4 into ofed_kernel
commit 3feae5270647e0c88f419f91d684eea807e61ad9
Author: Steve Wise <[EMAIL PROTECTED](none)>
Date: Thu Nov 13 18:37:55 2008 -0600
cxgb3: don't use multiple pages for the FL1 free list.
There are still bugs in the cxgb3 multi-page free list allocator, so
chelsio recommends turning off multi-page list altogether. This fixes
our RH4.x issues without introducing a page leak.
Signed-off-by: Steve Wise <[EMAIL PROTECTED]>
commit 4eea91e63b751ac34b946474dfbf41bcdf05a397
Author: Yevgeny Petrilin <[EMAIL PROTECTED]>
Date: Thu Nov 13 11:00:15 2008 +0200
mlx4_core: Sysfs entry for modifying both ports.
Added an option to notify the driver that the mode of all ports is
about to be modified. This way the driver would wait for all the new
port types to be set and only then will commit the change.
Since changing a port type is an operation that influences all ports,
this feature saves unnecessary stress.
Signed-off-by: Yevgeny Petrilin <[EMAIL PROTECTED]>
commit ef57d6ee60120d9267b67fda3626db3d5140afdf
Merge: 24e7496... 2d0ea1d...
Author: Vladimir Sokolovsky <[EMAIL PROTECTED]>
Date: Thu Nov 13 09:28:03 2008 +0200
Merge branch 'ofed_kernel' of ssh://openfabrics.org/home/ycote/scm/ofed_1_4/linux-2.6 into ofed_kernel
commit 2d0ea1d1c68b7fd60de682bc8d7ba49dc51f6565
Author: Dave Olson <[EMAIL PROTECTED]>
Date: Wed Nov 12 19:29:28 2008 -0800
IB/ipath: backport patch creating /dev/ipath device node.
Signed-off-by: Yannick Cote <[EMAIL PROTECTED]>
commit 24e749664fd913f5c58c3c6b2db5305fee00d058
Author: Chien Tung <[EMAIL PROTECTED]>
Date: Wed Nov 12 15:15:05 2008 -0800
From: Chien Tung <[EMAIL PROTECTED]>
RDMA/nes: Add loopback check to make_cm_node()
Check for loopback connection in make_cm_node()
Signed-off-by: Chien Tung <[EMAIL PROTECTED]>
commit 6f0c2b2c102d1e0ee8902934c33cc0a6269f476e
Author: Chien Tung <[EMAIL PROTECTED]>
Date: Wed Nov 12 15:14:23 2008 -0800
From: Chien Tung <[EMAIL PROTECTED]>
RDMA/nes: warning cleanup
Wrapped NES_DEBUG and assert macros with do while (0) to avoid ambiguous else.
Take out unused pointer returned from form_cm_frame().
drop_packet() should not increment reset counter on receiving a FIN.
Signed-off-by: Chien Tung <[EMAIL PROTECTED]>
commit e5b0a8db44136bb40827d16a604011d328769068
Author: Chien Tung <[EMAIL PROTECTED]>
Date: Wed Nov 12 15:13:39 2008 -0800
From: Faisal Latif <[EMAIL PROTECTED]>
RDMA/nes: Change accept_pend_cnt to atomic
There is a race condition on accept_pend_cnt. Change it to atomic.
Signed-off-by: Faisal Latif <[EMAIL PROTECTED]>
Signed-off-by: Chien Tung <[EMAIL PROTECTED]>
commit f53a13595203b55cd7912c906ed68c6370f3434d
Author: Chien Tung <[EMAIL PROTECTED]>
Date: Wed Nov 12 15:12:58 2008 -0800
From: Faisal Latif <[EMAIL PROTECTED]>
RDMA/nes: Check cqp_avail_reqs is empty after locking the list
Between the first empty list check and locking the list, the list
can change. Check it again after it is locked to make sure
the list is not empty.
Signed-off-by: Faisal Latif <[EMAIL PROTECTED]>
Signed-off-by: Chien Tung <[EMAIL PROTECTED]>
commit 5118290236d3b1ba9f0ba8c18be801881561d9f5
Author: Chien Tung <[EMAIL PROTECTED]>
Date: Wed Nov 12 15:01:03 2008 -0800
From: Faisal Latif <[EMAIL PROTECTED]>
RDMA/nes: Fix TCP compliance test failures
From ANVL testing, we are not handling all cm_node states during connection
establishment. Add missing state handlers.
Fixed sequence number
Send reset in handle_tcp_options()
Signed-off-by: Faisal Latif <[EMAIL PROTECTED]>
Signed-off-by: Chien Tung <[EMAIL PROTECTED]>
commit d6a73c4787593c23e5a17f3fb7ca19213a5d89a8
Author: Chien Tung <[EMAIL PROTECTED]>
Date: Wed Nov 12 14:59:30 2008 -0800
From: Faisal Latif <[EMAIL PROTECTED]>
RDMA/nes: Do not drop packets for a new connection with stale APBVT entry
Under heavy traffic, there is a small windows when an APBVT entry is not
yet removed and a new connection is established. Packets for the new
connection are dropped until APBVT entry is removed. This patch will
forward the packets instead of dropping them.
