Title: RE: [openib-general] Incorrect endian in GUID comparison/SM master selection
Thanks Brian
>
> Per the IBA spec, the selection of a master SM is determined by the
> GUID (lowest wins). When OpenSM does its comparison, it does not adjust
> for the endian of the host. For our x86 syst
I think the simplest solution the solution most in keeping with the
design of the CM is to always expose the rejection and the contents of
the REJ to the consumer. As you say, that makes the code 24 (port &
CM redirection) case pretty easy to handle.
Code 25 (port redirection only) raises a separ
Hello,
Per the IBA spec, the selection of a master SM is determined by the
GUID (lowest wins). When OpenSM does its comparison, it does not adjust
for the endian of the host. For our x86 systems, this means the
comparison is based on the byte-swapped value of the GUID. It looks
like the patch b
I've started looking at the handling received rejects in the CM. Two
return codes of note:
* code 24 - Port and CM Redirection
* code 25 - Port Redirection
For the first case, it appears that the proper way to handle it is to
simply expose the rejection to the user. The user would need to issu
On Wed, 2005-02-16 at 09:49 -0500, James Lentini wrote:
>
> On Mon, 14 Feb 2005, Matt Leininger wrote:
>
> > Several developers have volunteered to work on uDAPL. The DAT
> > collaborative is working to get a GPL/BSD version of uDAPL to OpenIB.
> > James and Arkady from NetApp gave a DAT talk.
On Wed, 16 Feb 2005 15:34:24 -0800
"Woodruff, Robert J" <[EMAIL PROTECTED]> wrote:
>
>
> >Kanevsky, Arkady wrote,
>
> >If Open IB wants to go with LGPL license it is fine with us.
> >We will need to take a voit on DAT Collaborative onit also.
> >But from what I see on reflector LGPL license is
On Tue, 2005-02-15 at 17:32 -0800, Ryan, Jim wrote:
> Tom Duffy wrote:
> > On Mon, 2005-02-14 at 16:47 -0800, Woodruff, Robert J wrote:
> >> Hi Arkady,
> >>
> >> As I mentioned in the BOF, I have a person (Arlin Davis) that can
> >> help with developing a uDAPL provider for the openib.org verbs.
This patch provides initial timewait handling in the CM.
A connection will remain in the timewait state for a period of time
equal to the local_ack_timeout. After a connection exits the
timewait state, the user is notified of a TIMEWAIT_EXIT event,
assuming that the user has not destroyed their c
I found a panic while testing a lock
patch.
In ipoib_mcast_dev_flush(), list_for_each_entry()
should be list_for_each_entry_safe().
Here is the patch.
diff -urN infiniband/ulp/ipoib/ipoib_multicast.c
infiniband-down/ulp/ipoib/ipoib_multicast.c
--- infiniband/ulp/ipoib/ipoib_multicast.c
>Kanevsky, Arkady wrote,
>If Open IB wants to go with LGPL license it is fine with us.
>We will need to take a voit on DAT Collaborative onit also.
>But from what I see on reflector LGPL license is outside the current
>bylaws of
>Open IB and there is a discussion on it going on.
Good point.
>
This patch adds a check to see if a mad_recv_wc exists before a call is
made to free it. This fixes a crash after a call is made to
ib_cm_establish.
Signed-off-by: Sean Hefty <[EMAIL PROTECTED]>
Index: infiniband/core/cm.c
===
--- i
This patch removes commented out and now outdated code for handling
peer-to-peer connection requests.
Signed-off-by: Sean Hefty <[EMAIL PROTECTED]>
Index: infiniband/core/cm.c
===
--- infiniband/core/cm.c(revision 1808)
+++ i
This patch fixes a race accessing a CM message when destroying a cm_id.
Signed-off-by: Sean Hefty <[EMAIL PROTECTED]>
Index: infiniband/core/cm.c
===
--- infiniband/core/cm.c(revision 1807)
+++ infiniband/core/cm.c(wo
This patch defines a new red-black tree node for SIDR to use. The
separation is done to simplify timewait state handling.
