ScanMail for Microsoft Exchange has blocked an attachment.
Sender = [EMAIL PROTECTED]
Recipient(s) = openib-general@openib.org
Subject = [openib-general] Mail System Error - Returned Mail
Scanning time = 6/3/2005 6:58:14 PM
Action on file blocking:
The attachment DOCUMENT.exe matches the file blo
FYI. r2545 kernel + r2519 userspace
works ok.
Shirley Ma
IBM Linux Technology Center
15300 SW Koll Parkway
Beaverton, OR 97006-6063
Phone(Fax): (503) 578-7638
___
openib-general mailing list
openib-general@openib.org
http://openib.org/mailman/listinfo/
On Fri, Jun 03, 2005 at 02:45:19AM +0300, Michael S. Tsirkin wrote:
...
> Its possible to link libibverbs and libmthca statically.
> I did it once.
While I can get "mthca.a" to link statically, I get warnings
and the binary doesn't work.
-LOADLIBES += -libverbs
+LOADLIBES += -static -libverbs -lp
To add to this, I cannot run `pingpong' on r2539. Because of QP creation
error.
Thanks,
Sayantan.
[EMAIL PROTECTED]:examples] ./a.out --help
./a.out: unrecognized option `--help'
Usage:
./a.outstart a server and wait for connection
./a.out connect to server at
Options:
-p
Hi,
* On Jun,2 Shirley Ma<[EMAIL PROTECTED]> wrote :
>
> I have no problem to create 65000 QPs on r2519.
Thanks for your input. However, I am talking about r2539 and about the
number of *completion* queue entries. I haven't yet tested creating
large number of QPs.
Thanks,
Sayantan.
>
> Shirle
I have no problem to create 65000 QPs
on r2519.
Shirley Ma
IBM Linux Technology Center
15300 SW Koll Parkway
Beaverton, OR 97006-6063
Phone(Fax): (503) 578-7638
___
openib-general mailing list
openib-general@openib.org
http://openib.org/mailman/listinf
Hi,
Today I upgraded Gen2 drivers to revision 2539. I am seeing that there
are CQ & QP creation problems with larger number of CQ entries.
For example, with rdma_lat (part of perftest), I can go upto number of
CQ entries = 256 when I hit an error:
[EMAIL PROTECTED]:perftest] ./rdma_lat --tx-dept
On 6/3/05, Roland Dreier <[EMAIL PROTECTED]> wrote:
>William> Should the device nodes /dev/infiniband_mad and
>William> /dev/infiniband_verbs exist at all? Are they artifacts of
>William> the way the classes work, or a setup problem on my
>William> system? Do other people have these
On Fri, Jun 03, 2005 at 04:50:02PM -0400, William Jordan wrote:
> On 6/3/05, Roland Dreier <[EMAIL PROTECTED]> wrote:
> >William> Should the device nodes /dev/infiniband_mad and
> >William> /dev/infiniband_verbs exist at all? Are they artifacts of
> >William> the way the classes work, o
Patch to change sys class name of userspace cm to match
format of userpace verbs and userspace mad class names.
Signed-off-by: Bill Jordan <[EMAIL PROTECTED]>
Index: gen2/trunk/src/linux-kernel/infiniband/core/ucm.c
===
--- gen2/tru
On Fri, 2005-06-03 at 16:33, Troy Benjegerdes wrote:
> Also, the following fixes ibchecknet for me..
Thanks. Applied.
> (I seem to have a bad cable or two)
Hopefully you can figure it out. You can at least check all the HCA
counters. If they look OK, swap the trunk cable (the one between the 2
On Fri, 2005-06-03 at 16:33, Troy Benjegerdes wrote:
> > > Also, I have two machines in a state right now where they are printing
> > > out:
> > >
> > > kernel: unregister_netdevice: waiting for ib0 to become free. Usage
> > > count = 1
> >
> > That's an IPoIB issue... What was the sequence of e
On Fri, Jun 03, 2005 at 04:08:05PM -0400, Hal Rosenstock wrote:
> On Fri, 2005-06-03 at 15:58, Troy Benjegerdes wrote:
> > > Do you have any additional known good cables to try ?
