Just to add there is a Lustre NAL over kDAPL in development
And few other application specific protocols done over kDAPL I know of
All those protocols Arkady mentioned can work on both RDMA technologies
and where designed in such a way (I'm familiar with their code and
architecture).
And another
On Fri, 2005-02-11 at 10:31, Bernhard Fischer wrote:
On Thu, Feb 10, 2005 at 03:26:45PM -0500, Hal Rosenstock wrote:
Initialization continues (not sure it should) and attempting to clear a
I'd say it should not but exit.
static object fails. Even if initialization didn't continue, opensm
On Fri, Feb 11, 2005 at 10:49:08AM -0500, Hal Rosenstock wrote:
On Fri, 2005-02-11 at 10:31, Bernhard Fischer wrote:
On Thu, Feb 10, 2005 at 03:26:45PM -0500, Hal Rosenstock wrote:
Initialization continues (not sure it should) and attempting to clear a
I'd say it should not but exit.
On Fri, 2005-02-11 at 11:04 -0500, Hal Rosenstock wrote:
There are many complex packages which use /usr/local/XXX rather than
just /usr/local as the standard path so we are not sure whether to
move this back to /usr/local. In any case, if this is to be done, we
would prefer to defer it until
Singed-off-by: Tom Duffy [EMAIL PROTECTED]
Index: drivers/infiniband/ulp/sdp/sdp_main.h
===
--- drivers/infiniband/ulp/sdp/sdp_main.h (revision 1766)
+++ drivers/infiniband/ulp/sdp/sdp_main.h (working copy)
@@ -61,6 +61,8
OpenSM: Scope of well known multicast group records needs to match the
scope in the MGID
Index: osm_sa.c
===
--- osm_sa.c(revision 1745)
+++ osm_sa.c(working copy)
@@ -531,7 +531,8 @@
mc_rec.rate = 0x3; /* 10Gb/sec */
On Thu, Feb 10, 2005 at 07:39:56PM -0800, Tom Duffy wrote:
On Thu, 2005-02-10 at 19:09 -0800, Libor Michalek wrote:
OK, I've gone with option number 3. Here is a one or two line
description of each file, and a TODO list. Both are checked in
along with the code in the infiniband/ulp/sdp
OpenSM/complib: Handle complib initialization failures better (e.g.
cleanup and exit gracefully)
Index: cl_complib.c
===
--- cl_complib.c(revision 1745)
+++ cl_complib.c(working copy)
@@ -86,7 +86,7 @@
{
On Fri, Feb 11, 2005 at 09:55:26AM -0800, Grant Grundler wrote:
I ignore references to Lustre since I'm not working on it.
But HPC community watching this list cares alot about Lustre
and more than a few are involved in Lustre developement. I don't
need to alienate them in order to get what I
On Fri, 2005-02-11 at 12:57, Libor Michalek wrote:
I was going to submit this as well, but decided to hold off since
the connection manager has yet to be plumbed into the build. Once
Sean adds the CM, then I'll apply this patch.
Don't we want to be a little bit careful here in terms of being
use C99, otherwise sparse complains
Signed-off-by: Tom Duffy [EMAIL PROTECTED]
Index: drivers/infiniband/ulp/sdp/sdp_proc.c
===
--- drivers/infiniband/ulp/sdp/sdp_proc.c (revision 1766)
+++
The first patch did get SDP building on sparc64, but it would not build
on x86_64. This is the correct fix that gets it to build on both
architectures (as well as x86_32).
Signed-off-by: Tom Duffy [EMAIL PROTECTED]
Index: drivers/infiniband/ulp/sdp/sdp_main.h
On Fri, Feb 11, 2005 at 08:47:08AM -0800, Tom Duffy wrote:
Singed-off-by: Tom Duffy [EMAIL PROTECTED]
Index: drivers/infiniband/ulp/sdp/sdp_main.h
===
--- drivers/infiniband/ulp/sdp/sdp_main.h (revision 1766)
+++
On Fri, Feb 11, 2005 at 11:07:19AM -0800, Grant Grundler wrote:
+#include asm/dma-mapping.h
please use linux/dma-mapping.h instead
nm...you already noticed...
grant
___
openib-general mailing list
openib-general@openib.org
On Fri, Feb 11, 2005 at 07:05:08PM +0100, Christoph Hellwig wrote:
On Fri, Feb 11, 2005 at 09:55:26AM -0800, Grant Grundler wrote:
I ignore references to Lustre since I'm not working on it.
But HPC community watching this list cares alot about Lustre
and more than a few are involved in
Sean Hefty wrote:
Roland Dreier wrote:
Hal Don't we want to be a little bit careful here in terms of
Hal being able to continue to derive patches on the pieces
Hal already submitted to kernel.org until we decide this is ready
Hal to go forward ? (Guess that might be when these
gcc 2.95 fails on the current tree with
cm.c:140: field `path' has incomplete type
Unfortunately, although the C spec says that using just [] to
declare arrays at the end of a struct is correct, gcc 2.95 insists on
[0]. This patch makes that change.
Signed-off-by: Roland Dreier [EMAIL
On Thu, Feb 10, 2005 at 08:15:22PM -0800, Tom Duffy wrote:
use module_param() instead of MODULE_PARM()
Thanks, I've applied and committed.
-Libor
Signed-off-by: Tom Duffy [EMAIL PROTECTED]
Index: drivers/infiniband/ulp/sdp/sdp_inet.c
On Thu, Feb 10, 2005 at 07:57:43PM -0800, Tom Duffy wrote:
Signed-off-by: Tom Duffy [EMAIL PROTECTED]
Thanks, applied and committed.