Signed-off-by: Faisal Latif <[EMAIL PROTECTED]>
Signed-off-by: Chien Tung <[EMAIL PROTECTED]>
commit ddb3671573b64bee8671558685ab597162c858eb
Author: Chien Tung <[EMAIL PROTECTED]>
Date: Wed Nov 12 14:57:07 2008 -0800
From: Faisal Latif <[EMAIL PROTECTED]>
RDMA/nes: Avoid race condition between MPA request and reset event to rdma_cm
In passive open after indicating MPA request to rdma_cm, an incoming RST
can cause rdma_cm to crash since the current state is not connected. The
solution is to wait for nes_accept() or nes_reject() before firing the
reset event. If nes_accept() or nes_reject() is already done, then the
reset event will fire when RST is processed.
Signed-off-by: Faisal Latif <[EMAIL PROTECTED]>
Signed-off-by: Chien Tung <[EMAIL PROTECTED]>
commit f0c90d3f43fd86c0bffd166acd9253ea3a394257
Author: Chien Tung <[EMAIL PROTECTED]>
Date: Wed Nov 12 14:56:02 2008 -0800
From: Faisal Latif <[EMAIL PROTECTED]>
RDMA/nes: Remove tx_free_list
There is no lock protecting tx_free_list thus causing a system crash
when skb_dequeue() is called and the list is empty. Since it did not give
any performance boost under heavy load, removing it to simplfy the code.
Change get_free_pkt call to allocate MAX_CM_BUFFER skb for connection
establishment/teardown as well as MPA request/response.
Signed-off-by: Faisal Latif <[EMAIL PROTECTED]>
Signed-off-by: Chien Tung <[EMAIL PROTECTED]>
commit 2ae853fbb241b99068614e8019ba3fec121064d9
Author: Chien Tung <[EMAIL PROTECTED]>
Date: Wed Nov 12 14:51:34 2008 -0800
From: Faisal Latif <[EMAIL PROTECTED]>
RDMA/nes: properly lock down connected_nodes list while processing it
While processing connected_nodes list, we would release the lock when
we need to send reset to remote partner. That created a window where
the list can be modified. Change this into a two step process. Place
nodes that need processing on a local list then process the local list.
Signed-off-by: Faisal Latif <[EMAIL PROTECTED]>
Signed-off-by: Chien Tung <[EMAIL PROTECTED]>
commit d04920ad8fc123102ae4694310869110a7b308fb
Author: Chien Tung <[EMAIL PROTECTED]>
Date: Wed Nov 12 14:44:25 2008 -0800
From: Faisal Latif <[EMAIL PROTECTED]>
RDMA/nes: cqp_request list cleanup
Use nes_free_cqp_request() from commit
1ff66e8c1faee7c2711b84b9c89e1c5fcd767839.
Change some continue to break in nes_cm_timer_tick. Send_entry
was a list processed in a loop, thus continue. Now it is a single
item, changing continue to break to be semantically correct.
Signed-off-by: Faisal Latif <[EMAIL PROTECTED]>
Signed-off-by: Chien Tung <[EMAIL PROTECTED]>
commit a05b368f89763a3274ba71c3bd6323b63e5f96e0
Author: Chien Tung <[EMAIL PROTECTED]>
Date: Wed Nov 12 14:39:46 2008 -0800
Revert "
"
This reverts commits
12fabeb7fab374d64d525a98d1780a0d53f22d9b
c96271276631172c2befab7746804c311991e51a
d7f8a8c86e089f58fc7a6461cac2ca46967927a2
91b02d4475322e568794547b89806a62c29f2c31
c910d0a08c69d84fc940538e80a64e3848b04e59
a46c5123d3e8ed4eb664aff3cd8ce30b9f48b6a2
3425f60e4103c1b604e6aa804d35e3389b3b76e9
b1532c55b6da7d6682ceb98905764b968babb112
3f1dd2345d9691cd54d414667836e99eddfe992c
fe8802ef870d22d623cc2c8e11a24211c06e0217
and will be replaced with patches in kernel_patches\fixes\nes_[0010-0100]*.patch
commit 2e6993b443fa5966061cc473c3079f93383cfb07
Author: Chien Tung <[EMAIL PROTECTED]>
Date: Wed Nov 12 10:05:19 2008 +0200
RDMA/CMA: Allocate PS_TCP ports from the host TCP port space.
This is needed for iwarp providers that support native and rdma
connections over the same interface.
Optionally turned on by the new rdma_cm unify_tcp_port_space module
option.