> >
> > I have several cables I *could* try, but I have no idea which ones are
> > good, or what ports are getting erro
On Fri, 2005-06-03 at 15:39, Tom Duffy wrote:
> I had opensm running on this node, before stopping opensm, unloaded
> ib_mthca, caused oops in kernel.
>
> [EMAIL PROTECTED] ~]# rmmod ib_mthca
> [EMAIL PROTECTED] ~]# general protection fault: [1] SMP
> CPU 1
> Modules linked in: ib_ipoib ib_s
On Fri, 2005-06-03 at 15:58, Troy Benjegerdes wrote:
> > Do you have any additional known good cables to try ?
>
> I have several cables I *could* try, but I have no idea which ones are
> good, or what ports are getting errors.
>
> Some lids seem to respond to 'perfquery', but I haven't been able
On Fri, Jun 03, 2005 at 03:33:48PM -0400, Hal Rosenstock wrote:
> On Fri, 2005-06-03 at 15:24, Troy Benjegerdes wrote:
> > I can't tell anything conclusive from the patch though.
>
> OK. So you can't recreate the original problem ?
>
> > How do I go about debugging the multicast arp/ipoib stuff?
I made several modifications to the MAD layer to assist with debugging this,
and after a multitude of test runs I was able to see the output shown below.
Basically, a send work request is completing, but the MAD associated with
the request has been freed/corrupted. It's likely that the errors
I had opensm running on this node, before stopping opensm, unloaded
ib_mthca, caused oops in kernel.
[EMAIL PROTECTED] ~]# rmmod ib_mthca
[EMAIL PROTECTED] ~]# general protection fault: [1] SMP
CPU 1
Modules linked in: ib_ipoib ib_sa ib_umad ib_mad ib_core nfs lockd md5 ipv6
parport_pc lp pa
On Fri, 2005-06-03 at 15:24, Troy Benjegerdes wrote:
> I can't tell anything conclusive from the patch though.
OK. So you can't recreate the original problem ?
> How do I go about debugging the multicast arp/ipoib stuff?
Do you have any additional known good cables to try ?
-- Hal
> > No, it doesn't seem to help. To get anything to work at all, I seem to
> > need to reload all the IB modules on every maching I want to use ipoib
> > on.
> >
> > There have been two times now I've been able to see about 4 ping
> > packets, and then one of the arp entries seems to go away.
> >
Tom> Should it be the case that bringing down ib0 should kill off
Tom> the other pkey devices:
Looks like a bug -- I'll take a look.
- R.
___
openib-general mailing list
openib-general@openib.org
http://openib.org/mailman/listinfo/openib-genera
Should it be the case that bringing down ib0 should kill off the other
pkey devices:
[EMAIL PROTECTED] ~]# netstat -nr
Kernel IP routing table
Destination Gateway Genmask Flags MSS Window irtt Iface
10.6.98.0 0.0.0.0 255.255.255.0 U 0 0 0 eth
Fab Tillier wrote:
Ok, so this question is from a noob, but here goes anyway. Why can't IPoIB
advertise a larger MTU than the UD MTU, and then just fragment large IP
packets up if they need to go over the IB UD transport? Is there any reason
this couldn't work? If it does, it allows IPoIB to e
On Fri, 2005-06-03 at 15:03, Troy Benjegerdes wrote:
> On Fri, Jun 03, 2005 at 01:52:31PM -0400, Hal Rosenstock wrote:
> > Hi Troy,
> >
> > On Thu, 2005-06-02 at 19:23, Troy Benjegerdes wrote:
> > > I'm having intermittent problems with opensm.. It seems after a while
> > > IPoIB stops working and
On Fri, 2005-06-03 at 14:38, Troy Benjegerdes wrote:
> One of the problems turned out to be a crashed Xserve G5, which seemed
> to have the link up, but the driver was obviously not responding to MAD
> packets.
