David Miller [EMAIL PROTECTED] wrote on 08/09/2007 09:57:27 AM:
Patrick had suggested calling dev_hard_start_xmit() instead of
conditionally calling the new API and to remove the new API
entirely. The driver determines whether batching is required or
not depending on (skb==NULL) or not.
Sean Hefty wrote:
I was leaning towards always including the SID with the PR query, but I want to
understand the consequences of doing so. The result is that it gives all of the
policy to the SA. Does anyone disagree with this approach?
I think its fine to let the SA do all the policy
Michael, OFED 1.2 is based on 2.6.21 and OFED 1.2.1 on 2.6.22, did you
do some 1.2 wrap up of returning fixes to the upstream code? from the
length of the patch list (specifically, the amount of ipoib and mthca
ones) it does not seem so, please correct me if I am wrong.
In 1.2.c most
This email was generated automatically, please do not reply
git_url: git://git.openfabrics.org/ofed_1_2/linux-2.6.git
git_branch: ofed_1_2
Common build parameters: --with-ipoib-mod --with-sdp-mod --with-srp-mod
--with-user_mad-mod --with-user_access-mod --with-mthca-mod --with-core-mod
Tziporet Koren wrote:
The patches that does not include sign-off and explanation were posted
by Eli, and he forgot this by mistake - to be fixed.
All the changes are going to be send for the list for review next week.
thanks for the clarification, however, I don't see a reason not to post
a
/ofa_1_2_c_kernel-20070809-0200_linux-2.6.19_powerpc_check/drivers/infiniband/hw/ehca/ehca_main.c:831:
error: invalid type argument of â-â
/home/vlad/tmp/ofa_1_2_c_kernel-20070809-0200_linux-2.6.19_powerpc_check/drivers/infiniband/hw/ehca/ehca_main.c:834:
error: invalid type argument of â-â
/home
Signed-off-by: Joachim Fenkes [EMAIL PROTECTED]
---
This patch should apply cleanly on top of Stefan's recent patchset. Please
review and apply for 2.6.23. Thanks.
drivers/infiniband/hw/ehca/ehca_hca.c | 10 +++---
1 files changed, 7 insertions(+), 3 deletions(-)
diff --git
Michael S. Tsirkin wrote:
In 1.2.c most patches are backported from 2.6.23.
Some have been submitted for 2.6.24.
Some have been carried from 1.2.
I see mistakes in the process here: for example patches that solve IB
bugs in 2.6.22 should have been submitted to -stable and this ofed
version
Hi,
For SGE tight integration we need the use of smpd rather than mpd.
Is it possible to have this as an ofed.conf configuration option?
For instance:
MVAPICH2_CONF_mp=smpd
I don't mind unpacking the tar from the SRPM and patching the
make.mvapich.* files once in a while but it would save us
Quoting Tang, Changqing [EMAIL PROTECTED]:
Subject: RE: RFCv2: SRC API
I have another question, when using SRQ, or SRC+SRQ, when a
completion
is returned by ibv_poll_cq(), is there an easy way to find
who sent
this message to me ? 'struct ibv_wc' does not say much about
Michael S. Tsirkin wrote:
Quoting Steve Wise [EMAIL PROTECTED]:
Subject: problems with ofed-1.2.c chelsio low level driver
Michael,
I've discovered that I was really testing the ofed-1.2 cxgb3.ko module
in my testing 1.2.c regression test. Upon getting a complete 1.2.c
installed, I'm
Hi Sean.
Valgrind is a great tool which helps many developers finds bugs in their
code
(and sometimes the bugs are found in the used libraries).
Do you plan to add support for valgrind in librdmacm and libibcm?
thanks
Dotan
___
general mailing list
Steve Wise wrote:
Tziporet,
Can we make the change to 1.2.5? This should probably include:
- change builds/connectx to builds/ofed-1.2.5 (or just add a link)
- change build names from 1.2.c to 1.2.5
I think we will go for this
Tziporet
___
Hi,
is it passible to create a virtual interface over a bonding interface?
We have created two bonding interfaces in this way:
cat /etc/infiniband/openib.conf
.# Enable the bonding driver on startup
IPOIBBOND_ENABLE=yes
# Set bond interface names
IPOIB_BONDS=ibbond0,ibbond1
# Set specific bond
Quoting Steve Wise [EMAIL PROTECTED]:
Subject: Re: problems with ofed-1.2.c chelsio low level driver
Michael S. Tsirkin wrote:
Quoting Steve Wise [EMAIL PROTECTED]:
Subject: problems with ofed-1.2.c chelsio low level driver
Michael,
I've discovered that I was really testing the
Hi All,
We wish to have the final build with latest fixes next Monday
This means ALL changes must be ready till end of Sunday (for US end of
the week).
Release target date is Wed 15-Aug
Steav Hoang-Nam - please send me RN for your drivers (the file itself,
or a patch against my git tree:
This is a query regarding IPoIB HA behaviour.
