Re: [ewg] OFA EWG Meeting: Monday, July 30, 2018 - Minutes

[Jeff Becker]

Hi Arlin,

What is the schedule for OFED-4.17 release? Thanks.

-jeff

On 07/30/2018 02:03 PM, Jeff Becker wrote:

Hi Arlin,

On 07/30/2018 01:15 PM, Davis, Arlin R wrote:


_Attendees     Company _

Devesh Sharma    Broadcom

Steve Wise   Chelsio

Arlin Davis   Intel

Sindhu Devale Intel – i40iw

Vladimir Sokolovsky    Mellanox

Rupert Dance   SW Forge

Yuval Bason Cavium

Minutes:
·Opens - none

·OFED-4.17 daily build (7-27): 
https://downloads.openfabrics.org/downloads/OFED/ofed-4.17-daily/latest.tgz


oDistro/backports (RH7.4, RH7.5, SLES12.3, SLES15)
§RH 7.5, all done and working except qmulex, qib, and vmw_pvrdma.
·Devesh is looking into Emulex, report back next meeting
·Arlin will check with Bill Schmidt (Intel) regarding qib support
·*Huawei status? Wei Hu?*
·*VMware status? Adit?*
§SLES 15 rdma_core, mlx drivers done, ready for other IHV’s
§RH 7.4, rdma_core done, working on rest.
§SLES12.3 in process
§Have commits to finalize backports in next 2-4 weeks.
§We are still committed/tracking to RC release in September
oLibrary and ULP status
§iSER, NVMe-oF – Vlad will cover these backports
§NFS-RDMA – *Jeff, do you have cycles to support this in OFED 4.17*?


Depending on the release schedule, I should.

-jeff


Regards,
Arlin





___
ewg mailing list
ewg@lists.openfabrics.org
https://lists.openfabrics.org/mailman/listinfo/ewg



___
ewg mailing list
ewg@lists.openfabrics.org
https://lists.openfabrics.org/mailman/listinfo/ewg

Re: [ewg] [Bug 2636] Unable to build OFED-4.8-1-20170905-1125 compat-rdma on SL 7.2

[Jeff Becker]

Hi Vlad,

My limited test resources only permit me to implement NFS/RDMA backports 
for SLES12SP2, and RHEL7.3. Please disable NFS/RDMA for older kernels. 
Thanks.


-jeff

On 09/06/2017 07:02 AM, bugzi...@openfabrics.org wrote:

http://bugs.openfabrics.org/show_bug.cgi?id=2636

Vladimir Sokolovsky  changed:

What|Removed |Added

  CC||rsda...@soft-forge.com
   Component|IB Core |NFSRDMA
  AssignedTo|v...@mellanox.co.il |bec...@nas.nasa.gov



___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/mailman/listinfo/ewg

Re: [ewg] NFS/RDMA backport for OFED 4.8-1 part 3

[Jeff Becker]

On 09/05/2017 12:51 PM, David Z. Dai wrote:

Hi, Jeff:
It has the same error when loading "rpmcrdma" directly.
[root@rhel73_host1_guest1_local_htx modules]# modprobe rpcrdma
modprobe: ERROR: could not insert 'rpcrdma': Exec format error

I also change the "rpcrdma.ko" mode from 644 to 744. It still has the
same issue.


Since you're on IBM Power architecture I suspect this may be a factor - 
I only have access to x86_64 for development and testing. Perhaps 
someone at UNHIOL could verify? Thanks.


-jeff


Thanks! - David Dai

On Tue, 2017-09-05 at 12:33 -0700, Jeff Becker wrote:

Hi,

I responded to the bugzilla 2633. In short, please try "modprobe
rpcrdma" directly, and just once as it now contains both server and
client. Thanks.

-jeff

On 09/05/2017 12:25 PM, David Z. Dai wrote:

Hi, Vlad, Jeff:
Today I got the latest daily build from:
https://urldefense.proofpoint.com/v2/url?u=http-3A__downloads.openfabrics.org_OFED_ofed-2D4.8-2D1-2Ddaily_OFED-2D4.8-2D1-2D20170905-2D0757.tgz=DwICaQ=jf_iaSHvJObTbx-siA1ZOg=jE7tivF653Vygp552zF5nqagPKJAuy1x71Qnyf3szIw=g18gQ6E3E67ica6Ua-VcSfLqUhqoidoQ6p9JBZX16ZM=7P1W9rvFy-ZeD1z4hfqzovK40umK372kBX0EKXyeYfQ=
This latest daily build includes the NFSoRDMA functionality now.

I did a full install on RHEL 7.3 LE OS:
[root@rhel73_host1_guest1_local_htx OFED-4.8-1-20170905-0757]# uname -a
Linux rhel73_host1_guest1_local_htx 3.10.0-514.el7.ppc64le #1 SMP Wed Oct 19
11:27:06 EDT 2016 ppc64le ppc64le ppc64le GNU/Linux

[root@rhel73_host1_guest1_local_htx OFED-4.8-1-20170905-0757]# ./install.pl  
--all

The installation finished successfully.

[root@rhel73_host1_guest1_local_htx OFED-4.8-1-20170905-0757]# ofed_info -s
OFED-4.8-1-20170905-0757:

[root@rhel73_host1_guest1_local_htx OFED-4.8-1-20170905-0757]# modinfo xprtrdma
filename:  
/lib/modules/3.10.0-514.el7.ppc64le/updates/net/sunrpc/xprtrdma/rpcrdma.ko
alias:  xprtrdma
alias:  svcrdma
license:Dual BSD/GPL
description:RPC/RDMA Transport
author: Open Grid Computing and Network Appliance, Inc.
rhelversion:7.3
srcversion: 9629610592FC8363A00B592
depends:sunrpc,compat,ib_core,rdma_cm
vermagic:   3.10.0-514.el7.ppc64le SMP mod_unload modversions

[root@rhel73_host1_guest1_local_htx OFED-4.8-1-20170905-0757]# modinfo svcrdma
filename:  
/lib/modules/3.10.0-514.el7.ppc64le/updates/net/sunrpc/xprtrdma/rpcrdma.ko
alias:  xprtrdma
alias:  svcrdma
license:Dual BSD/GPL
description:RPC/RDMA Transport
author: Open Grid Computing and Network Appliance, Inc.
rhelversion:7.3
srcversion: 9629610592FC8363A00B592n
depends:sunrpc,compat,ib_core,rdma_cm
vermagic:   3.10.0-514.el7.ppc64le SMP mod_unload modversions

They are not in "nfsrdma" rpm any more. They are now in "compat-rdma" rpm:
[root@rhel73_host1_guest1_local_htx OFED-4.8-1-20170905-0757]# rpm -qf
/lib/modules/3.10.0-514.el7.ppc64le/updates/net/sunrpc/xprtrdma/rpcrdma.ko
compat-rdma-4.8-0.20170905.0757.g63258b0.3.10.0_514.el7.ppc64le.ppc64le

[root@rhel73_host1_guest1_local_htx OFED-4.8-1-20170905-0757]# rpm -qf
/lib/modules/3.10.0-514.el7.ppc64le/updates/net/sunrpc/xprtrdma/rpcrdma.ko
compat-rdma-4.8-0.20170905.0757.g63258b0.3.10.0_514.el7.ppc64le.ppc64le

[root@rhel73_host1_guest1_local_htx OFED-4.8-1-20170905-0757]# ls -al 
/lib/modules/3.10.0-514.el7.ppc64le/updates/net/sunrpc/xprtrdma/rpcrdma.ko
-rw-r--r--. 1 root root 6031120 Sep  5 14:24 
/lib/modules/3.10.0-514.el7.ppc64le/updates/net/sunrpc/xprtrdma/rpcrdma.ko
[root@rhel73_host1_guest1_local_htx OFED-4.8-1-20170905-0757]# ls -al 
/lib/modules/3.10.0-514.el7.ppc64le/updates/net/sunrpc/xprtrdma/rpcrdma.ko
-rw-r--r--. 1 root root 6031120 Sep  5 14:24 
/lib/modules/3.10.0-514.el7.ppc64le/updates/net/sunrpc/xprtrdma/rpcrdma.ko

However I see the error during module load time:
[root@rhel73_host1_guest1_local_htx OFED-4.8-1-20170905-0757]# modprobe xprtrdma
modprobe: ERROR: could not insert 'rpcrdma': Exec format error
[root@rhel73_host1_guest1_local_htx OFED-4.8-1-20170905-0757]# modprobe svcrdma
modprobe: ERROR: could not insert 'rpcrdma': Exec format error


NOTE:
1) Both xprtrdma and svcrdma are using same "rpcrdma.ko"  (see modinfo output 
above)
2) The mode of "rpcrdma.ko" is 644. Should it be at least 744?

Could you please check on this?

Thanks! - David
On Sat, 2017-09-02 at 00:59 +0300, Vladimir Sokolovsky wrote:

Applied + build: OFED-4.8-1-20170901-1439

Regards,
Vladimir

On 09/02/2017 12:44 AM, Jeff Becker wrote:

Hi Vlad,

Please apply the attached patches to compat-rdma.git. Together with
the previous two patches, these have been tested with RHEL7.3 and
SLES12SP2. Thanks.

-jeff


___
ewg mailing list
ewg@lists.openfabrics.org
https://urldefense.proofpoint.com/v2/url?u=http-3A__lists.openfabrics.org_mailman_listinfo_ewg=DwIGaQ=jf_iaSHvJObTbx-siA1ZOg=jE7tivF653Vygp552zF5nqagPKJAuy

Re: [ewg] NFS/RDMA backport for OFED 4.8-1 part 3

[Jeff Becker]

Hi,

I responded to the bugzilla 2633. In short, please try "modprobe 
rpcrdma" directly, and just once as it now contains both server and 
client. Thanks.


-jeff

On 09/05/2017 12:25 PM, David Z. Dai wrote:

Hi, Vlad, Jeff:
Today I got the latest daily build from:
http://downloads.openfabrics.org/OFED/ofed-4.8-1-daily/OFED-4.8-1-20170905-0757.tgz
This latest daily build includes the NFSoRDMA functionality now.

I did a full install on RHEL 7.3 LE OS:
[root@rhel73_host1_guest1_local_htx OFED-4.8-1-20170905-0757]# uname -a
Linux rhel73_host1_guest1_local_htx 3.10.0-514.el7.ppc64le #1 SMP Wed Oct 19
11:27:06 EDT 2016 ppc64le ppc64le ppc64le GNU/Linux

[root@rhel73_host1_guest1_local_htx OFED-4.8-1-20170905-0757]# ./install.pl  
--all

The installation finished successfully.

[root@rhel73_host1_guest1_local_htx OFED-4.8-1-20170905-0757]# ofed_info -s
OFED-4.8-1-20170905-0757:

[root@rhel73_host1_guest1_local_htx OFED-4.8-1-20170905-0757]# modinfo xprtrdma
filename:  
/lib/modules/3.10.0-514.el7.ppc64le/updates/net/sunrpc/xprtrdma/rpcrdma.ko
alias:  xprtrdma
alias:  svcrdma
license:Dual BSD/GPL
description:RPC/RDMA Transport
author: Open Grid Computing and Network Appliance, Inc.
rhelversion:7.3
srcversion: 9629610592FC8363A00B592
depends:sunrpc,compat,ib_core,rdma_cm
vermagic:   3.10.0-514.el7.ppc64le SMP mod_unload modversions

[root@rhel73_host1_guest1_local_htx OFED-4.8-1-20170905-0757]# modinfo svcrdma
filename:  
/lib/modules/3.10.0-514.el7.ppc64le/updates/net/sunrpc/xprtrdma/rpcrdma.ko
alias:  xprtrdma
alias:  svcrdma
license:Dual BSD/GPL
description:RPC/RDMA Transport
author: Open Grid Computing and Network Appliance, Inc.
rhelversion:7.3
srcversion: 9629610592FC8363A00B592n
depends:sunrpc,compat,ib_core,rdma_cm
vermagic:   3.10.0-514.el7.ppc64le SMP mod_unload modversions

They are not in "nfsrdma" rpm any more. They are now in "compat-rdma" rpm:
[root@rhel73_host1_guest1_local_htx OFED-4.8-1-20170905-0757]# rpm -qf
/lib/modules/3.10.0-514.el7.ppc64le/updates/net/sunrpc/xprtrdma/rpcrdma.ko
compat-rdma-4.8-0.20170905.0757.g63258b0.3.10.0_514.el7.ppc64le.ppc64le

[root@rhel73_host1_guest1_local_htx OFED-4.8-1-20170905-0757]# rpm -qf
/lib/modules/3.10.0-514.el7.ppc64le/updates/net/sunrpc/xprtrdma/rpcrdma.ko
compat-rdma-4.8-0.20170905.0757.g63258b0.3.10.0_514.el7.ppc64le.ppc64le

[root@rhel73_host1_guest1_local_htx OFED-4.8-1-20170905-0757]# ls -al 
/lib/modules/3.10.0-514.el7.ppc64le/updates/net/sunrpc/xprtrdma/rpcrdma.ko
-rw-r--r--. 1 root root 6031120 Sep  5 14:24 
/lib/modules/3.10.0-514.el7.ppc64le/updates/net/sunrpc/xprtrdma/rpcrdma.ko
[root@rhel73_host1_guest1_local_htx OFED-4.8-1-20170905-0757]# ls -al 
/lib/modules/3.10.0-514.el7.ppc64le/updates/net/sunrpc/xprtrdma/rpcrdma.ko
-rw-r--r--. 1 root root 6031120 Sep  5 14:24 
/lib/modules/3.10.0-514.el7.ppc64le/updates/net/sunrpc/xprtrdma/rpcrdma.ko

However I see the error during module load time:
[root@rhel73_host1_guest1_local_htx OFED-4.8-1-20170905-0757]# modprobe xprtrdma
modprobe: ERROR: could not insert 'rpcrdma': Exec format error
[root@rhel73_host1_guest1_local_htx OFED-4.8-1-20170905-0757]# modprobe svcrdma
modprobe: ERROR: could not insert 'rpcrdma': Exec format error


NOTE:
1) Both xprtrdma and svcrdma are using same "rpcrdma.ko"  (see modinfo output 
above)
2) The mode of "rpcrdma.ko" is 644. Should it be at least 744?

Could you please check on this?

Thanks! - David
On Sat, 2017-09-02 at 00:59 +0300, Vladimir Sokolovsky wrote:

Applied + build: OFED-4.8-1-20170901-1439

Regards,
Vladimir

On 09/02/2017 12:44 AM, Jeff Becker wrote:

Hi Vlad,

Please apply the attached patches to compat-rdma.git. Together with
the previous two patches, these have been tested with RHEL7.3 and
SLES12SP2. Thanks.

-jeff


___
ewg mailing list
ewg@lists.openfabrics.org
https://urldefense.proofpoint.com/v2/url?u=http-3A__lists.openfabrics.org_mailman_listinfo_ewg=DwIGaQ=jf_iaSHvJObTbx-siA1ZOg=jE7tivF653Vygp552zF5nqagPKJAuy1x71Qnyf3szIw=antYpg37AAv4LKbRqOFw77idqtt-Kxw8KtFSKz_3UJE=RNm5Hv7MRR_y-c_JC9pR1ZHCbLAkAtQh5HJCBEOEfj0=




___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/mailman/listinfo/ewg

[ewg] NFS/RDMA backport for OFED 4.8-1 part 3

[Jeff Becker]

Hi Vlad,

Please apply the attached patches to compat-rdma.git. Together with the 
previous two patches, these have been tested with RHEL7.3 and SLES12SP2. 
Thanks.


-jeff

>From dcb076f7acbe1f2d34668d864a22dc780ae8a314 Mon Sep 17 00:00:00 2001
From: Jeff Becker <jeffrey.c.bec...@nasa.gov>
Date: Fri, 1 Sep 2017 13:25:35 -0700
Subject: [PATCH 2/2] NFS/RDMA backport patch for SLES12SP2

Signed-off-by: Jeff Becker <jeffrey.c.bec...@nasa.gov>
---
 patches/0017-BACKPORT-nfsrdma-sles12sp2.patch | 82 +++
 1 file changed, 82 insertions(+)
 create mode 100644 patches/0017-BACKPORT-nfsrdma-sles12sp2.patch

diff --git a/patches/0017-BACKPORT-nfsrdma-sles12sp2.patch b/patches/0017-BACKPORT-nfsrdma-sles12sp2.patch
new file mode 100644
index 000..ebc453f
--- /dev/null
+++ b/patches/0017-BACKPORT-nfsrdma-sles12sp2.patch
@@ -0,0 +1,82 @@
+diff --git a/net/sunrpc/xprtrdma/xprt_rdma.h b/net/sunrpc/xprtrdma/xprt_rdma.h
+index a71b0f5..4fd3708 100644
+--- a/net/sunrpc/xprtrdma/xprt_rdma.h
 b/net/sunrpc/xprtrdma/xprt_rdma.h
+@@ -52,6 +52,10 @@
+ #include  	/* RPC/RDMA protocol */
+ #include  	/* xprt parameters */
+ 
++#ifndef HAVE_XPRT_PUT_EXPORTED
++#include 
++#endif
++
+ #define RDMA_RESOLVE_TIMEOUT	(5000)	/* 5 seconds */
+ #define RDMA_CONNECT_RETRY_MAX	(2)	/* retries if no listener backlog */
+ 
+@@ -541,4 +541,30 @@ void xprt_rdma_bc_destroy(struct rpc_xprt *, unsigned int);
+ 
+ extern struct xprt_class xprt_rdma_bc;
+ 
++#ifndef HAVE_XPRT_PUT_EXPORTED
++static inline void backport_xprt_put(struct rpc_xprt *xprt)
++{
++	if (atomic_dec_and_test(>count)) {
++		printk("RPC:   destroying transport %p\n", xprt);
++
++		/* Exclude transport connect/disconnect handlers */
++		wait_on_bit_lock(>state, XPRT_LOCKED, TASK_UNINTERRUPTIBLE);
++
++		del_timer_sync(>timer);
++
++		debugfs_remove_recursive(xprt->debugfs);
++		xprt->debugfs = NULL;
++		rpc_destroy_wait_queue(>binding);
++		rpc_destroy_wait_queue(>pending);
++		rpc_destroy_wait_queue(>sending);
++		rpc_destroy_wait_queue(>backlog);
++		cancel_work_sync(>task_cleanup);
++		kfree(xprt->servername);
++		/*
++ 	 	 * Tear down transport state and free the rpc_xprt
++ 	 	 */
++		xprt->ops->destroy(xprt);
++	}
++}
++#endif
+ #endif/* _LINUX_SUNRPC_XPRT_RDMA_H */
+diff --git a/net/sunrpc/xprtrdma/svc_rdma_backchannel.c b/net/sunrpc/xprtrdma/svc_rdma_backchannel.c
+index a2a7519..1d2703c 100644
+--- a/net/sunrpc/xprtrdma/svc_rdma_backchannel.c
 b/net/sunrpc/xprtrdma/svc_rdma_backchannel.c
+@@ -350,7 +350,11 @@ xprt_setup_rdma_bc(struct xprt_create *args)
+ out_fail:
+ 	xprt_rdma_free_addresses(xprt);
+ 	args->bc_xprt->xpt_bc_xprt = NULL;
++#ifdef HAVE_XPRT_PUT_EXPORTED
+ 	xprt_put(xprt);
++#else
++	backport_xprt_put(xprt);
++#endif
+ 	xprt_free(xprt);
+ 	return ERR_PTR(-EINVAL);
+ }
+diff --git a/net/sunrpc/xprtrdma/svc_rdma_transport.c b/net/sunrpc/xprtrdma/svc_rdma_transport.c
+index dd94401..a9f812f 100644
+--- a/net/sunrpc/xprtrdma/svc_rdma_transport.c
 b/net/sunrpc/xprtrdma/svc_rdma_transport.c
+@@ -1223,7 +1223,11 @@ static void __svc_rdma_free(struct work_struct *work)
+ 
+ 	/* Final put of backchannel client transport */
+ 	if (xprt->xpt_bc_xprt) {
++#ifdef HAVE_XPRT_PUT_EXPORTED
+ 		xprt_put(xprt->xpt_bc_xprt);
++#else
++		backport_xprt_put(xprt->xpt_bc_xprt);
++#endif
+ 		xprt->xpt_bc_xprt = NULL;
+ 	}
+ 
+diff --git a/net/sunrpc/xprtrdma/xprt_rdma.h b/net/sunrpc/xprtrdma/xprt_rdma.h
+index a71b0f5..607c94c 100644
+--- a/net/sunrpc/xprtrdma/xprt_rdma.h
 b/net/sunrpc/xprtrdma/xprt_rdma.h
-- 
1.8.3.1

>From f33a8f4bfc55b57b45518ff00165f8d49fe58fc5 Mon Sep 17 00:00:00 2001
From: Jeff Becker <jeffrey.c.bec...@nasa.gov>
Date: Thu, 24 Aug 2017 10:06:10 -0700
Subject: [PATCH 1/2] NFS/RDMA backport patch to revert source files to 4.6
 kernel in order to facilitate dependency on distro SUNRPC. Include fix to use
 correct ib_map_mr_sg signature from OFED4.8.

Signed-off-by: Jeff Becker <jeffrey.c.bec...@nasa.gov>
---
 ofed_scripts/checkout_files|1 +
 patches/0016-BACKPORT-nfsrdma-to-4.6.patch | 3682 
 2 files changed, 3683 insertions(+)
 create mode 100644 patches/0016-BACKPORT-nfsrdma-to-4.6.patch

diff --git a/ofed_scripts/checkout_files b/ofed_scripts/checkout_files
index 21bc220..f1b6a76 100644
--- a/ofed_scripts/checkout_files
+++ b/ofed_scripts/checkout_files
@@ -25,3 +25,4 @@ drivers/nvme/
 include/linux/nvme.h
 include/linux/nvme-rdma.h
 include/uapi/linux/nvme_ioctl.h
+include/linux/sunrpc/
diff --git a/patches/0016-BACKPORT-nfsrdma-to-4.6.patch b/patches/0016-BACKPORT-nfsrdma-to-4.6.patch
new file mode 100644
index 000..e0846b2
--- /dev/null
+++ b/patches/0016-BACKPORT-nfsrdma-to-4.6.patch
@@ -0,0 +1,3682 @@
+diff --git a/net/sunrpc/xprtrdma/Makefile b/net/sunrpc/xprtrdma/Makefile
+index ef19fa4..dc9f3b5 100644
+--- a/net/sunrpc/xprtrdma/Makefile
 b/net/sunrp

[ewg] NFS/RDMA backport for OFED 4.8-1 part 2

[Jeff Becker]

Hi Vlad,

Please apply the attached patch to compat.git. Thanks.

-jeff

>From 96cfe9b9d341df91618e96549c0db2aa06c8f1f7 Mon Sep 17 00:00:00 2001
From: Jeff Becker <jeffrey.c.bec...@nasa.gov>
Date: Fri, 1 Sep 2017 13:35:23 -0700
Subject: [PATCH] NFS/RDMA backport: check if kernel exports xprt_put symbol.

Signed-off-by: Jeff Becker <jeffrey.c.bec...@nasa.gov>
---
 config/rdma.m4 | 6 ++
 1 file changed, 6 insertions(+)

diff --git a/config/rdma.m4 b/config/rdma.m4
index 0f13391..aaddbef 100644
--- a/config/rdma.m4
+++ b/config/rdma.m4
@@ -3499,6 +3499,12 @@ AC_DEFUN([LINUX_CONFIG_COMPAT],
 			[pat_enabled is exported by the kernel])],
 	[])
 
+	LB_CHECK_SYMBOL_EXPORT([xprt_put],
+		[net/sunrpc/xprt.c],
+		[AC_DEFINE(HAVE_XPRT_PUT_EXPORTED, 1,
+			[xprt_put is exported by the kernel])],
+	[])
+
 	AC_MSG_CHECKING([if configfs.h default_groups is list_head])
 	LB_LINUX_TRY_COMPILE([
 		#include 
-- 
1.8.3.1

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/mailman/listinfo/ewg

[ewg] NFS/RDMA backport for OFED 4.8-1 part 1

[Jeff Becker]

Hi Vlad,

Please apply the attached patch to linux-4.8.git. Thanks.

-jeff

>From 5568001640115eecfdc5925541bc86aa0c69810f Mon Sep 17 00:00:00 2001
From: Jeff Becker <jeffrey.c.bec...@nasa.gov>
Date: Thu, 24 Aug 2017 10:17:06 -0700
Subject: [PATCH] Add missing NFS/RDMA header files

Signed-off-by: Jeff Becker <jeffrey.c.bec...@nasa.gov>
---
 include/linux/sunrpc/rpc_rdma.h | 132 
 include/linux/sunrpc/svc_rdma.h | 267 
 2 files changed, 399 insertions(+)
 create mode 100644 include/linux/sunrpc/rpc_rdma.h
 create mode 100644 include/linux/sunrpc/svc_rdma.h

diff --git a/include/linux/sunrpc/rpc_rdma.h b/include/linux/sunrpc/rpc_rdma.h
new file mode 100644
index 000..3b1ff38
--- /dev/null
+++ b/include/linux/sunrpc/rpc_rdma.h
@@ -0,0 +1,132 @@
+/*
+ * Copyright (c) 2003-2007 Network Appliance, Inc. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the BSD-type
+ * license below:
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions
+ * are met:
+ *
+ *  Redistributions of source code must retain the above copyright
+ *  notice, this list of conditions and the following disclaimer.
+ *
+ *  Redistributions in binary form must reproduce the above
+ *  copyright notice, this list of conditions and the following
+ *  disclaimer in the documentation and/or other materials provided
+ *  with the distribution.
+ *
+ *  Neither the name of the Network Appliance, Inc. nor the names of
+ *  its contributors may be used to endorse or promote products
+ *  derived from this software without specific prior written
+ *  permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#ifndef _LINUX_SUNRPC_RPC_RDMA_H
+#define _LINUX_SUNRPC_RPC_RDMA_H
+
+#include 
+
+#define RPCRDMA_VERSION		1
+#define rpcrdma_version		cpu_to_be32(RPCRDMA_VERSION)
+
+struct rpcrdma_segment {
+	__be32 rs_handle;	/* Registered memory handle */
+	__be32 rs_length;	/* Length of the chunk in bytes */
+	__be64 rs_offset;	/* Chunk virtual address or offset */
+};
+
+/*
+ * read chunk(s), encoded as a linked list.
+ */
+struct rpcrdma_read_chunk {
+	__be32 rc_discrim;	/* 1 indicates presence */
+	__be32 rc_position;	/* Position in XDR stream */
+	struct rpcrdma_segment rc_target;
+};
+
+/*
+ * write chunk, and reply chunk.
+ */
+struct rpcrdma_write_chunk {
+	struct rpcrdma_segment wc_target;
+};
+
+/*
+ * write chunk(s), encoded as a counted array.
+ */
+struct rpcrdma_write_array {
+	__be32 wc_discrim;	/* 1 indicates presence */
+	__be32 wc_nchunks;	/* Array count */
+	struct rpcrdma_write_chunk wc_array[0];
+};
+
+struct rpcrdma_msg {
+	__be32 rm_xid;	/* Mirrors the RPC header xid */
+	__be32 rm_vers;	/* Version of this protocol */
+	__be32 rm_credit;	/* Buffers requested/granted */
+	__be32 rm_type;	/* Type of message (enum rpcrdma_proc) */
+	union {
+
+		struct {			/* no chunks */
+			__be32 rm_empty[3];	/* 3 empty chunk lists */
+		} rm_nochunks;
+
+		struct {			/* no chunks and padded */
+			__be32 rm_align;	/* Padding alignment */
+			__be32 rm_thresh;	/* Padding threshold */
+			__be32 rm_pempty[3];	/* 3 empty chunk lists */
+		} rm_padded;
+
+		struct {
+			__be32 rm_err;
+			__be32 rm_vers_low;
+			__be32 rm_vers_high;
+		} rm_error;
+
+		__be32 rm_chunks[0];	/* read, write and reply chunks */
+
+	} rm_body;
+};
+
+/*
+ * Smallest RPC/RDMA header: rm_xid through rm_type, then rm_nochunks
+ */
+#define RPCRDMA_HDRLEN_MIN	(sizeof(__be32) * 7)
+#define RPCRDMA_HDRLEN_ERR	(sizeof(__be32) * 5)
+
+enum rpcrdma_errcode {
+	ERR_VERS = 1,
+	ERR_CHUNK = 2
+};
+
+enum rpcrdma_proc {
+	RDMA_MSG = 0,		/* An RPC call or reply msg */
+	RDMA_NOMSG = 1,		/* An RPC call or reply msg - separate body */
+	RDMA_MSGP = 2,		/* An RPC call or reply msg with padding */
+	RDMA_DONE = 3,		/* Client signals reply completion */
+	RDMA_ERROR = 4		/* An RPC RDMA encodin

Re: [ewg] SLES12SP2 build error

[Jeff Becker]

On 08/25/2017 02:23 PM, Jeff Becker wrote:

On 08/25/2017 02:15 PM, Vladimir Sokolovsky wrote:

On 08/25/2017 11:56 PM, Jeff Becker wrote:

Hi Vlad,

OFED4.8-1 (including NFS/RDMA) builds fine on RHEL7.3. However, on 
SLES12SP2, I get the error below. Note that the initial warning 
about the extra CONFIG_QED token happens on RHEl7.3 as well.


Thanks for your help.

-jeff

# ./configure --with-core-mod --with-user_mad-mod 
--with-user_access-mod --with-addr_trans-mod --with-mlx4-mod 
--with-ipoib-mod --with-nfsrdma-mod

# make



Hi Jeff,
I removed the extra CONFIG_QED yesterday. Please rebase your git branch.


This morning I did:

% cd compat-rdma
% git checkout master
% git pull (got your new configure)
% git checkout nfsrdma
% git merge master

Does this not work?

To avoid compilation issue with Broadcom drivers add to your 
configure command: --without-bnxt_re-mod --without-bnxt_en-mod

They are still working on backports...


I tried adding those at the end of my configure command line, and I 
still get the error. Do they need to be in a certain order, e.g., before 
with-mlx4-mod --with-ipoib-mod


-jeff


OK. Thanks.

-jeff


Regards,
Vladimir


...

  CC [M] 
/root/jcbecker/OFED4.8-1/compat-rdma/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.o
In file included from 
/lib/modules/4.4.59-92.24-default/source/include/linux/kconfig.h:4:0,

 from :0:
/root/jcbecker/OFED4.8-1/compat-rdma/include/generated/autoconf.h:167:19: 
warning: extra tokens at end of #undef directive [enabled by default]

 #undef CONFIG_QED CONFIG_QED
   ^
/root/jcbecker/OFED4.8-1/compat-rdma/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c: 
In function ‘bnxt_grxclsrule’:
/root/jcbecker/OFED4.8-1/compat-rdma/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c:550:18: 
error: ‘struct flow_keys’ has no member named ‘ip_proto’

 if (fkeys->ip_proto == IPPROTO_TCP)
  ^
/root/jcbecker/OFED4.8-1/compat-rdma/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c:552:23: 
error: ‘struct flow_keys’ has no member named ‘ip_proto’

 else if (fkeys->ip_proto == IPPROTO_UDP)
   ^
/root/jcbecker/OFED4.8-1/compat-rdma/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c:557:44: 
error: ‘struct flow_keys’ has no member named ‘src’

 fs->h_u.tcp_ip4_spec.ip4src = fkeys->src;
^
/root/jcbecker/OFED4.8-1/compat-rdma/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c:560:44: 
error: ‘struct flow_keys’ has no member named ‘dst’

 fs->h_u.tcp_ip4_spec.ip4dst = fkeys->dst;
^
/root/jcbecker/OFED4.8-1/compat-rdma/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c:563:42: 
error: ‘struct flow_keys’ has no member named ‘port16’

 fs->h_u.tcp_ip4_spec.psrc = fkeys->port16[0];
  ^
/root/jcbecker/OFED4.8-1/compat-rdma/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c:566:42: 
error: ‘struct flow_keys’ has no member named ‘port16’

 fs->h_u.tcp_ip4_spec.pdst = fkeys->port16[1];
  ^
/usr/src/linux-4.4.59-92.24/scripts/Makefile.build:259: recipe for 
target 
'/root/jcbecker/OFED4.8-1/compat-rdma/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.o' 
failed








___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/mailman/listinfo/ewg

Re: [ewg] SLES12SP2 build error

[Jeff Becker]

On 08/25/2017 02:15 PM, Vladimir Sokolovsky wrote:

On 08/25/2017 11:56 PM, Jeff Becker wrote:

Hi Vlad,

OFED4.8-1 (including NFS/RDMA) builds fine on RHEL7.3. However, on 
SLES12SP2, I get the error below. Note that the initial warning about 
the extra CONFIG_QED token happens on RHEl7.3 as well.