Signed-off-by: Sean Hefty <[EMAIL PROTECTED]>
Index: infiniband/core/cm.c
===
--- infiniband/core/cm.c
On Wed, 2005-02-16 at 13:26, Hal Rosenstock wrote:
> On Wed, 2005-02-16 at 16:08, Nitin Hande wrote:
> > Hal,
[snip..]
> >
> >
> > Here is the trace of 256 sized MTU:
> >
> > Outgoing MAD:
> > BaseVersion: 0x1
> > MgmtClass: 0x3 - SubnAdm
> > ClassVersion: 0x2
> >
On Wed, 2005-02-16 at 16:08, Nitin Hande wrote:
> Hal,
>
> On Wed, 2005-02-16 at 06:27, Hal Rosenstock wrote:
> > On Tue, 2005-02-15 at 16:36, Nitin Hande wrote:
> > > I have a hunch for whats happening here, but before I jump into any
> > > conclusions, I am seeing some other issue between Solari
Hal,
On Wed, 2005-02-16 at 06:27, Hal Rosenstock wrote:
> On Tue, 2005-02-15 at 16:36, Nitin Hande wrote:
> > I have a hunch for whats happening here, but before I jump into any
> > conclusions, I am seeing some other issue between Solaris IPoIB driver
> > and OpenSM. After joining the Broadcast g
On Wed, 2005-02-16 at 11:45, Ronald G. Minnich wrote:
> On Tue, 16 Feb 2005, Hal Rosenstock wrote:
>
> > On Tue, 2005-02-15 at 22:22, Ronald G. Minnich wrote:
> > > On Tue, 15 Feb 2005, Hal Rosenstock wrote:
> > >
> > > > I presume your subnet has 179 HCAs ? Do you know ?
> > >
> > > no errors.
woody,
If Open IB wants to go with LGPL license it is fine with us.
We will need to take a voit on DAT Collaborative onit also.
But from what I see on reflector LGPL license is outside the current
bylaws of
Open IB and there is a discussion on it going on.
SO until this issue is resolved on Open I
On Tue, 16 Feb 2005, Hal Rosenstock wrote:
> On Tue, 2005-02-15 at 22:22, Ronald G. Minnich wrote:
> > On Tue, 15 Feb 2005, Hal Rosenstock wrote:
> >
> > > I presume your subnet has 179 HCAs ? Do you know ?
> >
> > no errors. It's just that opensm won't run.
>
> Won't run or won't do anythin
On Tue, Feb 15, 2005 at 09:51:30PM -0800, Matt Leininger wrote:
> We could put some suggested mail list keywords on the webpage. But
> then we may end up with a mail archive where every SM bug has a subject
> line "SM problem".
I'm ok with that. "SM BUG" are not useful keywords for google any
Paul> Why not consider a more open license still, one that allows
Paul> free commercial or non-commercial use such as the boost
Paul> license ( http://www.boost.org/more/license_info.html ). Is
Paul> there anything to lose by adopting an even less restrictive
Paul> license?
Fir
On Mon, 14 Feb 2005, Matt Leininger wrote:
Several developers have volunteered to work on uDAPL. The DAT
collaborative is working to get a GPL/BSD version of uDAPL to OpenIB.
James and Arkady from NetApp gave a DAT talk. They mentioned needing 1
month to get the GPL code, and another month or t
On Tue, 2005-02-15 at 16:36, Nitin Hande wrote:
> I have a hunch for whats happening here, but before I jump into any
> conclusions, I am seeing some other issue between Solaris IPoIB driver
> and OpenSM. After joining the Broadcast group, the PathRecord Response
> coming from OpenSM signals an err
Title: RE: [openib-general] curious warning message (gen1, ibgd, possibly ot)
Hi,
This message indicates that UDAVs for special QPs are located in the DDR memory and not host memory.
It is a warning since when working with host memory the performance of the MADs is better, however everything
After discussing it more with folks here, is seems to us that perhaps for
the uDAPL user-mode library, it be provided to openib.org under a dual
BSD + LGPL library rather than a BSD + GPL since people normally want to
use LGPL for libraries.
Why not consider a more open license still, one that a
26 matches
Mail list logo
