Hi Roland,
Thanks. I guess then the only neat solution is NAPI. I shall keep you posted if we can get NAPI implemented in the meanwhile.
thanks,
harishOn 8/29/06, Roland Dreier [EMAIL PROTECTED] wrote:
harish Hi Roland, I was hoping that the function will take careharish of that by introducing
Christoph Raisch wrote on 18.08.2006 17:35:54:
we'll change these EDEBs to a wrapper around dev_err, dev_dbg and
dev_warn as it's done in the mthca driver.
All EDEB_EN and EDEB_EX will be removed, that type of tracing can be
done if needed by kprobes.
There are a few cases where we won't get
On Wednesday 30 August 2006 11:13, Hoang-Nam Nguyen wrote:
Further comments/suggestions are appreciated!
There are a few places in the driver where you declare
external variables (mostly ehca_module and ehca_debug_level)
from C files instead of a header. This sometimes leads
to bugs when a type
Hi Suri,
On Thu, 2006-08-24 at 16:42, Suresh Shelvapille wrote:
Folks:
Is there a utility within the OFED1.0 package which can be used for
generating traffic on different SL (akin to the Voltaire perf_main utility)?
With OpenSM, you can configure an SL per IPoIB based partition and use
any
Hi.
john t wrote:
Hi,
In one of my multi-threaded application (simple send/recv application
written using uverbs), I am repeatedly getting an error code 12
(IB_WC_RETRY_EXC_ERR) from ibv_poll_cq. Not able to figure out
what is going wrong. Cam some one please give a suggestion so that
Sean Hefty wrote:
So I assume this to be a bug at least in the README file. I'm not the
kernel expert to tell where else this has to be changed.
Only the README file should need updating.
We also updated the README of OFED
Tziporet
___
Hoang-Nam Nguyen wrote:
We incorporated those changes throughout ehca code, which is accessible
from
Roland's git tree:
git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband.git
for-2.6.19
Hi Marcus Hoang-Nam,
RC3 is almost closed (going to be out tomorrow).
Following this
Hi Tziporet!
RC3 is almost closed (going to be out tomorrow).
Following this mail I wish to know if we need to update ehca for RC4.
I'm generating a patch against OFED git tree ehca_branch and could provide
you with a patch today if that makes sense to you. In the meanwhile we've
made some local
Hi Sasha,
On Tue, 2006-08-29 at 21:29, Sasha Khapyorsky wrote:
Hi Hal,
On 20:09 Tue 29 Aug , Hal Rosenstock wrote:
Hi Sasha,
On Fri, 2006-08-25 at 09:17, Sasha Khapyorsky wrote:
This provides RPC like API which may work with several ports.
I think you mean can work rather
On Fri, 2006-08-25 at 09:17, Sasha Khapyorsky wrote:
This provides RPC like API which may work with several ports.
Signed-off-by: Sasha Khapyorsky [EMAIL PROTECTED]
---
libibmad/include/infiniband/mad.h |9 +++
libibmad/src/libibmad.map |4 +
libibmad/src/register.c
Quoting r. Hoang-Nam Nguyen [EMAIL PROTECTED]:
Subject: Re: [PATCH 02/13] IB/ehca: includes
Hi Tziporet!
RC3 is almost closed (going to be out tomorrow).
Following this mail I wish to know if we need to update ehca for RC4.
I'm generating a patch against OFED git tree ehca_branch and
On 12:13 Wed 30 Aug , Hal Rosenstock wrote:
Hi Sasha,
On Tue, 2006-08-29 at 21:29, Sasha Khapyorsky wrote:
Hi Hal,
On 20:09 Tue 29 Aug , Hal Rosenstock wrote:
Hi Sasha,
On Fri, 2006-08-25 at 09:17, Sasha Khapyorsky wrote:
This provides RPC like API which may work
libibmad/src/rpc.c: Validate num_classes
Signed-off-by: Hal Rosenstock [EMAIL PROTECTED]
Index: libibmad/src/rpc.c
===
--- libibmad/src/rpc.c (revision 9192)
+++ libibmad/src/rpc.c (working copy)
@@ -306,6 +306,9 @@
Quoting r. Sean Hefty [EMAIL PROTECTED]:
Subject: Re: [PATCH ] RFC IB/cm do not track remote QPN in timewait state
Michael S. Tsirkin wrote:
And so can RTU, in which case again QP will be in RTR. So it seems
lost CM packets aren't protected by timewait.