Thanks for your help.

-jeff

# ./configure --with-core-mod --with-user_mad-mod 
--with-user_access-mod --with-addr_trans-mod --with-mlx4-mod 
--with-ipoib-mod --with-nfsrdma-mod

# make



Hi Jeff,
I removed the extra CONFIG_QED yesterday. Please rebase your git branch.


This morning I did:

% cd compat-rdma
% git checkout master
% git pull (got your new configure)
% git checkout nfsrdma
% git merge master

Does this not work?

To avoid compilation issue with Broadcom drivers add to your configure 
command: --without-bnxt_re-mod --without-bnxt_en-mod

They are still working on backports...


OK. Thanks.

-jeff


Regards,
Vladimir


...

  CC [M] 
/root/jcbecker/OFED4.8-1/compat-rdma/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.o
In file included from 
/lib/modules/4.4.59-92.24-default/source/include/linux/kconfig.h:4:0,

 from :0:
/root/jcbecker/OFED4.8-1/compat-rdma/include/generated/autoconf.h:167:19: 
warning: extra tokens at end of #undef directive [enabled by default]

 #undef CONFIG_QED CONFIG_QED
   ^
/root/jcbecker/OFED4.8-1/compat-rdma/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c: 
In function ‘bnxt_grxclsrule’:
/root/jcbecker/OFED4.8-1/compat-rdma/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c:550:18: 
error: ‘struct flow_keys’ has no member named ‘ip_proto’

 if (fkeys->ip_proto == IPPROTO_TCP)
  ^
/root/jcbecker/OFED4.8-1/compat-rdma/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c:552:23: 
error: ‘struct flow_keys’ has no member named ‘ip_proto’

 else if (fkeys->ip_proto == IPPROTO_UDP)
   ^
/root/jcbecker/OFED4.8-1/compat-rdma/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c:557:44: 
error: ‘struct flow_keys’ has no member named ‘src’

 fs->h_u.tcp_ip4_spec.ip4src = fkeys->src;
^
/root/jcbecker/OFED4.8-1/compat-rdma/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c:560:44: 
error: ‘struct flow_keys’ has no member named ‘dst’

 fs->h_u.tcp_ip4_spec.ip4dst = fkeys->dst;
^
/root/jcbecker/OFED4.8-1/compat-rdma/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c:563:42: 
error: ‘struct flow_keys’ has no member named ‘port16’

 fs->h_u.tcp_ip4_spec.psrc = fkeys->port16[0];
  ^
/root/jcbecker/OFED4.8-1/compat-rdma/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c:566:42: 
error: ‘struct flow_keys’ has no member named ‘port16’

 fs->h_u.tcp_ip4_spec.pdst = fkeys->port16[1];
  ^
/usr/src/linux-4.4.59-92.24/scripts/Makefile.build:259: recipe for 
target 
'/root/jcbecker/OFED4.8-1/compat-rdma/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.o' 
failed






___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/mailman/listinfo/ewg

[ewg] SLES12SP2 build error

[Jeff Becker]

Hi Vlad,

OFED4.8-1 (including NFS/RDMA) builds fine on RHEL7.3. However, on 
SLES12SP2, I get the error below. Note that the initial warning about 
the extra CONFIG_QED token happens on RHEl7.3 as well.


Thanks for your help.

-jeff

# ./configure --with-core-mod --with-user_mad-mod --with-user_access-mod 
--with-addr_trans-mod --with-mlx4-mod --with-ipoib-mod --with-nfsrdma-mod

# make

...

  CC [M] 
/root/jcbecker/OFED4.8-1/compat-rdma/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.o
In file included from 
/lib/modules/4.4.59-92.24-default/source/include/linux/kconfig.h:4:0,

 from :0:
/root/jcbecker/OFED4.8-1/compat-rdma/include/generated/autoconf.h:167:19: 
warning: extra tokens at end of #undef directive [enabled by default]

 #undef CONFIG_QED CONFIG_QED
   ^
/root/jcbecker/OFED4.8-1/compat-rdma/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c: 
In function ‘bnxt_grxclsrule’:
/root/jcbecker/OFED4.8-1/compat-rdma/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c:550:18: 
error: ‘struct flow_keys’ has no member named ‘ip_proto’

 if (fkeys->ip_proto == IPPROTO_TCP)
  ^
/root/jcbecker/OFED4.8-1/compat-rdma/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c:552:23: 
error: ‘struct flow_keys’ has no member named ‘ip_proto’

 else if (fkeys->ip_proto == IPPROTO_UDP)
   ^
/root/jcbecker/OFED4.8-1/compat-rdma/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c:557:44: 
error: ‘struct flow_keys’ has no member named ‘src’

 fs->h_u.tcp_ip4_spec.ip4src = fkeys->src;
^
/root/jcbecker/OFED4.8-1/compat-rdma/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c:560:44: 
error: ‘struct flow_keys’ has no member named ‘dst’

 fs->h_u.tcp_ip4_spec.ip4dst = fkeys->dst;
^
/root/jcbecker/OFED4.8-1/compat-rdma/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c:563:42: 
error: ‘struct flow_keys’ has no member named ‘port16’

 fs->h_u.tcp_ip4_spec.psrc = fkeys->port16[0];
  ^
/root/jcbecker/OFED4.8-1/compat-rdma/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c:566:42: 
error: ‘struct flow_keys’ has no member named ‘port16’

 fs->h_u.tcp_ip4_spec.pdst = fkeys->port16[1];
  ^
/usr/src/linux-4.4.59-92.24/scripts/Makefile.build:259: recipe for 
target 
'/root/jcbecker/OFED4.8-1/compat-rdma/drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.o' 
failed


___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/mailman/listinfo/ewg

Re: [ewg] OFED4.8 build error

[Jeff Becker]

On 08/21/2017 07:58 AM, Vladimir Sokolovsky wrote:

On 08/18/2017 12:14 AM, Jeff Becker wrote:
Hi. I was syncing up my git trees with the latest on the OFED server, 
and got an unexpected build error. Thus I tried freshly cloning trees 
to try the build again (without NFS/RDMA). I get the same error (see 
below). This is on RHEL7.3. Am I missing a step or did something 
change? Thanks for your help.


-jeff
...
Type 'make' to build kernel modules.
 1085$ make
Building kernel modules
Kernel version: 3.10.0-693.el7.x86_64


Hi Jeff,
This kernel belongs to RHEL7.4 which is not supported by OFED-4.8-1.


Yes, of course. I forgot that my desktop was upgraded to RHEL7.4 in the 
patching cycle the night before I did this build. Fortunately, I still 
have a RHEL7.3 test server. I'll redo the build there. Thanks., and 
sorry for the noise.


-jeff


Support for RHEL7.4 and SLES12 SP3 is planned for OFED-4.8-2.
So, please use the original kernel coming with RHEL7.3.

Regards,
Vladimir



___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/mailman/listinfo/ewg

Re: [ewg] EWG Meeting Minutes - 8-14-2017

[Jeff Becker]

Slight correction below...

On 08/14/2017 11:01 AM, Woodruff, Robert J wrote:

Agenda:

  * Opens
  * OFED 4.8-1 status:
  * User Package Update
  * Bug Report – Critical/blocking

  * 2631 – An SRP Reject from an SRP target causes SL7.3/OFED crash

  * Fails on OFED 4.8 out of the box, requires OFED+ib_srp_backports

  * 2632- SRP login failure for SL7.2 and OFED 4.8-rc4 (ib_srp:Sending
CM DREQ failed)

  * Fails on SL7.2, Works on SL7.3 and OFED-4.8

  * 2633 – NFS over RDMA kernel modules are not supported in OFED 4.8

Attendees:
Susan Coulter
Woody Woodruff
Rupert Dance
David ??? from IBM
Jeff Becker
Steve Wise
Jim Ryan
Rajalaxmi  Angadi
Notes:
OFED-4.8-1: There is a first daily build posted on 8-7-2017, 
_OFED-4.8-1-20170807-0808.tgz_ 
<http://downloads.openfabrics.org/OFED/ofed-4.8-1-daily/OFED-4.8-1-20170807-0808.tgz> 
that is available for early testing.
It contains the new rdma-core v14. Not sure if any of the backports 
for the Cavium or Broadcom drivers made it in yet as
there were no representatives present from Cavium or Broadcom. Jeff 
Becker said he is still working on the backports for NFS/RDMA.
He has it working on RHEL 7.3, but still fixing a bug he found on SLES 
12 SP 2 and will submit patches for those once he completes his debug

and updates for SLES 12 SP 2.
I've only tested module loading so far, so "working on RHEL 7.3" should 
really be: NFS/RDMA module loads correctly on RHEL7.3
No updates on if DDN is still working on any additional fixes for SRP. 
Rupert will ask Marty.
Open Item: RHEL 6.x support. There was a discussion on the ewg list if 
it would be possible to do a follow on to 3.18-2 to include
bug fixes, since OFED-4.8-x does not support that older RHEL version. 
Need more discussion on this, since Vlad could not attend this week
and is a key person needed to make a decision either way. We will 
defer this discussion/decision to the next meeting.



___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/mailman/listinfo/ewg



___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/mailman/listinfo/ewg

Re: [ewg] NFS/RDMA question

[Jeff Becker]

Hi Rupert,

On 08/12/2017 11:48 AM, RSD@SFI wrote:

Hi Check,

You are correct that OFED 4.x does not support any 6.x RHEL based distros
(see attached readme).

I will bring up this topic in tomorrow's EWG meeting.


Do you know the 800 number for this meeting? Thanks.

-jeff


Thanks

Rupert

-Original Message-
From: Chuck Lever [mailto:chuck.le...@oracle.com]
Sent: Friday, August 04, 2017 11:29 AM
To: RSD@SFI <rsda...@soft-forge.com>
Cc: Jeff Becker <jeffrey.c.bec...@nasa.gov>; ewg@lists.openfabrics.org
Subject: Re: [ewg] NFS/RDMA question

Thanks Rupert.

I'm thinking out loud here, but my interest is with the many many
deployments that are forced to remain on EL6-based distributions
(Scientific, CentOS) due to application certification constraints.
This includes many in the HPC community, and those who are doing chip CAD,
for example.

OFED / MOFED seems to be the only way to deliver working NFS/RDMA in that
class of environment, as it replaces the NFS/RDMA code that is in Red Hat's
EL6 kernel with newer 3.18 code.

As I understand it, OFED 4 is not supported on EL6-based systems.



On Aug 3, 2017, at 5:57 PM, RSD@SFI <rsda...@soft-forge.com> wrote:

OFED Bugzilla is located here: http://bugs.openfabrics.org/

We are now working on OFED 4.8-1 and hope to quickly move to OFED 4.14 so

I think it is unlikely that the EWG would spin a new release for OFED 3.18
but we would be glad to know of the issue so we can resolve it if it still
exist in OFED 4.8-1.

Rupert

-Original Message-
From: ewg [mailto:ewg-boun...@lists.openfabrics.org] On Behalf Of Jeff
Becker
Sent: Thursday, August 03, 2017 5:44 PM
To: Chuck Lever <chuck.le...@oracle.com>; ewg@lists.openfabrics.org
Subject: Re: [ewg] NFS/RDMA question

I think, e.g., 3.18.3 is still supported - forwarding to EWG to make sure.

On 08/03/2017 01:15 PM, Chuck Lever wrote:

I had one more question about OFED 3.x. Is it still "supported" --
meaning, if an admin applies OFED 3 to, say, CentOS 6, and encounters
a bug, who does she contact for help?


--
Chuck Lever




___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/mailman/listinfo/ewg


--
Chuck Lever






___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/mailman/listinfo/ewg

Re: [ewg] NFS/RDMA question

[Jeff Becker]

I think, e.g., 3.18.3 is still supported - forwarding to EWG to make sure.

On 08/03/2017 01:15 PM, Chuck Lever wrote:

I had one more question about OFED 3.x. Is it still "supported" --
meaning, if an admin applies OFED 3 to, say, CentOS 6, and
encounters a bug, who does she contact for help?


--
Chuck Lever





___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/mailman/listinfo/ewg

Re: [ewg] OFA EWG Meeting: Monday, July 31, 2017, 09:00 AM US Pacific Time (12pm EST) - Agenda

[Jeff Becker]

Hi,

I have a doctor's appt. this morning so I can't attend the meeting. I do 
have an update on NFSoRDMA. I've got a backport that builds on RHEL7.3, 
and I suspect will also build on SLES12SP2. I should be able to test 
these soon. My intention is to only support these distros for 4.8-1, as 
my testing resources are very limited. Please let me know if that's OK. 
Thanks.


-jeff

On 07/27/2017 02:40 PM, Davis, Arlin R wrote:


Join WebEx meeting 


meeting number (access code): 598 639 160

Dial 598639...@intel.webex.com 

_Join by phone_

*+1-210-795-1110* US Toll

*+1-866-662-9987* US Toll Free
Global call-in numbers 
| 
Toll-free calling restrictions 



Agenda:
·Opens
·OFED 4.8-1 status:
§Plan of Record: Cavium and Broadcom drivers, rdma-core v14, and SRP 
fixes - 30 day turn around if possible.


·User Package Update

§rdma-core v14;  other packages expected/needed?

·Bug Report – Critical/blocking

§2631 – An SRP Reject from an SRP target causes SL7.3/OFED crash

·Fails on OFED 4.8 out of the box, requires OFED+ib_srp_backports

§2632- SRP login failure for SL7.2 and OFED 4.8-rc4 (ib_srp:Sending CM 
DREQ failed)


·Fails on SL7.2, Works on SL7.3 and OFED-4.8

§2633 – NFS over RDMA kernel modules are not supported in OFED 4.8

Unless noted, Bob Woodruff will run meetings while I am out on 
Sabbatical (July 31-Oct 3^rd ) and Phil Cayton will publish 
agenda/minutes.


Regards,

Arlin



___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/mailman/listinfo/ewg



___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/mailman/listinfo/ewg

Re: [ewg] OFA EWG Meeting: Monday, Apr, 10, 2017, 09:00 AM US Pacific Time (12pm EST) - Minutes

[Jeff Becker]

Hi Arlin,

On 04/10/2017 10:19 AM, Davis, Arlin R wrote:


_Attendees   Company _

David DaiIBM

Arlin Davis Intel

Bill Schmidt Intel

Tatyana Nikolova  Intel

Robert Woodruff  Intel

Jim Ryan OFA

Rupert DanceSoftware Forge

Susan Coulterlanl

Minutes:
·Board approval process for OFED 4.8 should begin, need approval by 
May 15^th GA target.

§Woody/Arlin to draft formal email request to board
·OFED 4.8: http://downloads.openfabrics.org/OFED/ofed-4.8/OFED-4.8-rc1.tgz

§RC1 validation progress:

§Good progress by all, no major issues with most companies

§Intel Xeon Phi team hitting Bug #2619

·new libibscif released for v13 support, ready for RC2

§Intel qib team hitting Bug #2622

§Discussion regarding GA timelines

§RC level for interop/debug events preferred (Early May), ability to 
add fixes for GA


§Intel concerned about further delays, hoping for GA by May 15^th 
timeframe.


·General consensus reached for rdma-core v13

§Vlad, please pull in rdma-core v13 and roll RC2 (by end of week if 
possible).


§If IPoIB bug (#2619) can be resolved in next few days, pull into RC2

§Do we have a patch process for v13+ patches? (Arlin to work with Vlad)

§VMware, work with Vlad to remove out tree patches pushed into RC1.

·NFS-rdma backports – deferred until OFED 4.8-1 (fall 2017?).

§IBM would prefer NFS-rdma support in 4.8

§What is confidence level for NFS-rdma in 4.8-1?



First, it's nice to know that someone wants NFS-rdma for 4.8. When I 
asked Susan and others at the OFA workshop,  they couldn't name anyone 
who uses it (except possibly us (NASA) - unfortunately we don't despite 
my suggestions, although that may soon change).


As far as getting the backports done, I got as far as getting a 
successful build on both SLES12SP2 and RHEL7.3. Unfortunately, when I 
tried to run connectathon04 against it, only half the tests passed. 
After several discussions at OFA workshop with Chuck Lever (upstream 
NFS-rdma author) and Woody, I decided that the best way to proceed was 
to start with a base NFS-rdma source earlier than 4.8 (e.g., perhaps 
4.5). Then if I get that to build and pass tests, start adding Chuck's 
patches incrementally if they fix bugs or add important features. Then 
get code building and tested as patches are added. One big problem with 
NFS-rdma is that the patches potentially touch all of NFS, and it would 
not be desirable to backport all of NFS since the distro's have done 
extensive QA, and NFS is such a critical part of the system.


With that said, I can resume my backport efforts, and see how far I get. 
Given the amount of work involved, I'm not sure if I can make the May 15 
release, but I can try. What do you think?


Thanks,

-jeff


·User Package Update

§libfabric still at RC1, Sean Hefty to send Vlad GA packages for RC2.

·Bug Report

§2169 - P5 IPoIB ib0 interface fails with OPFED 4.8 on RH 7.2 
(xeon-phi only? P1?)


§2621 - P5 Unable to compile Lustre on daily builds (try RC1)

§2622 - P5 Can’t modprobe ib_qib after OS reboot/openibd reboot (P1?)

Regards,

Arlin



___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/mailman/listinfo/ewg



___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/mailman/listinfo/ewg

Re: [ewg] OFA EWG: current status needed on open items.

[Jeff Becker]

On 01/13/2017 02:25 PM, Davis, Arlin R wrote:


Since we haven’t meet since early December I would like to get some 
updates. Please provide status where noted. Thanks!


·Opens:

§ULPs:  srp (RH7.2, 7.3, SLES 12 SP2), iser (RH7.3, SLES 12 
SP2),STATUS? (Vlad)


nfsrdma(Jeff Becker may have time in late Dec, early Jan)STATUS? (Jeff)

Unfortunately, I'm tied up until the end of February. After that, if 
there's still time until the release, I'll be able to work on the 
RHEL7.3, and SLESSP2 nfsrdma backports. Thanks.


-jeff


*Next Meeting:*

**

*Monday, January, 23rd, 2016, 09:00 AM US Pacific Time***

*New Reservationless Bridge **(expires Jan 23, 2017):*

916-356-2663, 8-356-2663, Bridge: 2, Passcode: 7482911
Toll-free number in the United States and Canada is 1-888-875-9370

Regards,

Arlin

**



___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/mailman/listinfo/ewg



___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/mailman/listinfo/ewg

[ewg] Fix for bug 2540

[Jeff Becker]

Hi Vlad. Please apply the attached patch. Thanks.

-jeff
From 228a52f2032c41bcd1d4e69e3bb969ade3f5c116 Mon Sep 17 00:00:00 2001
From: root jeffrey.c.bec...@nasa.gov
Date: Thu, 28 May 2015 09:41:37 -0700
Subject: [PATCH] NFS/RDMA: fix for RHEL7.1 Suggested-by: Devesh Sharma
 devesh.sha...@avagotech.com Signed-off-by: Jeff Becker
 jeffrey.c.bec...@nasa.gov

---
 ofed_scripts/gen-compat-config.sh | 4 
 patches/0025-BACKPORT-nfsrdma-RHEL7.patch | 6 +++---
 2 files changed, 7 insertions(+), 3 deletions(-)

diff --git a/ofed_scripts/gen-compat-config.sh b/ofed_scripts/gen-compat-config.sh
index 47b5144..bf89c81 100755
--- a/ofed_scripts/gen-compat-config.sh
+++ b/ofed_scripts/gen-compat-config.sh
@@ -229,6 +229,10 @@ if [[ ${CONFIG_COMPAT_RHEL_6_6} = y ]]; then
 	set_config CONFIG_COMPAT_RPC_XPRT_OPS_HAS_ALLOC_SLOT y
 fi
 
+if [[ ${CONFIG_COMPAT_RHEL_7_1} = y ]]; then
+	set_config CONFIG_COMPAT_RPC_XPRT_OPS_HAS_SECURE_PORT y
+fi
+
 if [[ ${RHEL_MAJOR} -eq 6 ]]; then
 	set_config CONFIG_COMPAT_IS___SKB_TX_HASH y
 	set_config CONFIG_COMPAT_IS_BITMAP y
diff --git a/patches/0025-BACKPORT-nfsrdma-RHEL7.patch b/patches/0025-BACKPORT-nfsrdma-RHEL7.patch
index cec567c..2486795 100644
--- a/patches/0025-BACKPORT-nfsrdma-RHEL7.patch
+++ b/patches/0025-BACKPORT-nfsrdma-RHEL7.patch
@@ -6,7 +6,7 @@ index 4e61880..5620016 100644
  static void svc_rdma_detach(struct svc_xprt *xprt);
  static void svc_rdma_free(struct svc_xprt *xprt);
  static int svc_rdma_has_wspace(struct svc_xprt *xprt);
-+#if LINUX_VERSION_CODE = KERNEL_VERSION(3, 14, 0)
++#if (LINUX_VERSION_CODE = KERNEL_VERSION(3, 14, 0) || CONFIG_COMPAT_RPC_XPRT_OPS_HAS_SECURE_PORT)
  static int svc_rdma_secure_port(struct svc_rqst *);
 +#endif
  static void rq_cq_reap(struct svcxprt_rdma *xprt);
@@ -16,7 +16,7 @@ index 4e61880..5620016 100644
  	.xpo_prep_reply_hdr = svc_rdma_prep_reply_hdr,
  	.xpo_has_wspace = svc_rdma_has_wspace,
  	.xpo_accept = svc_rdma_accept,
-+#if LINUX_VERSION_CODE = KERNEL_VERSION(3, 14, 0)
++#if (LINUX_VERSION_CODE = KERNEL_VERSION(3, 14, 0) || CONFIG_COMPAT_RPC_XPRT_OPS_HAS_SECURE_PORT)
  	.xpo_secure_port = svc_rdma_secure_port,
 +#endif
  };
@@ -26,7 +26,7 @@ index 4e61880..5620016 100644
  	return 1;
  }
  
-+#if LINUX_VERSION_CODE = KERNEL_VERSION(3, 14, 0)
++#if (LINUX_VERSION_CODE = KERNEL_VERSION(3, 14, 0) || CONFIG_COMPAT_RPC_XPRT_OPS_HAS_SECURE_PORT)
  static int svc_rdma_secure_port(struct svc_rqst *rqstp)
  {
  	return 1;
-- 
1.8.3.1

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/mailman/listinfo/ewg

Re: [ewg] [Ofvwg] OpenFabrics Alliance websit back online

[Jeff Becker]

Typo on home page - deliverying.

-jeff

On 05/05/2015 07:48 PM, Bill Lee wrote:


All,

I wanted to let you all know that the OpenFabrics website 
https://www.openfabrics.org/ is back up.  The site had been down for 
a few weeks due to some corrupted code.  Rather than rebuild the site 
we decided to upgrade the security and update the overall structure.  
The site now services four categories of audiences.  Take a look at 
the home page for a complete description.


The working group portal has been complete revised to help you find 
what you are looking for in a click or two.  The next project will be 
to revise the structure of the OpenFabrics software download portal.


Please visit the site at your earliest convenience and let me know 
what you think.


I want to thank Ken Strandberg for making the upgrade and putting up 
with the back and forth between the content providers.  I also want to 
thank the content providers and reviewers for the improvements you all 
helped to build.


Regards,

Bill

-

Bill Lee

Chair, Marketing Working Group

The OpenFabrics Alliance

+1.408.916.0022



___
ofvwg mailing list
of...@lists.openfabrics.org
http://lists.openfabrics.org/mailman/listinfo/ofvwg


___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/mailman/listinfo/ewg

[ewg] NFSoRDMA RHEL 6.5 and 6.6. backport

[Jeff Becker]

Hi Vlad,

Please apply attached patch to compat-rdma for OFED 3.18. Thanks.

-jeff
From fee8373dd67c7bfc7a255906a341ec328463b185 Mon Sep 17 00:00:00 2001
From: Jeff Becker jeffrey.c.bec...@nasa.gov
Date: Wed, 1 Apr 2015 15:43:27 -0700
Subject: [PATCH] NFSoRDMA: backport for RHEL 6.5 and 6.6 Signed-off-by: Jeff
 Becker jeffrey.c.bec...@nasa.gov

---
 ofed_scripts/gen-compat-config.sh |5 +++
 patches/0027-BACKPORT-nfsrdma-SLES11SP3.patch |4 +-
 patches/0028-BACKPORT-nfsrdma-RHEL6.6.patch   |   41 +
 3 files changed, 48 insertions(+), 2 deletions(-)
 create mode 100644 patches/0028-BACKPORT-nfsrdma-RHEL6.6.patch

diff --git a/ofed_scripts/gen-compat-config.sh b/ofed_scripts/gen-compat-config.sh
index ab01d72..59d6b55 100755
--- a/ofed_scripts/gen-compat-config.sh
+++ b/ofed_scripts/gen-compat-config.sh
@@ -210,6 +210,7 @@ if [[ ${CONFIG_COMPAT_RHEL_6_3} = y ]]; then
 	set_config CONFIG_COMPAT_XPRT_ALLOC_4PARAMS y
 	set_config CONFIG_COMPAT_XPRT_RESERVE_XPRT_CONG_2PARAMS y
 	set_config CONFIG_COMPAT_FRAGS_SKB y
+	set_config CONFIG_COMPAT_KMAP_2PARAMS y
 fi
 
 if [[ ${CONFIG_COMPAT_RHEL_6_4} = y ]]; then
@@ -224,6 +225,10 @@ if [[ ${CONFIG_COMPAT_RHEL_6_4} = y ]]; then
 	set_config CONFIG_COMPAT_ETHTOOL_OPS_EXT y
 fi
 
+if [[ ${CONFIG_COMPAT_RHEL_6_6} = y ]]; then
+	set_config CONFIG_COMPAT_RPC_XPRT_OPS_HAS_ALLOC_SLOT y
+fi
+
 if [[ ${RHEL_MAJOR} -eq 6 ]]; then
 	set_config CONFIG_COMPAT_IS___SKB_TX_HASH y
 	set_config CONFIG_COMPAT_IS_BITMAP y
diff --git a/patches/0027-BACKPORT-nfsrdma-SLES11SP3.patch b/patches/0027-BACKPORT-nfsrdma-SLES11SP3.patch
index 2145c34..86103a3 100644
--- a/patches/0027-BACKPORT-nfsrdma-SLES11SP3.patch
+++ b/patches/0027-BACKPORT-nfsrdma-SLES11SP3.patch
@@ -112,7 +112,7 @@ diff -Naur a/net/sunrpc/xprtrdma/transport.c b/net/sunrpc/xprtrdma/transport.c
  	}
  
  	xprt = xprt_alloc(args-net, sizeof(struct rpcrdma_xprt),
-+#if (LINUX_VERSION_CODE = KERNEL_VERSION(3,1,0))
++#if (LINUX_VERSION_CODE = KERNEL_VERSION(3,1,0) || CONFIG_COMPAT_XPRT_ALLOC_4PARAMS)
  			xprt_rdma_slot_table_entries,
 +#endif
  			xprt_rdma_slot_table_entries);
@@ -138,7 +138,7 @@ diff -Naur a/net/sunrpc/xprtrdma/transport.c b/net/sunrpc/xprtrdma/transport.c
  static struct rpc_xprt_ops xprt_rdma_procs = {
  	.reserve_xprt		= xprt_reserve_xprt_cong,
  	.release_xprt		= xprt_release_xprt_cong, /* sunrpc/xprt.c */
-+#if (LINUX_VERSION_CODE = KERNEL_VERSION(3,6,0))
++#if (LINUX_VERSION_CODE = KERNEL_VERSION(3,6,0) || CONFIG_COMPAT_RPC_XPRT_OPS_HAS_ALLOC_SLOT)
  	.alloc_slot		= xprt_alloc_slot,
 +#endif
  	.release_request	= xprt_release_rqst_cong,   /* ditto */
diff --git a/patches/0028-BACKPORT-nfsrdma-RHEL6.6.patch b/patches/0028-BACKPORT-nfsrdma-RHEL6.6.patch
new file mode 100644
index 000..ab1a5c7
--- /dev/null
+++ b/patches/0028-BACKPORT-nfsrdma-RHEL6.6.patch
@@ -0,0 +1,41 @@
+diff --git a/net/sunrpc/xprtrdma/rpc_rdma.c b/net/sunrpc/xprtrdma/rpc_rdma.c
+index 6166c98..a43095e 100644
+--- a/net/sunrpc/xprtrdma/rpc_rdma.c
 b/net/sunrpc/xprtrdma/rpc_rdma.c
+@@ -357,9 +357,17 @@ rpcrdma_inline_pullup(struct rpc_rqst *rqst, int pad)
+ 			curlen = copy_len;
+ 		dprintk(RPC:   %s: page %d destp 0x%p len %d curlen %d\n,
+ 			__func__, i, destp, copy_len, curlen);
++#if defined(CONFIG_COMPAT_KMAP_2PARAMS)
++		srcp = kmap_atomic(ppages[i], KM_SKB_SUNRPC_DATA);
++#else
+ 		srcp = kmap_atomic(ppages[i]);
++#endif
+ 		memcpy(destp, srcp+page_base, curlen);
++#if defined(CONFIG_COMPAT_KMAP_2PARAMS)
++		kunmap_atomic(srcp, KM_SKB_SUNRPC_DATA);
++#else
+ 		kunmap_atomic(srcp);
++#endif
+ 		rqst-rq_svec[0].iov_len += curlen;
+ 		destp += curlen;
+ 		copy_len -= curlen;
+@@ -645,10 +653,18 @@ rpcrdma_inline_fixup(struct rpc_rqst *rqst, char *srcp, int copy_len, int pad)
+ 			dprintk(RPC:   %s: page %d
+  srcp 0x%p len %d curlen %d\n,
+ __func__, i, srcp, copy_len, curlen);
++#if defined(CONFIG_COMPAT_KMAP_2PARAMS)
++			destp = kmap_atomic(ppages[i], KM_SKB_SUNRPC_DATA);
++#else
+ 			destp = kmap_atomic(ppages[i]);
++#endif
+ 			memcpy(destp + page_base, srcp, curlen);
+ 			flush_dcache_page(ppages[i]);
++#if defined(CONFIG_COMPAT_KMAP_2PARAMS)
++			kunmap_atomic(destp, KM_SKB_SUNRPC_DATA);
++#else
+ 			kunmap_atomic(destp);
++#endif
+ 			srcp += curlen;
+ 			copy_len -= curlen;
+ 			if (copy_len == 0)
-- 
1.7.8.2

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/mailman/listinfo/ewg

[ewg] NFSRDMA for OFED 3.18

[Jeff Becker]

Hi Vlad,

Please apply the attached patch to compat-rdma. It allows RHEL7 to 
build. I also confirmed that SLES12 builds fine, as well as the base 
3.18 build. Thanks.


-jeff
From 1e9a23acf64b089698a5f2f07b71f260fd2fef76 Mon Sep 17 00:00:00 2001
From: Jeffrey C. Becker jeffrey.c.bec...@nasa.gov
Date: Fri, 16 Jan 2015 12:11:51 -0800
Subject: [PATCH] NFSRDMA: RHEL7 backport Signed-off-by: Jeff Becker
 jeffrey.c.bec...@nasa.gov

---
 patches/0025-BACKPORT-nfsrdma-RHEL7.patch |   37 +
 1 files changed, 37 insertions(+), 0 deletions(-)
 create mode 100644 patches/0025-BACKPORT-nfsrdma-RHEL7.patch

diff --git a/patches/0025-BACKPORT-nfsrdma-RHEL7.patch b/patches/0025-BACKPORT-nfsrdma-RHEL7.patch
new file mode 100644
index 000..cec567c
--- /dev/null
+++ b/patches/0025-BACKPORT-nfsrdma-RHEL7.patch
@@ -0,0 +1,37 @@
+diff --git a/net/sunrpc/xprtrdma/svc_rdma_transport.c b/net/sunrpc/xprtrdma/svc_rdma_transport.c
+index 4e61880..5620016 100644
+--- a/net/sunrpc/xprtrdma/svc_rdma_transport.c
 b/net/sunrpc/xprtrdma/svc_rdma_transport.c
+@@ -66,7 +66,9 @@ static void dto_tasklet_func(unsigned long data);
+ static void svc_rdma_detach(struct svc_xprt *xprt);
+ static void svc_rdma_free(struct svc_xprt *xprt);
+ static int svc_rdma_has_wspace(struct svc_xprt *xprt);
++#if LINUX_VERSION_CODE = KERNEL_VERSION(3, 14, 0)
+ static int svc_rdma_secure_port(struct svc_rqst *);
++#endif
+ static void rq_cq_reap(struct svcxprt_rdma *xprt);
+ static void sq_cq_reap(struct svcxprt_rdma *xprt);
+ 
+@@ -84,7 +86,9 @@ static struct svc_xprt_ops svc_rdma_ops = {
+ 	.xpo_prep_reply_hdr = svc_rdma_prep_reply_hdr,
+ 	.xpo_has_wspace = svc_rdma_has_wspace,
+ 	.xpo_accept = svc_rdma_accept,
++#if LINUX_VERSION_CODE = KERNEL_VERSION(3, 14, 0)
+ 	.xpo_secure_port = svc_rdma_secure_port,
++#endif
+ };
+ 
+ struct svc_xprt_class svc_rdma_class = {
+@@ -1205,10 +1209,12 @@ static int svc_rdma_has_wspace(struct svc_xprt *xprt)
+ 	return 1;
+ }
+ 
++#if LINUX_VERSION_CODE = KERNEL_VERSION(3, 14, 0)
+ static int svc_rdma_secure_port(struct svc_rqst *rqstp)
+ {
+ 	return 1;
+ }
++#endif
+ 
+ /*
+  * Attempt to register the kvec representing the RPC memory with the
-- 
1.7.7.4

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/mailman/listinfo/ewg

[ewg] fix for bug 2514

[Jeff Becker]

Hi Vlad,

Please apply attached patch to compat-rdma. Thanks.

