Quoting r. Sasha Khapyorsky <[EMAIL PROTECTED]>:
> > I think it is best to be running with --user-path=scm so that I can
> > put git trees under scm sub-directory and have that part exported.
>
> '--user-path=scm' in addition to '--base-path'? Looks good for me.
Could this be done please?
I even
Quoting r. Sean Hefty <[EMAIL PROTECTED]>:
> Subject: [PATCH] librdmacm: updated librdmacm to work with proposed 2.6.20
> kernel CMA
>
> Updates the librdmacm to work with ABI version 3, which is the proposed
> kernel changes for inclusion in 2.6.20.
>
> Test programs are also updated.
>
> Sign
On Thu, Oct 26, 2006 at 03:16:50PM +1300, vishal wrote:
> I am interested in moving data from memory in one PC to
> memory in the other using Infiniband hardware. Any suggestions on what
> would be the best way to do it ?
There are 2 main approaches, messages and RDMA. The best one depends
on t
I am interested in moving data from memory in one PC to
memory in the other using Infiniband hardware. Any suggestions on what
would be the best way to do it ?
Thanks!
Vishal
___
openib-general mailing list
openib-general@openib.org
http://openib.or
Hi Matt,
On 18:40 Wed 25 Oct , Matt Leininger wrote:
>
> Subversion is up and running on the new server at
> https://69.55.231.195/svn.
That is great.
What will be syncronization policy in staging period (with old, yet
current SVN repository)? Now I can see that commits >= r9957 are not
Subversion is up and running on the new server at
https://69.55.231.195/svn. Those who have write access to the current
svn should have write access on the new server. Let us know if any
issues arise.
Thanks,
- Matt
___
openib-general
You can now reference the new OpenFabrics server using
staging.openfabrics.org
Johann
___
openib-general mailing list
openib-general@openib.org
http://openib.org/mailman/listinfo/openib-general
To unsubscribe, please visit http://openib.org/mailman/lis
The patch for ib_sa_serv_notice_hdlr() handles multiple callers
registering for the same event by maintaining a linked list of callback
handlers. So I didn't think that the multiple user count was necessary.
The actual registration with the MAD layer happens only when the ib_sa
module is initia
All the patches are attached to the bugzilla bug reports.
Yes, it is possible to enhance ib_cm_init_qp_attr(), but it requires a
new parameter *alternate_path and changing it might break the
compatibility with the existing code. So I thought, it is better to add
new interface, so that it won't
patch14
Description: patch14
___
openib-general mailing list
openib-general@openib.org
http://openib.org/mailman/listinfo/openib-general
To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
patch13
Description: patch13
___
openib-general mailing list
openib-general@openib.org
http://openib.org/mailman/listinfo/openib-general
To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
patch12
Description: patch12
___
openib-general mailing list
openib-general@openib.org
http://openib.org/mailman/listinfo/openib-general
To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
patch11
Description: patch11
___
openib-general mailing list
openib-general@openib.org
http://openib.org/mailman/listinfo/openib-general
To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
patch9
Description: patch9
___
openib-general mailing list
openib-general@openib.org
http://openib.org/mailman/listinfo/openib-general
To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
patch7
Description: patch7
___
openib-general mailing list
openib-general@openib.org
http://openib.org/mailman/listinfo/openib-general
To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
patch6
Description: patch6
___
openib-general mailing list
openib-general@openib.org
http://openib.org/mailman/listinfo/openib-general
To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
patch3
Description: patch3
___
openib-general mailing list
openib-general@openib.org
http://openib.org/mailman/listinfo/openib-general
To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
patch2
Description: patch2
___
openib-general mailing list
openib-general@openib.org
http://openib.org/mailman/listinfo/openib-general
To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
The following set of patches contain the source code for the NetEffect
NE010 adapter. Patches 1 through 9 contain the kernel driver source and
patches 10 through 14 contain the userspace components. Due to license
restrictions the NetEffect TCP/IP stack module cannot be released under
the OFA lic
Updates the librdmacm to work with ABI version 3, which is the proposed
kernel changes for inclusion in 2.6.20.
Test programs are also updated.
Signed-off-by: Sean Hefty <[EMAIL PROTECTED]>
---
Index: include/rdma/rdma_cma_abi.h
===
Or Gerlitz wrote:
> Just making sure, to test user space multicast with these patches, i
> can apply the patches on top of 2.6.19.rcX kernel and use librdmacm
> (and the mckey app) from the svn, correct?