If I physically pull of the cable from IB port corresponding to primary
IPoIB interface, I notice that the configuration of primary IPoIB interface
is migrated to the secondary IPoIB interface. But, if the primary interface
is made down just by
Roland, Michael,
Could you please review this patch and provide your comments? I would like to
finish
up this work at the soonest and have this integrated into the for-2.6.24 tree.
Pradeep
___
general mailing list
general@lists.openfabrics.org
Quoting Pradeep Satyanarayana [EMAIL PROTECTED]:
Subject: Re: [ofa-general] [PATCH][RFC] Handle packet received on RQ of tx_qp
with NOSRQ
Roland, Michael,
Could you please review this patch and provide your comments? I would like to
finish
up this work at the soonest and have this
Do you plan to add support for valgrind in librdmacm and libibcm?
I have no objections to it being added, but have no plans to add it
myself at this time.
- Sean
___
general mailing list
general@lists.openfabrics.org
Please join us for Hot Interconnects: The 15th annual three-day
Symposium at Stanford University. IEEE Hot Interconnects brings together
computer architects and chip designers to discuss hardware and software
issues for next-generation network systems. Universities and companies
discuss the
+++ b/SRC.txt
@@ -0,0 +1,138 @@
+Here's some documentation on Scalable Reliable Connections.
I think this would be more useful in the man page for
ibv_open_src_domain() (which needs to be written ;).
You know, reading over the patch, it strikes me that src is a very
confusing acronym,
Roland: do you prefer that every fix will be posted in a separate
patch (with the warning it fixes) or one patch for
all of the fixes i have?
Either way is fine I guess.
- R.
___
general mailing list
general@lists.openfabrics.org
Quoting Roland Dreier [EMAIL PROTECTED]:
Subject: Re: [PATCHv4 RFC] Scalable Reliable Connection: API and documentation
+++ b/SRC.txt
@@ -0,0 +1,138 @@
+Here's some documentation on Scalable Reliable Connections.
I think this would be more useful in the man page for
You know, reading over the patch, it strikes me that src is a very
confusing acronym, since it really looks like an abbreviation for
source. I don't know what would be better off the top of my head
but ideas would be appreciated.
No idea.
How about something along the lines of connection
I actually took pains to avoid breaking the ABI, so I think we won't
need to change the version and/or create compat wrappers. No?
It doesn't seem so to me -- for example
@@ -408,6 +415,7 @@ struct ibv_qp_init_attr {
struct ibv_qp_cap cap;
enum ibv_qp_type
Quoting Roland Dreier [EMAIL PROTECTED]:
Subject: Re: [PATCHv4 RFC] Scalable Reliable Connection: API and documentation
I actually took pains to avoid breaking the ABI, so I think we won't
need to change the version and/or create compat wrappers. No?
It doesn't seem so to me -- for
I'd like to propose the following change as a simple solution for handling SA
scalability problems:
Modify the ib_sa module to support an SA LID that's separate from the SM LID.
This concept is supported by the spec through SA redirection; however, I propose
that we also allow the SA LID to be
Any more comments?
Steve Wise wrote:
Networking experts,
I'd like input on the patch below, and help in solving this bug
properly. iWARP devices that support both native stack TCP and iWARP
(aka RDMA over TCP/IP/Ethernet) connections on the same interface need
the fix below or some
Quoting Sean Hefty [EMAIL PROTECTED]:
Subject: Re: [ofa-general] Re: [PATCHv4 RFC] Scalable Reliable Connection:
API?and documentation
You know, reading over the patch, it strikes me that src is a very
confusing acronym, since it really looks like an abbreviation for
source. I don't know
All... a few questions on connectx support in OFED. I'm currently
trying to benchmark a machine with connectx HCAs. What version of OFED
properly supports connectx. Has anyone done any testing regarding which
MPI distro performs best [or at all] on connectx gear?
thanks in advance
--john
We get lots of kernel unaligned access warnings from
this one on ia64.
user_mad.c |3 ++-
1 files changed, 2 insertions(+), 1 deletion(-)
Signed-off-by: Arthur Kepner [EMAIL PROTECTED]
--
--- ofa_kernel-1.2.orig/drivers/infiniband/core/user_mad.c 2007-06-21
07:38:47.0 -0700
OFED 1.2.c supports ConnectX (OFED 1.2 does not).
Cisco is testing Open MPI, MVAPICH, MVAPICH2, Intel MPI, and HP MPI
successfully with ConnectX.
Scott Weitzenkamp
SQA and Release Manager
Server Virtualization Business Unit
Cisco Systems
-Original Message-
From: [EMAIL PROTECTED]
I thought we already dealt with this one in commit fa7252ed?
___
general mailing list
general@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/general
To unsubscribe, please visit
On 8/9/07, Amar Mudrankit [EMAIL PROTECTED] wrote:
This is a query regarding IPoIB HA behaviour.