-jeff
From e06555cfcd4867c9df90ede8612923f5258c52f7 Mon Sep 17 00:00:00 2001
From: Jeffrey C. Becker jeffrey.c.bec...@nasa.gov
Date: Tue, 23 Dec 2014 16:25:52 -0800
Subject: [PATCH] NFSRDMA: Changes to allow modules to build for 3.18
 Signed-off-by: Jeff Becker jeffrey.c.bec...@nasa.gov

---
 Makefile  |2 +-
 configure |   41 -
 makefile  |4 ++--
 3 files changed, 23 insertions(+), 24 deletions(-)

diff --git a/Makefile b/Makefile
index a0559be..79f6246 100644
--- a/Makefile
+++ b/Makefile
@@ -6,7 +6,7 @@ obj-$(CONFIG_MLX4_CORE) += drivers/net/ethernet/mellanox/mlx4/
 obj-$(CONFIG_MLX5_CORE) += drivers/net/ethernet/mellanox/mlx5/core/
 obj-$(CONFIG_BE2NET)+= drivers/net/ethernet/emulex/benet/
 obj-$(CONFIG_RDS)   += net/rds/
-obj-$(CONFIG_SUNRPC_XPRT_RDMA)  += net/sunrpc/xprtrdma/
+obj-$(CONFIG_SUNRPC_XPRT_RDMA_CLIENT)  += net/sunrpc/xprtrdma/
 obj-$(CONFIG_SCSI_CXGB3_ISCSI)  += drivers/scsi/cxgbi/
 obj-$(CONFIG_SCSI_CXGB4_ISCSI)  += drivers/scsi/cxgbi/
 obj-$(CONFIG_SCSI_SRP_ATTRS)+= drivers/scsi/
diff --git a/configure b/configure
index 84d3a36..0681716 100755
--- a/configure
+++ b/configure
@@ -187,7 +187,7 @@ Usage:  `basename $0` [options]
 --with-debug-info  make CONFIG_DEBUG_INFO=y [yes]
 --without-debug-info [no]
 
---with-nfsrdma-modmake CONFIG_SUNRPC_XPRT_RDMA=m [no]
+--with-nfsrdma-modmake CONFIG_SUNRPC_XPRT_RDMA_CLIENT=m CONFIG_SUNRPC_XPRT_RDMA_SERVER=m [no]
 --without-nfsrdma-mod[yes]
 
 --help - print out options
@@ -659,12 +659,12 @@ main()
 CONFIG_BE2NET=
 ;;
 --with-nfsrdma-mod)
-CONFIG_SUNRPC_XPRT_RDMA=m
-CONFIG_NFSD_RDMA=y
+CONFIG_SUNRPC_XPRT_RDMA_CLIENT=m
+CONFIG_SUNRPC_XPRT_RDMA_SERVER=m
 ;;
 --without-nfsrdma-mod)
-CONFIG_SUNRPC_XPRT_RDMA=
-CONFIG_NFSD_RDMA=
+CONFIG_SUNRPC_XPRT_RDMA_CLIENT=
+CONFIG_SUNRPC_XPRT_RDMA_SERVER=
 ;;
 --with-modprobe|--without-modprobe)
 ;;
@@ -771,7 +771,8 @@ CONFIG_INFINIBAND_NES=${CONFIG_INFINIBAND_NES:-''}
 CONFIG_INFINIBAND_OCRDMA=${CONFIG_INFINIBAND_OCRDMA:-''}
 CONFIG_BE2NET=${CONFIG_BE2NET:-''}
 
-CONFIG_SUNRPC_XPRT_RDMA=${CONFIG_SUNRPC_XPRT_RDMA:-''} 
+CONFIG_SUNRPC_XPRT_RDMA_CLIENT=${CONFIG_SUNRPC_XPRT_RDMA_CLIENT:-''} 
+CONFIG_SUNRPC_XPRT_RDMA_SERVER=${CONFIG_SUNRPC_XPRT_RDMA_SERVER:-''} 
 CONFIG_SUNRPC=${CONFIG_SUNRPC:-''} 
 CONFIG_SUNRPC_GSS=${CONFIG_SUNRPC_GSS:-''} 
 CONFIG_RPCSEC_GSS_KRB5=${CONFIG_RPCSEC_GSS_KRB5:-''} 
@@ -794,7 +795,6 @@ CONFIG_NFSD_V2_ACL=${CONFIG_NFSD_V2_ACL:-''}
 CONFIG_NFSD_V3=${CONFIG_NFSD_V3:-''}
 CONFIG_NFSD_V3_ACL=${CONFIG_NFSD_V3_ACL:-''}
 CONFIG_NFSD_V4=${CONFIG_NFSD_V4:-''}
-CONFIG_NFSD_RDMA=${CONFIG_NFSD_RDMA:-''}
 
 CONFIG_INFINIBAND_IPOIB_DEBUG_DATA=${CONFIG_INFINIBAND_IPOIB_DEBUG_DATA:-''}
 CONFIG_INFINIBAND_SDP_SEND_ZCOPY=${CONFIG_INFINIBAND_SDP_SEND_ZCOPY:-''}
@@ -903,7 +903,8 @@ CONFIG_INFINIBAND_NES=${CONFIG_INFINIBAND_NES}
 CONFIG_INFINIBAND_OCRDMA=${CONFIG_INFINIBAND_OCRDMA}
 CONFIG_BE2NET=${CONFIG_BE2NET}
 
-CONFIG_SUNRPC_XPRT_RDMA=${CONFIG_SUNRPC_XPRT_RDMA}
+CONFIG_SUNRPC_XPRT_RDMA_CLIENT=${CONFIG_SUNRPC_XPRT_RDMA_CLIENT}
+CONFIG_SUNRPC_XPRT_RDMA_SERVER=${CONFIG_SUNRPC_XPRT_RDMA_SERVER}
 CONFIG_SUNRPC=${CONFIG_SUNRPC}
 CONFIG_SUNRPC_GSS=${CONFIG_SUNRPC_GSS}
 CONFIG_RPCSEC_GSS_KRB5=${CONFIG_RPCSEC_GSS_KRB5}
@@ -923,8 +924,6 @@ CONFIG_NFSD_V2_ACL=${CONFIG_NFSD_V2_ACL}
 CONFIG_NFSD_V3=${CONFIG_NFSD_V3}
 CONFIG_NFSD_V3_ACL=${CONFIG_NFSD_V3_ACL}
 CONFIG_NFSD_V4=${CONFIG_NFSD_V4}
-CONFIG_NFSD_RDMA=${CONFIG_NFSD_RDMA}
-
 
 CONFIG_INFINIBAND_IPOIB_DEBUG_DATA=${CONFIG_INFINIBAND_IPOIB_DEBUG_DATA}
 CONFIG_INFINIBAND_SDP_SEND_ZCOPY=${CONFIG_INFINIBAND_SDP_SEND_ZCOPY}
@@ -1228,8 +1227,9 @@ if [ X${CONFIG_MLX4_EN_DCB} == Xy ]; then
 else
 DEFINE_MLX4_EN_DCB=#undef CONFIG_MLX4_EN_DCB
 fi
-if [ X${CONFIG_SUNRPC_XPRT_RDMA} == Xm ]; then
-DEFINE_SUNRPC_XPRT_RDMA=#undef CONFIG_SUNRPC_XPRT_RDMA\n#define CONFIG_SUNRPC_XPRT_RDMA 1
+if [ X${CONFIG_SUNRPC_XPRT_RDMA_CLIENT} == Xm ]; then
+DEFINE_SUNRPC_XPRT_RDMA_CLIENT=#undef CONFIG_SUNRPC_XPRT_RDMA_CLIENT\n#define CONFIG_SUNRPC_XPRT_RDMA_CLIENT 1
+DEFINE_SUNRPC_XPRT_RDMA_SERVER=#undef CONFIG_SUNRPC_XPRT_RDMA_SERVER\n#define CONFIG_SUNRPC_XPRT_RDMA_SERVER 1
 DEFINE_SUNRPC=#undef CONFIG_SUNRPC\n#define CONFIG_SUNRPC 1
 DEFINE_SUNRPC_GSS=#undef CONFIG_SUNRPC_GSS\n#define CONFIG_SUNRPC_GSS 1
 DEFINE_RPCSEC_GSS_KRB5=#undef CONFIG_RPCSEC_GSS_KRB5\n#define CONFIG_RPCSEC_GSS_KRB5 1
@@ -1248,9 +1248,9 @@ if [ X

[ewg] NFSRDMA backport for SLES11sp3

[Jeff Becker]

Hi Vlad. Please apply the attached patches to compat-rdma. Thanks.

-jeff
From 157b0a91ae127f710606bba550a6983619be3519 Mon Sep 17 00:00:00 2001
From: Jeffrey C. Becker jeffrey.c.bec...@nasa.gov
Date: Wed, 19 Feb 2014 15:42:08 -0800
Subject: [PATCH 1/2] NFSRDMA: SLES11sp3 backport
 Signed-off-by: Jeff Becker jeffrey.c.bec...@nasa.gov

---
 patches/0024-nfsrdma-backport-for-SLES11sp3.patch |   42 +
 1 files changed, 42 insertions(+), 0 deletions(-)
 create mode 100644 patches/0024-nfsrdma-backport-for-SLES11sp3.patch

diff --git a/patches/0024-nfsrdma-backport-for-SLES11sp3.patch b/patches/0024-nfsrdma-backport-for-SLES11sp3.patch
new file mode 100644
index 000..228b9e0
--- /dev/null
+++ b/patches/0024-nfsrdma-backport-for-SLES11sp3.patch
@@ -0,0 +1,42 @@
+diff --git a/net/sunrpc/xprtrdma/transport.c b/net/sunrpc/xprtrdma/transport.c
+index 285dc08..ab74500 100644
+--- a/net/sunrpc/xprtrdma/transport.c
 b/net/sunrpc/xprtrdma/transport.c
+@@ -279,7 +281,9 @@ xprt_setup_rdma(struct xprt_create *args)
+ 	}
+ 
+ 	xprt = xprt_alloc(args-net, sizeof(struct rpcrdma_xprt),
++#if (LINUX_VERSION_CODE = KERNEL_VERSION(3,1,0))
+ 			xprt_rdma_slot_table_entries,
++#endif
+ 			xprt_rdma_slot_table_entries);
+ 	if (xprt == NULL) {
+ 		dprintk(RPC:   %s: couldn't allocate rpcrdma_xprt\n,
+@@ -448,8 +452,15 @@ xprt_rdma_connect(struct rpc_xprt *xprt, struct rpc_task *task)
+ }
+ 
+ static int
++#if (LINUX_VERSION_CODE = KERNEL_VERSION(3,1,0))
+ xprt_rdma_reserve_xprt(struct rpc_xprt *xprt, struct rpc_task *task)
++#else
++xprt_rdma_reserve_xprt(struct rpc_task *task)
++#endif
+ {
++#if (LINUX_VERSION_CODE  KERNEL_VERSION(3,1,0))
++	struct rpc_xprt *xprt = task-tk_xprt;
++#endif
+ 	struct rpcrdma_xprt *r_xprt = rpcx_to_rdmax(xprt);
+ 	int credits = atomic_read(r_xprt-rx_buf.rb_credits);
+ 
+@@ -461,7 +472,11 @@ xprt_rdma_reserve_xprt(struct rpc_xprt *xprt, struct rpc_task *task)
+ 		BUG_ON(r_xprt-rx_buf.rb_cwndscale = 0);
+ 	}
+ 	xprt-cwnd = credits * r_xprt-rx_buf.rb_cwndscale;
++#if (LINUX_VERSION_CODE = KERNEL_VERSION(3,1,0))
+ 	return xprt_reserve_xprt_cong(xprt, task);
++#else
++	return xprt_reserve_xprt_cong(task);
++#endif
+ }
+ 
+ /*
-- 
1.7.1

From 54a4e0b324d63d41c95b3d25709dc66c4c709cf0 Mon Sep 17 00:00:00 2001
From: Jeffrey C. Becker jeffrey.c.bec...@nasa.gov
Date: Thu, 20 Feb 2014 09:51:06 -0800
Subject: [PATCH 2/2] NFSRDMA: Renumber SLES11sp3 patch
 Signed-off-by: Jeff Becker jeffrey.c.bec...@nasa.gov

---
 patches/0024-nfsrdma-backport-for-SLES11sp3.patch |   42 -
 patches/0025-nfsrdma-backport-for-SLES11sp3.patch |   42 +
 2 files changed, 42 insertions(+), 42 deletions(-)
 delete mode 100644 patches/0024-nfsrdma-backport-for-SLES11sp3.patch
 create mode 100644 patches/0025-nfsrdma-backport-for-SLES11sp3.patch

diff --git a/patches/0024-nfsrdma-backport-for-SLES11sp3.patch b/patches/0024-nfsrdma-backport-for-SLES11sp3.patch
deleted file mode 100644
index 228b9e0..000
--- a/patches/0024-nfsrdma-backport-for-SLES11sp3.patch
+++ /dev/null
@@ -1,42 +0,0 @@
-diff --git a/net/sunrpc/xprtrdma/transport.c b/net/sunrpc/xprtrdma/transport.c
-index 285dc08..ab74500 100644
 a/net/sunrpc/xprtrdma/transport.c
-+++ b/net/sunrpc/xprtrdma/transport.c
-@@ -279,7 +281,9 @@ xprt_setup_rdma(struct xprt_create *args)
- 	}
- 
- 	xprt = xprt_alloc(args-net, sizeof(struct rpcrdma_xprt),
-+#if (LINUX_VERSION_CODE = KERNEL_VERSION(3,1,0))
- 			xprt_rdma_slot_table_entries,
-+#endif
- 			xprt_rdma_slot_table_entries);
- 	if (xprt == NULL) {
- 		dprintk(RPC:   %s: couldn't allocate rpcrdma_xprt\n,
-@@ -448,8 +452,15 @@ xprt_rdma_connect(struct rpc_xprt *xprt, struct rpc_task *task)
- }
- 
- static int
-+#if (LINUX_VERSION_CODE = KERNEL_VERSION(3,1,0))
- xprt_rdma_reserve_xprt(struct rpc_xprt *xprt, struct rpc_task *task)
-+#else
-+xprt_rdma_reserve_xprt(struct rpc_task *task)
-+#endif
- {
-+#if (LINUX_VERSION_CODE  KERNEL_VERSION(3,1,0))
-+	struct rpc_xprt *xprt = task-tk_xprt;
-+#endif
- 	struct rpcrdma_xprt *r_xprt = rpcx_to_rdmax(xprt);
- 	int credits = atomic_read(r_xprt-rx_buf.rb_credits);
- 
-@@ -461,7 +472,11 @@ xprt_rdma_reserve_xprt(struct rpc_xprt *xprt, struct rpc_task *task)
- 		BUG_ON(r_xprt-rx_buf.rb_cwndscale = 0);
- 	}
- 	xprt-cwnd = credits * r_xprt-rx_buf.rb_cwndscale;
-+#if (LINUX_VERSION_CODE = KERNEL_VERSION(3,1,0))
- 	return xprt_reserve_xprt_cong(xprt, task);
-+#else
-+	return xprt_reserve_xprt_cong(task);
-+#endif
- }
- 
- /*
diff --git a/patches/0025-nfsrdma-backport-for-SLES11sp3.patch b/patches/0025-nfsrdma-backport-for-SLES11sp3.patch
new file mode 100644
index 000..228b9e0
--- /dev/null
+++ b/patches/0025-nfsrdma-backport-for-SLES11sp3.patch
@@ -0,0 +1,42 @@
+diff --git a/net/sunrpc/xprtrdma/transport.c b/net/sunrpc/xprtrdma/transport.c
+index 285dc08..ab74500 100644
+--- a/net/sunrpc/xprtrdma/transport.c
 b/net/sunrpc/xprtrdma/transport.c
+@@ -279,7 +281,9 @@ xprt_setup_rdma(struct xprt_create *args)
+ 	}
+ 
+ 	xprt

[ewg] NFSRDMA RHEL6.5 backport

[Jeff Becker]

Hi Vlad. Please apply the attached patch to compat-rdma.git. Thanks.

-jeff
From 6a782d522f1783657615ce57cda6da84ed134535 Mon Sep 17 00:00:00 2001
From: Jeffrey C. Becker jeffrey.c.bec...@nasa.gov
Date: Mon, 3 Feb 2014 16:17:34 -0800
Subject: [PATCH] NFSRDMA: Add RHEL6.5 backport
 Signed-off-by: Jeff Becker jeffrey.c.bec...@nasa.gov

---
 patches/0023-nfsrdma-backport-for-RHEL6.5 |  138 +
 1 files changed, 138 insertions(+), 0 deletions(-)
 create mode 100644 patches/0023-nfsrdma-backport-for-RHEL6.5

diff --git a/patches/0023-nfsrdma-backport-for-RHEL6.5 b/patches/0023-nfsrdma-backport-for-RHEL6.5
new file mode 100644
index 000..17bf6f4
--- /dev/null
+++ b/patches/0023-nfsrdma-backport-for-RHEL6.5
@@ -0,0 +1,138 @@
+diff --git a/net/sunrpc/xprtrdma/rpc_rdma.c b/net/sunrpc/xprtrdma/rpc_rdma.c
+index e03725b..410b551 100644
+--- a/net/sunrpc/xprtrdma/rpc_rdma.c
 b/net/sunrpc/xprtrdma/rpc_rdma.c
+@@ -338,9 +338,17 @@ rpcrdma_inline_pullup(struct rpc_rqst *rqst, int pad)
+ 			curlen = copy_len;
+ 		dprintk(RPC:   %s: page %d destp 0x%p len %d curlen %d\n,
+ 			__func__, i, destp, copy_len, curlen);
++#if (LINUX_VERSION_CODE  KERNEL_VERSION(3,4,0))
++srcp = kmap_atomic(ppages[i], KM_SKB_SUNRPC_DATA);
++#else
+ 		srcp = kmap_atomic(ppages[i]);
++#endif
+ 		memcpy(destp, srcp+page_base, curlen);
++#if (LINUX_VERSION_CODE  KERNEL_VERSION(3,4,0))
++kunmap_atomic(srcp, KM_SKB_SUNRPC_DATA);
++#else
+ 		kunmap_atomic(srcp);
++#endif
+ 		rqst-rq_svec[0].iov_len += curlen;
+ 		destp += curlen;
+ 		copy_len -= curlen;
+@@ -639,10 +647,18 @@ rpcrdma_inline_fixup(struct rpc_rqst *rqst, char *srcp, int copy_len, int pad)
+ 			dprintk(RPC:   %s: page %d
+  srcp 0x%p len %d curlen %d\n,
+ __func__, i, srcp, copy_len, curlen);
++#if (LINUX_VERSION_CODE  KERNEL_VERSION(3,4,0))
++destp = kmap_atomic(ppages[i], KM_SKB_SUNRPC_DATA);
++#else
+ 			destp = kmap_atomic(ppages[i]);
++#endif
+ 			memcpy(destp + page_base, srcp, curlen);
+ 			flush_dcache_page(ppages[i]);
++#if (LINUX_VERSION_CODE  KERNEL_VERSION(3,4,0))
++kunmap_atomic(destp, KM_SKB_SUNRPC_DATA);
++#else
+ 			kunmap_atomic(destp);
++#endif
+ 			srcp += curlen;
+ 			copy_len -= curlen;
+ 			if (copy_len == 0)
+diff --git a/net/sunrpc/xprtrdma/svc_rdma_recvfrom.c b/net/sunrpc/xprtrdma/svc_rdma_recvfrom.c
+index 0ce7552..afd46ee 100644
+--- a/net/sunrpc/xprtrdma/svc_rdma_recvfrom.c
 b/net/sunrpc/xprtrdma/svc_rdma_recvfrom.c
+@@ -524,8 +524,13 @@ next_sge:
+ 	 * Detach res pages. If svc_release sees any it will attempt to
+ 	 * put them.
+ 	 */
++#if (LINUX_VERSION_CODE  KERNEL_VERSION(3,8,0))
++while (rqstp-rq_resused)
++rqstp-rq_respages[--rqstp-rq_resused] = NULL;
++#else
+ 	while (rqstp-rq_next_page != rqstp-rq_respages)
+ 		*(--rqstp-rq_next_page) = NULL;
++#endif
+ 
+ 	return err;
+ }
+@@ -550,7 +555,11 @@ static int rdma_read_complete(struct svc_rqst *rqstp,
+ 
+ 	/* rq_respages starts after the last arg page */
+ 	rqstp-rq_respages = rqstp-rq_arg.pages[page_no];
++#if (LINUX_VERSION_CODE  KERNEL_VERSION(3,8,0))
++rqstp-rq_resused = 0;
++#else
+ 	rqstp-rq_next_page = rqstp-rq_arg.pages[page_no];
++#endif
+ 
+ 	/* Rebuild rq_arg head and tail. */
+ 	rqstp-rq_arg.head[0] = head-arg.head[0];
+diff --git a/net/sunrpc/xprtrdma/svc_rdma_sendto.c b/net/sunrpc/xprtrdma/svc_rdma_sendto.c
+index c1d124d..59d462d 100644
+--- a/net/sunrpc/xprtrdma/svc_rdma_sendto.c
 b/net/sunrpc/xprtrdma/svc_rdma_sendto.c
+@@ -548,7 +548,9 @@ static int send_reply(struct svcxprt_rdma *rdma,
+ 	int sge_no;
+ 	int sge_bytes;
+ 	int page_no;
++#if (LINUX_VERSION_CODE = KERNEL_VERSION(3,8,0))
+ 	int pages;
++#endif
+ 	int ret;
+ 
+ 	/* Post a recv buffer to handle another request. */
+@@ -612,8 +614,12 @@ static int send_reply(struct svcxprt_rdma *rdma,
+ 	 * respages array. They are our pages until the I/O
+ 	 * completes.
+ 	 */
++#if (LINUX_VERSION_CODE  KERNEL_VERSION(3,8,0))
++for (page_no = 0; page_no  rqstp-rq_resused; page_no++) {
++#else
+ 	pages = rqstp-rq_next_page - rqstp-rq_respages;
+ 	for (page_no = 0; page_no  pages; page_no++) {
++#endif
+ 		ctxt-pages[page_no+1] = rqstp-rq_respages[page_no];
+ 		ctxt-count++;
+ 		rqstp-rq_respages[page_no] = NULL;
+diff --git a/net/sunrpc/xprtrdma/svc_rdma_transport.c b/net/sunrpc/xprtrdma/svc_rdma_transport.c
+index 62e4f9b..f4d65b1 100644
+--- a/net/sunrpc/xprtrdma/svc_rdma_transport.c
 b/net/sunrpc/xprtrdma/svc_rdma_transport.c
+@@ -445,7 +445,11 @@ static struct svcxprt_rdma *rdma_create_xprt(struct svc_serv *serv,
+ 
+ 	if (!cma_xprt)
+ 		return NULL;
++#if (LINUX_VERSION_CODE  KERNEL_VERSION(3,3,0))
++	svc_xprt_init(svc_rdma_class, cma_xprt-sc_xprt, serv);
++#else
+ 	svc_xprt_init(init_net, svc_rdma_class, cma_xprt-sc_xprt, serv);
++#endif
+ 	INIT_LIST_HEAD(cma_xprt-sc_accept_q);
+ 	INIT_LIST_HEAD(cma_xprt-sc_dto_q);
+ 	INIT_LIST_HEAD(cma_xprt-sc_rq_dto_q

[ewg] compat-rdma compile problem on OFED-3.5.2-rc3, on SLES11SP2

[Jeff Becker]

Does this look familiar to anyone? Thanks.

-jeff

/var/tmp/OFED_topdir/BUILD/compat-rdma-3.5/drivers/infiniband/core/addr.c: 
In function 'dst_fetch_ha':
/var/tmp/OFED_topdir/BUILD/compat-rdma-3.5/drivers/infiniband/core/addr.c:197: 
error: implicit declaration of function 'dst_get_neighbour'


___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [ewg] compat-rdma compile problem on OFED-3.5.2-rc3, on SLES11SP2

[Jeff Becker]

Hi Bill,

On 12/13/2013 04:23 PM, Bill Snapko wrote:

Hi Jeff,

I had trouble building on something other than the default SLES 11sp2 kernel. I 
think the default kernel version for 11sp2 is 3.0.13-0.27-default.

I'm just guessing that you are trying to build the package with a kernel 
version something other than 3.0.13-0.27-default.

I can build rc3 on SLES 11sp2 with the 3.0.13-0.27-default kernel. Using 
another kernel version like 3.0.80-0.7-default doesn't work. I get the same or 
similar build errors.


Exactly. I found the problem in compat-rdma-3.5/config.mk. As you can 
see below, the build only thinks that it has SP2 if the KERNEL_SUBLEVEL 
is 13, otherwise it thinks it has SP3. I'm working on a fix, but the 
problem is that according to a Novell wiki page I was looking at, SP3 
had an original kernel version of 3.0.76. Hence the valid kernel version 
ranges for SP2 and SP3 overlap.


-jeff

/* excerpt from compat-rdma-3.5/config.mk */
...
 KLIB_SOURCE := $(subst build,source,$(KLIB_BUILD))
 NAME := $(shell grep ^NAME $(KLIB_SOURCE)/Makefile | sed -n 's/.*= 
*\(.*\)/\1/p')

 ifneq ($(NAME),)
 ifeq ($(strip $(NAME)),Sneaky Weasel)
 SLES_MAJOR := 11
 KERNEL_SUBLEVEL := $(shell grep ^SUBLEVEL $(KLIB_SOURCE)/Makefile | 
sed -n 's/.*= *\(.*\)/\1/p')

 ifeq ($(KERNEL_SUBLEVEL),13) - this is the problem!!
 SLES_MINOR := 2
 CONFIG_COMPAT_SLES_11_2 := y
 else
 SLES_MINOR := 3
 CONFIG_COMPAT_SLES_11_3 := y
 endif
 endif
 endif


Bill

-Original Message-
From: ewg-boun...@lists.openfabrics.org 
[mailto:ewg-boun...@lists.openfabrics.org] On Behalf Of Jeff Becker
Sent: Friday, December 13, 2013 1:38 PM
To: ewg@lists.openfabrics.org; Rupert Dance
Subject: [ewg] compat-rdma compile problem on OFED-3.5.2-rc3, on SLES11SP2

Does this look familiar to anyone? Thanks.

-jeff

/var/tmp/OFED_topdir/BUILD/compat-rdma-3.5/drivers/infiniband/core/addr.c:
In function 'dst_fetch_ha':
/var/tmp/OFED_topdir/BUILD/compat-rdma-3.5/drivers/infiniband/core/addr.c:197:
error: implicit declaration of function 'dst_get_neighbour'

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg


___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [ewg] [ANNOUNCE] OFED-3.5-rc2 is available

[Jeff Becker]

Hi Vlad. Happy New Year (belated) ! One additional note below...

On 10/03/2012 08:54 AM, Vladimir Sokolovsky wrote:

Hi,
OFED 3.5-rc2 is available.

The tarball is available on:
http://www.openfabrics.org/downloads/OFED/ofed-3.5/OFED-3.5-rc2.tgz

To get BUILD_ID run ofed_info

Please report any issues in bugzilla https://bugs.openfabrics.org/ for OFED 3.5

Regards,
Vladimir


OFED-3.5-rc2 Main Changes from OFED 3.5-rc1
---
compat-rdma: Add SRP backport
compat-rdma: /etc/init.d/openibd: Fix LSB header
compat-rdma: IB/qib: linux-3.6 patches backported
compat-rdma: iw_cxgb4: Fix bug 2369 in OFED bugzilla
compat-rdma: IB/qib: fix compliance regression in 3.5
compat-rdma: RDMA/nes: Added linux-next-pending patches
compat-rdma: RDMA/nes: Updated backports
compat-rdma: NFSRDMA RHEL6.3 backport
compat-rdma: NFSRDMA SLES11SP2 backport
compat-rdma: linux-next-cherry-picks: RDMA/ucma.c: Different fix for ucma 
context uid=0, causing iWarp RDMA applications to fail in connection 
establishment


Although it didn't require additional patches, I have verified that 
NFSRDMA is also supported on RHEL6.2. Thanks.


-jeff


Updated packages:
  infinipath-psm-3.0.1-115.1015_open
  perftest-1.4.0-0.80.gd1763bd
  qperf-0.4.7-0.2.gf3f7001

Supported Platforms and Operating Systems
-
o   CPU architectures:
  - x86_64
  - x86
  - ppc64
  - ia64

o   Linux Operating Systems:
  - RedHat EL6.2  2.6.32-220.el6
  - RedHat EL6.3  2.6.32-279.el6
  - SLES11 SP23.0.13-0.27-default
  - kernel.org3.5*

* Minimal QA for these versions.


OFED_release_notes.txt

Note: See the release notes of each component for additional issues.



___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

[ewg] [PATCH] NFSRDMA RHEL6.3 backport

[Jeff Becker]

Hi Vlad. Please apply the attached patch to compat-rdma/compat-rdma.git. 
Thanks.