Actually, you will need my out of tree version of librdmacm. Reporting the
connection parame
On 10/25/06, Sean Hefty <[EMAIL PROTECTED]> wrote:
> The following set of patches expand the rdma_cm support to include
> UD and multicast, and expose the rdma_cm to userspace. I would like to
> target the 2.6.20 kernel, but at least getting them into one or more
> branches would be helpful for ot
On 12:50 Wed 25 Oct , Michael S. Tsirkin wrote:
>
> Now, what's the URL to access git from outside?
> I do:
> git clone git://69.55.231.195/~mst/linux-2.6
> and this fails.
git-daemon runs with '--base-path=/pub/scm', so
git clone git://69.55.231.195/linux-2.6.18
will clone /pub/scm/linux
Jack> Why is this any worse than having a situation where a kernel
Jack> module has a bug in which it returns an incorrect value in
Jack> an ABI field in which no name change is involved?
Jack> If we fix such a bug, do we need to increment the ABI as well?
Obviously fixing the ker
Quoting r. Matthew Wilcox <[EMAIL PROTECTED]>:
> Subject: Re: Ordering between PCI config space writes and MMIO reads?
>
> On Wed, Oct 25, 2006 at 12:30:22AM -0600, Grant Grundler wrote:
> > Can someone provide a quote of the PCI Local bus spec that allows this?
> > (Or at least a reference to a s
> So, what we had for ofed 1.1, was a common git tree that had the addons and
> additionally pulled from upstream from time to time. This means, however,
> that
> life was harder for people working just on backports or just scripts as they
> always
> had to check out the full linux tree.
CB.
Looks like we better update git to 1.4.3.2 on the new server.
- Forwarded message from Junio C Hamano <[EMAIL PROTECTED]> -
Subject: Re: [git failure] failure pulling latest Linus tree
Date: Wed, 25 Oct 2006 19:35:03 +0200
From: "Junio C Hamano" <[EMAIL PROTECTED]>
"H. Peter Anvin" <
On Wed, Oct 25, 2006 at 08:18:59AM -0600, Matthew Wilcox wrote:
> PCI-PCI bridges are allowed to do it. If you look in table E-1 of PCI
> 2.3, or table 8-3 of PCI-X 2.0, you'll see that a Posted Memory Write
> can pass a Delayed Write Request (or in PCI-X, a Memory Write can pass a
> Split Write
Tang, Changqing wrote:
> Is there a way (not limited to cm) to know that path 1 is back again and
> reload it as the new alternate path ? If path 1 is down, we can not set
> it as alternate path, right ?
>
> If we can not bring a path back on fly, the usage of APM is limited.
Yes - the usage is c
Yes, I plan on doing this moving forward.
-- Hal
From: Tziporet Koren [mailto:[EMAIL PROTECTED]
Sent: Wed 10/25/2006 11:30 AM
To: Hal Rosenstock; Sean Hefty
Cc: OPENIB
Subject: creating releases for the libraries you own
Hi Hal & Sean,
I want to suggest that
Thanks, In situation that path 1 is primary, path 2 is alternate, when
path 1 failed, traffic migrates to path 2. Now we can think that path 2
is primary, but there is no alternate path.
Is there a way (not limited to cm) to know that path 1 is back again and
reload it as the new alternate path ?
On Wednesday 25 October 2006 16:09, Roland Dreier wrote:
> > Finally, the ABI structure change is only a name change -
> > no need to increment the ABI version.
>
> I don't believe this is true unfortunately. If you have a new
> libibverbs and an old kernel, then you will return the value of
>
> How do you know that the old path is back ? Do you have a notification
> handler called, or you need to query periodically ?
The cm doesn't know that it's back. It records the original path used when
sending the REQ and uses that same path for all future messages.
- Sean
Tziporet Koren wrote:
> I want to suggest that you will create releases to the libraries you own
> (like Roland maintains for libibverbs).
> This will help us in OFED integration since we will be able to start the
> release from a known stable version, instead of taking code with unknown
> stabi
Hi Hal & Sean,
I want to suggest that you will create releases to the libraries you own
(like Roland maintains for libibverbs).
This will help us in OFED integration since we will be able to start the
release from a known stable version, instead of taking code with unknown
stability from svn.