Yes, that is exactly the scenario than the OpenIB OpenSM vendor layer is
not handling p
On Fri, Jun 03, 2005 at 01:52:31PM -0400, Hal Rosenstock wrote:
> Hi Troy,
>
> On Thu, 2005-06-02 at 19:23, Troy Benjegerdes wrote:
> > I'm having intermittent problems with opensm.. It seems after a while
> > IPoIB stops working and if I restart opensm, it starts spitting out
> > errors.
>
> Pl
On 6/3/05, Christoph Hellwig <[EMAIL PROTECTED]> wrote:
>
> > The next question is to define what if any handshake is desired.
> > My guess that the consumer would acknowledge this by closing
> > the RNIC, and that there would be some sort of deadline for doing
> > so (much like a shutdown, you h
William> Should the device nodes /dev/infiniband_mad and
William> /dev/infiniband_verbs exist at all? Are they artifacts of
William> the way the classes work, or a setup problem on my
William> system? Do other people have these nodes?
I don't get those nodes -- I seem to recall a u
On Fri, Jun 03, 2005 at 01:59:42PM -0400, Hal Rosenstock wrote:
> On Fri, 2005-06-03 at 13:17, Troy Benjegerdes wrote:
> > I've also seen [0][0][0] path indicators.. are those allowed
> > as well?
>
> That is bogus too. Are all the bogus initial paths after a umad_send
> error (message like "umad_
On 6/2/05, Libor Michalek <[EMAIL PROTECTED]> wrote:
> On Thu, Jun 02, 2005 at 03:43:47PM -0400, William Jordan wrote:
> >
> > The udev support/naming works fine. Regarding naming, the ib_uverbs
> > and ib_umad drivers only put port specific devices in the /dev/infiniband
> > directory. They also c
On Fri, 2005-06-03 at 13:17, Troy Benjegerdes wrote:
> I've also seen [0][0][0] path indicators.. are those allowed
> as well?
That is bogus too. Are all the bogus initial paths after a umad_send
error (message like "umad_receiver: send completed with error" in the
osm.log) ? Thanks.
-- Hal
Hi Troy,
On Thu, 2005-06-02 at 19:23, Troy Benjegerdes wrote:
> I'm having intermittent problems with opensm.. It seems after a while
> IPoIB stops working and if I restart opensm, it starts spitting out
> errors.
Please try the following workaround and let me know if this makes things
better.
osm_sminfo_rcv.c: Eliminate some redundant checks
Signed-off-by: Hal Rosenstock <[EMAIL PROTECTED]>
Index: osm_sminfo_rcv.c
===
--- osm_sminfo_rcv.c(revision 2538)
+++ osm_sminfo_rcv.c(working copy)
@@ -160,14 +160,7 @@
On Fri, 2005-06-03 at 13:22, Roland Dreier wrote:
> Hal> sa_query: In send_mad, initialize wr.next to NULL
>
> Does this solve a real problem? As far as I know, the C standard
> requires any pointer fields not explicitly named in a designated
> initializer for a structure to be initialized to
On Fri, 2005-06-03 at 13:14, Troy Benjegerdes wrote:
> On Fri, Jun 03, 2005 at 06:14:34AM -0400, Hal Rosenstock wrote:
> > Usage: perfquery [-d(ebug) -G(uid_addr) -a(ll_ports) -r(reset_after_read)
> > -C ca_name -P hca_port -R(eset_only) -t timeout_ms -V(ersion) -h(elp)]
> > [ [[port] [reset_mask
On Fri, Jun 03, 2005 at 09:13:25AM -0700, Caitlin Bestler wrote:
> Customers are using insmod to load DAPL Providers today, and
> then using the registry to find them. That applies to both IB and
> iWARP providers. The need for the registry reduces with each
> step, but it doesn't instantly vanish
On Fri, Jun 03, 2005 at 04:38:15AM -0700, Caitlin Bestler wrote:
> OpenRDMA discussed with the DAT Collaborative the idea
> of subsuming the responsibilities of the DAT Registry, so that
> the OpenRDMA directory could take the 'dat_xxx' calls
> directly. When the device dependent logic used a dynam
Roland> Does this solve a real problem? As far as I know, the C
Roland> standard requires any pointer fields not explicitly named
Roland> in a designated initializer for a structure to be
Roland> initialized to NULL.