-jeff
commit 63a9a0f587f99bc3ef7b527727090b2b603fa053
Author: Jeffrey C. Becker jeffrey.c.bec...@nasa.gov
Date:   Fri Aug 24 11:08:57 2012 -0700

NFSRDMA: RHEL 6.3 backport
Signed-off-by: Jeff Becker jeffrey.c.bec...@nasa.gov

diff --git a/config.mk b/config.mk
index 696d26c..6b65701 100644
--- a/config.mk
+++ b/config.mk
@@ -65,3 +65,7 @@ endif #CONFIG_COMPAT_KERNEL_2_6_38
 ifdef CONFIG_COMPAT_SLES_11_2
  NEED_MIN_DUMP_ALLOC_ARG=y
 endif
+
+ifdef CONFIG_COMPAT_RHEL_6_3
+ CONFIG_COMPAT_XPRTRDMA_NEEDED=y
+endif
diff --git a/makefile b/makefile
index 8d04b9c..c286c6b 100644
--- a/makefile
+++ b/makefile
@@ -27,6 +27,11 @@ CFLAGS += \
-DCONFIG_COMPAT_SLES_11_2
 endif
 
+ifneq ($(CONFIG_COMPAT_XPRTRDMA_NEEDED),)
+CFLAGS += \
+   -DCONFIG_COMPAT_XPRTRDMA_NEEDED
+endif
+
 ifneq ($(NEED_MIN_DUMP_ALLOC_ARG),)
 CFLAGS += \
-DNEED_MIN_DUMP_ALLOC_ARG
diff --git 
a/patches/0020-nfsrdma-Backports-for-RHEL6.2-RHEL6.3-and-SLES11-SP2.patch 
b/patches/0020-nfsrdma-Backports-for-RHEL6.2-RHEL6.3-and-SLES11-SP2.patch
index 4f79a56..22a9564 100644
--- a/patches/0020-nfsrdma-Backports-for-RHEL6.2-RHEL6.3-and-SLES11-SP2.patch
+++ b/patches/0020-nfsrdma-Backports-for-RHEL6.2-RHEL6.3-and-SLES11-SP2.patch
@@ -19,7 +19,7 @@
}
  
xprt = xprt_alloc(args-net, sizeof(struct rpcrdma_xprt),
-+#if (LINUX_VERSION_CODE = KERNEL_VERSION(3,1,0))
++#if (LINUX_VERSION_CODE = KERNEL_VERSION(3,1,0) || 
CONFIG_COMPAT_XPRTRDMA_NEEDED)
xprt_rdma_slot_table_entries,
 +#endif
xprt_rdma_slot_table_entries);
@@ -29,7 +29,7 @@
  }
  
  static int
-+#if (LINUX_VERSION_CODE = KERNEL_VERSION(3,1,0))
++#if (LINUX_VERSION_CODE = KERNEL_VERSION(3,1,0) || 
CONFIG_COMPAT_XPRTRDMA_NEEDED)
  xprt_rdma_reserve_xprt(struct rpc_xprt *xprt, struct rpc_task *task)
  {
 +#else
@@ -44,7 +44,7 @@
BUG_ON(r_xprt-rx_buf.rb_cwndscale = 0);
}
xprt-cwnd = credits * r_xprt-rx_buf.rb_cwndscale;
-+#if (LINUX_VERSION_CODE = KERNEL_VERSION(3,1,0))
++#if (LINUX_VERSION_CODE = KERNEL_VERSION(3,1,0) || 
CONFIG_COMPAT_XPRTRDMA_NEEDED)
return xprt_reserve_xprt_cong(xprt, task);
 +#else
 +  return xprt_reserve_xprt_cong(task);
@@ -52,3 +52,43 @@
  }
  
  /*
+
+--- a/net/sunrpc/xprtrdma/rpc_rdma.c   2012-08-10 14:24:49.029166778 -0700
 b/net/sunrpc/xprtrdma/rpc_rdma.c   2012-08-20 15:10:21.838979969 -0700
+@@ -338,9 +338,17 @@
+   curlen = copy_len;
+   dprintk(RPC:   %s: page %d destp 0x%p len %d curlen %d\n,
+   __func__, i, destp, copy_len, curlen);
++#if (LINUX_VERSION_CODE = KERNEL_VERSION(3,4,0))
+   srcp = kmap_atomic(ppages[i]);
++#else
++  srcp = kmap_atomic(ppages[i], KM_SKB_SUNRPC_DATA);
++#endif
+   memcpy(destp, srcp+page_base, curlen);
++#if (LINUX_VERSION_CODE = KERNEL_VERSION(3,4,0))
+   kunmap_atomic(srcp);
++#else
++  kunmap_atomic(srcp, KM_SKB_SUNRPC_DATA);
++#endif
+   rqst-rq_svec[0].iov_len += curlen;
+   destp += curlen;
+   copy_len -= curlen;
+@@ -639,10 +647,18 @@
+   dprintk(RPC:   %s: page %d
+srcp 0x%p len %d curlen %d\n,
+   __func__, i, srcp, copy_len, curlen);
++#if (LINUX_VERSION_CODE = KERNEL_VERSION(3,4,0))
+   destp = kmap_atomic(ppages[i]);
++#else
++  destp = kmap_atomic(ppages[i], KM_SKB_SUNRPC_DATA);
++#endif
+   memcpy(destp + page_base, srcp, curlen);
+   flush_dcache_page(ppages[i]);
++#if (LINUX_VERSION_CODE = KERNEL_VERSION(3,4,0))
+   kunmap_atomic(destp);
++#else
++  kunmap_atomic(destp, KM_SKB_SUNRPC_DATA);
++#endif
+   srcp += curlen;
+   copy_len -= curlen;
+   if (copy_len == 0)
___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

[ewg] [PATCH] NFSRDMA SLES11SP2 backport

[Jeff Becker]

Hi Vlad. Please apply attached patch to compat-rdma/compat-rdma.git. Thanks.

-jeff
diff --git 
a/patches/0020-nfsrdma-Backports-for-RHEL6.2-RHEL6.3-and-SLES11-SP2.patch 
b/patches/0020-nfsrdma-Backports-for-RHEL6.2-RHEL6.3-and-SLES11-SP2.patch
new file mode 100644
index 000..4f79a56
--- /dev/null
+++ b/patches/0020-nfsrdma-Backports-for-RHEL6.2-RHEL6.3-and-SLES11-SP2.patch
@@ -0,0 +1,54 @@
+--- a/net/sunrpc/xprtrdma/svc_rdma_transport.c 2012-08-10 14:24:49.030166769 
-0700
 b/net/sunrpc/xprtrdma/svc_rdma_transport.c 2012-08-13 11:07:38.566980365 
-0700
+@@ -445,7 +445,11 @@
+ 
+   if (!cma_xprt)
+   return NULL;
++#if (LINUX_VERSION_CODE = KERNEL_VERSION(3,3,0))
+   svc_xprt_init(init_net, svc_rdma_class, cma_xprt-sc_xprt, serv);
++#else
++  svc_xprt_init(svc_rdma_class, cma_xprt-sc_xprt, serv);
++#endif
+   INIT_LIST_HEAD(cma_xprt-sc_accept_q);
+   INIT_LIST_HEAD(cma_xprt-sc_dto_q);
+   INIT_LIST_HEAD(cma_xprt-sc_rq_dto_q);
+
+--- a/net/sunrpc/xprtrdma/transport.c  2012-08-10 14:24:49.030166769 -0700
 b/net/sunrpc/xprtrdma/transport.c  2012-08-15 10:30:17.125653683 -0700
+@@ -283,7 +283,9 @@
+   }
+ 
+   xprt = xprt_alloc(args-net, sizeof(struct rpcrdma_xprt),
++#if (LINUX_VERSION_CODE = KERNEL_VERSION(3,1,0))
+   xprt_rdma_slot_table_entries,
++#endif
+   xprt_rdma_slot_table_entries);
+   if (xprt == NULL) {
+   dprintk(RPC:   %s: couldn't allocate rpcrdma_xprt\n,
+@@ -453,8 +455,14 @@
+ }
+ 
+ static int
++#if (LINUX_VERSION_CODE = KERNEL_VERSION(3,1,0))
+ xprt_rdma_reserve_xprt(struct rpc_xprt *xprt, struct rpc_task *task)
+ {
++#else
++xprt_rdma_reserve_xprt(struct rpc_task *task)
++{
++  struct rpc_xprt *xprt = task-tk_xprt;
++#endif
+   struct rpcrdma_xprt *r_xprt = rpcx_to_rdmax(xprt);
+   int credits = atomic_read(r_xprt-rx_buf.rb_credits);
+ 
+@@ -466,7 +474,11 @@
+   BUG_ON(r_xprt-rx_buf.rb_cwndscale = 0);
+   }
+   xprt-cwnd = credits * r_xprt-rx_buf.rb_cwndscale;
++#if (LINUX_VERSION_CODE = KERNEL_VERSION(3,1,0))
+   return xprt_reserve_xprt_cong(xprt, task);
++#else
++  return xprt_reserve_xprt_cong(task);
++#endif
+ }
+ 
+ /*
___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [ewg] ipoib build failures ... RE: [GIT PULL OFED-3.5] qib/ipath backporting

[Jeff Becker]

Hi Tom.

On 08/01/2012 11:08 AM, Elken, Tom wrote:

Vlad , and others on the list,

Thanks for merging Mike's backports.

Has anyone else seen build failures with IPoIB on RHEL 6.2 with the daily 
OFED-3.5 builds?
  -- Both Dean (CCed, a new Intel driver engineer) and Mike have (details on 
the failure in the P.S. below).


I saw the same build failure against kernel.org 3.2.


If so, are there patches in progress by anyone?


Vlad told me he's working on it.

-jeff


Just wondering, partly because we are close to the target date for release of 
RC1.

Regards,
Tom

Details 

Another build failure.  This time ulp/ipoib.

==
/var/tmp/OFED_topdir/BUILD/compat-rdma-
3.5/drivers/infiniband/ulp/ipoib/ipoib_ma
in.c: In function 'ipoib_start_xmit':
/var/tmp/OFED_topdir/BUILD/compat-rdma-
3.5/drivers/infiniband/ulp/ipoib/ipoib_main.c:718: error: implicit
declaration of function 'dst_get_neighbour_noref'
/var/tmp/OFED_topdir/BUILD/compat-rdma-
3.5/drivers/infiniband/ulp/ipoib/ipoib_main.c:718: warning: assignment
makes pointer from integer without a cast
/var/tmp/OFED_topdir/BUILD/compat-rdma-
3.5/drivers/infiniband/ulp/ipoib/ipoib_main.c: At top level:
/var/tmp/OFED_topdir/BUILD/compat-rdma-
3.5/drivers/infiniband/ulp/ipoib/ipoib_main.c:989: error: unknown
field 'ndo_fix_features' specified in initializer
/var/tmp/OFED_topdir/BUILD/compat-rdma-
3.5/drivers/infiniband/ulp/ipoib/ipoib_main.c:989: warning:
initialization from incompatible pointer type
/var/tmp/OFED_topdir/BUILD/compat-rdma-
3.5/drivers/infiniband/ulp/ipoib/ipoib_main.c: In function
'ipoib_set_dev_features':
/var/tmp/OFED_topdir/BUILD/compat-rdma-
3.5/drivers/infiniband/ulp/ipoib/ipoib_main.c:1170: error: 'struct net_device'
has no member named 'hw_features'
/var/tmp/OFED_topdir/BUILD/compat-rdma-
3.5/drivers/infiniband/ulp/ipoib/ipoib_main.c:1174: error: 'struct net_device'
has no member named 'hw_features'
/var/tmp/OFED_topdir/BUILD/compat-rdma-
3.5/drivers/infiniband/ulp/ipoib/ipoib_main.c:1176: error: 'struct net_device'
has no member named 'hw_features'
/var/tmp/OFED_topdir/BUILD/compat-rdma-
3.5/drivers/infiniband/ulp/ipoib/ipoib_main.c: In function 'ipoib_add_port':
/var/tmp/OFED_topdir/BUILD/compat-rdma-
3.5/drivers/infiniband/ulp/ipoib/ipoib_main.c:1208: error: 'struct net_device'
has no member named 'neigh_priv_len'
make[4]: *** [/var/tmp/OFED_topdir/BUILD/compat-rdma-
3.5/drivers/infiniband/ulp/ipoib/ipoib_main.o] Error 1 ==

-Tom



-Original Message-
From: Vladimir Sokolovsky [mailto:v...@dev.mellanox.co.il]
Sent: Wednesday, August 01, 2012 6:54 AM
To: Marciniszyn, Mike
Cc: ewg@lists.openfabrics.org; Elken, Tom; Mascarenhas, Edward
Subject: Re: [ewg] [GIT PULL OFED-3.5] qib/ipath backporting

Hi Mike,
I merged your backports.
Please send your git URL with the pull request next time.

Regards,
Vladimir

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg


___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [ewg] [PATCH] NFSRDMA: Makefile fix

[Jeff Becker]

Hi Steve.

On 07/09/2012 06:54 AM, Steve Wise wrote:

Hey Jeff,

Can you tell us the status of NFSRDMA in OFED-3.x?  IE what backports, 
if any, are complete, and what testing have you done on various distros?


I was waiting for the 3.5 trees to get set up so I'm just getting 
started. I intend to get RHEL 6.2 and 6.3 tested, as well as SLES 11 
SP2. I'll post successful test results (i.e., cthon 04 passes) as I 
achieve them. Thanks.


-jeff



Thanks,

Steve.


On 7/5/2012 11:40 AM, Jeff Becker wrote:
Hi Vlad. Please apply the attached patch to compat-rdma. It alllows 
NFSRDMA to build correctly. Thanks.


-jeff


___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg





___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

[ewg] [PATCH] NFSRDMA: Makefile fix

[Jeff Becker]

Hi Vlad. Please apply the attached patch to compat-rdma. It alllows 
NFSRDMA to build correctly. Thanks.


-jeff
From 59e0fae9bfbc100fb5a10677cdccc1f67fb5ce7a Mon Sep 17 00:00:00 2001
From: Jeffrey C. Becker jeffrey.c.bec...@nasa.gov
Date: Thu, 5 Jul 2012 09:24:41 -0700
Subject: [PATCH] NFSRDMA: only build sunrpc/xprtrdma
 Signed-off-by: Jeff Becker jeffrey.c.bec...@nasa.gov

---
 Makefile |7 ---
 1 files changed, 0 insertions(+), 7 deletions(-)

diff --git a/Makefile b/Makefile
index 945ffc0..e6e07da 100644
--- a/Makefile
+++ b/Makefile
@@ -4,11 +4,4 @@ obj-$(CONFIG_CHELSIO_T3)+= 
drivers/net/ethernet/chelsio/cxgb3/
 obj-$(CONFIG_CHELSIO_T4)+= drivers/net/ethernet/chelsio/cxgb4/
 obj-$(CONFIG_MLX4_CORE) += drivers/net/ethernet/mellanox/mlx4/
 obj-$(CONFIG_RDS)   += net/rds/
-obj-$(CONFIG_SUNRPC_XPRT_RDMA)  += net/sunrpc/
-obj-$(CONFIG_SUNRPC_XPRT_RDMA)  += net/sunrpc/auth_gss/
 obj-$(CONFIG_SUNRPC_XPRT_RDMA)  += net/sunrpc/xprtrdma/
-obj-$(CONFIG_SUNRPC_XPRT_RDMA)  += fs/nfs/
-obj-$(CONFIG_SUNRPC_XPRT_RDMA)  += fs/lockd/
-obj-$(CONFIG_SUNRPC_XPRT_RDMA)  += fs/exportfs/
-obj-$(CONFIG_SUNRPC_XPRT_RDMA)  += fs/nfs_common/
-obj-$(CONFIG_SUNRPC_XPRT_RDMA)  += fs/nfsd/
-- 
1.7.1

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

[ewg] Please apply attached patch to ofed_kernel_1_5

[Jeff Becker]

It fixes an NFSRDMA performance issue. Thanks.

-jeff
From dea5ead8be2c42e0a7d793190bbc65e760f9b0f0 Mon Sep 17 00:00:00 2001
From: Jeff Becker jeffrey.c.bec...@nasa.gov
Date: Mon, 5 Dec 2011 16:09:59 -0800
Subject: [PATCH] NFSRDMA: Increasing RPCRDMA_MAX_DATA_SEGS sometimes causes NFS mount to
 hang.
 Signed-off-by: Jeff Becker jeffrey.c.bec...@nasa.gov

---
 ...013-RDMA-Increasing-RPCRDMA_MAX_DATA_SEGS.patch |   31 
 1 files changed, 0 insertions(+), 31 deletions(-)
 delete mode 100644 kernel_patches/fixes/nfsrdma_0013-RDMA-Increasing-RPCRDMA_MAX_DATA_SEGS.patch

diff --git a/kernel_patches/fixes/nfsrdma_0013-RDMA-Increasing-RPCRDMA_MAX_DATA_SEGS.patch b/kernel_patches/fixes/nfsrdma_0013-RDMA-Increasing-RPCRDMA_MAX_DATA_SEGS.patch
deleted file mode 100644
index 2674c86..000
--- a/kernel_patches/fixes/nfsrdma_0013-RDMA-Increasing-RPCRDMA_MAX_DATA_SEGS.patch
+++ /dev/null
@@ -1,31 +0,0 @@
-From 2773395b34883fe54418de188733a63bb38e0ad6 Mon Sep 17 00:00:00 2001
-From: Steve Dickson ste...@redhat.com
-Date: Thu, 21 Jul 2011 13:49:02 -0400
-Subject: [PATCH] RDMA: Increasing RPCRDMA_MAX_DATA_SEGS
-
-Our performance team has noticed that increasing
-RPCRDMA_MAX_DATA_SEGS from 8 to 64 significantly
-increases throughput when using the RDMA transport.
-
-Signed-off-by: Steve Dickson ste...@redhat.com
-Signed-off-by: Trond Myklebust trond.mykleb...@netapp.com

- net/sunrpc/xprtrdma/xprt_rdma.h |2 +-
- 1 files changed, 1 insertions(+), 1 deletions(-)
-
-diff --git a/net/sunrpc/xprtrdma/xprt_rdma.h b/net/sunrpc/xprtrdma/xprt_rdma.h
-index cae761a..5d1cfe5 100644
 a/net/sunrpc/xprtrdma/xprt_rdma.h
-+++ b/net/sunrpc/xprtrdma/xprt_rdma.h
-@@ -109,7 +109,7 @@ struct rpcrdma_ep {
-  */
- 
- /* temporary static scatter/gather max */
--#define RPCRDMA_MAX_DATA_SEGS	(8)	/* max scatter/gather */
-+#define RPCRDMA_MAX_DATA_SEGS	(64)	/* max scatter/gather */
- #define RPCRDMA_MAX_SEGS 	(RPCRDMA_MAX_DATA_SEGS + 2) /* head+tail = 2 */
- #define MAX_RPCRDMAHDR	(\
- 	/* max supported RPC/RDMA header */ \
--- 
-1.7.3.4
-
-- 
1.7.3.4

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

[ewg] NFS/RDMA release notes.

[Jeff Becker]

Hi Vlad. Please apply attached patch to ofed_1_5/docs. Thanks.

-jeff
From 9df98d45089e0e06046cc757572b6336e1fcebc7 Mon Sep 17 00:00:00 2001
From: Jeff Becker jeffrey.c.bec...@nasa.gov
Date: Sun, 4 Dec 2011 11:29:39 -0800
Subject: [PATCH] NFSRDMA: release notes
 Signed-off-by: Jeff Becker jeffrey.c.bec...@nasa.gov

---
 release_notes/nfs-rdma.release-notes.txt |   20 +++-
 1 files changed, 11 insertions(+), 9 deletions(-)

diff --git a/release_notes/nfs-rdma.release-notes.txt b/release_notes/nfs-rdma.release-notes.txt
index ef00e66..7a933c7 100644
--- a/release_notes/nfs-rdma.release-notes.txt
+++ b/release_notes/nfs-rdma.release-notes.txt
@@ -9,6 +9,8 @@
  Adapted for OFED 1.5.1 (from linux-2.6.30/Documentation/filesystems/nfs-rdma.txt)
  by Jon Mason
 
+ Further adapted for OFED 1.5.4 by Jeff Becker
+
 Table of Contents
 ~
  - Overview
@@ -33,6 +35,13 @@ Overview
   the full Connectathon test suite and operates over both Infiniband and iWARP
   RDMA adapters.
 
+OFED 1.5.4 limitations:
+~
+  NFS/RDMA has been ported and tested on the following kernels:
+  - Redhat Enterprise Linux (RHEL) version 6.1
+  - SUSE Linux Enterprise Server (SLES) version 11 service pack 1
+  - kernel.org 2.6.30-2.6.38
+
 OFED 1.5.3 limitations:
 ~
   NFSoRDMA was removed from the list of packages in the install script to
@@ -82,10 +91,8 @@ Installation
 Testing has been performed using several Mellanox-based IB cards and 
 the Chelsio cxgb3 iWARP adapter.
 
-  - Install OFED 1.5.1
+  - Install OFED 1.5.4
 
-NFS/RDMA has been tested on RHEL5.2, RHEL 5.3, RHEL5.4, SLES11,
-kernels 2.6.22, 2.6.25, and 2.6.30.  On these kernels,
 NFS-RDMA will be installed by default if you simply select install all,
 and can be specifically included by a custom install.
 
@@ -211,12 +218,7 @@ NFS/RDMA Setup
 
 Mount the NFS/RDMA server:
 
-$ mount.rnfs IPoIB-server-name-or-address:/export /mnt -o proto=rdma,port=20049 
-
-NOTE: For kernels  2.6.23, the -i flag must be passed into mount.rnfs.
-This option allows the mount command to ignore the kernel version check.  If
-not disabled, the check will prevent passing arguments to the kernel and not
-allow the updated version of NFS to accept the rdma NFS option.
+$ mount -t nfs4 IPoIB-server-name-or-address:/ /mnt -o rdma,port=20049 
 
 To verify that the mount is using RDMA, run cat /proc/mounts and check
 the proto field for the given mount.
-- 
1.7.3.4

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [ewg] [ANNOUNCE] OFED-1.5.4-rc5 is available

[Jeff Becker]

Hi Vlad. I have yet to use rnfs-utils in any of my testing. Modern
versions of mount (at least the one that comes with SLES11SP1 and
RHEL6.1) support NFSRDMA, so I don't think rnfs-utils is needed anymore
(it provided an upstream version of mount for older OFED versions).
Maybe we should at least remove the dependency between ofed-kernel
nfsrdma and rnfs-utils, or remove it from OFED? What do you think? Thanks.

-jeff

On 12/01/11 06:01, Vladimir Sokolovsky wrote:
 Hi,
 OFED 1.5.4-rc5 is available.

 The tarball is available on:
 http://www.openfabrics.org/downloads/OFED/ofed-1.5.4/OFED-1.5.4-rc5.tgz

 To get BUILD_ID run ofed_info

 Please report any issues in bugzilla https://bugs.openfabrics.org/ for OFED 
 1.5.4

 Regards,
 Vladimir

 Main changes from RC4:
 --
 Updated packages:
 libcxgb4-1.2.0
 libnes-1.1.3
 mstflint-1.4-1.18.g1adcfbf
 perftest-1.3.0-0.56.gd6f6093

 MPI:
 mvapich2-1.7-5.2011svn5031

 Supported Platforms and Operating Systems
 -
   o   CPU architectures:
 - x86_64
 - x86
 - ppc64
 - ia64

   o   Linux Operating Systems:
 - RedHat EL5 up62.6.18-238.el5
 - RedHat EL5 up72.6.18-274.el5
 - RedHat EL62.6.32-71.el6
 - RedHat EL6.1  2.6.32-131.0.15.el6
 - Scientific Linux 6.1  2.6.32-131.0.15.el6
 - SLES10 SP42.6.16.60-0.85.1-smp
 - SLES112.6.27.19-5-default
 - SLES11 SP12.6.32.12-0.7-default
 - kernel.org2.6.30 - 3.0






___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [ewg] EWG/OFED meeting minutes for Nov 29, 2011

[Jeff Becker]

Hi Tziporet

On 12/01/11 09:42, Tziporet Koren wrote:
 Meeting minutes for EWG/OFED meeting for Nov 29, 2011:
 

 1. OFED 1.5.4:

 We have decide that all issues are not critical except for the compilation 
 issue
 Rc5 - Thu Dec 1 - was released by Vlad
 GA - next Mon - Dec 5


I still need to do the NFSRDMA ports for kernel.org 2.6.39 and 3.0. I
don't envision any issues getting these done before Monday. However, I
am seeing a performance issue in one of the NFS Connectathon tests for
2.6.37, and I've tracked it down to a slow MMAP call response on the
server. The test completes correctly however. I am currently debugging
this, but if it is more important, I can submit the remaining ports
(although the bug may affect them as well) and resume debugging. Please
advise. Thanks.

-jeff
___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [ewg] EWG/OFED meeting minutes for Nov 29, 2011

[Jeff Becker]

Hi Woody

On 12/01/11 11:20, Woodruff, Robert J wrote:
 Are these items absolutely required for OFED-1.5.4 ? Personally, I
 have some people that really need to have this release out ASAP and do not
 want to hold it up anymore if possible. The release has already slipped 
 a lot from our original plan.

So you would prefer to see the additional ports (since OFED is supposed
to be supported on those kernels)? Please clarify. Thanks.

-jeff

 woody

 -Original Message-
 From: ewg-boun...@lists.openfabrics.org 
 [mailto:ewg-boun...@lists.openfabrics.org] On Behalf Of Jeff Becker
 Sent: Thursday, December 01, 2011 11:12 AM
 To: Tziporet Koren
 Cc: ewg@lists.openfabrics.org
 Subject: Re: [ewg] EWG/OFED meeting minutes for Nov 29, 2011

 Hi Tziporet

 On 12/01/11 09:42, Tziporet Koren wrote:
 Meeting minutes for EWG/OFED meeting for Nov 29, 2011:
 

 1. OFED 1.5.4:

 We have decide that all issues are not critical except for the compilation 
 issue
 Rc5 - Thu Dec 1 - was released by Vlad
 GA - next Mon - Dec 5


 I still need to do the NFSRDMA ports for kernel.org 2.6.39 and 3.0. I
 don't envision any issues getting these done before Monday. However, I
 am seeing a performance issue in one of the NFS Connectathon tests for
 2.6.37, and I've tracked it down to a slow MMAP call response on the
 server. The test completes correctly however. I am currently debugging
 this, but if it is more important, I can submit the remaining ports
 (although the bug may affect them as well) and resume debugging. Please
 advise. Thanks.

 -jeff
 ___
 ewg mailing list
 ewg@lists.openfabrics.org
 http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [ewg] EWG/OFED meeting minutes for Nov 29, 2011

[Jeff Becker]

On 12/01/11 11:30, Woodruff, Robert J wrote:
 Well, we all agreed in the last EWG meeting that RC5 was to 
 be the last RC, so if those backports are not in RC5, they will have to
 wait for the next release. 

OK. I will work on the performance issue then. When I fix the release
notes I'll mention only kernel.org 2.6.30-38 is supported. Is that
alright with everyone? Thanks.

-jeff


 -Original Message-
 From: Jeff Becker [mailto:jeffrey.c.bec...@nasa.gov] 
 Sent: Thursday, December 01, 2011 11:27 AM
 To: Woodruff, Robert J
 Cc: Tziporet Koren; ewg@lists.openfabrics.org
 Subject: Re: [ewg] EWG/OFED meeting minutes for Nov 29, 2011

 Hi Woody

 On 12/01/11 11:20, Woodruff, Robert J wrote:
 Are these items absolutely required for OFED-1.5.4 ? Personally, I
 have some people that really need to have this release out ASAP and do not
 want to hold it up anymore if possible. The release has already slipped 
 a lot from our original plan.
 So you would prefer to see the additional ports (since OFED is supposed
 to be supported on those kernels)? Please clarify. Thanks.

 -jeff
 woody

 -Original Message-
 From: ewg-boun...@lists.openfabrics.org 
 [mailto:ewg-boun...@lists.openfabrics.org] On Behalf Of Jeff Becker
 Sent: Thursday, December 01, 2011 11:12 AM
 To: Tziporet Koren
 Cc: ewg@lists.openfabrics.org
 Subject: Re: [ewg] EWG/OFED meeting minutes for Nov 29, 2011

 Hi Tziporet

 On 12/01/11 09:42, Tziporet Koren wrote:
 Meeting minutes for EWG/OFED meeting for Nov 29, 2011:
 

 1. OFED 1.5.4:

 We have decide that all issues are not critical except for the compilation 
 issue
 Rc5 - Thu Dec 1 - was released by Vlad
 GA - next Mon - Dec 5


 I still need to do the NFSRDMA ports for kernel.org 2.6.39 and 3.0. I
 don't envision any issues getting these done before Monday. However, I
 am seeing a performance issue in one of the NFS Connectathon tests for
 2.6.37, and I've tracked it down to a slow MMAP call response on the
 server. The test completes correctly however. I am currently debugging
 this, but if it is more important, I can submit the remaining ports
 (although the bug may affect them as well) and resume debugging. Please
 advise. Thanks.

 -jeff
 ___
 ewg mailing list
 ewg@lists.openfabrics.org
 http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

[ewg] Please apply attached patch to ofed_kernel_1_5

[Jeff Becker]

It contains NFSRDMA port for 2.6.38. Thanks.

-jeff
From d35ba9ddcc3231ec2a104d9fea6e0aa017ea92d9 Mon Sep 17 00:00:00 2001
From: Jeff Becker jeffrey.c.bec...@nasa.gov
Date: Mon, 28 Nov 2011 11:18:26 -0800
Subject: [PATCH] NFSRDMA: Pull in patches for 2.6.38
 Signed-off-by: Jeff Becker jeffrey.c.bec...@nasa.gov

---
 ...ve-duplicate-universal-address-generation.patch |   42 +
 ...rpc_ntop-for-constructing-transport-addre.patch |   74 +
 .../nfsrdma-0003-SUNRPC-Kill-RPC_DISPLAY_ALL.patch |   28 +++
 ...move-struct-file-argument-of-proc_handler.patch |   37 
 ...5-headers-remove-sched.h-from-interrupt.h.patch |   23 +++
 ...sctl-net-Remove-unused-binary-sysctl-code.patch |  168 +++
 ...ysctl-Drop-in-front-of-every-proc_handler.patch |  174 
 ...008-sunrpc-Clean-never-used-include-files.patch |   32 
 ...entralise-most-calls-to-svc_xprt_received.patch |   61 +++
 ...010-sunrpc-Remove-uses-of-NIPQUAD-use-pI4.patch |   28 +++
 ...-sunrpc-Convert-void-snprintf-to-snprintf.patch |   38 +
 ...-the-test-for-XPRT_CONNECTING-into-xprt_c.patch |   49 ++
 ...-Fail-over-more-quickly-on-connect-errors.patch |   25 +++
 ...-the-task-tk_bytes_sent-and-tk_rtt-to-str.patch |   24 +++
 ...anup-Update-gfp.h-and-slab.h-includes-to-.patch |   66 
 ...-sunrpc-xprtrdma-clean-up-workqueue-usage.patch |  135 +++
 ...017-sunrpc-Factor-out-rpc_xprt-allocation.patch |   38 +
 ...a-0018-sunrpc-Factor-out-rpc_xprt-freeing.patch |   33 
 ...nfsrdma-0019-sunrpc-Tag-rpc_xprt-with-net.patch |   25 +++
 ...ull-net-argument-downto-svc_create_socket.patch |   33 
 20 files changed, 1133 insertions(+), 0 deletions(-)
 create mode 100644 kernel_patches/backport/2.6.38/nfsrdma-0001-SUNRPC-Remove-duplicate-universal-address-generation.patch
 create mode 100644 kernel_patches/backport/2.6.38/nfsrdma-0002-SUNRPC-Use-rpc_ntop-for-constructing-transport-addre.patch
 create mode 100644 kernel_patches/backport/2.6.38/nfsrdma-0003-SUNRPC-Kill-RPC_DISPLAY_ALL.patch
 create mode 100644 kernel_patches/backport/2.6.38/nfsrdma-0004-sysctl-remove-struct-file-argument-of-proc_handler.patch
 create mode 100644 kernel_patches/backport/2.6.38/nfsrdma-0005-headers-remove-sched.h-from-interrupt.h.patch
 create mode 100644 kernel_patches/backport/2.6.38/nfsrdma-0006-sysctl-net-Remove-unused-binary-sysctl-code.patch
 create mode 100644 kernel_patches/backport/2.6.38/nfsrdma-0007-sysctl-Drop-in-front-of-every-proc_handler.patch
 create mode 100644 kernel_patches/backport/2.6.38/nfsrdma-0008-sunrpc-Clean-never-used-include-files.patch
 create mode 100644 kernel_patches/backport/2.6.38/nfsrdma-0009-sunrpc-centralise-most-calls-to-svc_xprt_received.patch
 create mode 100644 kernel_patches/backport/2.6.38/nfsrdma-0010-sunrpc-Remove-uses-of-NIPQUAD-use-pI4.patch
 create mode 100644 kernel_patches/backport/2.6.38/nfsrdma-0011-sunrpc-Convert-void-snprintf-to-snprintf.patch
 create mode 100644 kernel_patches/backport/2.6.38/nfsrdma-0012-SUNRPC-Move-the-test-for-XPRT_CONNECTING-into-xprt_c.patch
 create mode 100644 kernel_patches/backport/2.6.38/nfsrdma-0013-SUNRPC-Fail-over-more-quickly-on-connect-errors.patch
 create mode 100644 kernel_patches/backport/2.6.38/nfsrdma-0014-SUNRPC-Move-the-task-tk_bytes_sent-and-tk_rtt-to-str.patch
 create mode 100644 kernel_patches/backport/2.6.38/nfsrdma-0015-include-cleanup-Update-gfp.h-and-slab.h-includes-to-.patch
 create mode 100644 kernel_patches/backport/2.6.38/nfsrdma-0016-sunrpc-xprtrdma-clean-up-workqueue-usage.patch
 create mode 100644 kernel_patches/backport/2.6.38/nfsrdma-0017-sunrpc-Factor-out-rpc_xprt-allocation.patch
 create mode 100644 kernel_patches/backport/2.6.38/nfsrdma-0018-sunrpc-Factor-out-rpc_xprt-freeing.patch
 create mode 100644 kernel_patches/backport/2.6.38/nfsrdma-0019-sunrpc-Tag-rpc_xprt-with-net.patch
 create mode 100644 kernel_patches/backport/2.6.38/nfsrdma-0020-sunrpc-Pull-net-argument-downto-svc_create_socket.patch

diff --git a/kernel_patches/backport/2.6.38/nfsrdma-0001-SUNRPC-Remove-duplicate-universal-address-generation.patch b/kernel_patches/backport/2.6.38/nfsrdma-0001-SUNRPC-Remove-duplicate-universal-address-generation.patch
new file mode 100644
index 000..de9c3a8
--- /dev/null
+++ b/kernel_patches/backport/2.6.38/nfsrdma-0001-SUNRPC-Remove-duplicate-universal-address-generation.patch
@@ -0,0 +1,42 @@
+From ba809130bc260fce04141aca01ef9e068d32af2a Mon Sep 17 00:00:00 2001
+From: Chuck Lever chuck.le...@oracle.com
+Date: Sun, 9 Aug 2009 15:09:35 -0400
+Subject: [PATCH] SUNRPC: Remove duplicate universal address generation
+
+RPC universal address generation is currently done in several places:
+rpcb_clnt.c, nfs4proc.c xprtsock.c, and xprtrdma.c.  Remove the
+redundant cases that convert a socket address to a universal
+address.  The nfs4proc.c case takes a pre-formatted presentation
+address string, not a socket address, so we'll leave that one.
+
+Because the new uaddr constructor uses the recently

[ewg] Please apply attached patches for ofed_kernel_1_5

[Jeff Becker]

They contain NFSRDMA fixes for 2.6.37. Thanks.