Maybe we just try to deal with
Hi,
There are a few places in the driver where you declare
external variables (mostly ehca_module and ehca_debug_level)
from C files instead of a header. This sometimes leads
to bugs when a type changes and is therefore considered
bad style.
Good point. See patch attached below.
Moreover,
Quoting r. Michael S. Tsirkin [EMAIL PROTECTED]:
Subject: Re: [PATCH ] RFC IB/cm do not track remote QPN in timewait state
Quoting r. Sean Hefty [EMAIL PROTECTED]:
Subject: Re: [PATCH ] RFC IB/cm do not track remote QPN in timewait state
Michael S. Tsirkin wrote:
And so can RTU, in
We've been testing an application that archives large quantities of data
from a Linux system onto a Windows-based server (64bit server 2003 R2).
As part of the investigation into relatively modest transfer speeds in the
win-linux configuration, we configured a Linux-Linux transfer via IpoIB
Michael S. Tsirkin wrote:
Apparently, list-prev pointer in CMA id_priv structure is NULL
which causes a crash in list_del.
I note that rdma_destroy_id tests outside the mutex lock.
Could that be the problem?
The problem is not unfortunately easily reproducible.
I think I see one bug, but
Hi Paul,
On 8/30/06, Paul Baxter [EMAIL PROTECTED] wrote:
Are there any other inter-operable windows-linux solutions now?
(cross-platform NFS over RDMA or SRP initiator/target?)
There is an SRP initiator for Windows, but not a target. There is a
Linux SRP target reference implementation, but
On Wed, 2006-08-30 at 10:35 -0700, Roland Dreier wrote:
OK, getting closer to finishing the merge...
anyway, why is iw_cm_private.h in include/rdma where it is visible
everywhere? As far as I can tell drivers/infiniband/core/iwcm.c is
the only place it's included. So why not just put this
-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Paul Baxter
Sent: Wednesday, August 30, 2006 2:13 PM
To: openib-general@openib.org
Cc: openib-windows@openib.org
Subject: [openib-general] File transfer performance options
We've been testing an
Quoting r. Sean Hefty [EMAIL PROTECTED]:
Subject: Re: [openib-general] CMA oops
Michael S. Tsirkin wrote:
Apparently, list-prev pointer in CMA id_priv structure is NULL
which causes a crash in list_del.
I note that rdma_destroy_id tests outside the mutex lock.
Could that be the
Michael S. Tsirkin wrote:
I'm trying to come up with a fix for this, but I'm not convinced it's the
problem that you're seeing.
Could be what you describe leads to a memory corruption.
I believe so. If this were the cause of the crash, I would expect to see an
issue with list-prev-prev or
Are you familiar with NFSoRDMA? We get 600+MB/s on read and about
150-200MB/s on write with an XFS filesystem on a stripped software raid
with four spindles.
This is Linux -- Linux.
On Wed, 2006-08-30 at 20:13 +0100, Paul Baxter wrote:
We've been testing an application that archives large
While merging this, I uninlined rdma_node_get_transport, since I don't
think there's any reason to make it inline:
add/remove: 1/0 grow/shrink: 7/16 up/down: 65/-146 (-81)
function old new delta
rdma_node_get_transport- 33
Are you familiar with NFSoRDMA? We get 600+MB/s on read and about
150-200MB/s on write with an XFS filesystem on a stripped software raid
with four spindles.
umm:
[Whilst for a real Linux-Linux configuration I would look for the RDMA
over
NFS solution, this wouldn't translate
Michael S. Tsirkin wrote:
It exposed a race in SDP. The patch itself does not lead to crashes -
I re-attach it here for reference.
As we discussed, this needs to be extended to handle DREQ retries
properly.
I've committed this patch to SVN 9193.
The CM should already handle DREQ retries
Roland Dreier wrote:
While merging this, I uninlined rdma_node_get_transport, since I don't
think there's any reason to make it inline:
I've committed the patch to svn to sync as well.
- Sean
___
openib-general mailing list
openib-general@openib.org
Roland Dreier rdreier at cisco.com writes:
harish Hi, The interruptThresholdRate module parameter allows you
harish to control the maximum number of interrupts/sec for an
harish e1000 Intel NIC for example. Is there an equivalent
harish parameter for Infiniband NICs. I am
The MVAPICH team is pleased to announce the availability of MVAPICH2
0.9.5 with the following NEW features:
- Shared Receive Queue (SRQ) and Adaptive RDMA support: These
features reduce memory usage of the MPI library significantly to
provide scalability without any degradation in
30 matches
Mail list logo