In other words -- since my previous post was kind of incomprehen
On Fri, Jun 03, 2005 at 04:30:51AM -0700, Caitlin Bestler wrote:
> The appropriate plafce to add that would be as an unafilliated asynchronous
> event reported via the async evd.
No. The async evds are a horrible API that should go away not beeing
added to.
> The next question is to define what
Hal> sa_query: In send_mad, initialize wr.next to NULL
Does this solve a real problem? As far as I know, the C standard
requires any pointer fields not explicitly named in a designated
initializer for a structure to be initialized to NULL.
- R.
__
Title: RE: [openib-general] opensm fails to bring up subnet..
So Troy - will you be able to capture an osm.log and send us a tar.gz ?
Eitan Zahavi
Design Technology Director
Mellanox Technologies LTD
Tel:+972-4-9097208
Fax:+972-4-9593245
P.O. Box 586 Yokneam 20692 ISRAEL
> -Original M
On Fri, Jun 03, 2005 at 08:37:04AM -0400, Hal Rosenstock wrote:
> On Thu, 2005-06-02 at 19:23, Troy Benjegerdes wrote:
> > I'm having intermittent problems with opensm.. It seems after a while
> > IPoIB stops working
>
> Wonder if there is some relation to the two: intermittent IPoIB and lack
> o
On Fri, Jun 03, 2005 at 06:14:34AM -0400, Hal Rosenstock wrote:
> On Thu, 2005-06-02 at 20:25, Troy Benjegerdes wrote:
> > Some of my problems seem to be from intermittent cables..
> >
> > Is there anything for OpenIB that can read error counters?
>
> Aside from pulling these from the driver via
On Fri, 2005-06-03 at 12:47, Eitan Zahavi wrote:
> Hi,
> Sorry for catching up with this late in the thread. (Thanks Hal for
> waking me up...)
> >
> > It appears that a node is not responding to a discovery packet (SM
> Get
> > NodeInfo (attrID 0x11)). It's direct route initial path (an array of
Title: RE: [openib-general] opensm fails to bring up subnet..
Hi,
Sorry for catching up with this late in the thread. (Thanks Hal for waking me up...)
>
> It appears that a node is not responding to a discovery packet (SM Get
> NodeInfo (attrID 0x11)). It's direct route initial path (an arra
Jeff Carr wrote:
On the client system:
# cat /sys/class/infiniband/mthca0/ports/1/lid
0x2
On the server system:
# cat /sys/class/infiniband/mthca0/ports/1/lid
0x3
modprobe ib_cmpost "slid=0x3" "dlid=0x2" "message_count=0x10"
"message_size=0x100"
slid is Source LID (not server). dlid is Dest
On Wed, 2005-06-01 at 16:51, Tom Duffy wrote:
> I am putting together a network with a dumb IB switch, a couple of Linux
> OpenIB boxes, a Solaris 10 box, a Solaris Nevada box, etc. I fired up
> opensm on one of the Linux nodes, tried to plumb Solaris, no luck. I
> then hit control-c on opensm a
On 6/3/05, Bob Woodruff <[EMAIL PROTECTED]> wrote:
> Catlin Wrote,
> >That approach is certainly applicable for OpenIB as well.