-jeff
From 4145755c7cfe4b1d2ad16e140bafb21f8286ec50 Mon Sep 17 00:00:00 2001
From: Jeff Becker jeffrey.c.bec...@nasa.gov
Date: Mon, 21 Nov 2011 17:29:54 -0800
Subject: [PATCH 5/5] NFSRDMA: And yet another patch for 2.6.37
 Signed-off-by: Jeff Becker jeffrey.c.bec...@nasa.gov

---
 ...ull-net-argument-downto-svc_create_socket.patch |   33 
 1 files changed, 33 insertions(+), 0 deletions(-)
 create mode 100644 kernel_patches/backport/2.6.37/nfsrdma-0020-sunrpc-Pull-net-argument-downto-svc_create_socket.patch

diff --git a/kernel_patches/backport/2.6.37/nfsrdma-0020-sunrpc-Pull-net-argument-downto-svc_create_socket.patch b/kernel_patches/backport/2.6.37/nfsrdma-0020-sunrpc-Pull-net-argument-downto-svc_create_socket.patch
new file mode 100644
index 000..4ca1e81
--- /dev/null
+++ b/kernel_patches/backport/2.6.37/nfsrdma-0020-sunrpc-Pull-net-argument-downto-svc_create_socket.patch
@@ -0,0 +1,33 @@
+From 62832c039eab9d03cd28a66427ce8276988f28b0 Mon Sep 17 00:00:00 2001
+From: Pavel Emelyanov xe...@parallels.com
+Date: Wed, 29 Sep 2010 16:04:18 +0400
+Subject: [PATCH] sunrpc: Pull net argument downto svc_create_socket
+
+After this the socket creation in it knows the context.
+
+Signed-off-by: Pavel Emelyanov xe...@openvz.org
+Signed-off-by: J. Bruce Fields bfie...@redhat.com
+---
+diff --git a/net/sunrpc/xprtrdma/svc_rdma_transport.c b/net/sunrpc/xprtrdma/svc_rdma_transport.c
+index edea15a..950a206 100644
+--- a/net/sunrpc/xprtrdma/svc_rdma_transport.c
 b/net/sunrpc/xprtrdma/svc_rdma_transport.c
+@@ -52,6 +52,7 @@
+ #define RPCDBG_FACILITY	RPCDBG_SVCXPRT
+ 
+ static struct svc_xprt *svc_rdma_create(struct svc_serv *serv,
++	struct net *net,
+ 	struct sockaddr *sa, int salen,
+ 	int flags);
+ static struct svc_xprt *svc_rdma_accept(struct svc_xprt *xprt);
+@@ -670,6 +671,7 @@ static int rdma_cma_handler(struct rdma_cm_id *cma_id,
+  * Create a listening RDMA service endpoint.
+  */
+ static struct svc_xprt *svc_rdma_create(struct svc_serv *serv,
++	struct net *net,
+ 	struct sockaddr *sa, int salen,
+ 	int flags)
+ {
+-- 
+1.7.3.4
+
-- 
1.7.3.4

From 83845d49329ccad931962ebd5dfd221e39f398cb Mon Sep 17 00:00:00 2001
From: Jeff Becker jeffrey.c.bec...@nasa.gov
Date: Mon, 21 Nov 2011 17:20:15 -0800
Subject: [PATCH 4/5] NFSRDMA: forgot a patch for 2.6.37
 Signed-off-by: Jeff Becker jeffrey.c.bec...@nasa.gov

---
 ...nfsrdma-0019-sunrpc-Tag-rpc_xprt-with-net.patch |   25 
 1 files changed, 25 insertions(+), 0 deletions(-)
 create mode 100644 kernel_patches/backport/2.6.37/nfsrdma-0019-sunrpc-Tag-rpc_xprt-with-net.patch

diff --git a/kernel_patches/backport/2.6.37/nfsrdma-0019-sunrpc-Tag-rpc_xprt-with-net.patch b/kernel_patches/backport/2.6.37/nfsrdma-0019-sunrpc-Tag-rpc_xprt-with-net.patch
new file mode 100644
index 000..6b43edc
--- /dev/null
+++ b/kernel_patches/backport/2.6.37/nfsrdma-0019-sunrpc-Tag-rpc_xprt-with-net.patch
@@ -0,0 +1,25 @@
+From 37aa2133731d9231eb834f700119f0d3f1ed2664 Mon Sep 17 00:00:00 2001
+From: Pavel Emelyanov xe...@parallels.com
+Date: Wed, 29 Sep 2010 16:05:43 +0400
+Subject: [PATCH] sunrpc: Tag rpc_xprt with net
+
+The net is known from the xprt_create and this tagging will also
+give un the context in the conntection workers where real sockets
+are created.
+
+Signed-off-by: Pavel Emelyanov xe...@openvz.org
+Signed-off-by: J. Bruce Fields bfie...@redhat.com
+---
+diff --git a/net/sunrpc/xprtrdma/transport.c b/net/sunrpc/xprtrdma/transport.c
+index 0f7a1b9..2da32b4 100644
+--- a/net/sunrpc/xprtrdma/transport.c
 b/net/sunrpc/xprtrdma/transport.c
+@@ -283,7 +283,7 @@ xprt_setup_rdma(struct xprt_create *args)
+ 		return ERR_PTR(-EBADF);
+ 	}
+ 
+-	xprt = xprt_alloc(sizeof(struct rpcrdma_xprt),
++	xprt = xprt_alloc(args-net, sizeof(struct rpcrdma_xprt),
+ 			xprt_rdma_slot_table_entries);
+ 	if (xprt == NULL) {
+ 		dprintk(RPC:   %s: couldn't allocate rpcrdma_xprt\n,
-- 
1.7.3.4

From a5e436d6ebd78fbed9ed64fdc21b1f01c54936b6 Mon Sep 17 00:00:00 2001
From: Jeff Becker jeffrey.c.bec...@nasa.gov
Date: Mon, 21 Nov 2011 17:05:24 -0800
Subject: [PATCH 3/5] NFSRDMA: Still a few more patches for 2.6.37
 Signed-off-by: Jeff Becker jeffrey.c.bec...@nasa.gov

---
 ...017-sunrpc-Factor-out-rpc_xprt-allocation.patch |   38 
 ...a-0018-sunrpc-Factor-out-rpc_xprt-freeing.patch |   33 +
 2 files changed, 71 insertions(+), 0 deletions(-)
 create mode 100644 kernel_patches/backport/2.6.37/nfsrdma-0017-sunrpc-Factor-out-rpc_xprt-allocation.patch
 create mode 100644 kernel_patches/backport/2.6.37/nfsrdma-0018-sunrpc-Factor-out-rpc_xprt-freeing.patch

diff --git a/kernel_patches/backport/2.6.37/nfsrdma-0017-sunrpc-Factor-out-rpc_xprt-allocation.patch b/kernel_patches/backport/2.6.37/nfsrdma-0017-sunrpc-Factor-out-rpc_xprt-allocation.patch
new file mode 100644
index 000..b21a33c
--- /dev/null
+++ b/kernel_patches/backport/2.6.37/nfsrdma-0017

[ewg] Please apply attached patches to ofed_kernel_1_5

[Jeff Becker]

They comprise NFSRDMA fixes for 2.6.35. Thanks.

-jeff
From 88b5fb066cdeddb7e0057a43ed44635eb4e31745 Mon Sep 17 00:00:00 2001
From: Jeff Becker jeffrey.c.bec...@nasa.gov
Date: Wed, 9 Nov 2011 14:16:19 -0800
Subject: [PATCH] NFSRDMA: A few more patches for 2.6.35
 Signed-off-by: Jeff Becker jeffrey.c.bec...@nasa.gov

---
 ...-the-test-for-XPRT_CONNECTING-into-xprt_c.patch |   49 
 ...entralise-most-calls-to-svc_xprt_received.patch |   61 
 2 files changed, 110 insertions(+), 0 deletions(-)
 create mode 100644 
kernel_patches/backport/2.6.35/nfsrdma-0004-SUNRPC-Move-the-test-for-XPRT_CONNECTING-into-xprt_c.patch
 create mode 100644 
kernel_patches/backport/2.6.35/nfsrdma-sunrpc-centralise-most-calls-to-svc_xprt_received.patch

diff --git 
a/kernel_patches/backport/2.6.35/nfsrdma-0004-SUNRPC-Move-the-test-for-XPRT_CONNECTING-into-xprt_c.patch
 
b/kernel_patches/backport/2.6.35/nfsrdma-0004-SUNRPC-Move-the-test-for-XPRT_CONNECTING-into-xprt_c.patch
new file mode 100644
index 000..5a1d1ea
--- /dev/null
+++ 
b/kernel_patches/backport/2.6.35/nfsrdma-0004-SUNRPC-Move-the-test-for-XPRT_CONNECTING-into-xprt_c.patch
@@ -0,0 +1,49 @@
+From 0b9e79431377df452348e78262dd5a3dc359eeef Mon Sep 17 00:00:00 2001
+From: Trond Myklebust trond.mykleb...@netapp.com
+Date: Fri, 16 Apr 2010 16:41:57 -0400
+Subject: [PATCH] SUNRPC: Move the test for XPRT_CONNECTING into xprt_connect()
+
+This fixes a bug with setting xprt-stat.connect_start.
+
+Reviewed-by: Chuck Lever chuck.le...@oracle.com
+Signed-off-by: Trond Myklebust trond.mykleb...@netapp.com
+---
+diff --git a/net/sunrpc/xprtrdma/transport.c b/net/sunrpc/xprtrdma/transport.c
+index 187257b..0607b9a 100644
+--- a/net/sunrpc/xprtrdma/transport.c
 b/net/sunrpc/xprtrdma/transport.c
+@@ -449,21 +449,19 @@ xprt_rdma_connect(struct rpc_task *task)
+   struct rpc_xprt *xprt = (struct rpc_xprt *)task-tk_xprt;
+   struct rpcrdma_xprt *r_xprt = rpcx_to_rdmax(xprt);
+ 
+-  if (!xprt_test_and_set_connecting(xprt)) {
+-  if (r_xprt-rx_ep.rep_connected != 0) {
+-  /* Reconnect */
+-  schedule_delayed_work(r_xprt-rdma_connect,
+-  xprt-reestablish_timeout);
+-  xprt-reestablish_timeout = 1;
+-  if (xprt-reestablish_timeout  (30 * HZ))
+-  xprt-reestablish_timeout = (30 * HZ);
+-  else if (xprt-reestablish_timeout  (5 * HZ))
+-  xprt-reestablish_timeout = (5 * HZ);
+-  } else {
+-  schedule_delayed_work(r_xprt-rdma_connect, 0);
+-  if (!RPC_IS_ASYNC(task))
+-  flush_scheduled_work();
+-  }
++  if (r_xprt-rx_ep.rep_connected != 0) {
++  /* Reconnect */
++  schedule_delayed_work(r_xprt-rdma_connect,
++  xprt-reestablish_timeout);
++  xprt-reestablish_timeout = 1;
++  if (xprt-reestablish_timeout  (30 * HZ))
++  xprt-reestablish_timeout = (30 * HZ);
++  else if (xprt-reestablish_timeout  (5 * HZ))
++  xprt-reestablish_timeout = (5 * HZ);
++  } else {
++  schedule_delayed_work(r_xprt-rdma_connect, 0);
++  if (!RPC_IS_ASYNC(task))
++  flush_scheduled_work();
+   }
+ }
+ 
diff --git 
a/kernel_patches/backport/2.6.35/nfsrdma-sunrpc-centralise-most-calls-to-svc_xprt_received.patch
 
b/kernel_patches/backport/2.6.35/nfsrdma-sunrpc-centralise-most-calls-to-svc_xprt_received.patch
new file mode 100644
index 000..dea25f7
--- /dev/null
+++ 
b/kernel_patches/backport/2.6.35/nfsrdma-sunrpc-centralise-most-calls-to-svc_xprt_received.patch
@@ -0,0 +1,61 @@
+From b48fa6b99100dc7772af3cd276035fcec9719ceb Mon Sep 17 00:00:00 2001
+From: Neil Brown ne...@suse.de
+Date: Mon, 1 Mar 2010 16:51:14 +1100
+Subject: [PATCH] sunrpc: centralise most calls to svc_xprt_received
+
+svc_xprt_received must be called when -xpo_recvfrom has finished
+receiving a message, so that the XPT_BUSY flag will be cleared and
+if necessary, requeued for further work.
+
+This call is currently made in each -xpo_recvfrom function, often
+from multiple different points.  In each case it is the earliest point
+on a particular path where it is known that the protection provided by
+XPT_BUSY is no longer needed.
+
+However there are (still) some error paths which do not call
+svc_xprt_received, and requiring each -xpo_recvfrom to make the call
+does not encourage robustness.
+
+So: move the svc_xprt_received call to be made just after the
+call to -xpo_recvfrom(), and move it of the various -xpo_recvfrom
+methods.
+
+This means that it may not be called at the earliest possible instant,
+but this is unlikely to be a measurable performance issue.
+
+Note that there are still other calls to svc_xprt_received as it is
+also needed

[ewg] Please apply attached patch for ofed_kernel_1_5

[Jeff Becker]

It contains NFSRDMA fixes for 2.6.34. Thanks.

-jeff
From 510efe6af21226125d139008226014ce1673d9a2 Mon Sep 17 00:00:00 2001
From: Jeff Becker jeffrey.c.bec...@nasa.gov
Date: Thu, 3 Nov 2011 10:14:26 -0700
Subject: [PATCH] NFSRDMA: patches for 2.6.34
 Signed-off-by: Jeff Becker jeffrey.c.bec...@nasa.gov

---
 ...ve-duplicate-universal-address-generation.patch |   42 +
 ...rpc_ntop-for-constructing-transport-addre.patch |   74 +
 .../nfsrdma-0003-SUNRPC-Kill-RPC_DISPLAY_ALL.patch |   28 +++
 ...sctl-net-Remove-unused-binary-sysctl-code.patch |  168 +++
 ...ysctl-Drop-in-front-of-every-proc_handler.patch |  174 
 ...a-headers-remove-sched.h-from-interrupt.h.patch |   23 +++
 ...dma-sunrpc-Clean-never-used-include-files.patch |   32 
 ...move-struct-file-argument-of-proc_handler.patch |   37 
 8 files changed, 578 insertions(+), 0 deletions(-)
 create mode 100644 
kernel_patches/backport/2.6.34/nfsrdma-0001-SUNRPC-Remove-duplicate-universal-address-generation.patch
 create mode 100644 
kernel_patches/backport/2.6.34/nfsrdma-0002-SUNRPC-Use-rpc_ntop-for-constructing-transport-addre.patch
 create mode 100644 
kernel_patches/backport/2.6.34/nfsrdma-0003-SUNRPC-Kill-RPC_DISPLAY_ALL.patch
 create mode 100644 
kernel_patches/backport/2.6.34/nfsrdma-0004-sysctl-net-Remove-unused-binary-sysctl-code.patch
 create mode 100644 
kernel_patches/backport/2.6.34/nfsrdma-0005-sysctl-Drop-in-front-of-every-proc_handler.patch
 create mode 100644 
kernel_patches/backport/2.6.34/nfsrdma-headers-remove-sched.h-from-interrupt.h.patch
 create mode 100644 
kernel_patches/backport/2.6.34/nfsrdma-sunrpc-Clean-never-used-include-files.patch
 create mode 100644 
kernel_patches/backport/2.6.34/nfsrdma-sysctl-remove-struct-file-argument-of-proc_handler.patch

diff --git 
a/kernel_patches/backport/2.6.34/nfsrdma-0001-SUNRPC-Remove-duplicate-universal-address-generation.patch
 
b/kernel_patches/backport/2.6.34/nfsrdma-0001-SUNRPC-Remove-duplicate-universal-address-generation.patch
new file mode 100644
index 000..de9c3a8
--- /dev/null
+++ 
b/kernel_patches/backport/2.6.34/nfsrdma-0001-SUNRPC-Remove-duplicate-universal-address-generation.patch
@@ -0,0 +1,42 @@
+From ba809130bc260fce04141aca01ef9e068d32af2a Mon Sep 17 00:00:00 2001
+From: Chuck Lever chuck.le...@oracle.com
+Date: Sun, 9 Aug 2009 15:09:35 -0400
+Subject: [PATCH] SUNRPC: Remove duplicate universal address generation
+
+RPC universal address generation is currently done in several places:
+rpcb_clnt.c, nfs4proc.c xprtsock.c, and xprtrdma.c.  Remove the
+redundant cases that convert a socket address to a universal
+address.  The nfs4proc.c case takes a pre-formatted presentation
+address string, not a socket address, so we'll leave that one.
+
+Because the new uaddr constructor uses the recently introduced
+rpc_ntop(), it now supports proper :: shorthanding for IPv6
+addresses.  This allows the kernel to register properly formed
+universal addresses with the local rpcbind service, in _all_ cases.
+
+The kernel can now also send properly formed universal addresses in
+RPCB_GETADDR requests, and support link-local properly when
+encoding and decoding IPv6 addresses.
+
+Signed-off-by: Chuck Lever chuck.le...@oracle.com
+Signed-off-by: Trond Myklebust trond.mykleb...@netapp.com
+---
+diff --git a/net/sunrpc/xprtrdma/transport.c b/net/sunrpc/xprtrdma/transport.c
+index 1dd6123..537c210 100644
+--- a/net/sunrpc/xprtrdma/transport.c
 b/net/sunrpc/xprtrdma/transport.c
+@@ -202,14 +202,6 @@ xprt_rdma_format_addresses(struct rpc_xprt *xprt)
+   snprintf(buf, 8, %4hx, ntohs(addr-sin_port));
+   xprt-address_strings[RPC_DISPLAY_HEX_PORT] = buf;
+ 
+-  buf = kzalloc(30, GFP_KERNEL);
+-  if (buf)
+-  snprintf(buf, 30, %pI4.%u.%u,
+-  addr-sin_addr.s_addr,
+-  ntohs(addr-sin_port)  8,
+-  ntohs(addr-sin_port)  0xff);
+-  xprt-address_strings[RPC_DISPLAY_UNIVERSAL_ADDR] = buf;
+-
+   /* netid */
+   xprt-address_strings[RPC_DISPLAY_NETID] = rdma;
+ }
diff --git 
a/kernel_patches/backport/2.6.34/nfsrdma-0002-SUNRPC-Use-rpc_ntop-for-constructing-transport-addre.patch
 
b/kernel_patches/backport/2.6.34/nfsrdma-0002-SUNRPC-Use-rpc_ntop-for-constructing-transport-addre.patch
new file mode 100644
index 000..67d3f7b
--- /dev/null
+++ 
b/kernel_patches/backport/2.6.34/nfsrdma-0002-SUNRPC-Use-rpc_ntop-for-constructing-transport-addre.patch
@@ -0,0 +1,74 @@
+From c877b849d302d1275452af80b7221a2555dc02e1 Mon Sep 17 00:00:00 2001
+From: Chuck Lever chuck.le...@oracle.com
+Date: Sun, 9 Aug 2009 15:09:36 -0400
+Subject: [PATCH] SUNRPC: Use rpc_ntop() for constructing transport address 
strings
+
+Clean up:  In addition to using the new generic rpc_ntop() and
+rpc_get_port() functions, have the RPC client compute the presentation
+address buffer sizes dynamically using kstrdup().
+
+Signed-off-by: Chuck Lever chuck.le...@oracle.com

[ewg] Please apply attached patches to ofed_kernel_1_5

[Jeff Becker]

They fix NFSRDMA for 2.6.33. Thanks.

-jeff
From 241e76f31129fd2c7d19c0cca7a9a86268519500 Mon Sep 17 00:00:00 2001
From: Jeff Becker jeffrey.c.bec...@nasa.gov
Date: Thu, 27 Oct 2011 14:44:04 -0700
Subject: [PATCH 1/2] NFSRDMA: pull in patches for 2.6.33
 Signed-off-by: Jeff Becker jeffrey.c.bec...@nasa.gov

---
 ...ve-duplicate-universal-address-generation.patch |   42 +
 ...rpc_ntop-for-constructing-transport-addre.patch |   74 +
 .../nfsrdma-0003-SUNRPC-Kill-RPC_DISPLAY_ALL.patch |   28 +++
 ...sctl-net-Remove-unused-binary-sysctl-code.patch |  168 +++
 ...ysctl-Drop-in-front-of-every-proc_handler.patch |  174 
 ...dma-sunrpc-Clean-never-used-include-files.patch |   32 
 ...move-struct-file-argument-of-proc_handler.patch |   37 
 7 files changed, 555 insertions(+), 0 deletions(-)
 create mode 100644 
kernel_patches/backport/2.6.33/nfsrdma-0001-SUNRPC-Remove-duplicate-universal-address-generation.patch
 create mode 100644 
kernel_patches/backport/2.6.33/nfsrdma-0002-SUNRPC-Use-rpc_ntop-for-constructing-transport-addre.patch
 create mode 100644 
kernel_patches/backport/2.6.33/nfsrdma-0003-SUNRPC-Kill-RPC_DISPLAY_ALL.patch
 create mode 100644 
kernel_patches/backport/2.6.33/nfsrdma-0004-sysctl-net-Remove-unused-binary-sysctl-code.patch
 create mode 100644 
kernel_patches/backport/2.6.33/nfsrdma-0005-sysctl-Drop-in-front-of-every-proc_handler.patch
 create mode 100644 
kernel_patches/backport/2.6.33/nfsrdma-sunrpc-Clean-never-used-include-files.patch
 create mode 100644 
kernel_patches/backport/2.6.33/nfsrdma-sysctl-remove-struct-file-argument-of-proc_handler.patch

diff --git 
a/kernel_patches/backport/2.6.33/nfsrdma-0001-SUNRPC-Remove-duplicate-universal-address-generation.patch
 
b/kernel_patches/backport/2.6.33/nfsrdma-0001-SUNRPC-Remove-duplicate-universal-address-generation.patch
new file mode 100644
index 000..de9c3a8
--- /dev/null
+++ 
b/kernel_patches/backport/2.6.33/nfsrdma-0001-SUNRPC-Remove-duplicate-universal-address-generation.patch
@@ -0,0 +1,42 @@
+From ba809130bc260fce04141aca01ef9e068d32af2a Mon Sep 17 00:00:00 2001
+From: Chuck Lever chuck.le...@oracle.com
+Date: Sun, 9 Aug 2009 15:09:35 -0400
+Subject: [PATCH] SUNRPC: Remove duplicate universal address generation
+
+RPC universal address generation is currently done in several places:
+rpcb_clnt.c, nfs4proc.c xprtsock.c, and xprtrdma.c.  Remove the
+redundant cases that convert a socket address to a universal
+address.  The nfs4proc.c case takes a pre-formatted presentation
+address string, not a socket address, so we'll leave that one.
+
+Because the new uaddr constructor uses the recently introduced
+rpc_ntop(), it now supports proper :: shorthanding for IPv6
+addresses.  This allows the kernel to register properly formed
+universal addresses with the local rpcbind service, in _all_ cases.
+
+The kernel can now also send properly formed universal addresses in
+RPCB_GETADDR requests, and support link-local properly when
+encoding and decoding IPv6 addresses.
+
+Signed-off-by: Chuck Lever chuck.le...@oracle.com
+Signed-off-by: Trond Myklebust trond.mykleb...@netapp.com
+---
+diff --git a/net/sunrpc/xprtrdma/transport.c b/net/sunrpc/xprtrdma/transport.c
+index 1dd6123..537c210 100644
+--- a/net/sunrpc/xprtrdma/transport.c
 b/net/sunrpc/xprtrdma/transport.c
+@@ -202,14 +202,6 @@ xprt_rdma_format_addresses(struct rpc_xprt *xprt)
+   snprintf(buf, 8, %4hx, ntohs(addr-sin_port));
+   xprt-address_strings[RPC_DISPLAY_HEX_PORT] = buf;
+ 
+-  buf = kzalloc(30, GFP_KERNEL);
+-  if (buf)
+-  snprintf(buf, 30, %pI4.%u.%u,
+-  addr-sin_addr.s_addr,
+-  ntohs(addr-sin_port)  8,
+-  ntohs(addr-sin_port)  0xff);
+-  xprt-address_strings[RPC_DISPLAY_UNIVERSAL_ADDR] = buf;
+-
+   /* netid */
+   xprt-address_strings[RPC_DISPLAY_NETID] = rdma;
+ }
diff --git 
a/kernel_patches/backport/2.6.33/nfsrdma-0002-SUNRPC-Use-rpc_ntop-for-constructing-transport-addre.patch
 
b/kernel_patches/backport/2.6.33/nfsrdma-0002-SUNRPC-Use-rpc_ntop-for-constructing-transport-addre.patch
new file mode 100644
index 000..67d3f7b
--- /dev/null
+++ 
b/kernel_patches/backport/2.6.33/nfsrdma-0002-SUNRPC-Use-rpc_ntop-for-constructing-transport-addre.patch
@@ -0,0 +1,74 @@
+From c877b849d302d1275452af80b7221a2555dc02e1 Mon Sep 17 00:00:00 2001
+From: Chuck Lever chuck.le...@oracle.com
+Date: Sun, 9 Aug 2009 15:09:36 -0400
+Subject: [PATCH] SUNRPC: Use rpc_ntop() for constructing transport address 
strings
+
+Clean up:  In addition to using the new generic rpc_ntop() and
+rpc_get_port() functions, have the RPC client compute the presentation
+address buffer sizes dynamically using kstrdup().
+
+Signed-off-by: Chuck Lever chuck.le...@oracle.com
+Signed-off-by: Trond Myklebust trond.mykleb...@netapp.com
+---
+diff --git a/net/sunrpc/xprtrdma/transport.c b/net/sunrpc/xprtrdma/transport.c
+index 537c210..5f9b867

[ewg] OFED 1.5: kernel_patches/fixes/iw_cxgb3_1500_zbread_fix.patch fails to apply

[Jeff Becker]

Importing patch
/home/jcbecker/OFED/ofed_build/kernel_patches/fixes/iw_cxgb3_1500_zbread_fix.patch
(stored as iw_cxgb3_1500_zbread_fix.patch)
/usr/bin/quilt --quiltrc /home/jcbecker/OFED/ofed_build/patches/quiltrc
push patches/iw_cxgb3_1500_zbread_fix.patch
Applying patch iw_cxgb3_1500_zbread_fix.patch
patching file drivers/infiniband/hw/cxgb3/iwch_cm.c
Hunk #1 FAILED at 913.
Hunk #2 FAILED at 1077.
Hunk #3 FAILED at 1088.
3 out of 3 hunks FAILED -- rejects in file
drivers/infiniband/hw/cxgb3/iwch_cm.c
patching file drivers/infiniband/hw/cxgb3/iwch_provider.h
Hunk #1 FAILED at 332.
1 out of 1 hunk FAILED -- rejects in file
drivers/infiniband/hw/cxgb3/iwch_provider.h
patching file drivers/infiniband/hw/cxgb3/iwch_qp.c
Hunk #1 FAILED at 738.
Hunk #2 FAILED at 761.
2 out of 2 hunks FAILED -- rejects in file
drivers/infiniband/hw/cxgb3/iwch_qp.c
Patch iw_cxgb3_1500_zbread_fix.patch does not apply (enforce with -f)

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [ewg] EWG/OFED meeting Oct 17, 2011 - Meeting Minutes

[Jeff Becker]

On 10/17/11 09:38, Woodruff, Robert J wrote:
 Here are a few notes I took from today's meeting.

 We reviewed the OFED-1.5.4 feature list. 
 All of the features that were targeted at OFED-1.5.4 are now available.
 MVAPICH2 1.7 was release this weekend and will need to be pulled into the 
 builds.
 Patches have been submitted to allow OFED-1.5.4 to build on RHEL EL 6.2 beta.

Hi all. Just wanted to update NFSRDMA status. I have tested and
confirmed that it works on SLES 11 SP1 and kernel.org 2.6.30-32. I'm
currently tracking down a bug in RHEL 6.1 but I anticipate this should
be solved soon. When that is done, I will work on the remaining targeted
kernel.org kernel ports. Thanks.

-jeff

 Rupert from the interop group requested that we release an RC2 this week so 
 it 
 can be used for the interop event that starts at the end of the week. 

 People should start pulling and testing the RCs on their hardware and filing
 any bugs found in bugzilla. We will start to review the blocker bugs starting 
 in the 
 next meeting. 

 OFED-2.0 upstream kernel work .

 Sean is working with Roland to get the XRC patches upstream. Most likely
 Roland will add these to his tree to be included in kernel.org 3.2. Roland
 found a couple bugs in the patch that Sean is fixing. 

 There are some other minor patches that were identified, such as raw QP 
 support that
 are in OFED but not yet upstream. 


___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

[ewg] Please apply attached patch for ofed_kernel_1_5

[Jeff Becker]

It contains the NFSRDMA fixes for SLES11 SP1. Thanks.