Signed-off-by: Steve Wise <[EMAIL PROTECTED]>
Signed-off-by: Faisal Latif <[EMAIL PROTECTED]>
Signed-off-by: Chien Tung <[EMAIL PROTECTED]>
commit 71d450105df5214c8c8ffd0952ae9506d33bdcde
Merge: f75ccb5... 02874f0...
Author: Vladimir Sokolovsky <[EMAIL PROTECTED]>
Date: Wed Nov 12 09:57:20 2008 +0200
Merge branch 'ofed_kernel' of ssh://openfabrics.org/home/ycote/scm/ofed_1_4/linux-2.6 into ofed_kernel
commit f75ccb5151f7b8e138fdda57704dbe05c53ddc03
Author: Steve Wise <[EMAIL PROTECTED]>
Date: Tue Nov 11 16:40:32 2008 -0600
RDMA/cxgb3: fix crash on rhel 4.x when doing 'rmmod cxgb3'
RHEL4.x kernels crash due to over-dereferencing.
Signed-off-by: Steve Wise <[EMAIL PROTECTED]>
commit eb56f4f7c4f6e53b4461021645037c4e862a3400
Author: Steve Wise <[EMAIL PROTECTED]>
Date: Tue Nov 11 16:40:30 2008 -0600
cxgb3 - fix eeprom content access.
From: Divy Le Ray <[EMAIL PROTECTED]>
- fix eeprom content access
Signed-off-by: Divy Le Ray <[EMAIL PROTECTED]>
Signed-off-by: Steve Wise <[EMAIL PROTECTED]>
commit 41480e9068c2ecd3a6c0422f6a88f31d82cf797c
Author: Steve Wise <[EMAIL PROTECTED]>
Date: Tue Nov 11 16:10:36 2008 -0600
RDMA/cxgb3: deadlock in iw_cxgb3 can cause hang when configuring interface.
When the iw_cxgb3 module's cxgb3_client "add" func gets called by the
cxgb3 module, the iwarp driver ends up calling the ethtool ops get_drvinfo
function in cxgb3 to get the fw version and other info. Currently the
iwarp driver grabs the rtnl lock around this down call to serialize.
As of 2.6.27 or so, things changed such that the rtnl lock is held around
the call to the netdev driver open function. Also the cxgb3_client "add"
function doesn't get called if the device is down.
So, if you load cxgb3, then load iw_cxgb3, then ifconfig up the device,
the iw_cxgb3 add func gets called with the rtnl_lock held. If you
load cxgb3, ifconfig up the device, then load iw_cxgb3, the add func
gets called without the rtnl_lock held. The former causes the deadlock,
the latter does not.
In addition, there are iw_cxgb3 sysfs handlers that also can call
down into cxgb3 to gather the fw and hw versions. These can be called
concurrently on different processors and at any time. Thus we need to
push this serialization down in the cxgb3 driver get_drvinfo func.
The fix is to remove rtnl lock usage, and use a per-device lock in cxgb3.
Signed-off-by: Steve Wise <[EMAIL PROTECTED]>
commit 02874f02dde51edcedfe14e345b0542449070f5e
Author: Yannick Cote <[EMAIL PROTECTED]>
Date: Tue Nov 11 09:36:10 2008 -0800
IB/ipath: backport patch creating /dev/ipath device node.
Signed-off-by: Yannick Cote <[EMAIL PROTECTED]>
"bug_id","bug_severity","priority","op_sys","assigned_to","bug_status","resolution","short_short_desc"
1346,"normal","P4","SLES 10","[EMAIL PROTECTED]","RESOLVED","FIXED","NFS-RDMA: nfs-utils source installation not being detected by OFED installation script"
1373,"major","P3","All","[EMAIL PROTECTED]","RESOLVED","FIXED","osmtest fails in multicast flow"
1323,"blocker","P1","All","[EMAIL PROTECTED]","CLOSED","FIXED","IB/ehca: possibillity of kernel panic under certain circumstances"
1364,"critical","P2","RHEL 4","[EMAIL PROTECTED]","RESOLVED","FIXED","system hang on rmmod cxgb3 in rhel4.7"
1366,"critical","P2","RHEL 4","[EMAIL PROTECTED]","RESOLVED","FIXED","Panic during boot-up after an OFED install in RHEL 4.5"
1183,"major","P1","SLES 10","[EMAIL PROTECTED]","RESOLVED","FIXED","higest priority doesn't become the MASTER opensm"
1370,"blocker","P1","All","[EMAIL PROTECTED]","RESOLVED","FIXED","Ping over IPoIB I/F fails after ifconfig down and up"
1390,"trivial","P5","RHEL 5","[EMAIL PROTECTED]","RESOLVED","FIXED","RPM errors when erasing native RedHat openmpi packages"
_______________________________________________
general mailing list
[email protected]
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/general
To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