T
How do you know that the old path is back ? Do you have a notification
handler called, or you need to query periodically ?
Thanks.
--CQ
> another question: If one brings back the old path (after the first
> failure) and use the old path record
> to do lap/apr then reenabling migration using L
On Wed, Oct 25, 2006 at 12:30:22AM -0600, Grant Grundler wrote:
> Can someone provide a quote of the PCI Local bus spec that allows this?
> (Or at least a reference to a spec version and section number)
PCI-PCI bridges are allowed to do it. If you look in table E-1 of PCI
2.3, or table 8-3 of PCI
> One thing is that we definitely don't want to fix this by,
> for example, reading back the PCI_COMMAND register or something
> like that. That causes two problems:
>
> 1) Some PCI config writes shut the device down and make it
>no respond to some kinds of PCI config transactions.
>
i tried to built the OFED-1.1 and got the following error when building
ibutils.
What is going wrong?
If this is not the appropiate mailing list pls let me know where to inform.
Thanks.
Sanjeewa
updating cache /var/tmp/OFED/ibutils.cache
configure: creating ./config.status
config.status: creati
> I'm looking at arch/ia64/pci/pci.c.
> Wouldn't it be reasonable to include memory barriers around calls
> to SAL config space access functions?
It's reasonable, but is there a memory barrier strong enough to
guarantee that a config write has actually completed?
- R.
___
> Finally, the ABI structure change is only a name change -
> no need to increment the ABI version.
I don't believe this is true unfortunately. If you have a new
libibverbs and an old kernel, then you will return the value of
en_sqd_async_notify to the consumer but tell the consumer that the
va
> I'll check if there is any additional reordering that can occur AFTER the
> PIO_WRITE_COUNT goes to zero. If so, it would be at bus level - not in
> shub or routers.
Unfortunately, at least in theory, the reordering can occur. For
example a bridge on some card plugged into an SN slot is all
> I'm not sure that can work either. The PCI-X spec is very clear, you
> must wait for a non-posted completion if you care about order. Doing a
> config read in the driver as a surrogate flush is not good enough in
> the general case. Like you say, a pci bridge is free to reorder all
> in flig
Roland (as probably our most experienced git user), all, would like to ask
your ideas about ofed git layout.
Basically, here's what ofed kernel code needs to include:
- upstream kernel (we really only used the infiniband/rdma subdirectories
from there)
- ofed addons, including:
- "backpo
Quoting r. Michael S. Tsirkin <[EMAIL PROTECTED]>:
> Subject: Re: new server up and running
>
> OK, I tested this a bit.
> The machine and the inet connection seem very fast, thanks for that!
>
> An issue I've run into:
>
> [EMAIL PROTECTED]:~$ git clone -n
> 'git://git.kernel.org/pub/scm/linux/
FYI.
Of interest to openib is the change fixing backwards-compatibility -
this might help people running old git clients.
--
MST
--- Begin Message ---
The latest maintenance release GIT 1.4.3.2 is available at the
usual places:
http://www.kernel.org/pub/software/scm/git/
git-1.4.3.2.tar.{gz
Return the sq_draining value back to user space for query_qp instead
of the en_sqd_async notify value. This last is valid only for
modify_qp. For query_qp, the draining status should returned.
Signed-off-by: Jack Morgenstein <[EMAIL PROTECTED]>
---
Roland,
Without this patch, can't manage sq dr
Return the sq_draining value back to user space for query_qp instead
of the en_sqd_async notify value. This last is valid only for
modify_qp. For query_qp (according to the IB Spec V1.2), the
draining status should returned.
Signed-off-by: Jack Morgenstein <[EMAIL PROTECTED]>
---
Roland,
Witho
OK, I tested this a bit.
The machine and the inet connection seem very fast, thanks for that!
An issue I've run into:
[EMAIL PROTECTED]:~$ git clone -n
'git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git'
fatal: read error (Connection reset by peer)
fetch-pack from
'git://git.ke
http://openib.org/bugzilla/show_bug.cgi?id=286
--- Comment #1 from [EMAIL PROTECTED] 2006-10-25 00:30 ---
Can you check it with OFED 1.1?
Also please upgrade FW version to 4.7.600 since this is the version we
qualified OFED 1.1 with
Thanks,
Tziporet
--- You are receiving this
50 matches
Mail list logo