-jeff
diff --git 
a/kernel_patches/backport/2.6.32_sles11sp1/mlx4_semaphore_include.patch 
b/kernel_patches/backport/2.6.32_sles11sp1/mlx4_semaphore_include.patch
index f2141f1..011d364 100644
--- a/kernel_patches/backport/2.6.32_sles11sp1/mlx4_semaphore_include.patch
+++ b/kernel_patches/backport/2.6.32_sles11sp1/mlx4_semaphore_include.patch
@@ -1,14 +1,14 @@
-From: Sven-Thorsten Dietrich sdietr...@suse.de
-Subject: Include semaphore.h explicitly.
-
-Signed-off-by: Sven-Thorsten Dietrich sdietr...@suse.de
 a/drivers/net/mlx4/cmd.c   2012-09-01 10:14:58.0 -0700
-+++ b/drivers/net/mlx4//cmd.c  2012-09-01 13:19:46.0 -0700
-@@ -35,6 +35,7 @@
- #include linux/sched.h
- #include linux/pci.h
- #include linux/errno.h
-+#include linux/semaphore.h
- 
- #include linux/mlx4/cmd.h
- 
+From: Sven-Thorsten Dietrich sdietr...@suse.de
+Subject: Include semaphore.h explicitly.
+
+Signed-off-by: Sven-Thorsten Dietrich sdietr...@suse.de
+--- a/drivers/net/mlx4/cmd.c   2012-09-01 10:14:58.0 -0700
 b/drivers/net/mlx4//cmd.c  2012-09-01 13:19:46.0 -0700
+@@ -35,6 +35,7 @@
+ #include linux/sched.h
+ #include linux/pci.h
+ #include linux/errno.h
++#include linux/semaphore.h
+ 
+ #include linux/mlx4/cmd.h
+ 
diff --git 
a/kernel_patches/backport/2.6.32_sles11sp1/net_skb-dst_accessors.patch 
b/kernel_patches/backport/2.6.32_sles11sp1/net_skb-dst_accessors.patch
index 36d612b..c742dad 100644
--- a/kernel_patches/backport/2.6.32_sles11sp1/net_skb-dst_accessors.patch
+++ b/kernel_patches/backport/2.6.32_sles11sp1/net_skb-dst_accessors.patch
@@ -1,149 +1,149 @@
-From: Sven-Thorsten Dietrich sdietr...@suse.de
-Subject: Forward-port new accessor for net: dst.
-
-See Kernel.org:
-commit adf30907d63893e4208dfe3f5c88ae12bc2f25d5
-Author: Eric Dumazet eric.duma...@gmail.com
-Date:   Tue Jun 2 05:19:30 2009 +
-Subject: net: skb-dst accessors
-
-Signed-off-by: Sven-Thorsten Dietrich sdietr...@suse.de
 a/drivers/infiniband/ulp/ipoib/ipoib_cm.c
-+++ b/drivers/infiniband/ulp/ipoib/ipoib_cm.c
-@@ -1394,8 +1394,8 @@ void ipoib_cm_skb_too_long(struct net_device *dev, 
struct sk_buff *skb,
-   struct ipoib_dev_priv *priv = netdev_priv(dev);
-   int e = skb_queue_empty(priv-cm.skb_queue);
- 
--  if (skb-dst)
--  skb-dst-ops-update_pmtu(skb-dst, mtu);
-+  if (skb_dst(skb))
-+  skb_dst(skb)-ops-update_pmtu(skb_dst(skb), mtu);
- 
-   skb_queue_tail(priv-cm.skb_queue, skb);
-   if (e)
 a/drivers/infiniband/ulp/ipoib/ipoib_main.c
-+++ b/drivers/infiniband/ulp/ipoib/ipoib_main.c
-@@ -561,7 +561,7 @@ static void neigh_add_path(struct sk_buff *skb, struct 
net_device *dev)
-   struct ipoib_neigh *neigh;
-   unsigned long flags;
- 
--  neigh = ipoib_neigh_alloc(skb-dst-neighbour, skb-dev);
-+  neigh = ipoib_neigh_alloc(skb_dst(skb)-neighbour, skb-dev);
-   if (!neigh) {
-   ++dev-stats.tx_dropped;
-   dev_kfree_skb_any(skb);
-@@ -570,9 +570,9 @@ static void neigh_add_path(struct sk_buff *skb, struct 
net_device *dev)
- 
-   spin_lock_irqsave(priv-lock, flags);
- 
--  path = __path_find(dev, skb-dst-neighbour-ha + 4);
-+  path = __path_find(dev, skb_dst(skb)-neighbour-ha + 4);
-   if (!path) {
--  path = path_rec_create(dev, skb-dst-neighbour-ha + 4);
-+  path = path_rec_create(dev, skb_dst(skb)-neighbour-ha + 4);
-   if (!path)
-   goto err_path;
- 
-@@ -605,7 +605,7 @@ static void neigh_add_path(struct sk_buff *skb, struct 
net_device *dev)
-   goto err_drop;
-   }
-   } else
--  ipoib_send(dev, skb, path-ah, 
IPOIB_QPN(skb-dst-neighbour-ha));
-+  ipoib_send(dev, skb, path-ah, 
IPOIB_QPN(skb_dst(skb)-neighbour-ha));
-   } else {
-   neigh-ah  = NULL;
- 
-@@ -635,15 +635,15 @@ static void ipoib_path_lookup(struct sk_buff *skb, 
struct net_device *dev)
-   struct ipoib_dev_priv *priv = netdev_priv(skb-dev);
- 
-   /* Look up path record for unicasts */
--  if (skb-dst-neighbour-ha[4] != 0xff) {
-+  if (skb_dst(skb)-neighbour-ha[4] != 0xff) {
-   neigh_add_path(skb, dev);
-   return;
-   }
- 
-   /* Add in the P_Key for multicasts */
--  skb-dst-neighbour-ha[8] = (priv-pkey  8)  0xff;
--  skb-dst-neighbour-ha[9] = priv-pkey  0xff;
--  ipoib_mcast_send(dev, skb-dst-neighbour-ha + 4, skb);
-+  skb_dst(skb)-neighbour-ha[8] = (priv-pkey  8)  0xff;
-+  skb_dst(skb)-neighbour-ha[9] = priv-pkey  0xff;
-+  ipoib_mcast_send(dev, skb_dst(skb)-neighbour-ha + 4, skb);
- }
- 
- static void unicast_arp_send(struct sk_buff *skb, struct net_device *dev,
-@@ -708,16 +708,16 @@ static int ipoib_start_xmit(struct sk_buff *skb, struct 
net_device *dev)
-   struct ipoib_neigh *neigh;
-   unsigned long flags;
- 
--  if 

[ewg] Please apply attached patches for ofed_kernel_1_5

[Jeff Becker]

NFSRDMA now builds/runs on 2.6.32. Thanks.

-jeff
From a7e736a75223d343c3ef92e4f505202c17e49a82 Mon Sep 17 00:00:00 2001
From: Jeff Becker jeffrey.c.bec...@nasa.gov
Date: Tue, 6 Sep 2011 12:42:58 -0700
Subject: [PATCH 2/2] NFSRDMA: fixes for 2.6.32
 Signed-off-by: Jeff Becker jeffrey.c.bec...@nasa.gov

---
 .../backport/2.6.32/nfsrdma_to_2_6_32.patch|   77 
 1 files changed, 77 insertions(+), 0 deletions(-)
 create mode 100644 kernel_patches/backport/2.6.32/nfsrdma_to_2_6_32.patch

diff --git a/kernel_patches/backport/2.6.32/nfsrdma_to_2_6_32.patch 
b/kernel_patches/backport/2.6.32/nfsrdma_to_2_6_32.patch
new file mode 100644
index 000..f4f83e5
--- /dev/null
+++ b/kernel_patches/backport/2.6.32/nfsrdma_to_2_6_32.patch
@@ -0,0 +1,77 @@
+diff --git a/net/sunrpc/xprtrdma/svc_rdma_transport.c 
b/net/sunrpc/xprtrdma/svc_rdma_transport.c
+index 5151f9f..2df5734 100644
+--- a/net/sunrpc/xprtrdma/svc_rdma_transport.c
 b/net/sunrpc/xprtrdma/svc_rdma_transport.c
+@@ -42,6 +42,7 @@
+ #include linux/sunrpc/svc_xprt.h
+ #include linux/sunrpc/debug.h
+ #include linux/sunrpc/rpc_rdma.h
++#include linux/sched.h
+ #include linux/spinlock.h
+ #include rdma/ib_verbs.h
+ #include rdma/rdma_cm.h
+diff --git a/net/sunrpc/xprtrdma/transport.c b/net/sunrpc/xprtrdma/transport.c
+index 1dd6123..9a63f66 100644
+--- a/net/sunrpc/xprtrdma/transport.c
 b/net/sunrpc/xprtrdma/transport.c
+@@ -168,47 +168,25 @@ static struct rpc_xprt_ops xprt_rdma_procs;  /* 
forward reference */
+ static void
+ xprt_rdma_format_addresses(struct rpc_xprt *xprt)
+ {
+-  struct sockaddr_in *addr = (struct sockaddr_in *)
++  struct sockaddr *sap = (struct sockaddr *)
+   rpcx_to_rdmad(xprt).addr;
+-  char *buf;
++  struct sockaddr_in *sin = (struct sockaddr_in *)sap;
++  char buf[64];
+ 
+-  buf = kzalloc(20, GFP_KERNEL);
+-  if (buf)
+-  snprintf(buf, 20, %pI4, addr-sin_addr.s_addr);
+-  xprt-address_strings[RPC_DISPLAY_ADDR] = buf;
++  (void)rpc_ntop(sap, buf, sizeof(buf));
++  xprt-address_strings[RPC_DISPLAY_ADDR] = kstrdup(buf, GFP_KERNEL);
+ 
+-  buf = kzalloc(8, GFP_KERNEL);
+-  if (buf)
+-  snprintf(buf, 8, %u, ntohs(addr-sin_port));
+-  xprt-address_strings[RPC_DISPLAY_PORT] = buf;
++  (void)snprintf(buf, sizeof(buf), %u, rpc_get_port(sap));
++  xprt-address_strings[RPC_DISPLAY_PORT] = kstrdup(buf, GFP_KERNEL);
+ 
+   xprt-address_strings[RPC_DISPLAY_PROTO] = rdma;
+ 
+-  buf = kzalloc(48, GFP_KERNEL);
+-  if (buf)
+-  snprintf(buf, 48, addr=%pI4 port=%u proto=%s,
+-  addr-sin_addr.s_addr,
+-  ntohs(addr-sin_port), rdma);
+-  xprt-address_strings[RPC_DISPLAY_ALL] = buf;
+-
+-  buf = kzalloc(10, GFP_KERNEL);
+-  if (buf)
+-  snprintf(buf, 10, %02x%02x%02x%02x,
+-  NIPQUAD(addr-sin_addr.s_addr));
+-  xprt-address_strings[RPC_DISPLAY_HEX_ADDR] = buf;
+-
+-  buf = kzalloc(8, GFP_KERNEL);
+-  if (buf)
+-  snprintf(buf, 8, %4hx, ntohs(addr-sin_port));
+-  xprt-address_strings[RPC_DISPLAY_HEX_PORT] = buf;
+-
+-  buf = kzalloc(30, GFP_KERNEL);
+-  if (buf)
+-  snprintf(buf, 30, %pI4.%u.%u,
+-  addr-sin_addr.s_addr,
+-  ntohs(addr-sin_port)  8,
+-  ntohs(addr-sin_port)  0xff);
+-  xprt-address_strings[RPC_DISPLAY_UNIVERSAL_ADDR] = buf;
++  (void)snprintf(buf, sizeof(buf), %02x%02x%02x%02x,
++  NIPQUAD(sin-sin_addr.s_addr));
++  xprt-address_strings[RPC_DISPLAY_HEX_ADDR] = kstrdup(buf, GFP_KERNEL);
++
++  (void)snprintf(buf, sizeof(buf), %4hx, rpc_get_port(sap));
++  xprt-address_strings[RPC_DISPLAY_HEX_PORT] = kstrdup(buf, GFP_KERNEL);
+ 
+   /* netid */
+   xprt-address_strings[RPC_DISPLAY_NETID] = rdma;
-- 
1.7.3.4

From dec6453797550afa086617e27ea77ffd6b196d99 Mon Sep 17 00:00:00 2001
From: Jeff Becker jeffrey.c.bec...@nasa.gov
Date: Tue, 6 Sep 2011 09:43:38 -0700
Subject: [PATCH 1/2] Remove carriage returns from DOS formatted patches so they 
can apply
 Signed-off-by: Jeff Becker jeffrey.c.bec...@nasa.gov

---
 .../backport/2.6.32/mlx4_semaphore_include.patch   |   28 +-
 .../backport/2.6.32/net_skb-dst_accessors.patch|  298 ++--
 .../backport/2.6.32/new_frags_interface.patch  |  262 +-
 3 files changed, 294 insertions(+), 294 deletions(-)

diff --git a/kernel_patches/backport/2.6.32/mlx4_semaphore_include.patch 
b/kernel_patches/backport/2.6.32/mlx4_semaphore_include.patch
index f2141f1..011d364 100644
--- a/kernel_patches/backport/2.6.32/mlx4_semaphore_include.patch
+++ b/kernel_patches/backport/2.6.32/mlx4_semaphore_include.patch
@@ -1,14 +1,14 @@
-From: Sven-Thorsten Dietrich sdietr...@suse.de
-Subject: Include semaphore.h explicitly.
-
-Signed-off-by: Sven-Thorsten Dietrich sdietr

[ewg] Please apply attached patch

[Jeff Becker]

It fixes some more DOS formatted patches that could not be applied for
2.6.31. Thanks.

-jeff
From c114c8b2066beaa37822e9d3c0e3098f370a2cb7 Mon Sep 17 00:00:00 2001
From: Jeff Becker jeffrey.c.bec...@nasa.gov
Date: Mon, 22 Aug 2011 14:02:52 -0700
Subject: [PATCH 10/10] Remove carriage returns from DOS formatted patches.
 Signed-off-by: Jeff Becker jeffrey.c.bec...@nasa.gov

---
 .../backport/2.6.31/mlx4_semaphore_include.patch   |   28 +-
 .../backport/2.6.31/net_skb-dst_accessors.patch|  298 ++--
 .../backport/2.6.31/new_frags_interface.patch  |  262 +-
 3 files changed, 294 insertions(+), 294 deletions(-)

diff --git a/kernel_patches/backport/2.6.31/mlx4_semaphore_include.patch 
b/kernel_patches/backport/2.6.31/mlx4_semaphore_include.patch
index f2141f1..011d364 100644
--- a/kernel_patches/backport/2.6.31/mlx4_semaphore_include.patch
+++ b/kernel_patches/backport/2.6.31/mlx4_semaphore_include.patch
@@ -1,14 +1,14 @@
-From: Sven-Thorsten Dietrich sdietr...@suse.de
-Subject: Include semaphore.h explicitly.
-
-Signed-off-by: Sven-Thorsten Dietrich sdietr...@suse.de
 a/drivers/net/mlx4/cmd.c   2012-09-01 10:14:58.0 -0700
-+++ b/drivers/net/mlx4//cmd.c  2012-09-01 13:19:46.0 -0700
-@@ -35,6 +35,7 @@
- #include linux/sched.h
- #include linux/pci.h
- #include linux/errno.h
-+#include linux/semaphore.h
- 
- #include linux/mlx4/cmd.h
- 
+From: Sven-Thorsten Dietrich sdietr...@suse.de
+Subject: Include semaphore.h explicitly.
+
+Signed-off-by: Sven-Thorsten Dietrich sdietr...@suse.de
+--- a/drivers/net/mlx4/cmd.c   2012-09-01 10:14:58.0 -0700
 b/drivers/net/mlx4//cmd.c  2012-09-01 13:19:46.0 -0700
+@@ -35,6 +35,7 @@
+ #include linux/sched.h
+ #include linux/pci.h
+ #include linux/errno.h
++#include linux/semaphore.h
+ 
+ #include linux/mlx4/cmd.h
+ 
diff --git a/kernel_patches/backport/2.6.31/net_skb-dst_accessors.patch 
b/kernel_patches/backport/2.6.31/net_skb-dst_accessors.patch
index 36d612b..c742dad 100644
--- a/kernel_patches/backport/2.6.31/net_skb-dst_accessors.patch
+++ b/kernel_patches/backport/2.6.31/net_skb-dst_accessors.patch
@@ -1,149 +1,149 @@
-From: Sven-Thorsten Dietrich sdietr...@suse.de
-Subject: Forward-port new accessor for net: dst.
-
-See Kernel.org:
-commit adf30907d63893e4208dfe3f5c88ae12bc2f25d5
-Author: Eric Dumazet eric.duma...@gmail.com
-Date:   Tue Jun 2 05:19:30 2009 +
-Subject: net: skb-dst accessors
-
-Signed-off-by: Sven-Thorsten Dietrich sdietr...@suse.de
 a/drivers/infiniband/ulp/ipoib/ipoib_cm.c
-+++ b/drivers/infiniband/ulp/ipoib/ipoib_cm.c
-@@ -1394,8 +1394,8 @@ void ipoib_cm_skb_too_long(struct net_device *dev, 
struct sk_buff *skb,
-   struct ipoib_dev_priv *priv = netdev_priv(dev);
-   int e = skb_queue_empty(priv-cm.skb_queue);
- 
--  if (skb-dst)
--  skb-dst-ops-update_pmtu(skb-dst, mtu);
-+  if (skb_dst(skb))
-+  skb_dst(skb)-ops-update_pmtu(skb_dst(skb), mtu);
- 
-   skb_queue_tail(priv-cm.skb_queue, skb);
-   if (e)
 a/drivers/infiniband/ulp/ipoib/ipoib_main.c
-+++ b/drivers/infiniband/ulp/ipoib/ipoib_main.c
-@@ -561,7 +561,7 @@ static void neigh_add_path(struct sk_buff *skb, struct 
net_device *dev)
-   struct ipoib_neigh *neigh;
-   unsigned long flags;
- 
--  neigh = ipoib_neigh_alloc(skb-dst-neighbour, skb-dev);
-+  neigh = ipoib_neigh_alloc(skb_dst(skb)-neighbour, skb-dev);
-   if (!neigh) {
-   ++dev-stats.tx_dropped;
-   dev_kfree_skb_any(skb);
-@@ -570,9 +570,9 @@ static void neigh_add_path(struct sk_buff *skb, struct 
net_device *dev)
- 
-   spin_lock_irqsave(priv-lock, flags);
- 
--  path = __path_find(dev, skb-dst-neighbour-ha + 4);
-+  path = __path_find(dev, skb_dst(skb)-neighbour-ha + 4);
-   if (!path) {
--  path = path_rec_create(dev, skb-dst-neighbour-ha + 4);
-+  path = path_rec_create(dev, skb_dst(skb)-neighbour-ha + 4);
-   if (!path)
-   goto err_path;
- 
-@@ -605,7 +605,7 @@ static void neigh_add_path(struct sk_buff *skb, struct 
net_device *dev)
-   goto err_drop;
-   }
-   } else
--  ipoib_send(dev, skb, path-ah, 
IPOIB_QPN(skb-dst-neighbour-ha));
-+  ipoib_send(dev, skb, path-ah, 
IPOIB_QPN(skb_dst(skb)-neighbour-ha));
-   } else {
-   neigh-ah  = NULL;
- 
-@@ -635,15 +635,15 @@ static void ipoib_path_lookup(struct sk_buff *skb, 
struct net_device *dev)
-   struct ipoib_dev_priv *priv = netdev_priv(skb-dev);
- 
-   /* Look up path record for unicasts */
--  if (skb-dst-neighbour-ha[4] != 0xff) {
-+  if (skb_dst(skb)-neighbour-ha[4] != 0xff) {
-   neigh_add_path(skb, dev);
-   return;
-   }
- 
-   /* Add in the P_Key for multicasts */
--  skb-dst-neighbour-ha[8] = (priv-pkey  8)  0xff;
--  skb-dst-neighbour-ha[9] = priv-pkey

[ewg] 2.6.31 backport commit introduced dos formatted patches which cannot be applied

[Jeff Becker]

Hi Vlad. I started working on the NFSRDMA 2.6.31 port for OFED 1.5.4,
and noticed that the following commit contained patches with carriage
returns which prevents them from being applied.

Add 2.6.31 kernel support
authorSven-Thorsten Dietrich sdietr...@novell.com   
Wed, 2 Dec 2009 15:27:44 + (17:27 +0200)
committerVladimir Sokolovsky v...@mellanox.co.il   
Wed, 2 Dec 2009 15:27:44 + (17:27 +0200)
commitbc9e3a889fad30b18160ee72790c6fdef3e61831

These patches were then copied to several other backport directories
(e.g. 2.6.32, 2.6.33 etc.) which will then have the same problem. Can
you fix these or should I? Please let me know. Thanks.

-jeff
___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [ewg] ofa_1_5_kernel 20110809-0417 daily build status

[Jeff Becker]

Hi Vlad.

On 08/09/11 04:35, Vladimir Sokolovsky (Mellanox) wrote:
 This email was generated automatically, please do not reply


 git_url: git://git.openfabrics.org/ofed_1_5/linux-2.6.git
 git_branch: ofed_kernel_1_5

 Common build parameters:   --with-ipoib-mod --with-sdp-mod 
 --with-user_mad-mod --with-user_access-mod --with-mthca-mod --with-mlx4-mod 
 --with-mlx4_en-mod --with-core-mod --with-addr_trans-mod --with-rds-mod 
 --with-cxgb3-mod --with-srp-mod --with-nes-mod 

 Passed:
 Passed on i686 with 2.6.32-24-generic-pae
 Passed on i686 with 2.6.32-71.el6.i686
 Passed on i686 with linux-2.6.32
 Passed on i686 with linux-2.6.30
 Passed on i686 with linux-2.6.31
 Passed on i686 with linux-2.6.36
 Passed on i686 with linux-2.6.34
 Passed on i686 with linux-2.6.35
 Passed on i686 with linux-2.6.33
 Passed on x86_64 with linux-2.6.18-194.el5
 Passed on x86_64 with linux-2.6.18-164.el5
 Passed on x86_64 with linux-2.6.18-238.el5

 Failed:
 Build failed on x86_64 with linux-2.6.16.60-0.21-smp
 Log:
 /home/vlad/tmp/ofa_1_5_kernel-20110809-0417_linux-2.6.16.60-0.21-smp_x86_64_check/drivers/infiniband/core/addr.c:444:
  error: implicit declaration of function 'register_netevent_notifier'
 /home/vlad/tmp/ofa_1_5_kernel-20110809-0417_linux-2.6.16.60-0.21-smp_x86_64_check/drivers/infiniband/core/addr.c:
  In function 'addr_cleanup':
 /home/vlad/tmp/ofa_1_5_kernel-20110809-0417_linux-2.6.16.60-0.21-smp_x86_64_check/drivers/infiniband/core/addr.c:450:
  error: implicit declaration of function 'unregister_netevent_notifier'
 make[4]: *** 
 [/home/vlad/tmp/ofa_1_5_kernel-20110809-0417_linux-2.6.16.60-0.21-smp_x86_64_check/drivers/infiniband/core/addr.o]
  Error 1
 make[3]: *** 
 [/home/vlad/tmp/ofa_1_5_kernel-20110809-0417_linux-2.6.16.60-0.21-smp_x86_64_check/drivers/infiniband/core]
  Error 2
 make[2]: *** 
 [/home/vlad/tmp/ofa_1_5_kernel-20110809-0417_linux-2.6.16.60-0.21-smp_x86_64_check/drivers/infiniband]
  Error 2
 make[1]: *** 
 [_module_/home/vlad/tmp/ofa_1_5_kernel-20110809-0417_linux-2.6.16.60-0.21-smp_x86_64_check]
  Error 2
 make[1]: Leaving directory 
 `/home/vlad/kernel.org/x86_64/linux-2.6.16.60-0.21-smp'
 make: *** [kernel] Error 2
 --

The reason you get the error below is because your configure.mk.kernel has

CONFIG_SUNRPC_XPRT_RDMA=m.

When I try to configure with your config options against the same kernel
here, I get
CONFIG_SUNRPC_XPRT_RDMA=,

and net/sunrpc/xprtrdma is not built. This is expected since
--with-nfsrdma-mod is not set. So I'm not sure how
CONFIG_SUNRPC_XPRT_RDMA got set for you. Please check into it, and let
me know. Thanks.

-jeff

 Build failed on x86_64 with linux-2.6.16.60-0.54.5-smp
 Log:
 /home/vlad/tmp/ofa_1_5_kernel-20110809-0417_linux-2.6.16.60-0.54.5-smp_x86_64_check/net/sunrpc/xprtrdma/svc_rdma.c:248:
  error: implicit declaration of function 'svc_unreg_xprt_class'
 /home/vlad/tmp/ofa_1_5_kernel-20110809-0417_linux-2.6.16.60-0.54.5-smp_x86_64_check/net/sunrpc/xprtrdma/svc_rdma.c:
  In function 'svc_rdma_init':
 /home/vlad/tmp/ofa_1_5_kernel-20110809-0417_linux-2.6.16.60-0.54.5-smp_x86_64_check/net/sunrpc/xprtrdma/svc_rdma.c:255:
  error: 'RPCDBG_SVCXPRT' undeclared (first use in this function)
 /home/vlad/tmp/ofa_1_5_kernel-20110809-0417_linux-2.6.16.60-0.54.5-smp_x86_64_check/net/sunrpc/xprtrdma/svc_rdma.c:289:
  error: implicit declaration of function 'svc_reg_xprt_class'
 make[3]: *** 
 [/home/vlad/tmp/ofa_1_5_kernel-20110809-0417_linux-2.6.16.60-0.54.5-smp_x86_64_check/net/sunrpc/xprtrdma/svc_rdma.o]
  Error 1
 make[2]: *** 
 [/home/vlad/tmp/ofa_1_5_kernel-20110809-0417_linux-2.6.16.60-0.54.5-smp_x86_64_check/net/sunrpc/xprtrdma]
  Error 2
 make[1]: *** 
 [_module_/home/vlad/tmp/ofa_1_5_kernel-20110809-0417_linux-2.6.16.60-0.54.5-smp_x86_64_check]
  Error 2
 make[1]: Leaving directory 
 `/home/vlad/kernel.org/x86_64/linux-2.6.16.60-0.54.5-smp'
 make: *** [kernel] Error 2
 --
 Build failed on x86_64 with linux-2.6.32-131.0.15.el6
 Log:
 /home/vlad/tmp/ofa_1_5_kernel-20110809-0417_linux-2.6.32-131.0.15.el6_x86_64_check/drivers/infiniband/core/addr.c:
  At top level:
 /home/vlad/tmp/ofa_1_5_kernel-20110809-0417_linux-2.6.32-131.0.15.el6_x86_64_check/drivers/infiniband/core/addr.c:66:
  error: initializer element is not constant
 /home/vlad/tmp/ofa_1_5_kernel-20110809-0417_linux-2.6.32-131.0.15.el6_x86_64_check/drivers/infiniband/core/addr.c:66:
  error: (near initialization for 'lock.wait_lock')
 make[4]: *** 
 [/home/vlad/tmp/ofa_1_5_kernel-20110809-0417_linux-2.6.32-131.0.15.el6_x86_64_check/drivers/infiniband/core/addr.o]
  Error 1
 make[3]: *** 
 [/home/vlad/tmp/ofa_1_5_kernel-20110809-0417_linux-2.6.32-131.0.15.el6_x86_64_check/drivers/infiniband/core]
  Error 2
 make[2]: *** 
 [/home/vlad/tmp/ofa_1_5_kernel-20110809-0417_linux-2.6.32-131.0.15.el6_x86_64_check/drivers/infiniband]
  Error 2
 make[1]: *** 
 

Re: [ewg] Please apply attached patches to ofed_kernel_1_5

[Jeff Becker]

Hi Steve. I only tested these against 2.6.30 (OFED 1.5.4 base). I haven't 
gotten to backports yet. The main change is not to check out and build NFS, 
since that conflicted with the distros. Anyway, any help is appreciated. Thanks.

Jeff

Steve Wise sw...@opengridcomputing.com wrote:

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

[ewg] [PATCH] Malformed patch

[Jeff Becker]

Hi Vlad. The mlx4_x_0150_Fix_mlx4_ib_reg_xrc_rcv_qp_locking patch
appears to be in DOS format and does not apply. I fixed it with the
attached patch. Please apply. Thanks.

-jeff
From cc336a7859c4cbab4e7e91b471a1848f8bcb42e4 Mon Sep 17 00:00:00 2001
From: Jeff Becker jeffrey.c.bec...@nasa.gov
Date: Mon, 11 Jul 2011 10:11:13 -0700
Subject: [PATCH] Patch was formatted as DOS and did not apply - convert to UNIX 
format
 Signed-off-by: Jeff Becker jeffrey.c.bec...@nasa.gov

---
 ...x_0150_Fix_mlx4_ib_reg_xrc_rcv_qp_locking.patch |   82 ++--
 1 files changed, 41 insertions(+), 41 deletions(-)

diff --git 
a/kernel_patches/fixes/mlx4_x_0150_Fix_mlx4_ib_reg_xrc_rcv_qp_locking.patch 
b/kernel_patches/fixes/mlx4_x_0150_Fix_mlx4_ib_reg_xrc_rcv_qp_locking.patch
index 1ca3e71..0442493 100644
--- a/kernel_patches/fixes/mlx4_x_0150_Fix_mlx4_ib_reg_xrc_rcv_qp_locking.patch
+++ b/kernel_patches/fixes/mlx4_x_0150_Fix_mlx4_ib_reg_xrc_rcv_qp_locking.patch
@@ -1,41 +1,41 @@
-mlx4_ib XRC RCV: Fix mlx4_ib_reg_xrc_rcv_qp() locking
-
-In mlx4_ib_reg_xrc_rcv_qp(), we need to take the xrc_reg_list_lock spinlock
-when walking the xrc_reg_list.
-
-Also, I guess spin_lock_irqsave() could be replaced by spin_lock_irq() in
-that function as we know for sure we're in process context.
-
-Signed-off-by: Sébastien Dugué sebastien.du...@bull.net
-
---
-
- qp.c |3 +++
- 1 file changed, 3 insertions(+)
-
-dIndex: kernel-ib/drivers/infiniband/hw/mlx4/qp.c
-===
 kernel-ib.orig/drivers/infiniband/hw/mlx4/qp.c 2011-01-31 
16:52:11.0 +0100
-+++ kernel-ib/drivers/infiniband/hw/mlx4/qp.c  2011-02-11 15:24:27.0 
+0100
-@@ -2549,13 +2549,16 @@
-   }
- 
-   mutex_lock(mibqp-mutex);
-+  spin_lock_irqsave(mibqp-xrc_reg_list_lock, flags);
-   list_for_each_entry(tmp, mibqp-xrc_reg_list, list)
-   if (tmp-context == context) {
-+  spin_unlock_irqrestore(mibqp-xrc_reg_list_lock, 
flags);
-   mutex_unlock(mibqp-mutex);
-   kfree(ctx_entry);
-   mutex_unlock(to_mdev(xrcd-device)-xrc_reg_mutex);
-   return 0;
-   }
-+  spin_unlock_irqrestore(mibqp-xrc_reg_list_lock, flags);
- 
-   ctx_entry-context = context;
-   spin_lock_irqsave(mibqp-xrc_reg_list_lock, flags);
---
-To unsubscribe from this list: send the line unsubscribe linux-rdma in
-the body of a message to majord...@vger.kernel.org
-More majordomo info at  http://vger.kernel.org/majordomo-info.html
-
+mlx4_ib XRC RCV: Fix mlx4_ib_reg_xrc_rcv_qp() locking
+
+In mlx4_ib_reg_xrc_rcv_qp(), we need to take the xrc_reg_list_lock spinlock
+when walking the xrc_reg_list.
+
+Also, I guess spin_lock_irqsave() could be replaced by spin_lock_irq() in
+that function as we know for sure we're in process context.
+
+Signed-off-by: Sébastien Dugué sebastien.du...@bull.net
+
+--
+
+ qp.c |3 +++
+ 1 file changed, 3 insertions(+)
+
+dIndex: kernel-ib/drivers/infiniband/hw/mlx4/qp.c
+===
+--- kernel-ib.orig/drivers/infiniband/hw/mlx4/qp.c 2011-01-31 
16:52:11.0 +0100
 kernel-ib/drivers/infiniband/hw/mlx4/qp.c  2011-02-11 15:24:27.0 
+0100
+@@ -2549,13 +2549,16 @@
+   }
+ 
+   mutex_lock(mibqp-mutex);
++  spin_lock_irqsave(mibqp-xrc_reg_list_lock, flags);
+   list_for_each_entry(tmp, mibqp-xrc_reg_list, list)
+   if (tmp-context == context) {
++  spin_unlock_irqrestore(mibqp-xrc_reg_list_lock, 
flags);
+   mutex_unlock(mibqp-mutex);
+   kfree(ctx_entry);
+   mutex_unlock(to_mdev(xrcd-device)-xrc_reg_mutex);
+   return 0;
+   }
++  spin_unlock_irqrestore(mibqp-xrc_reg_list_lock, flags);
+ 
+   ctx_entry-context = context;
+   spin_lock_irqsave(mibqp-xrc_reg_list_lock, flags);
+--
+To unsubscribe from this list: send the line unsubscribe linux-rdma in
+the body of a message to majord...@vger.kernel.org
+More majordomo info at  http://vger.kernel.org/majordomo-info.html
+
-- 
1.7.3.4

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

[ewg] NFSRDMA patches for OFED 1.6

[Jeff Becker]

Hi Vlad. I cleaned up the tree a bit, and checked that my changes build
against 2.6.38. I also verified that NFS connectathon tests work. These
will be the basis for upcoming backports. Please apply the attached
patches (numbered) to OFED 1.6. Thanks.

-jeff

P.S. Sorry about the silly file names. They were picked by git format-patch.
From bc76477a624ad0551b4fca0b87931f0a70f1d92f Mon Sep 17 00:00:00 2001
From: Jeff Becker jeffrey.c.bec...@nasa.gov
Date: Thu, 12 May 2011 14:59:37 -0700
Subject: [PATCH 1/3] Initial NFSRDMA cleanup. We don't need to checkout all of 
nfs, only
 xprtrdma.