-Libor
Index: drivers/infiniband/ulp/sdp/sdp_inet.c
===
---
On Fri, Feb 11, 2005 at 10:58:18AM -0800, Tom Duffy wrote:
The first patch did get SDP building on sparc64, but it would not build
on x86_64. This is the correct fix that gets it to build on both
architectures (as well as x86_32).
Thanks, checked in.
-Libor
Signed-off-by: Tom Duffy [EMAIL
On Fri, 2005-02-11 at 20:08, Tom Duffy wrote:
[ Putting on OpenIB ]
On Fri, 2005-02-11 at 19:43 -0500, Hal Rosenstock wrote:
Hey Tom,
On Fri, 2005-02-11 at 19:35, Tom Duffy wrote:
Solaris, you need to ifconfig ibd0 plumb to see it in ifconfig -a. You
are further than me since you
On Fri, Feb 11, 2005 at 10:10:51AM -0800, Tom Duffy wrote:
use C99, otherwise sparse complains
Signed-off-by: Tom Duffy [EMAIL PROTECTED]
Thanks, applied and committed.
-Libor
___
openib-general mailing list
openib-general@openib.org
On Thu, Feb 10, 2005 at 07:39:56PM -0800, Tom Duffy wrote:
On Thu, 2005-02-10 at 19:09 -0800, Libor Michalek wrote:
OK, I've gone with option number 3. Here is a one or two line
description of each file, and a TODO list. Both are checked in
along with the code in the infiniband/ulp/sdp
Remove sparse warnings:
warning: Using plain integer as NULL pointer
Signed-off-by: Tom Duffy [EMAIL PROTECTED]
Index: drivers/infiniband/ulp/sdp/sdp_send.c
===
--- drivers/infiniband/ulp/sdp/sdp_send.c (revision 1778)
+++
On Thu, Feb 10, 2005 at 08:01:01PM -0800, Tom Duffy wrote:
Oops. I only changed it one place. Why is it defined in two places?
You're right, it doesn't need to be in two places. I'll delete
sdp_inet.h since it's redundant, plus make the AF_INET_SDP change.
Here's the patch.
-Libor
Index:
On Fri, Feb 11, 2005 at 05:44:29PM -0800, Tom Duffy wrote:
Remove sparse warnings:
warning: Using plain integer as NULL pointer
Thanks, commited the change.
-Libor
Signed-off-by: Tom Duffy [EMAIL PROTECTED]
Index: drivers/infiniband/ulp/sdp/sdp_send.c
Annotate __user pointers in sdp_inet.c.
Signed-off-by: Tom Duffy [EMAIL PROTECTED]
Index: drivers/infiniband/ulp/sdp/sdp_inet.c
===
--- drivers/infiniband/ulp/sdp/sdp_inet.c (revision 1781)
+++
If no one objects, a patch to clean up compile warnings on x86_64. Most
of the warnings are a result of print format mismatches, the most common
being the need to use %Zu for size_t/sizeof.
For the curious, I'm seeing about 720 MB/s and 17us latency on a pair
of dual 3.4 GHz 64bit Xeon PCIe
On Fri, 2005-02-11 at 18:40 -0800, Libor Michalek wrote:
If no one objects, a patch to clean up compile warnings on x86_64. Most
of the warnings are a result of print format mismatches, the most common
being the need to use %Zu for size_t/sizeof.
Looks good. This patch gets rid of the
Tom Duffy wrote:
On Fri, 2005-02-11 at 20:11 -0500, Hal Rosenstock wrote:
The log indicates that the MTU is 4 which is 2048. I also saw this in
the IB trace.
Ok, I will ask some other Solaris IB guys as well...
Yes, I would like the patch.
BTW, Solaris does work now (Yippie),
On Fri, 2005-02-11 at 20:21 -0800, Nitin Hande wrote:
Allright, so I am not seeing this anymore on the test setup here. I will
keep a watch and conduct some more experiments over weekend if time permits.
Does that mean you fixed it? Or can't reproduce it? Can you ping or
put any traffic
On Fri, 2005-02-11 at 00:58 -0700, Eric W. Biederman wrote:
Matt Leininger [EMAIL PROTECTED] writes:
On Thu, 2005-02-10 at 12:27 -0800, Grant Grundler wrote:
On Thu, Feb 10, 2005 at 12:05:58PM -0800, Matt Leininger wrote:
uDAPL - Oracle, MPI
kDAPL - iSER, NFS over RDMA, Lustre?
Tom Duffy wrote:
On Fri, 2005-02-11 at 20:21 -0800, Nitin Hande wrote:
Allright, so I am not seeing this anymore on the test setup here. I will
keep a watch and conduct some more experiments over weekend if time permits.
Does that mean you fixed it? Or can't reproduce it? Can you ping
On Fri, Feb 11, 2005 at 06:40:58PM -0800, Libor Michalek wrote:
If no one objects, a patch to clean up compile warnings on x86_64.
I haven't applied this patch yet - I read mail on the other side of
a firewall where my machines are. But I don't think it will fix
the compile error below.
I'm
The presentations given at the 2005 OpenIB Developers Workshop in
Sonoma this week are available at www.openib.org/workshop.html. We
still have a few more presentations to track down, but most of them are
there.
- Matt
___
openib-general
On Fri, Feb 11, 2005 at 09:58:27PM -0800, Grant Grundler wrote:
I'm guessing _sdp_iocb_page_save() is just broken.
The ia64 kernel builds fine.
And every other use of pmd_offset() looks like this:
mm/rmap.c: pmd = pmd_offset(pud, address);
The following patch makes
36 matches
Mail list logo