Its unclear from your question if you use the IPoIB HA daemon or the IPoIB
bonding. The first one is not going to be supported in OFED 1.3 and at least
two companies (Cisco and Voltaire) has
On 8/9/07, Peter Selzner [EMAIL PROTECTED] wrote:
is it passible to create a virtual interface over a bonding interface?
what is a virtual interface? do you mean to an alias?
We have created two bonding interfaces in this way:
cat /etc/infiniband/openib.conf
ibbond0_IP=10.252.10.21
Steve Wise wrote:
Any more comments?
Does anyone have ideas on how to reserve the port space without using a
struct socket?
- Sean
___
general mailing list
general@lists.openfabrics.org
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/general
From: Sean Hefty [EMAIL PROTECTED]
Date: Thu, 09 Aug 2007 14:40:16 -0700
Steve Wise wrote:
Any more comments?
Does anyone have ideas on how to reserve the port space without using a
struct socket?
How about we just remove the RDMA stack altogether? I am not at all
kidding. If you guys
How about we just remove the RDMA stack altogether? I am not at all
kidding. If you guys can't stay in your sand box and need to cause
problems for the normal network stack, it's unacceptable. We were
told all along the if RDMA went into the tree none of this kind of
stuff would be an issue.
On Mon, 2007-08-06 at 15:30 +0200, Hoang-Nam Nguyen wrote:
Hello Doug and Scott!
On Thursday 02 August 2007 18:08, Michael S. Tsirkin wrote:
ehca backports for kernel.org kernels seem to be broken.
1. Does anyone care enough to fix them? If not we'll disable
ehca in build for these
On Thu, Aug 09, 2007 at 01:54:31PM -0700, Roland Dreier wrote:
I thought we already dealt with this one in commit fa7252ed?
I didn't see anything which obviously addresses this in
any tree I found at kernel.org or openfabrics.org, but
maybe I didn't look in the right tree.
--
Arthur
Hey Sasha, Dave,
See below...
On Thu, 9 Aug 2007 00:42:39 +0300
Sasha Khapyorsky [EMAIL PROTECTED] wrote:
Hi,
On 09:41 Wed 01 Aug , [EMAIL PROTECTED] wrote:
Add ibcli program and man page to infiniband-diags.
[snip...]
+.SH SYNOPSIS
+.B ibcli
+get [archive]
From 3c78f9f7797a9029b33a1a80476f82929f8a7ad7 Mon Sep 17 00:00:00 2001
From: Ira K. Weiny [EMAIL PROTECTED]
Date: Thu, 9 Aug 2007 18:51:09 -0700
Subject: [PATCH] Change the name of set_mthca_nodedesc.sh to set_nodedesc.sh
because it should
work for all HCA types now.
Signed-off-by: Ira K. Weiny
From 3abf9ac10ee90f7272969486704c2989d254a2ac Mon Sep 17 00:00:00 2001
From: Ira K. Weiny [EMAIL PROTECTED]
Date: Thu, 9 Aug 2007 16:07:38 -0700
Subject: [PATCH] Modify set_mthca_nodedesc to set the node_desc for any hca
type found in sysfs.
This has been tested for the MLX4 and mthca cards but
I have changed set_mthca_nodedesc.sh to look for all HCA's in sysfs and set the
nodedesc to the host name.
I have tested it on MLX4 and mthca cards and it works fine. I don't know if
other dirvers allow writes to their nodedesc sysfs entry. If they do this
should work for all cards.
I thought we already dealt with this one in commit fa7252ed?
I didn't see anything which obviously addresses this in
any tree I found at kernel.org or openfabrics.org, but
maybe I didn't look in the right tree.
this is the commit I'm talking about; I think it went into 2.6.21 but
On 8/9/07, Ira Weiny [EMAIL PROTECTED] wrote:
From 3abf9ac10ee90f7272969486704c2989d254a2ac Mon Sep 17 00:00:00 2001
From: Ira K. Weiny [EMAIL PROTECTED]
Date: Thu, 9 Aug 2007 16:07:38 -0700
Subject: [PATCH] Modify set_mthca_nodedesc to set the node_desc for any hca
type found in sysfs.
OSM Simulation Regression Summary
[Generated mail - please do NOT reply]
OpenSM rev = Wed_Aug_8_22:48:15_2007 [83d1f5ea339704f30e26918a897f9ead5a10678e]
ibutils rev = Tue_Mar_13_14:36:32_2007
[80aaff94f0eb65117db39b9db7d609ffdcc055de]
Total=520 Pass=461 Fail=59
Pass:
39 LidMgr
Michael S. Tsirkin wrote:
Quoting Pradeep Satyanarayana [EMAIL PROTECTED]:
Subject: Re: [ofa-general] [PATCH][RFC] Handle packet received on RQ of
tx_qp with NOSRQ
Roland, Michael,
Could you please review this patch and provide your comments? I would like
to finish
up this work at the
49 matches
Mail list logo