Committed-by: Jeff Becker jeffrey.c.bec...@nasa.gov
---
 ofed_scripts/checkout_files|7 +--
 ofed_scripts/nfs_header_checkout_files |   22 --
 ofed_scripts/ofed_checkout.sh  |   10 --
 3 files changed, 1 insertions(+), 38 deletions(-)
 delete mode 100644 ofed_scripts/nfs_header_checkout_files

diff --git a/ofed_scripts/checkout_files b/ofed_scripts/checkout_files
index 15de37f..232e030 100644
--- a/ofed_scripts/checkout_files
+++ b/ofed_scripts/checkout_files
@@ -9,11 +9,6 @@ drivers/scsi/iscsi_tcp.c
 drivers/scsi/iscsi_tcp.h
 drivers/scsi/libiscsi.c
 drivers/scsi/scsi_transport_iscsi.c
-fs/exportfs/
-fs/lockd/
-fs/nfs/
-fs/nfs_common/
-fs/nfsd/
 include/linux/mlx4/
 include/rdma/
 include/scsi/iscsi_if.h
@@ -28,5 +23,5 @@ lib/klist.c
 net/rds/
 include/linux/rds.h
 include/linux/socket.h
-net/sunrpc/
+net/sunrpc/xprtrdma
 ofed_scripts/
diff --git a/ofed_scripts/nfs_header_checkout_files 
b/ofed_scripts/nfs_header_checkout_files
deleted file mode 100644
index 29dd7ff..000
--- a/ofed_scripts/nfs_header_checkout_files
+++ /dev/null
@@ -1,22 +0,0 @@
-include/linux/exportfs.h
-include/linux/lockd/
-include/linux/nfs.h
-include/linux/nfs2.h
-include/linux/nfs3.h
-include/linux/nfs4.h
-include/linux/nfs4_acl.h
-include/linux/nfs4_mount.h
-include/linux/nfs_fs.h
-include/linux/nfs_fs_i.h
-include/linux/nfs_fs_sb.h
-include/linux/nfs_idmap.h
-include/linux/nfs_iostat.h
-include/linux/nfs_mount.h
-include/linux/nfs_page.h
-include/linux/nfs_xdr.h
-include/linux/nfsacl.h
-include/linux/nfsd/
-include/linux/nfsd_idmap.h
-include/linux/pipe_fs_i.h
-include/linux/splice.h
-include/linux/sunrpc/
diff --git a/ofed_scripts/ofed_checkout.sh b/ofed_scripts/ofed_checkout.sh
index 7766f98..fa91d08 100755
--- a/ofed_scripts/ofed_checkout.sh
+++ b/ofed_scripts/ofed_checkout.sh
@@ -29,16 +29,6 @@ xex()
 # branch defaults to ofed_kernel
 branch=${1:-ofed_kernel}
 
-#checkout nfs headerfiles and put them in their new location
-if [ ! -e ofed_scripts/nfs_header_checkout_files ]; then
-   ex git checkout ${branch} ofed_scripts/nfs_header_checkout_files
-fi
-
-xex xargs -rt git ls-tree -r --name-only ${branch} 
ofed_scripts/nfs_header_checkout_files |
-   xex xargs -rt git checkout ${branch}
-
-ex mv include include-nfs
-
 # make sure we have the list of all the other files and dirs
 if [ ! -e ofed_scripts/checkout_files ]; then
ex git checkout ${branch} ofed_scripts/checkout_files
-- 
1.7.3.4

From a975e42fa7bc043e2490088a4f3b0b1410762403 Mon Sep 17 00:00:00 2001
From: Jeff Becker jeffrey.c.bec...@nasa.gov
Date: Thu, 12 May 2011 16:10:22 -0700
Subject: [PATCH 2/3] NFSRDMA: forgot a few headers

Signed-off-by: Jeff Becker jeffrey.c.bec...@nasa.gov
---
 ofed_scripts/checkout_files |3 +++
 1 files changed, 3 insertions(+), 0 deletions(-)

diff --git a/ofed_scripts/checkout_files b/ofed_scripts/checkout_files
index 232e030..54e76ef 100644
--- a/ofed_scripts/checkout_files
+++ b/ofed_scripts/checkout_files
@@ -10,6 +10,9 @@ drivers/scsi/iscsi_tcp.h
 drivers/scsi/libiscsi.c
 drivers/scsi/scsi_transport_iscsi.c
 include/linux/mlx4/
+include/linux/sunrpc/rpc_rdma.h
+include/linux/sunrpc/svc_rdma.h
+include/linux/sunrpc/xprtrdma.h
 include/rdma/
 include/scsi/iscsi_if.h
 include/scsi/iscsi_proto.h
-- 
1.7.3.4

From 1bbe0ae80e6efbbb781412ed0d60acb108461709 Mon Sep 17 00:00:00 2001
From: Jeff Becker jeffrey.c.bec...@nasa.gov
Date: Thu, 12 May 2011 16:54:14 -0700
Subject: [PATCH 3/3] NFSRDMA: only need xprtrdma modules - remove stuff to 
build rest of nfs

Signed-off-by: Jeff Becker jeffrey.c.bec...@nasa.gov
---
 ofed_scripts/Makefile  |7 ---
 ofed_scripts/configure |   36 
 ofed_scripts/makefile  |   10 ++
 3 files changed, 2 insertions(+), 51 deletions(-)

diff --git a/ofed_scripts/Makefile b/ofed_scripts/Makefile
index f86eb61..9ec90a9 100644
--- a/ofed_scripts/Makefile
+++ b/ofed_scripts/Makefile
@@ -16,11 +16,4 @@ obj-$(CONFIG_CHELSIO_T4)+= drivers/net/cxgb4/
 obj-$(CONFIG_MLX4_CORE) += drivers/net/mlx4/
 obj-$(CONFIG_RDS)   += net/rds/
 obj-$(CONFIG_MEMTRACK)  += drivers/infiniband/debug/
-obj-$(CONFIG_SUNRPC_XPRT_RDMA)  += net/sunrpc/
-obj-$(CONFIG_SUNRPC_XPRT_RDMA)  += net/sunrpc/auth_gss/
 obj-$(CONFIG_SUNRPC_XPRT_RDMA)  += net/sunrpc/xprtrdma/
-obj-$(CONFIG_SUNRPC_XPRT_RDMA)  += fs/nfs

Re: [ewg] EWG/Meeting agenda for today - 28-Mar, 2011

[Jeff Becker]

On 03/28/11 09:14, Tziporet Koren wrote:
 EWG/OFED Agenda for today:

 1. OFED 1.6 schedule:
 --
 - Move to kernel 2.6.38 (since its GA already, and we have not started 
 backports)
 - Ongoing work on backports - during Q2
 - First RC - end of June
   RCs every 2 weeks
 - GA - End of Aug

 2. OFED 1.6 main features:
 
 - Mellanox: CX3 support
 - SRIOV support for mlx4 with CX2  CX3
 - FDR support
 - New OSes support: As usual the latest OSes will be supported
 - Remove MPI packages from OFED
 - Ne management package: Alex - please send details

 - All - please add more features you expect to be in

NFSRDMA will be supported with limited backports (RHEL6, SLES11sp1).

-jeff
 3. Open discussion



 Tziporet
 ___
 ewg mailing list
 ewg@lists.openfabrics.org
 http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [ewg] Power maintenance this weekend, service upgrade

[Jeff Becker]

 Hi. I would like to introduce Ken Strandberg as our new admin. I will
continue in a supporting role. In the meantime please remove
joh...@georgex.org  joh...@openfabrics.org ,dani...@georgex.org ,
dani...@openfabrics.org , i...@uchicago.edu  from your list of contacts.

On 11/01/10 17:06, JC Support wrote:
 Re: sofa.openfabrics.org

 Hi,

 This weekend we will be performing some power maintenance and
 infrastructure upgrades that will benefit your server. This upgrade will
 include the ability to securely power cycle your server from a remote
 location. In addition to our IP KVM, this will give you complete remote
 control over your server from anywhere in the world.

 We have two maintenance windows available for you to choose from:

 Saturday Nov 6th from 9pm to 12am Pacific time

 or

 Sunday Nov 7th from 9pm to 12am Pacific time

 Please respond and let us know which date you prefer. If we do not hear
 from you we will schedule your maintenance for Saturday.

That actually is best. Most of our developers are in Israel,  so their
work week starts on Sunday.

Thanks.

-jeff

 Total downtime during the maintenance will last approximately 30mins. We
 will use control-alt-delete to initiate a shutdown/reboot on your
 server. If you wish to perform a shutdown from your shell, please let us
 know if you'd like to coordinate a time. You can always power down in
 advance of the maintenance window as well. We will power your server back
 up at the conclusion of the maintenance.

 We beg your pardon for any inconvenience this may cause and we appreciate
 your patience as we bring these new features online and perform this


 -Dave

 JohnCompanies Support
 supp...@johncompanies.com


___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [ewg] Where is the latest GIT for opensm?

[Jeff Becker]

 Hi Jason.

On 10/06/10 19:33, Jason Gunthorpe wrote:
 Hey Sasha,

 Can you tell me where the latest git repository for OpenSM is these
 days? There doesn't seem to be anything recent on the OFA server,
 ie

 git://git.openfabrics.org/~sashak/management.git

 Is quite old.

Yes. The problem is that most of the developers have not yet moved to
the new server. For your example, I had to do:

git clone git://69.55.239.13/~sashak/management.git

to get the up to date repo. on the old server.

It would be great if developers could move their git repo's to the new
server so OFA can stop using (and paying) for the old server. Vlad sent
out a note about this earlier.

Thanks.

-jeff
 Thanks,
 Jason

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

[ewg] test

[Jeff Becker]

 Please ignore

-jeff
___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [ewg] Infiniband Interoperability

[Jeff Becker]

Hi Matt

On 06/29/10 10:15, Matt Breitbach wrote:
 So I know that this message isn't about starting a new group.  I've actually
 tried to join one of the mailing lists but it failed to sign me up.
   
You probably tried to sign up for gene...@lists.openfabrics.org, which
doesn't really exist anymore (although the archives are still there for
searching). It sounds like you should post your question on
e...@lists.openfabrics.org. I'll forward it for you.

Jeff Becker
OpenFabrics Server admin

 I'm working on getting an InfiniBand setup working in a mixed environment of
 Windows, Linux, and OpenSolaris.  I'm having huge difficulties getting the
 IB network to be stable.  We're working with IPoIB mainly, and get drops
 between the Windows and Linux/OpenSolaris systems.

 We had a professional take a look at our configuration, and he thought that
 OpenSM was configured properly, and actually had the network stable for
 about 2 days, but it degraded severely after that.  We are now to the point
 that most times the OpenSolaris box is unreachable from the Windows systems,
 and sometimes from the Linux system (which is running OpenSM).

 Is there any direction that you would be able to point me in to get some
 advice on this or some high-end consulting?  We've invested nearly 3 months
 into this project only to have a bladecenter with 4 IB Capable blades and a
 SuperMicro 4U server that can't reliably communicate over the InfiniBand
 network.

 -Matt Breitbach
 zfsbuild.com (our storage writeup about OpenSolaris ZFS over InfiniBand)


   

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [ewg] Updates to OFA Web site

[Jeff Becker]

Hi all. In conjunction with the update below, we will slowly begin
migrating to sofa.openfabrics.org as our server (finally :-) ). I have
already set up gitweb, so http://sofa.openfabrics.org/gitweb shows all
the git trees on sofa. It might be a good idea to start migrating your
trees over there. Also, Vlad, perhaps we should start doing the nightly
builds there.

In addition, I have set up mediawiki
(http://sofa.openfabrics.org/mediawiki) so if you want to add wiki
content (there's nothing there yet), go ahead, but please coordinate
with Cheri at Owen Media, as their team is looking at moving our old
wiki content over. In particular, tell them about out-of-date content we
don't want moved.

Thanks, and let me know if you have any questions/issues.

-jeff

On 06/23/10 18:52, Cheri Winterberg wrote:

 EWG – the OFA is in the early stages of updating several elements of
 the OFA Web site. We'll be looking at the look  feel, organization
 and content of the site. We're also considering adding a members-only
 area with password-protected pages.

 We'd like input from the technical working groups on the following:

 · What are the features of the current OFA Web site that you are using
 and would like to see remain relatively unchanged?

 · What are we missing on the current OFA Web site that you would like
 to see added?

 I'm with the marketing agency working on the site. If you have any
 input or feedback on potential changes, please e-mail me directly.


 Thanks!

 Cheri

 Cheri Winterberg │ Owen Media, Inc. │ 978.660.6405 │ che...@owenmedia.com


___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

[ewg] Testing today's build

[Jeff Becker]

Hi. I installed it on my Dell Cluster:

SLES 11 SP1/ConnectX
RHEL 5.3/Infinihost III
CentOS 5.4/Infinihost III

Installation went smoothly on all three platforms, and I was able to
bring up IPOIB on all three and do pings. Also, I tried rdma_{lat,bw} on
random pairs of nodes and these completed.In particular SLES 11 SP1
seems to work. Thanks.

-jeff
___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [ewg] 2.6.32 patch has carriage returns in it that are causing rejects

[Jeff Becker]

Hi Vlad

On 06/15/10 06:42, Vladimir Sokolovsky wrote:
 Jeff Becker wrote:
   
 Some good news for SLES 11 SP1 below...

 
 Hi Jeff,
 Please send me updated patches or your git tree to pull from.
   

Sure - if I need to change anything. Modulo the carriage return problem,
OFED 1.5.2 built fine. Hence, there aren't any patches yet.
 Do you have quilt installed? I think that quilt knows how to handle such 
 issues.
 At least, ofed_makedist.sh passed successfully on the OFA server.
   

You're right - quilt was not installed. That's probably it, since I had
another independent reply from someone who didn't see the carriage
return problem either. I'll install it, and hopefully it fixes things.
Thanks.

-jeff
 Thanks,
 Vladimir

   
 On 06/14/10 12:55, Jeff Becker wrote:
 
 On 06/14/10 11:16, Jeff Becker wrote:
   
 I cloned ofed_1_5/linux-2.6.git and tried to do:

 ./ofed_scripts/ofed_patch.sh --with-backport=2.6.32

 It rejected
 kernel_patches/backport/2.6.32/mlx4_semaphore_include.patch because
 the latter ends lines with carriage return and newline. I checked
 other patches and they only have newline. Should I submit a patch to
 the patch? Thanks.
 
 I find the ^M's in:

 kernel_patches/backport/2.6.32/new_frags_interface.patch
 kernel_patches/backport/2.6.32/net_skb-dst_accessors.patch

 as well. Is there a special option I need to give to ofed_checkout.sh?
 Thanks.
   
 After fixing the ^M's and getting ofed_patch.sh to complete, a basic build:

 configure --with-core-mod --with-user_mad-mod --with-user_access-mod
 --with-addr_trans-mod --with-mlx4-mod

 builds fine against the SLES 11 SP1 kernel. Now to test it...

 -jeff
 
 -jeff

   
 -jeff
 


 

 ___
 ewg mailing list
 ewg@lists.openfabrics.org
 http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
 
   

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

[ewg] 2.6.32 patch has carriage returns in it that are causing rejects

[Jeff Becker]

I cloned ofed_1_5/linux-2.6.git and tried to do:

./ofed_scripts/ofed_patch.sh --with-backport=2.6.32

It rejected kernel_patches/backport/2.6.32/mlx4_semaphore_include.patch
because the latter ends lines with carriage return and newline. I
checked other patches and they only have newline. Should I submit a
patch to the patch? Thanks.

-jeff
___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [ewg] 2.6.32 patch has carriage returns in it that are causing rejects

[Jeff Becker]

On 06/14/10 11:16, Jeff Becker wrote:
 I cloned ofed_1_5/linux-2.6.git and tried to do:

 ./ofed_scripts/ofed_patch.sh --with-backport=2.6.32

 It rejected
 kernel_patches/backport/2.6.32/mlx4_semaphore_include.patch because
 the latter ends lines with carriage return and newline. I checked
 other patches and they only have newline. Should I submit a patch to
 the patch? Thanks.

I find the ^M's in:

kernel_patches/backport/2.6.32/new_frags_interface.patch
kernel_patches/backport/2.6.32/net_skb-dst_accessors.patch

as well. Is there a special option I need to give to ofed_checkout.sh?
Thanks.

-jeff


 -jeff

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

[ewg] Server upgrade

[Jeff Becker]

Hi. I've upgraded sofa.openfabrics.org to Ubuntu 10.04LTS. System came
up fine after the upgrade. Please let me know if you have any issues.
Thanks.

-jeff
___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

[ewg] Upgrading sofa.openfabrics.org

[Jeff Becker]

Hi all. In preparation for moving entirely off the old server (and
fixing the wiki, etc.), I would like to upgrade from Ubuntu 8.04 to
Ubuntu 10.04. This will give us a very recent distro. and kernel to
start with. Please let me know if you have any concerns or questions
about this. Barring any issues, I plan to do the upgrade next week.

Thanks.

-jeff
___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [ewg] broken link on http://www.openfabrics.org/download_linux.htm

[Jeff Becker]

David N. Lombard wrote:
 Jeff Becker wrote:
   
 David N. Lombard wrote:
 
 The utils link, http://www.openfabrics.org/downloads/utils, on 
 http://www.openfabrics.org/download_linux.htm is 404.

   
   
 It appears the utils directory has been removed. Should we remove it
 from the downloads page, or . ? Thanks.
 

 It looks like the directory was renamed to ibutils
   

Good catch - fixed!

-jeff

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [ewg] broken link on http://www.openfabrics.org/download_linux.htm

[Jeff Becker]

David N. Lombard wrote:
 The utils link, http://www.openfabrics.org/downloads/utils, on 
 http://www.openfabrics.org/download_linux.htm is 404.

   
It appears the utils directory has been removed. Should we remove it
from the downloads page, or . ? Thanks.

-jeff
___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [ewg] OFED bug 1981

[Jeff Becker]

Jeff Becker wrote:
 Hi Tom. Slight update...

   
 --snippage--

 So upon further investigation, if you just load xprtrdma on the client,
 and then do /etc/init.d/openibd stop,  you get the expected resutl:

 Unloading rdma_cm  [FAILED]
 ERROR: Module rdma_cm is in use by xprtrdma

 However, doing the mount.rnfs and then trying to stop openibd gives the
 hanging behavior. I'll go and see if I can figure out what dependency
 the mount is affecting.

   

As I was looking at mount and friends, I realized the version of
nfs-utils included with OFED 1.5.1 is 1.1.5, which is about a year old.
Since the kernel base has moved to 2.6.30, would it be worth looking
into using a newer version of nfs-utils, e.g. 2.x? Thanks.

-jeff
 -jeff

   
   
 
   
 Ok, thanks...

   
 

   

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [ewg] mailman upgraded on server

[Jeff Becker]

On 01/04/2010 11:47 AM, Jeff Becker wrote:
 Apparently, something in the upgrade caused mailman to start erring out,
 causing messages to get shunted. Upgrading mailman seems to have fixed
 that. The lists should all be working now. Thanks.
   

One more thing. I had to clear the mail queues to do the upgrade. Please
resubmit any important mail from the past week. Thanks.

-jeff

 -jeff
 ___
 ewg mailing list
 ewg@lists.openfabrics.org
 http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
   

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

[ewg] Re: End user documentation for Infiniband?

[Jeff Becker]

On 12/09/2009 06:14 AM, Ross Smith wrote:
 Hi guys,

 Sorry if this is a bit of a newbie question, or posted to the wrong
 place, but I've been struggling to find any documentation for
 configuring infiniband, and the openfabrics mailing lists appear well
 and truly dead.
   

The Enterprise Working Group List (ewg@lists.openfabrics.org) is the
list for end users. Thanks.

-jeff

 I have a very basic network running now, with three hosts, ipoib and
 opensm.  It's running at 2.5Gbps which I understand is the default,
 but I cannot for the life of me find any documentation on how to
 configure it to run at higher speeds.

 Any network benchmarks are peaking at around 290MB/s, so just under 2.5Gbps.
 ibstat tells me that the link is running at 20.
 ibdiagnet tells me that the ipoib subnet is configured for 10Gbps, but
 that it has no members.

 I'm guessing that I need to configure the ipoib subnet for 20Gbps, and
 join the machines to it.  However I can't find documentation on how to
 do either of these.

 Can anybody help, or point me in the direction of some documentation?

 thanks,

 Ross
 --
 To unsubscribe from this list: send the line unsubscribe linux-rdma in
 the body of a message to majord...@vger.kernel.org
 More majordomo info at  http://vger.kernel.org/majordomo-info.html
   

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [ewg] ofa_1_5_kernel 20091104-0200 daily build status

[Jeff Becker]

Hi.

Vladimir Sokolovsky wrote:
 Brian J. Murrell wrote:
   
 On Thu, 2009-11-19 at 16:28 +0200, Vladimir Sokolovsky wrote: 

 You recall that this was a change you suggested in response to the
 previous failure:

 http://www.mail-archive.com/ewg@lists.openfabrics.org/msg07773.html

 which Jeff Becker is also reporting here:

 http://www.mail-archive.com/ewg@lists.openfabrics.org/msg07854.html

 Jeff includes more detail about the commit that actually broke this:

 committer Jack Morgenstein ja...@dev.mellanox.co.il 
 Thu, 11 Jun 2009 13:17:33 + (16:17 +0300)
 commit 1f462241bd18d9b5727ddea90459e7763b69e11c
 backports: 2.6.16_sles10_sp2: patches and add-ons based on kernel 2.6.18 
 backport

 

Jon Mason suggested adding a #ifndef ipv6_addr_loopback around the
function definition in
kernel_addons/backport/2.6.16_sles10_sp2/include/net/ipv6.h . I'll look
into this today.

-jeff

 Then this kernel requires another backport directory based on 
 2.6.16_sles10_sp2 under
 kernel_patches/backport/ and kernel_addons/backport/ with corresponding 
 change in ofed_scripts/get_backport_dir.sh
 (E.g. 2.6.16_sles10_sp2_lustre).
   
   ^^
 This breakage has got nothing to do with Lustre, per Jeff's report.  In
 any case, creating an entire new and mostly duplicate backport for a
 single change that is simply not portable seems like a sledgehammer of a
 solution, no?

 
 Please prepare backports for this kernel and I'll apply it to the OFED's 
 1.5 kernel git tree.
   
 I don't think I am going to have to time to prepare an entire backport
 (nor do I believe it's the correct solution) for this problem, but it
 seems it must be fixed before GA per Jeff's independent report of the
 same failure or anyone using this newer SLES10_SP2 kernel will run into
 this problem.

 Cheers,
 b.

 

 Ok, I will provide some solution next week.

 Regards,
 Vladimir


   

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [ewg] SLES10sp2 build break

[Jeff Becker]

I recently upgraded to the latest released SLES10SP2 kernel
(2.6.16.60-0.42.7) and I see the problem below as well. However, in
addition, I see:

In file included from
/usr/src/linux-2.6.16.60-0.42.7/include/net/addrconf.h:51,
 from
/var/tmp/OFED_topdir/BUILD/ofa_kernel-1.5/drivers/infiniband/core/addr.c:43:
/var/tmp/OFED_topdir/BUILD/ofa_kernel-1.5/kernel_addons/backport/2.6.16_sles10_sp2/include/net/ipv6.h:
At top level:
/var/tmp/OFED_topdir/BUILD/ofa_kernel-1.5/kernel_addons/backport/2.6.16_sles10_sp2/include/net/ipv6.h:22:
error: redefinition of 'ipv6_addr_loopback'
/usr/src/linux-2.6.16.60-0.42.7/include/net/ipv6.h:361: error: previous
definition of 'ipv6_addr_loopback' was here

The 'ipv6_addr_loopback' was added to
kernel_addons/backport/2.6.16_sles10_sp2/include/net/ipv6.h by:

committer Jack Morgenstein ja...@dev.mellanox.co.il
Thu, 11 Jun 2009 13:17:33 + (16:17 +0300)
commit 1f462241bd18d9b5727ddea90459e7763b69e11c
backports: 2.6.16_sles10_sp2: patches and add-ons based on kernel 2.6.18 
backport

Although this works for the base SLES10SP2 kernel (2.6.16.60-0.21),
people trying to install OFED on recent SLES10SP2 kernels will have the
same problem. Vlad, do you have any suggestions on how to fix this? Thanks.

-jeff

Jon Mason wrote:
 It looks like a recent commit broke the SLES10sp2 build.  I am seeing:

 In file included from /usr/src/linux-2.6.16.60-0.21/include/linux/aio.h:5,
  from /usr/src/linux-2.6.16.60-0.21/include/linux/sched.h:255,
  from 
 /root/ofed_kernel-2.6.16_sles10_sp2/kernel_addons/backport/2.6.16_sles10_sp2/include/linux/sched.h:4,
  from /usr/src/linux-2.6.16.60-0.21/include/linux/mm.h:4,
  from 
 /root/ofed_kernel-2.6.16_sles10_sp2/kernel_addons/backport/2.6.16_sles10_sp2/include/linux/mm.h:4,
  from /usr/src/linux-2.6.16.60-0.21/include/linux/skbuff.h:26,
  from 
 /root/ofed_kernel-2.6.16_sles10_sp2/kernel_addons/backport/2.6.16_sles10_sp2/include/linux/skbuff.h:4,
  from 
 /usr/src/linux-2.6.16.60-0.21/include/linux/if_ether.h:111,
  from 
 /root/ofed_kernel-2.6.16_sles10_sp2/kernel_addons/backport/2.6.16_sles10_sp2/include/linux/if_ether.h:4,
  from 
 /usr/src/linux-2.6.16.60-0.21/include/linux/netdevice.h:29,
  from 
 /root/ofed_kernel-2.6.16_sles10_sp2/kernel_addons/backport/2.6.16_sles10_sp2/include/linux/netdevice.h:4,
  from 
 /usr/src/linux-2.6.16.60-0.21/include/linux/inetdevice.h:7,
  from 
 /root/ofed_kernel-2.6.16_sles10_sp2/kernel_addons/backport/2.6.16_sles10_sp2/include/linux/inetdevice.h:4,
  from 
 /root/ofed_kernel-2.6.16_sles10_sp2/drivers/infiniband/core/addr.c:37:
 /root/ofed_kernel-2.6.16_sles10_sp2/kernel_addons/backport/2.6.16_sles10_sp2/include/linux/workqueue.h:
  In function ‘backport_delayed_work_pending’:
 /root/ofed_kernel-2.6.16_sles10_sp2/kernel_addons/backport/2.6.16_sles10_sp2/include/linux/workqueue.h:31:
  error: implicit declaration of function ‘delayed_work_pending’
 make[6]: *** 
 [/root/ofed_kernel-2.6.16_sles10_sp2/drivers/infiniband/core/addr.o] Error 1
 make[5]: *** [/root/ofed_kernel-2.6.16_sles10_sp2/drivers/infiniband/core] 
 Error 2
 make[4]: *** [/root/ofed_kernel-2.6.16_sles10_sp2/drivers/infiniband] Error 2
 make[3]: *** [_module_/root/ofed_kernel-2.6.16_sles10_sp2] Error 2
 make[2]: *** [modules] Error 2
 make[1]: *** [modules] Error 2
 make: *** [kernel] Error 2
 r...@demo2 ~/ofed_kernel-2.6.16_sles10_sp2 # uname -r
 2.6.16.60-0.21-smp


 kernel_addons/backport/2.6.16_sles10_sp2/include/linux/workqueue.h was last 
 touched in

 commit 4e0958d8db84ec233a75d289f7d636f974741d9d
 Author: Yevgeny Petrilin yevge...@mellanox.co.il
 Date:   Tue Nov 17 18:20:22 2009 +0200

 mlx4_core: Updated 'cancel_delayed_work_synch' backport
 
 Signed-off-by: Yevgeny Petrilin yevge...@mellanox.co.il

 In that patch, references to delayed_work_pending were added.

 Am I missing something?  Let me know if you want me to open a bug.

 Thanks,
 Jon
 ___
 ewg mailing list
 ewg@lists.openfabrics.org
 http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
   

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

[ewg] Re: [Fwd: FW: OFED installation question re-AltLinux]

[Jeff Becker]

Or Gerlitz wrote:
 Bill Boas wrote:
   
 Jeff, Please re-open openfabrics-general just as you describe. Tell the 
 developers on the linux-rdma list that we'd like them to subscribe, and that 
 we may from time to time forward an email to their list if we think it is
 appropriate.
 
 Bill,

 I don't see any reason to revive the general list, it served the 
 developer community and now migrated to some other place. The question 
 raised in this thread and its like has nothing to do with development 
 and hence matches the ewg list.
   

This makes sense. Although it seems like we should have a more
descriptive statement about ewg on the mail lists web page, so folks
know this is where to ask questions about OFED usage. Any takers? Thanks.

-jeff

 On top of that, the decisions on the mailing list definition / location 
 / move was never taken by a single individual sending an email to 
 another individual, but rather taken through a discussion in a work 
 group or over email. If you want to open/revive a mailing list, please 
 bring it somewhere and not over a private thread. BTW - some months ago 
 ofa opened bunch of user mailing lists without a single email sent 
 over them, so we already have too many mailing lists, aren't we?



 Or.

   

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [ewg] Perftest Maintainer.

[Jeff Becker]

Hi.

Tziporet Koren wrote:
 Jeff Squyres wrote:
   
 Are there any plans to make all the perftest tools use RDMA CM for 
 connections?

 
I'm happy to take a crack at this. I recently volunteered (with Bill
Boas)  to write an RDMA programming howto, so it's relevant.
Jeff - you'd like to have all the perftest tools look like
rdma_[bw,lat].c with the --cma option, right? Do you also want to get
rid of this option, and just have it on by default in these two tests?
Please let me know. Thanks.

-jeff

 We don't have such a plan
 If someone will submit patches we will take them

 Tziporet
 ___
 ewg mailing list
 ewg@lists.openfabrics.org
 http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
   

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

[ewg] NFSRDMA backport for 2.6.29

[Jeff Becker]

Hi Vlad. The attached patch contains the NFSRDMA backport for 2.6.29. It
passes Connectathon. Please apply to  ofed_1_5/linux-2.6.git. Thanks.

-jeff


diff --git a/kernel_addons/backport/2.6.29/include/linux/fs.h b/kernel_addons/backport/2.6.29/include/linux/fs.h
new file mode 100644
index 000..6c88678
--- /dev/null
+++ b/kernel_addons/backport/2.6.29/include/linux/fs.h
@@ -0,0 +1,56 @@
+#ifndef BACKPORT_LINUX_FS_H
+#define BACKPORT_LINUX_FS_H
+
+#include_next linux/fs.h
+#include linux/sched.h
+#include linux/fs_struct.h
+
+static inline int current_umask(void)
+{
+	return current-fs-umask;
+}
+
+static inline void free_fs_struct(struct fs_struct *fs)
+{
+	struct task_struct *tsk;
+
+	tsk = kzalloc(sizeof(struct task_struct), GFP_KERNEL);
+	if (!tsk)
+		return;
+
+	spin_lock_init(tsk-alloc_lock);
+	tsk-fs = fs;
+
+	exit_fs(tsk);
+	kfree(tsk);
+}
+
+static inline int unshare_fs_struct(void)
+{
+	struct fs_struct *fs = current-fs;
+	struct fs_struct *new_fs = copy_fs_struct(fs);
+	int kill;
+
+	if (!new_fs)
+		return -ENOMEM;
+
+	task_lock(current);
+	write_lock(fs-lock);
+	kill = atomic_read(fs-count) == 1;
+	current-fs = new_fs;
+	write_unlock(fs-lock);
+	task_unlock(current);
+
+	if (kill)
+		free_fs_struct(fs);
+
+	return 0;
+}
+
+static inline void deactivate_locked_super(struct super_block *s)
+{
+	up_write(s-s_umount);
+	deactivate_super(s);
+}
+
+#endif
diff --git a/kernel_addons/backport/2.6.29/include/linux/fscache.h b/kernel_addons/backport/2.6.29/include/linux/fscache.h
new file mode 100644
index 000..f758384
--- /dev/null
+++ b/kernel_addons/backport/2.6.29/include/linux/fscache.h
@@ -0,0 +1,4 @@
+#ifndef BACKPORT_LINUX_FSCACHE_H
+#define BACKPORT_LINUX_FSCACHE_H
+
+#endif
diff --git a/kernel_patches/backport/2.6.29/rnfs_fs.patch b/kernel_patches/backport/2.6.29/rnfs_fs.patch
new file mode 100644
index 000..5572f3a
--- /dev/null
+++ b/kernel_patches/backport/2.6.29/rnfs_fs.patch
@@ -0,0 +1,40 @@
+diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c
+index b660435..cf09e1d 100644
+--- a/fs/nfsd/vfs.c
 b/fs/nfsd/vfs.c
+@@ -1005,9 +1005,7 @@ nfsd_vfs_write(struct svc_rqst *rqstp, struct svc_fh *fhp, struct file *file,
+ 	if (!EX_ISSYNC(exp))
+ 		stable = 0;
+ 	if (stable  !EX_WGATHER(exp)) {
+-		spin_lock(file-f_lock);
+ 		file-f_flags |= O_SYNC;
+-		spin_unlock(file-f_lock);
+ 	}
+ 
+ 	/* Write the data. */
+diff --git a/fs/nfs/file.c b/fs/nfs/file.c
+index ec7e27d..0850bd8 100644
+--- a/fs/nfs/file.c
 b/fs/nfs/file.c
+@@ -484,9 +484,8 @@ const struct address_space_operations nfs_file_aops = {
+  * writable, implying that someone is about to modify the page through a
+  * shared-writable mapping
+  */
+-static int nfs_vm_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf)
++static int nfs_vm_page_mkwrite(struct vm_area_struct *vma, struct page *page)
+ {
+-	struct page *page = vmf-page;
+ 	struct file *filp = vma-vm_file;
+ 	struct dentry *dentry = filp-f_path.dentry;
+ 	unsigned pagelen;
+@@ -517,9 +516,9 @@ static int nfs_vm_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf)
+ 
+ 	ret = nfs_updatepage(filp, page, 0, pagelen);
+ out_unlock:
++	unlock_page(page);
+ 	if (!ret)
+ 		return VM_FAULT_LOCKED;
+-	unlock_page(page);
+ 	return VM_FAULT_SIGBUS;
+ }
+ 
___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [ewg] NFSRDMA backport for 2.6.29

[Jeff Becker]

Hi Vlad. I generated this patch by doing git diff on my branch compared
to your master, and I forgot to add a commit message. You can use the
following:

NFSRDMA 2.6.29 backport: passes Connectathon
Signed-off-by Jeff Becker jeffrey.c.bec...@nasa.gov

Sorry about the confusion. Thanks.