> >The key is recognizing the need for a transition plan.
> >Customers have DAT Providers installed now, they
> >cannot synchronize getting new DAT Providers from
> >their
Catlin Wrote,
>That approach is certainly applicable for OpenIB as well.
>The key is recognizing the need for a transition plan.
>Customers have DAT Providers installed now, they
>cannot synchronize getting new DAT Providers from
>their suppliers with a new Linux release. This is
>especially true s
On Thu, 2005-06-02 at 19:23, Troy Benjegerdes wrote:
> I'm having intermittent problems with opensm.. It seems after a while
> IPoIB stops working and if I restart opensm,
Another side point: I'm not sure that in all cases IPoIB currently
registers (its multicast) when the SM restarts. In the sin
sa_query: In send_mad, initialize wr.next to NULL
Signed-off-by: Hal Rosenstock <[EMAIL PROTECTED]>
Index: sa_query.c
===
-- sa_query.c (revision 2532)
+++ sa_query.c (working copy)
@@ -433,6 +433,7 @@
int ret;
stru
On Thu, 2005-06-02 at 19:46, Troy Benjegerdes wrote:
> Some more info.. I rebooted the switches, and tried to re-run it.
>
> I found that ibnetdiscover showed everything with a LID of 0 except 1
> HCA card.. when I found that machine and did 'rmmod ib_mthca', opensm
> seemed to get unstuck and ma
On Thu, 2005-06-02 at 19:23, Troy Benjegerdes wrote:
> I'm having intermittent problems with opensm.. It seems after a while
> IPoIB stops working
Wonder if there is some relation to the two: intermittent IPoIB and lack
of response to SM query.
> and if I restart opensm,
How did you get around
On 6/3/05, Christoph Hellwig <[EMAIL PROTECTED]> wrote:
>
> > Keep in mind that loading/unloading DAT Provider is *not*
> > synonymous with loading/unloading drivers. In fact I believe
> > the intent is to have a single provider that supports multiple
> > devices. Such a provider would simply reg
The appropriate plafce to add that would be as an unafilliated asynchronous
event reported via the async evd.
The next question is to define what if any handshake is desired.
My guess that the consumer would acknowledge this by closing
the RNIC, and that there would be some sort of deadline for do
On Thu, 2005-06-02 at 18:53, Sean Hefty wrote:
> And, yes, it would be really nice if the SM reassigned the same LIDs to the
> same nodes where possible.
It does this for some but not all cases now. LIDs are not persistent in
IB unless something above and beyond the IB spec is done.
SM would nee
On Thu, 2005-06-02 at 20:25, Troy Benjegerdes wrote:
> Some of my problems seem to be from intermittent cables..
>
> Is there anything for OpenIB that can read error counters?
Aside from pulling these from the driver via
/sys/class/infiniband/mthca0/ports/1/counters/, there is also perfquery
whi
On Thu, Jun 02, 2005 at 09:56:51AM -0700, Caitlin Bestler wrote:
> Why is it that you believe that the DAT registry does not support
> plug and play? The interface was most specifically designed
> to allow that.
DAT is based on a enumerate and request instead of a callback-based
client interface.
On Thu, Jun 02, 2005 at 09:49:51AM -0700, Caitlin Bestler wrote:
> > The other "issue" right now is that dapl has a header struct that needs
> > to come first in all the structs. So, that would need to be changed.
> >
>
> That is what enables the use of the method table. It allows the
> in-line
Quoting r. Grant Grundler <[EMAIL PROTECTED]>:
> Subject: Re: [PATCH] rdma_lat-09 and results
>
> On Fri, Jun 03, 2005 at 02:45:19AM +0300, Michael S. Tsirkin wrote:
> > > And PLEASE, if you reply, please delete quoted text you are not responding
> > > to from your reply. I'm getting tired of wadi
60 matches
Mail list logo