-jeff


Jeff Becker wrote:
 Hi Vlad. The attached patch contains the NFSRDMA backport for 2.6.29. It
 passes Connectathon. Please apply to  ofed_1_5/linux-2.6.git. Thanks.

 -jeff


   

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [ewg] OFA server fs is full

[Jeff Becker]

Hi Tziporet. I believe Ido is working on moving us to the new server.

-jeff

Tziporet Koren wrote:
 Sasha Khapyorsky wrote:
   
 Now there is:

 Filesystem   1K-blocks  Used Available Use% Mounted on
 /dev/sda1151873632 139556036   4602784  97% /


 We will have a next overflow just in few days.


   
 
 Jeff
 When will we move to the new server?
 Can Ido help with this?

 Tziporet

   

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [ewg] EWG/OFED meeting meeting minutes for May 18, 09

[Jeff Becker]

Hi Tziporet

Tziporet Koren wrote:

 These are the OFED meeting minutes for May 18 09:

 Summary:

 

 1. OFED 1.4.1 is delayed:

 RC6 is planed for this week once bug 1596 is resolved

 GA will be 5 days later

 2. OFED 1.5: We will close schedule in next meeting.

 Kernel tree is available to start backports

 Reminder:

 OFED roadmap on the web:
 _http://www.openfabrics.org/txt/woody/roadmap.txt_

 EWG meeting minutes:
 _http://www.openfabrics.org/txt/documentation/linux/EWG_meeting_minutes/_

 Details:

 ==

 1. OFED 1.4.1 bugs status:

 * 1596cri jeffrey.c.bec...@nasa.gov   openibd stop
   failed when nfs is loaded -  RHEL 5.x are done, wait for SLES10 SP2


I committed this last night and Vlad pulled it into his tree in time for
the most recent build. Please test. Thanks.

-jeff

 * MVAPICH will be replaced for a bug fix - under work

 2. New memory registration API

 * Need to follow the list and see what will be the resolution

 3. OFED 1.5 status:

 * Kernel git tree: URL:
   git://git.openfabrics.org/ofed_1_5/linux-2.6.git
   Branch: ofed_kernel_1_5
 * We already have backports of kernel.org kernels 2.6.24 -- 2.6.29.
   The backports compile for core modules, mthca, mlx4, ipoib, and
   sdp. (not run yet)
 * We may need to push freeze for June 20 (plan was Jun 7) - will
   decide it in next meeting on June 1
 * All user space packages (as we decided at Sonoma) must be ready
   for the beta release (will not be required for the alpha)


 Tziporet


___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

[ewg] Fixes for bug 1596 (for 2.6.25, 2.6.22 and RHEL5.3)

[Jeff Becker]

Hi Vlad. Please pull from:
git://git.openfabrics.org/~jbecker/nfs_rdma_ofed_1_4.git

Thanks.

-jeff

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

[ewg] Re: What is the status of backport fix for bug 1596

[Jeff Becker]

Hi Tziporet

Tziporet Koren wrote:
 This is the only bug we wait for its fix for 1.4.1-rc6
 Do you expect to fix it by end of this week (including testing)?
   

I've completed 2.6.25 and 2.6.22, and I'm currently working on RHEL5.3.
Although
the first of these was straightforward, the 2.6.22 port took longer due
to several
unforeseen dependences. Now that I understand these, the rest of the
backports should
go quicker. I will try to have something ready for Vlad when he comes in
on Monday (i.e., I
will try to be done by the end of Sunday my time), but I honestly can't
guarantee it. Definitely
early next week though.

 If not - can we go without it?
   

It's OK with me. I was only trying to satisfy Betsy's request.  Thanks.

-jeff

 Please advise since I wish to push 1.4.1 forward

 Thanks
 Tziporet
   

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [ewg] Re: 2.6.25 backport fix for bug 1596

[Jeff Becker]

Hi Tziporet

Tziporet Koren wrote:
 Jeff
 What is the status of this?
 We wait for this bug fix with RC6
   

Regarding the fix for bug 1596, I know what needs to be fixed. The
2.6.25 backport is done, and I'm testing 2.6.22. Then
I will do the RHEL and SLES backports. Thanks.

-jeff

 Thanks
 Tziporet 

 -Original Message-
 From: ewg-boun...@lists.openfabrics.org
 [mailto:ewg-boun...@lists.openfabrics.org] On Behalf Of Jeff Becker
 Sent: Tuesday, May 12, 2009 1:35 AM
 To: Steve Wise; Vladimir Sokolovsky
 Cc: OpenFabrics EWG
 Subject: [ewg] Re: 2.6.25 backport fix for bug 1596

 So is it OK for Vlad to apply? Vlad if you prefer, I can fix all the
 backports (actually 2.6.25 was fine as it turned out)
 and resubmit, rather than commit each file individually. Please let me
 know. Thanks.

 -jeff

 Steve Wise wrote:
   
 Hard to read these sorts of patches to patch files. :)  But it looks
 
 ok 
   
 I think.  And cthon runs!

 Steve.


 Jeff Becker wrote:
   
 
 I just checked this passes connectathon. Please review. Thanks.

 -jeff

 NFSRDMA: remove dependence of 2.6.25 backport ib_core on nfs:
   
 fixes
   
 Bug 1596
 Signed-off-by: Jeff Becker jeffrey.c.bec...@nasa.gov

 diff --git a/kernel_patches/backport/2.6.25/rnfs_fs.patch
 b/kernel_patches/backport/2.6.25/rnfs_fs.patch
 index 342bd8a..c6cd2de 100644
 --- a/kernel_patches/backport/2.6.25/rnfs_fs.patch
 +++ b/kernel_patches/backport/2.6.25/rnfs_fs.patch
 @@ -1,18 +1,21 @@
 -diff --git a/drivers/infiniband/core/Makefile
 b/drivers/infiniband/core/Makefile
 -index f646040..9508889 100644
  a/drivers/infiniband/core/Makefile
 -+++ b/drivers/infiniband/core/Makefile
 -@@ -30,3 +30,5 @@ ib_umad-y :=user_mad.o
 - ib_ucm-y :=ucm.o
 +diff --git a/fs/nfs/Makefile b/fs/nfs/Makefile
 +index ac6170c..0bae18e 100644
 +--- a/fs/nfs/Makefile
  b/fs/nfs/Makefile
 +@@ -6,7 +6,7 @@ obj-$(CONFIG_NFS_FS) += nfs.o
  
 - ib_uverbs-y :=uverbs_main.o uverbs_cmd.o
   
 uverbs_marshall.o
   
 -+
 -+ib_core-y +=namespace.o
 -diff --git a/drivers/infiniband/core/namespace.c
 b/drivers/infiniband/core/namespace.c
 + nfs-y := client.o dir.o file.o getroot.o inode.o
   
 super.o
   
 nfs2xdr.o \
 +direct.o pagelist.o proc.o read.o symlink.o unlink.o
   
 \
   
 +-   write.o namespace.o mount_clnt.o
 ++   write.o namespace.o mount_clnt.o
   
 backport-namespace.o
   
 + nfs-$(CONFIG_ROOT_NFS)+= nfsroot.o
 + nfs-$(CONFIG_NFS_V3)+= nfs3proc.o nfs3xdr.o
 + nfs-$(CONFIG_NFS_V3_ACL)+= nfs3acl.o
 +diff --git a/fs/nfs/backport-namespace.c
   
 b/fs/nfs/backport-namespace.c
   
  new file mode 100644
  index 000..de57f8b
  --- /dev/null
 -+++ b/drivers/infiniband/core/namespace.c
  b/fs/nfs/backport-namespace.c
  @@ -0,0 +1 @@
  +#include src/namespace.c
  diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
   
 
   
   
 

 ___
 ewg mailing list
 ewg@lists.openfabrics.org
 http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
   

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [ewg] ofed-1.4.1 rc5 and sles10sp2

[Jeff Becker]

Hi Steve

Steve Wise wrote:
 I installed RC5 on sles10sp2 and none of the ofed modules would load due 
 to symbol mismatches.  For instance, ib_core won't load and this is logged:

 May 11 13:32:13 rac4 kernel: ib_core: Unknown symbol init_mnt_writers


 I backed up to RC4 and things work ok.

 Something bad happened to RC5.
   

My bad - somehow I introduced a dependency loop in the modules. Working
on a fix.

-jeff

 Anybody else seeing this?


 Steve.

 ___
 ewg mailing list
 ewg@lists.openfabrics.org
 http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
   

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [ewg] ofed-1.4.1 rc5 and sles10sp2

[Jeff Becker]

Steve Wise wrote:
 Jeff Becker wrote:
   
 Hi Steve

 Steve Wise wrote:
   
 
 I installed RC5 on sles10sp2 and none of the ofed modules would load due 
 to symbol mismatches.  For instance, ib_core won't load and this is logged:

 May 11 13:32:13 rac4 kernel: ib_core: Unknown symbol init_mnt_writers


 I backed up to RC4 and things work ok.

 Something bad happened to RC5.
   
 
   
 My bad - somehow I introduced a dependency loop in the modules. Working
 on a fix.

 -jeff

   
 

 Tziporet, I suggest you recall RC5.  It is DOA. 

 Jeff, please please regression test things before pushing upstream. 
   

I'm very sorry. I was rushing to get this into rc5 to fix bug 1596, and
I didn't think
the changes looked very drastic, as I basically just moved some stuff
around. Thus
I only build tested it. Jon I now understand your concern about fixing
this bug.

If it makes sense, revert my patch and re-release rc5. I'm working on a
fix right now.

Again, my apologies. This won't happen again.

-jeff


   

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

[Fwd: Re: [ewg] ofed-1.4.1 rc5 and sles10sp2]

[Jeff Becker]

Vlad, please revert my last commit. It appears to be broken. I will
resubmit fixed patches for each backport OS individually after testing
and further review by Steve and Jon. Thanks.

-jeff

 Original Message 
Subject:Re: [ewg] ofed-1.4.1 rc5 and sles10sp2
Date:   Mon, 11 May 2009 15:59:43 -0500
From:   Jeff Becker jeffrey.c.bec...@nasa.gov
To: Steve Wise sw...@opengridcomputing.com
CC: OpenFabrics EWG ewg@lists.openfabrics.org
References: 4a088053.60...@opengridcomputing.com
4a088f75.9060...@nasa.gov 4a089049.80...@opengridcomputing.com



Steve Wise wrote:
 Jeff Becker wrote:
   
 Hi Steve

 Steve Wise wrote:
   
 
 I installed RC5 on sles10sp2 and none of the ofed modules would load due 
 to symbol mismatches.  For instance, ib_core won't load and this is logged:

 May 11 13:32:13 rac4 kernel: ib_core: Unknown symbol init_mnt_writers


 I backed up to RC4 and things work ok.

 Something bad happened to RC5.
   
 
   
 My bad - somehow I introduced a dependency loop in the modules. Working
 on a fix.

 -jeff

   
 

 Tziporet, I suggest you recall RC5.  It is DOA. 

 Jeff, please please regression test things before pushing upstream. 
   

I'm very sorry. I was rushing to get this into rc5 to fix bug 1596, and
I didn't think
the changes looked very drastic, as I basically just moved some stuff
around. Thus
I only build tested it. Jon I now understand your concern about fixing
this bug.

If it makes sense, revert my patch and re-release rc5. I'm working on a
fix right now.

Again, my apologies. This won't happen again.

-jeff


   

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

[ewg] Re: 2.6.25 backport fix for bug 1596

[Jeff Becker]

So is it OK for Vlad to apply? Vlad if you prefer, I can fix all the
backports (actually 2.6.25 was fine as it turned out)
and resubmit, rather than commit each file individually. Please let me
know. Thanks.

-jeff

Steve Wise wrote:
 Hard to read these sorts of patches to patch files. :)  But it looks ok 
 I think.  And cthon runs!

 Steve.


 Jeff Becker wrote:
   
 I just checked this passes connectathon. Please review. Thanks.

 -jeff

 NFSRDMA: remove dependence of 2.6.25 backport ib_core on nfs: fixes
 Bug 1596
 Signed-off-by: Jeff Becker jeffrey.c.bec...@nasa.gov

 diff --git a/kernel_patches/backport/2.6.25/rnfs_fs.patch
 b/kernel_patches/backport/2.6.25/rnfs_fs.patch
 index 342bd8a..c6cd2de 100644
 --- a/kernel_patches/backport/2.6.25/rnfs_fs.patch
 +++ b/kernel_patches/backport/2.6.25/rnfs_fs.patch
 @@ -1,18 +1,21 @@
 -diff --git a/drivers/infiniband/core/Makefile
 b/drivers/infiniband/core/Makefile
 -index f646040..9508889 100644
  a/drivers/infiniband/core/Makefile
 -+++ b/drivers/infiniband/core/Makefile
 -@@ -30,3 +30,5 @@ ib_umad-y :=user_mad.o
 - ib_ucm-y :=ucm.o
 +diff --git a/fs/nfs/Makefile b/fs/nfs/Makefile
 +index ac6170c..0bae18e 100644
 +--- a/fs/nfs/Makefile
  b/fs/nfs/Makefile
 +@@ -6,7 +6,7 @@ obj-$(CONFIG_NFS_FS) += nfs.o
  
 - ib_uverbs-y :=uverbs_main.o uverbs_cmd.o uverbs_marshall.o
 -+
 -+ib_core-y +=namespace.o
 -diff --git a/drivers/infiniband/core/namespace.c
 b/drivers/infiniband/core/namespace.c
 + nfs-y := client.o dir.o file.o getroot.o inode.o super.o
 nfs2xdr.o \
 +direct.o pagelist.o proc.o read.o symlink.o unlink.o \
 +-   write.o namespace.o mount_clnt.o
 ++   write.o namespace.o mount_clnt.o backport-namespace.o
 + nfs-$(CONFIG_ROOT_NFS)+= nfsroot.o
 + nfs-$(CONFIG_NFS_V3)+= nfs3proc.o nfs3xdr.o
 + nfs-$(CONFIG_NFS_V3_ACL)+= nfs3acl.o
 +diff --git a/fs/nfs/backport-namespace.c b/fs/nfs/backport-namespace.c
  new file mode 100644
  index 000..de57f8b
  --- /dev/null
 -+++ b/drivers/infiniband/core/namespace.c
  b/fs/nfs/backport-namespace.c
  @@ -0,0 +1 @@
  +#include src/namespace.c
  diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
   
 

   

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

[ewg] NFSRDMA backport fixes for [Bug 1596] openibd stop failed when nfs is loaded

[Jeff Becker]

Hi Vlad. Please pull from:
git://git.openfabrics.org/~jbecker/nfs_rdma_ofed_1_4.git

If the fixes look OK, please revert or fix the last patch to 
ofed_scripts/openibd.


Thanks.

-jeff
___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

[ewg] Re: [Fwd: [Bug 1596] openibd stop failed when nfs is loaded]

[Jeff Becker]

Hi all. I haven't heard any further discussion on this. Should we fix
this for 1.4.1? I'm happy to work on this. Thanks.

-jeff

Jeff Becker wrote:
 Jon Mason wrote:
   
 On Wed, May 06, 2009 at 09:28:49AM -0700, Jeff Becker wrote:
   
 
 Hi Jon. I can take a crack at this, or do you have a solution already in
 mind? Thanks.
 
   
 The issue is whether or not the user will unmount the NFS dirs before
 stopping the openibd.  If it is an NFSRDMA mount, they'll have to
 anyway.  So I do not see the problem.
   
 

 I think the problem is that the dependence is there even if they are running
 plain NFS (not RDMA) and IB, and (for whatever reason) the OFED NFS modules
 got built (replacing the standard ones). That's what needs to be fixed.

 -jeff

   
   
 
 -jeff

  Original Message 
 Subject:[Bug 1596] openibd stop failed when nfs is loaded
 Date:   Tue, 5 May 2009 20:47:09 -0500
 From:   bugzilla-dae...@lists.openfabrics.org
 bugzilla-dae...@lists.openfabrics.org
 To: Becker, Jeffrey C. (ARC-TN)[COMPUTER SCIENCES CORPORATION]
 jeffrey.c.bec...@nasa.gov



 https://bugs.openfabrics.org/show_bug.cgi?id=1596


 betsy.zel...@qlogic.com changed:

What|Removed |Added
 
Severity|major   |critical
Priority|P3  |P2




 --- Comment #10 from betsy.zel...@qlogic.com  2009-05-05 18:47 ---
 I know we discussed this in Monday's meeting, but it is a really bad issue.
 Although we don't expect customers to often reload their driver, it does 
 happen
 sometimes, and forcing them to reboot isn't really a reasonable workaround. 

 It also makes testing almost unmanageable.

 I believe we have to fix this for 1.4.1, so I'm raising the priority and
 severity on this. -Betsy


 -- 
 Configure bugmail: https://bugs.openfabrics.org/userprefs.cgi?tab=email
 --- You are receiving this mail because: ---
 You are the assignee for the bug, or are watching the assignee.

 
   


   

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

[ewg] NFSRDMA SLES10SP2 backport: fixes for bug 1607

[Jeff Becker]

Hi Vlad. Please pull from:
git://git.openfabrics.org/~jbecker/nfs_rdma_ofed_1_4.git

Thanks.

-jeff
___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [ewg] OFED 1.4.1 RC4 is delayed to Thursday

[Jeff Becker]

Hi Tziporet. My update is that I believe I know what is causing bug
1607, and I'm working on a fix. Thanks.

-jeff

Steve Wise wrote:
 Status update:

 I cleaned up some NFSRDMA server crashes that happen when there are 
 asynchronous WR failures.  That might help Vu figure out 1571.  I think 
 there is a FW issue causing the async failure.  But the code shouldn't 
 crash anymore with my latest fix.

 But I'd also like 1613, 1616 into ofed-1.4.1:

 1613: dapl regression that UNH uncovered.  Arlin has a fix.
 1616: nfsrdma ppc64 issue uncovered today.  Hopefully we can nail this 
 one by EOB friday

 Should we crank RC4 tomorrow and plan an RC5?  Or hold off for a few 
 more days with RC4? 


 Steve.



 Tziporet Koren wrote:
   
 Jon Mason wrote:
 
 On Mon, Apr 27, 2009 at 05:43:05PM +0300, Tziporet Koren wrote:
  
   
 Hi All
 Since there are still few open critical bugs we delay OFED 1.4.1-RC4
 build to Thursday.
 Note that we are on vacation on Wed this week (Israel Independence Day)

 The bugs that must be fixed:
 1607blo  SLES  jeffrey.c.bec...@nasa.gov  kernel oops
 during login on sles10 sp2 with OFED-1.4.1-20...
 1609 cri RHEL sw...@opengridcomputing.com kernel panic
 running iozone on x86 system
 
 
 This was fixed by the patch Steve pushed on Friday.  I'll close the bug
 for him.

   
   
 Well - its too late now for us to build and test it
 What about bug 1571 ?

 Tziporet
 ___
 ewg mailing list
 ewg@lists.openfabrics.org
 http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
 

 ___
 ewg mailing list
 ewg@lists.openfabrics.org
 http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
   

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [ewg] website really slow

[Jeff Becker]

Hi all.

Steve Wise wrote:
 There is something wrong with the site.  bugzirra is taking forever too.
   

Turns out our web server is getting DOS'd. I'm patching our server right
now. Reboot to upgraded OS (and apache and...) coming shortly. Also
planning to crank down on iptables.

-jeff


 Brian J. Murrell wrote:
   
 FWIW, this did finally complete and download rate was good but it sure
 took a while to get it started.

 b.

   
 

 ___
 ewg mailing list
 ewg@lists.openfabrics.org
 http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
 

   

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

[ewg] NFSRDMA - pull in Tom Talpey's upstream patches, fix a few missing symbols for SLES10SP2 backport

[Jeff Becker]

Hi Vlad. Please pull from:
git://git.openfabrics.org/~jbecker/nfs_rdma_ofed_1_4.git

Thanks.

-jeff
___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [ewg] website really slow

[Jeff Becker]

Please try again. Thanks.

-jeff

Ryan, Jim wrote:
 I'm unable to log on to approve list requests before it times out, so things 
 are backing up

 Jim 

 -Original Message-
 From: ewg-boun...@lists.openfabrics.org 
 [mailto:ewg-boun...@lists.openfabrics.org] On Behalf Of Jeff Becker
 Sent: Wednesday, April 15, 2009 10:14 AM
 To: Steve Wise
 Cc: ewg@lists.openfabrics.org
 Subject: Re: [ewg] website really slow

 Hi all.

 Steve Wise wrote:
   
 There is something wrong with the site.  bugzirra is taking forever too.
   
 

 Turns out our web server is getting DOS'd. I'm patching our server right
 now. Reboot to upgraded OS (and apache and...) coming shortly. Also
 planning to crank down on iptables.

 -jeff

   
 Brian J. Murrell wrote:
   
 
 FWIW, this did finally complete and download rate was good but it sure
 took a while to get it started.

 b.

   
 

 ___
 ewg mailing list
 ewg@lists.openfabrics.org
 http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
 
   
   
 

 ___
 ewg mailing list
 ewg@lists.openfabrics.org
 http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
   

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

[ewg] NFSRDMA SLES10SP2 backport: update based on Jon Mason's RHEL5.x backports

[Jeff Becker]

Hi Vlad. Please pull from: 
git://git.openfabrics.org/~jbecker/nfs_rdma_ofed_1_4.git

Thanks.

-jeff

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: Fwd: [Fwd: [ewg] compile failure on SLES11 in NFS]

[Jeff Becker]

Hi Jon

Jon Mason wrote:
 Hey Jeff,
 I just wanted to make you aware of the issue below.  I thought I broke
 something until I saw the SLES11.  Did you already do the SLES11
 backport, or are they trying to build something that is not completed?
   

I did see this. The problem is that I don't have a powerpc to test
builds on. The x86-64 SLES11 backport is done. Can someone with access
to powerpc help out here? Thanks.

-jeff

 Thanks,
 Jon

 - Forwarded message from Steve Wise sw...@opengridcomputing.com -

 User-Agent: Thunderbird 2.0.0.21 (Windows/20090302)
 From: Steve Wise sw...@opengridcomputing.com
 To: Jon Mason j...@opengridcomputing.com
 Date: Thu, 02 Apr 2009 19:56:18 -0500
 Subject: [Fwd: [ewg] compile failure on SLES11 in NFS]

 Maybe another compile issue?


  Original Message 
 Subject:  [ewg] compile failure on SLES11 in NFS
 Date: Thu, 2 Apr 2009 18:06:22 +0200
 From: Alexander Schmidt al...@linux.vnet.ibm.com
 To:   of-ewg ewg@lists.openfabrics.org
 CC:   Hoang-Nam Nguyen hngu...@linux.vnet.ibm.com, Christoph Raisch  
 rai...@de.ibm.com, Stefan Roscher stefan.rosc...@de.ibm.com



 Hi everybody,

 with the latest daily build (OFED-1.4.1-20090402-0600) I'm getting the compile
 problem below. I'm running SLES11 on a powerpc box. Could someone take a look
 at this?

 Appreciate your help!

 Alex

 -Iarch/powerpc  -I/var/tmp/OFED_topdir/BUILD/ofa_kernel-1.4.1/fs/nfs -Wall 
 -Wundef -Wstrict-prototypes -Wno-trigraphs -fno-strict-aliasing -fno-common 
 -Werror-implicit-function-declaration -Os -msoft-float -pipe 
 -I/usr/src/linux-2.6.27.19-5/arch/powerpc -Iarch/powerpc -mminimal-toc 
 -mtraceback=none -mcall-aixdesc -mcpu=power4 -mtune=cell -mno-altivec 
 -mno-spe -funit-at-a-time -mno-string -Wa,-maltivec -fno-stack-protector 
 -fomit-frame-pointer -g -Wdeclaration-after-statement -Wno-pointer-sign 
 -DMODULE -DKBUILD_STR(s)=#s -DKBUILD_BASENAME=KBUILD_STR(file)  
 -DKBUILD_MODNAME=KBUILD_STR(nfs) -DDEBUG_HASH=3 -DDEBUG_HASH2=61 -c -o 
 /var/tmp/OFED_topdir/BUILD/ofa_kernel-1.4.1/fs/nfs/.tmp_file.o 
 /var/tmp/OFED_topdir/BUILD/ofa_kernel-1.4.1/fs/nfs/file.c
 /var/tmp/OFED_topdir/BUILD/ofa_kernel-1.4.1/fs/nfs/file.c: In function 
 ‘nfs_write_begin’:
 /var/tmp/OFED_topdir/BUILD/ofa_kernel-1.4.1/fs/nfs/file.c:354: error: 
 implicit declaration of function ‘__grab_cache_page’
 /var/tmp/OFED_topdir/BUILD/ofa_kernel-1.4.1/fs/nfs/file.c:354: warning: 
 assignment makes pointer from integer without a cast
 make[5]: *** [/var/tmp/OFED_topdir/BUILD/ofa_kernel-1.4.1/fs/nfs/file.o] 
 Error 1
 make[4]: *** [/var/tmp/OFED_topdir/BUILD/ofa_kernel-1.4.1/fs/nfs] Error 2
 make[3]: *** [_module_/var/tmp/OFED_topdir/BUILD/ofa_kernel-1.4.1] Error 2
 make[2]: *** [sub-make] Error 2
 make[1]: *** [all] Error 2
 make[1]: Leaving directory `/usr/src/linux-2.6.27.19-5-obj/ppc64/ppc64'
 make: *** [kernel] Error 2
 error: Bad exit status from /var/tmp/rpm-tmp.24575 (%build)
 ___
 ewg mailing list
 ewg@lists.openfabrics.org
 http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg


 - End forwarded message -
   

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [Fwd: Re: [ewg] Moving to new server]

[Jeff Becker]

Hi all. Thanks to Jeff Squyres and the Indiana U DNS maintainers, the
new server is in DNS as sofa.openfabrics.org. Note sofa stands for
Server of the OpenFabrics Alliance (and hopefully it's also a
comfortable place to do development :-) ). Please move your git trees to
it, and after I set up gitweb, we will have git.openfabrics.org point to
the new server.  Thanks.

-jeff

Jeff Squyres wrote:
 BTW, I see 69.55.231.68 is the IP address of the new server.

 Do you want a [temporary] DNS name for this?  (e.g., I see the  
 hostname is set to sofa -- so sofa.openfabrics.org?)

 The Indiana U. DNS maintainers are very receptive to whatever DNS  
 changes you want...


 On Mar 31, 2009, at 12:36 PM, Alexander Schmidt wrote:

   
 On Sun, 29 Mar 2009 11:06:00 -0700
 Jeff Becker jeffrey.c.bec...@nasa.gov wrote:

 
 Hi Yannick. Please try again. On Friday, I realized that the home
 directory ownership was wrong, and fixed these. Thanks.

   
 Hi Jeff,

 I'm trying to move the libehca.git tree and I'm also not able to  
 login, on
 the old server everything works fine... Could you take a look at our  
 account?

 [alexs...@bl3d1974 ~]$ ssh hngu...@69.55.231.68
 hngu...@69.55.231.68's password:

 Thanks,
 Alex

 ___
 ewg mailing list
 ewg@lists.openfabrics.org
 http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
 


   

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [Fwd: Re: [ewg] Moving to new server]

[Jeff Becker]

Hi Tziporet

Tziporet Koren wrote:
 Jeff
 Do I have to move my staff under 
 /data/www/openfabrics.org/txt/documentation/linux/ too?
 We hold there the release roadmap and the EWG meeting minutes.
   

For now, we are just moving git, so please just move whatever git trees
you have. Thanks.

-jeff

 Thanks
 Tziporet

   

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

[Fwd: Re: [ewg] Moving to new server]

[Jeff Becker]

Hi Yannick. Please try again. On Friday, I realized that the home 
directory ownership was wrong, and fixed these. Thanks.


-jeff

 Original Message 
Subject:Re: [ewg] Moving to new server
Date:   Fri, 27 Mar 2009 20:34:55 -0500
From:   Yannick Cote yannick.c...@qlogic.com
To: 	Becker, Jeffrey C. (ARC-TN)[COMPUTER SCIENCES CORPORATION] 
jeffrey.c.bec...@nasa.gov
References: 	497e00fa.1020...@nasa.gov 
49a2c7ef.8050...@dev.mellanox.co.il 49cd6037.4090...@nasa.gov




Hi Jeff,

I tried to logon couple times in the past and it didn't work. I told
myself that accounts haven't been moved yet.

yc...@69.55.231.68's password: 
Permission denied, please try again.


yep still doesn't work. I wouldn't be surprised if I missed some
instructions to migrate though ;-)

--
Yannick Cote
yannick.c...@qlogic.com

On Fri, 27 Mar 2009, Jeff Becker wrote:


Hi all. Except for Vlad, no-one has moved their git trees from
openfabrics.org to the new server. I'm happy to go and copy ~/scm/* from
everyone's home directory, but I wanted to give you a chance to pick
what git trees you want and move them yourself (or tell me, and I can).
Just thought this would be a good chance to clean up.  Thanks.

-jeff

Vladimir Sokolovsky wrote:

Jeff Becker wrote:


Hi Vlad. As requested in today's EWG telecon, can we move the build
environment and git trees to the new server (69.55.231.68) this week?
Please let me know what help you need if any. Thanks.

-jeff




Hi Jeff,
I copied and tested cross compilation environment and OFED build
environment  on the new server  (69.55.231.68).
Remember to add links under /pub/scm directory (required by git):
~user - /home/user/scm

Regards,
Vladimir



___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg



___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [ewg] Moving to new server

[Jeff Becker]

Hi Steve

Steve Wise wrote:

What exactly should I be moving over from the other server?

Just git trees?
  


For now, that's fine. The main thing we want to do right now is migrate 
git.openfabrics.org to the new server. Eventually, we will move 
completely to the new server, but Tziporet said this was not a priority. 
Thanks.


-jeff





Jeff Becker wrote:
  

Hi all. Except for Vlad, no-one has moved their git trees from
openfabrics.org to the new server. I'm happy to go and copy ~/scm/* from
everyone's home directory, but I wanted to give you a chance to pick
what git trees you want and move them yourself (or tell me, and I can).
Just thought this would be a good chance to clean up.  Thanks.

-jeff

Vladimir Sokolovsky wrote:
  


Jeff Becker wrote:
  

  

Hi Vlad. As requested in today's EWG telecon, can we move the build
environment and git trees to the new server (69.55.231.68) this week?
Please let me know what help you need if any. Thanks.

-jeff

  

  


Hi Jeff,
I copied and tested cross compilation environment and OFED build 
environment  on the new server  (69.55.231.68).

Remember to add links under /pub/scm directory (required by git):
~user - /home/user/scm

Regards,
Vladimir
  

  

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
  



  


___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [ewg] Moving to new server

[Jeff Becker]

Hi Sasha

Sasha Khapyorsky wrote:

Hi Jeff,

On 16:24 Fri 27 Mar , Jeff Becker wrote:
  

Just thought this would be a good chance to clean up.



Thanks for your work.

I copied my trees and basically it works. Few things are missing:

1. ~username symbolic links under /pub/scm. So things like:

  git clone git://69.55.231.68/~sashak/management

, don't work yet.

2. Mail server (guess postfix) is not configured and mailx is not
installed - both can be useful (and was used on current server) for
git commit notification.

3. No gitweb yet.
  


I plan on setting these up after a good number of developers have moved 
their git trees - just a handful right now. Thanks.


-jeff


Sasha
  


___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Re: [ewg] Moving to new server

[Jeff Becker]

Hi all. Except for Vlad, no-one has moved their git trees from
openfabrics.org to the new server. I'm happy to go and copy ~/scm/* from
everyone's home directory, but I wanted to give you a chance to pick
what git trees you want and move them yourself (or tell me, and I can).
Just thought this would be a good chance to clean up.  Thanks.

-jeff

Vladimir Sokolovsky wrote:
 Jeff Becker wrote:
   
 Hi Vlad. As requested in today's EWG telecon, can we move the build
 environment and git trees to the new server (69.55.231.68) this week?
 Please let me know what help you need if any. Thanks.

 -jeff

   
 
 Hi Jeff,
 I copied and tested cross compilation environment and OFED build 
 environment  on the new server  (69.55.231.68).
 Remember to add links under /pub/scm directory (required by git):
 ~user - /home/user/scm

 Regards,
 Vladimir
   

___
ewg mailing list
ewg@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg