ere is one issue I need to fix.
Let me know if you have additional questions.
-- Hal
-Forwarded Message-----
From: Hal Rosenstock <[EMAIL PROTECTED]>
To: "Chahande, Takshak" <[EMAIL PROTECTED]>
Cc: openib-general@openib.org
Subject: Re: [openib-general] Data Structures a
On Fri, 2006-04-07 at 17:26, Roland Dreier wrote:
> Hal> The simplest way I can think of to add this from an ABI/API
> Hal> perspective is to add an ioctl to user_mad for this. Prior to
> Hal> the kernel supporting this, libibumad can just return not
> Hal> supported for the "is dua
Hi,
I am in the process of adding support for SA MultiPathRecord to OpenSM.
Dual sided RMPP is not supported in the current RMPP support nor is the
use of dual sided RMPP by OpenSM (or the vendor SA client).
It can be added but may require an API change and possibly an ABI
change. It seems that
Hi Takshak,
On Thu, 2006-04-06 at 19:27, Chahande, Takshak wrote:
> Hi Hal and others,
>
> I find that, there are no standard header files exists at userspace
> which can define structure for PORT_INFO, NODE_INFO and other elements
> like Path Records, Service record etc.
There is no user spac
On Thu, 2006-04-06 at 17:56, Sean Hefty wrote:
> >There's also code in opensm/osm_sa_mcmember_record.c that you can
> >peruse.
>
> Thanks - that's helpful.
>
> >> The code uses a promotion/demotion mechanism based on a reference count of
> >> membership types. The restriction is that only a sing
On Thu, 2006-04-06 at 17:21, Sean Hefty wrote:
> >> I don't think it's needed. MGID and PortGID together form the record
> >> identifier for multicast groups.
> >
> >I see your point so perhaps it is not needed. For IPoIB, that clearly
> >works as the PKey is embedded in the MGID. I didn't think t
On Thu, 2006-04-06 at 17:17, Roland Dreier wrote:
> Hal> I see your point so perhaps it is not needed. For IPoIB, that
> Hal> clearly works as the PKey is embedded in the MGID. I didn't
> Hal> think that was the case with generalized IB multicast.
>
> A multicast group is uniquely iden
On Thu, 2006-04-06 at 17:01, Sean Hefty wrote:
> >1. My main initial comment is that I think that cmp_rec needs to be more
> >complicated that the matching which is there. The selectors include
> >things like greater than, less than, and largest available in addition
> >to equal to which is what is
On Thu, 2006-04-06 at 16:58, Roland Dreier wrote:
> Hal> Why did the PKey component get removed from the leave ?
>
> I don't think it's needed. MGID and PortGID together form the record
> identifier for multicast groups.
I see your point so perhaps it is not needed. For IPoIB, that clearly
w
On Thu, 2006-04-06 at 14:13, Sean Hefty wrote:
Another question:
> +static int send_leave(struct mcast_group *group, u8 leave_state)
> +{
> + struct mcast_port *port = group->port;
> + struct ib_sa_mcmember_rec rec;
> + int ret;
> +
> + rec = group->rec;
> + rec.join_state = l
On Thu, 2006-04-06 at 14:13, Sean Hefty wrote:
> Add kernel support that tracks joining and leaving multicast groups.
> The SA tracks join/leave operations on a per port basis. In order
> to support multiple users of the same multicast group, we need to
> track join / leave requests locally.
On i
On Thu, 2006-04-06 at 16:16, Sean Hefty wrote:
> Roland Dreier wrote:
> > > +dev = kmalloc(sizeof *dev + device->phys_port_cnt * sizeof
> > *port,
> > > + GFP_KERNEL);
> > > +if (!dev)
> > > +return;
> > > +
> > > +for (i = 1; i <=
On Thu, 2006-04-06 at 15:51, Roland Dreier wrote:
> > + dev = kmalloc(sizeof *dev + device->phys_port_cnt * sizeof *port,
> > +GFP_KERNEL);
> > + if (!dev)
> > + return;
> > +
> > + for (i = 1; i <= device->phys_port_cnt; i++) {
>
> Seems like this is implicitly a
On Thu, 2006-04-06 at 09:57, Tziporet Koren wrote:
> Bryan O'Sullivan wrote:
> > Unfortunately, the coordination of this with the 1.0 process has thus
> > far not been very effective, so I am spending a lot of time manually
> > filtering diffs to see what has changed between the first EWG software
On Thu, 2006-04-06 at 09:57, Tziporet Koren wrote:
> Bryan O'Sullivan wrote:
> > Unfortunately, the coordination of this with the 1.0 process has thus
> > far not been very effective, so I am spending a lot of time manually
> > filtering diffs to see what has changed between the first EWG software
Hi Yael,
On Thu, 2006-04-06 at 05:01, Yael Kalka wrote:
> Hi Hal,
>
> > -Original Message-
> > From: Hal Rosenstock [mailto:[EMAIL PROTECTED]
> > Sent: Wednesday, April 05, 2006 1:59 PM
> > To: Yael Kalka
> > Cc: openib-general@openib.org; Eitan Z
On Thu, 2006-04-06 at 07:52, Yael Kalka wrote:
> Hi Hal,
> The patch fixes the problem I saw.
> Please apply it.
Thanks. Applied to both trunk and 1.0 branch.
-- Hal
> Thanks,
> Yael
>
> > -Original Message-
> > From: Hal Rosenstock [mailto:[EMAIL PROTECT
On Thu, 2006-04-06 at 01:28, Krishna Kumar2 wrote:
> --- a/drivers/infiniband/core/ping.c2006-03-28 18:01:31.0
> +0530
> +++ b/drivers/infiniband/core/ping.c2006-03-28 18:06:09.0
> +0530
> @@ -259,10 +259,11 @@ static void ib_ping_init_device(struct i
> }
>
>
for GSI classes assumes RMPP. There are two
groups of GSI classes: those which support RMPP and those which don't.
This patch handles them properly in osm_vendor_send.
Problem pointed out by Yael Kalka <[EMAIL PROTECTED]>
Signed-off-by: Hal Rosenstock <[EMAIL PROTECTED]>
I
Hal
>
> Yael
>
> > -Original Message-
> > From: Hal Rosenstock [mailto:[EMAIL PROTECTED]
> > Sent: Wednesday, April 05, 2006 2:12 PM
> > To: Yael Kalka
> > Cc: openib-general@openib.org; Eitan Zahavi; Sasha Khapyorsky; Ofer
> Gigi
> > Su
GSI classes: those which support RMPP and those which don't.
This patch handles them properly in osm_vendor_send.
Problem pointed out by Yael Kalka <[EMAIL PROTECTED]>
Signed-off-by: Hal Rosenstock <[EMAIL PROTECTED]>
Index: inclu
On Wed, 2006-04-05 at 07:11, Hal Rosenstock wrote:
> Hi Yael,
>
> On Wed, 2006-04-05 at 04:54, Yael Kalka wrote:
> > Hi Hal,
> >
> > We saw the following problem in the osm_vendor_send mad (in
> > osm_vendor_ibumad.c). Currently, there is a case on the Manag
Hi Yael,
On Wed, 2006-04-05 at 04:54, Yael Kalka wrote:
> Hi Hal,
>
> We saw the following problem in the osm_vendor_send mad (in
> osm_vendor_ibumad.c). Currently, there is a case on the Management
> Class values, where the cases are
> IB_MCLASS_SUBN_DIR/IB_MCLASS_SUBN_LID and default, when the
Hi Yael,
On Wed, 2006-04-05 at 02:24, Yael Kalka wrote:
> Hi Hal,
>
> I saw that the complib patch (removal of constructor and destructor
> attribute), wasn't fully added to the branch.
> Attached is a patch for the branch.
Is this needed for 1.0 ? Is this safe to add ? Was there more to it tha
On Wed, 2006-04-05 at 02:31, Yael Kalka wrote:
> Hi Hal,
>
> Attached is a patch for branch 1.0, for the osm_vendor_mlx_svc.h that
> was applied on the trunk but not on the branch 1.0.
>
> Yael
>
> OpenSM/osm_vendor_mlx_svc.h: Identify RMPP MADs
>
> RMPP mads can only sent in 4 MAD classes:
>
On Tue, 2006-04-04 at 16:37, Roland Dreier wrote:
> Sean> When does ipoib join a multicast group for send and receive
> Sean> membership? It looks like it joins the broadcast group for
> Sean> send/receive, but all other groups are send-only. Is this
> Sean> correct?
>
> Not quit
On Mon, 2006-04-03 at 14:55, Roland Dreier wrote:
> Hal> Hi Roland, I have a port which only has the full default
> Hal> partition configured but ifconfig allows an IPoIB interface
> Hal> with a PKey which is not in the Pkey table. Shouldn't the
> Hal> ifconfig fail for this (rather
Hi Mark,
On Mon, 2006-04-03 at 16:08, Uncle Fester wrote:
> I am interested in getting my hands messy with Infiniband--and perhaps
> in eventually contributing in some small way to its progress. I have
> done very rudimentary research, but would like to confirm my
> assumptions with the experts h
Hi Roland,
I have a port which only has the full default partition configured but
ifconfig allows an IPoIB interface with a PKey which is not in the Pkey
table. Shouldn't the ifconfig fail for this (rather than the subsequent
ping) ?
-- Hal
smpquery pkeys 1 1
0: 0x 0x 0x 0x 0
OpenSM/osm_sa_portinfo_record.c: Fix precedence in bitwise compare of
PortInfo capability mask
Signed-off-by: Hal Rosenstock <[EMAIL PROTECTED]>
Index: opensm/osm_sa_portinfo_record.c
===
--- opensm/osm_sa_portinfo_re
On Mon, 2006-04-03 at 03:11, Eitan Zahavi wrote:
> Hi Hal
>
> This patch adds support for the following 1.2 errata MGTWG8372.
> This should be useful for scalability of:
> * SRP target discovery and
> * Queries for all SM ports.
>
> Reference ID: 4291
>
> Add to table: 186 SA-Specific ClassPor
On Mon, 2006-04-03 at 04:01, Tziporet Koren wrote:
> Eitan Zahavi wrote:
> > Hi Hal
> >
> > This patch adds support for the following 1.2 errata MGTWG8372.
> > This should be useful for scalability of:
> > * SRP target discovery and
> > * Queries for all SM ports.
> >
> >
> Hal,
>
> Please che
On Wed, 2006-03-29 at 14:47, Sasha Khapyorsky wrote:
> Hello,
>
> There are fixes in MTU/rate verification/setup when MC Group is created.
> Most of them were discussed in this thread:
>
> http://openib.org/pipermail/openib-general/2006-March/01.html
>
> Sasha.
>
>
> Fixes in MTU/rate veri
On Wed, 2006-03-29 at 18:54, Rolf Manderscheid wrote:
> Hi Hal,
>
> opensm attempts to set pkey table entries on external switch ports
> even if the switch declares a PartitionEnforcementCap of zero. The
> consequence is ERR 4108.
>
> The decision to set the block is based on the size of the
> p
On Fri, 2006-03-31 at 13:55, Rolf Manderscheid wrote:
> Hi Hal,
>
> Trivial patch to add Obsidian's official vendor id.
>
> Signed-off-by: Rolf Manderscheid <[EMAIL PROTECTED]>
>
> ---
>
> Rolf
Thanks. Applied.
-- Hal
___
openib-general mailing
On Thu, 2006-03-30 at 15:28, Eitan Zahavi wrote:
> Hi Hal
>
> Just found out that if the user provides --with-umad-libs and not
> --with-umad-includes
> the compilation breaks. This patch fixes this.
>
> Eitan
>
> Signed-off-by: Eitan Zahavi <[EMAIL PROTECTED]>
Thanks. Applied to both trunk a
On Thu, 2006-03-30 at 15:31, Eitan Zahavi wrote:
> Hi Hal
>
> For some reason this function that is useful for finding the CA index
> and port by GUID was missing in the libosmvendor.map
>
> Eitan
>
> Signed-off-by: Eitan Zahavi <[EMAIL PROTECTED]>
Thanks. Applied.
___
On Thu, 2006-03-30 at 11:26, Sasha Khapyorsky wrote:
> Hello Hal,
>
> On 06:41 Thu 30 Mar , Hal Rosenstock wrote:
> > On Wed, 2006-03-29 at 18:54, Rolf Manderscheid wrote:
> > > Hi Hal,
> > >
> > > opensm attempts to set pkey table entries on exte
On Thu, 2006-03-30 at 13:38, Rolf Manderscheid wrote:
> >>
> >>> Also, is PortInfo:PartitionEnforcementInbound/Outbound set properly (0)
> >>> for these ports which don't support partition enforcement ? (That might
> >>> be another issue).
> >>>
> >>>
> >> Yes, those components are s
On Thu, 2006-03-30 at 13:04, Rolf Manderscheid wrote:
> Hi Hal,
> > I presume this is with the trunk version, right ?
> Right.
> >
> > This appears like it would get past the ERR 4108 but wouldn't OpenSM
> > then insert the block several lines down in
> > osm_port.c::osm_physp_set_pkey_tbl where it
On Thu, 2006-03-30 at 13:04, Rolf Manderscheid wrote:
> Hi Hal,
> > I presume this is with the trunk version, right ?
> Right.
> >
> > This appears like it would get past the ERR 4108 but wouldn't OpenSM
> > then insert the block several lines down in
> > osm_port.c::osm_physp_set_pkey_tbl where it
On Thu, 2006-03-30 at 12:31, Sasha Khapyorsky wrote:
> On 12:20 Thu 30 Mar , Hal Rosenstock wrote:
> > >
> > > > Also, is PortInfo:PartitionEnforcementInbound/Outbound set properly (0)
> > > > for these ports which don't support partition enforceme
On Thu, 2006-03-30 at 11:26, Sasha Khapyorsky wrote:
> Hello Hal,
>
> On 06:41 Thu 30 Mar , Hal Rosenstock wrote:
> > On Wed, 2006-03-29 at 18:54, Rolf Manderscheid wrote:
> > > Hi Hal,
> > >
> > > opensm attempts to set pkey table entries on exte
Hi Jack,
On Thu, 2006-03-30 at 09:51, Jack Morgenstein wrote:
> management/diags: updating branch 1.0 from main trunk.
The ibnetdiscover/grouping changes are still meant only for the trunk
and are not yet ready for 1.0.
Also, saquery source has been checked in but there is a pending issue
and th
On Thu, 2006-03-30 at 05:40, Eitan Zahavi wrote:
> Hi Hal
>
> A previous patch adding explicit set remote port state DOWN in case the
> local one is recycled was not safe. In a case of irresponsive remote
> port it would actually try and traverse a null pointer of the port info.
>
> This simple
Hi Rolf,
On Wed, 2006-03-29 at 18:54, Rolf Manderscheid wrote:
> Hi Hal,
>
> opensm attempts to set pkey table entries on external switch ports
> even if the switch declares a PartitionEnforcementCap of zero. The
> consequence is ERR 4108.
I presume this is with the trunk version, right ?
> Th
On Thu, 2006-03-30 at 03:33, Eitan Zahavi wrote:
> I would rather have the PartitionManager inspect the IB attributes
> defining the port P_Key table length then change the pre-allocation of
> PKey entry. Even if this does not break anything.
Can you state the reason for this preference ?
> So I
On Wed, 2006-03-29 at 17:29, Eitan Zahavi wrote:
> Hi Hal
>
> This simple patch allows easy recognition of the vendor type built.
> A new link is created with teh vendor type postfix.
>
> Eitan
>
> Signed-off-by: Eitan Zahavi <[EMAIL PROTECTED]>
Thanks. Applied.
-- Hal
_
On Fri, 2006-03-24 at 15:11, Sasha Khapyorsky wrote:
> Hello,
>
> This removes #ifdef __KERNEL__ code sections from osmtest.
>
> Sasha.
>
>
> Remove #ifdef __KERNEL__ from osmtest.
>
> Signed-off-by: Sasha Khapyorsky <[EMAIL PROTECTED]>
Thanks. Applied.
-- Hal
__
On Sun, 2006-03-26 at 18:02, Sasha Khapyorsky wrote:
> Hello,
>
> There is improvements to Pkey support in SA PathRecord - there is
> functional P_Key values now instead of IB_DEFAULT_PKEY.
>
> Sasha.
>
>
> P_Key support in SA PathRecord (due to functional partition manager).
>
> Also there ar
On Wed, 2006-03-29 at 10:55, Bryan O'Sullivan wrote:
> On Wed, 2006-03-29 at 14:44 +0200, Aviram Gutman wrote:
> > I thought the intent was to merge 1.0 rc2 with IBED rc3.
>
> Yes.
What's IBED ?
-- hal
___
openib-general mailing list
openib-general@op
On Tue, 2006-03-28 at 18:16, Roland Dreier wrote:
> Sean> Can (or should it be permitted for) a userspace process join
> Sean> the same group as one used by ipoib?
>
> Well, that's the question. I don't think there's not any strong
> reason to forbid it, except to make implementation easi
On Tue, 2006-03-28 at 12:09, Roland Dreier wrote:
> Tom> Newbie questions... Is the max_rdd attribute the max RDMA
> Tom> READ limit, or something else. Are initiator and responder
> Tom> resource limits symmetric?
>
> max reliable datagram domain. irrelevant nowadays that RD transpo
ib_mad.h: Add remaining defined management classes
Signed-off-by: Hal Rosenstock <[EMAIL PROTECTED]>
Index: include/rdma/ib_mad.h
===
--- include/rdma/ib_mad.h (revision 6041)
+++ include/rdma/ib_mad.h (workin
On Mon, 2006-03-27 at 14:56, Roland Dreier wrote:
> As we enter the second week of the 2.6.17 merge window, I thought it
> might be helpful to give an update on what I'm thinking about the
> remaining pending pieces. This is really to stimulate discussion --
> if there's something that you think r
On Sun, 2006-03-26 at 18:08, Sasha Khapyorsky wrote:
> Hello,
>
> There is small cleanup - remove complib's unused cl_signal_osd.h.
>
> Sasha.
>
>
> Remove unused cl_signal_osd.h file.
>
> Signed-off-by: Sasha Khapyorsky <[EMAIL PROTECTED]>
Thanks. Applied to trunk only.
-- Hal
On Fri, 2006-03-24 at 15:42, Sasha Khapyorsky wrote:
> Hello,
>
> This makes local functions static in osm_sa_path_record_ctrl.c and
> osm_sa_service_record_ctrl.c.
>
> Sasha.
>
>
> Make some local functions static.
>
> Signed-off-by: Sasha Khapyorsky <[EMAIL PROTECTED]>
Thanks. Applied to bo
Hi Ofer,
On Thu, 2006-03-23 at 05:01, Ofer Gigi wrote:
> Hi Hal,
>
> The fix below fixes the retrieve of the mcmember records according to the
> Errata MGTWG3280.
>
> Quoting from MGTWG3280:
>
> SA can be queried for multicast groups by sending a SubnAdmGet() or a
> SubnAdmGetTable() request
On Sat, 2006-03-25 at 11:59, Michael S. Tsirkin wrote:
> Quoting r. Hal Rosenstock <[EMAIL PROTECTED]>:
> > > However we have this, towards the end of page 355.
> > >
> > >
> > >
> > > "Even a responder which does generate end-
On Fri, 2006-03-24 at 07:08, Michael S. Tsirkin wrote:
> Quoting r. Hal Rosenstock <[EMAIL PROTECTED]>:
> >
> > I did some spec reading to find this and found the following which I
> > think makes the current requirement clear:
> >
> > p.347 line 37 state
On Thu, 2006-03-23 at 11:38, Hal Rosenstock wrote:
> Hi Ofer,
>
> On Thu, 2006-03-23 at 11:04, Ofer Gigi wrote:
> > Hi Hal,
> >
> > The fix below fixes the retrieve of the mcmember records according to
> > the
> > Errata MGTWG3280.
> >
> > Quo
On Fri, 2006-03-24 at 09:37, Sasha Khapyorsky wrote:
> Hello Hal,
>
> On 09:22 Fri 24 Mar , Hal Rosenstock wrote:
> > Hi Sasha,
> >
> > On Thu, 2006-03-23 at 14:41, Sasha Khapyorsky wrote:
> > > Hello,
> > >
> > > There is updated pa
On Thu, 2006-03-23 at 14:21, Sasha Khapyorsky wrote:
> On 13:16 Thu 23 Mar , Hal Rosenstock wrote:
> > >
> > > > The realizability is when the port joins not when the group is created.
> > > > This is significant for the precreated groups (as other group
Hi Sasha,
On Thu, 2006-03-23 at 14:41, Sasha Khapyorsky wrote:
> Hello,
>
> There is updated patch.
>
> Sasha.
>
>
> Disregard subn->ca_min_mtu and subn->ca_min_rate when new MC group is
> created and exact MTU and/or rate values are specified.
>
> Signed-off-by: Sasha Khapyorsky <[EMAIL PROT
Hi Ira,
On Thu, 2006-03-23 at 20:19, Ira Weiny wrote:
> Our ibnetdiscover output was getting hosed by a '\r' in the switches nodedesc.
>
> Before:
>
> # odev0 /home/weiny2 > ibswitches
> " port 0 lid 208f104003f00a1 ports 8 "SW-12CIB4 Voltaire
> ...
>
> After:
>
> # odev0 /home/weiny2 > ./ibs
On Thu, 2006-03-23 at 12:34, Michael S. Tsirkin wrote:
> Quoting r. Hal Rosenstock <[EMAIL PROTECTED]>:
> > > > > > > > > >Sean, just to wrap it up, the API at the verbs layer will
> > > > > > > > > >look
> > &
Hi again Sasha,
On Thu, 2006-03-23 at 13:17, Sasha Khapyorsky wrote:
> On 11:48 Thu 23 Mar , Hal Rosenstock wrote:
> > > > But is that field set to the max rate/MTU ? (I didn't check the code for
> > > > this). Is it just a name thing or more ?
> >
On Thu, 2006-03-23 at 12:09, Michael S. Tsirkin wrote:
> Quoting r. Hal Rosenstock <[EMAIL PROTECTED]>:
> > Subject: Re: RFC: e2e credits
> >
> > On Thu, 2006-03-23 at 11:41, Michael S. Tsirkin wrote:
> > > Quoting r. Hal Rosenstock <[EMAIL PROTECTED]&
On Thu, 2006-03-23 at 11:41, Michael S. Tsirkin wrote:
> Quoting r. Hal Rosenstock <[EMAIL PROTECTED]>:
> > Subject: Re: RFC: e2e credits
> >
> > On Thu, 2006-03-23 at 11:13, Michael S. Tsirkin wrote:
> > > Quoting r. Hal Rosenstock <[EMAIL PROTECTED]>
On Thu, 2006-03-23 at 11:22, Sasha Khapyorsky wrote:
> On 10:24 Thu 23 Mar , Hal Rosenstock wrote:
> >
> > On Thu, 2006-03-23 at 09:57, Eitan Zahavi wrote:
> > > Now I get it. Its my bug.
> > > What I meant was to check that the request is realiza
Hi Suri,
On Thu, 2006-03-23 at 11:04, Suresh Shelvapille wrote:
> Folks:
>
> I am happy to announce that the Switch implementation of OpenIB stack is
> working very well.
Great!
> I have tested this with Voltaire and Silver Storm SM.
What about OpenSM ?
> Of course, this is a very limited im
Hi Ofer,
On Thu, 2006-03-23 at 11:04, Ofer Gigi wrote:
> Hi Hal,
>
> The fix below fixes the retrieve of the mcmember records according to
> the
> Errata MGTWG3280.
>
> Quoting from MGTWG3280:
>
> SA can be queried for multicast groups by sending a SubnAdmGet() or a
> SubnAdmGetTable() reques
On Thu, 2006-03-23 at 11:13, Michael S. Tsirkin wrote:
> Quoting r. Hal Rosenstock <[EMAIL PROTECTED]>:
> > > >Sean, just to wrap it up, the API at the verbs layer will look like the
> > > >below, and then ULPs just put the value they want in the CM and CM wi
On Wed, 2006-03-22 at 18:26, Sean Hefty wrote:
> >Sean, just to wrap it up, the API at the verbs layer will look like the
> >below,
> >and then ULPs just put the value they want in the CM and CM will pass
> >it in to low level.
>
> I'm fine with this, but I do think that it's a minor spec
> viol
< / > / =)
> And so on
Yes.
-- Hal
> Eitan Zahavi
> Design Technology Director
> Mellanox Technologies LTD
> Tel:+972-4-9097208
> Fax:+972-4-9593245
> P.O. Box 586 Yokneam 20692 ISRAEL
>
>
> > -Original Message-
> > From: Hal Rosensto
Hi Yael,
On Thu, 2006-03-23 at 04:50, Yael Kalka wrote:
> Hi Hal,
>
> There was an error in osmt_multicast.c that matched the 1.1 spec and
> wasn't updated to the 1.2 version of the spec. When checking
> unrealistic rate, the MCMemberRecord was sent with rate 30BG/sec and
> rateSelector set to 0.
am not sure.
>
> Eitan Zahavi
> Design Technology Director
> Mellanox Technologies LTD
> Tel:+972-4-9097208
> Fax:+972-4-9593245
> P.O. Box 586 Yokneam 20692 ISRAEL
>
>
> > -Original Message-
> > From: Hal Rosenstock [mailto:[EMAIL PROTECTED]
> &
Hi Or,
On Wed, 2006-03-22 at 11:03, Or Gerlitz wrote:
> Hi Sean, Hal
>
> Since updating to 2.6.16 and drivers/infiniband replaced by openib svn trunk
> r5919 i am getting ETIMEDOUT (-110) cma event on the iser initiator side,
> about 4/5 of my reconnect tries.
>
> > iser:iser_cma_handler:event
On Wed, 2006-03-22 at 08:25, Sasha Khapyorsky wrote:
> Hello,
>
> Now at MC Group creation when exact rate or MTU values are requested and
> those values are greater than rate (or mtu) of slowest port on the subnet
> then MC group creation fails. It is likely not desired behaviour.
Yes, if there
On Wed, 2006-03-22 at 06:54, Yael Kalka wrote:
> Hi Hal,
>
> Attached is the patch for adding honor_guid2lid flag for the 1.0 branch.
>
> Thanks,
> Yael
>
> Signed-off-by: Yael Kalka <[EMAIL PROTECTED]>
Thanks. Applied to 1.0 branch.
-- Hal
___
ope
On Wed, 2006-03-22 at 06:52, Yael Kalka wrote:
> Hi Hal,
>
> Attached is the patch for adding guid to opensm opts for the 1.0 branch.
>
> Thanks,
> Yael
Thanks. Applied to 1.0 branch.
-- Hal
___
openib-general mailing list
openib-general@openib.org
h
On Tue, 2006-03-14 at 19:47, Sasha Khapyorsky wrote:
> Hello,
>
> There are changes in IPoIB MC Groups pre-creation - now it is on
> per-partition basis.
>
> This patch is for trunk.
>
> Sasha.
>
>
> Changes in IPoIB MC Groups pre-creation:
>
> - IPoIB MC Group precreation is done on per-par
Hi Yael,
On Wed, 2006-03-22 at 02:55, Yael Kalka wrote:
> Hi Hal,
>
> Attached is another fix for windows - again some casting issue.
> And also a comment naming fix in ib_types.h.
>
> Thanks,
> Yael
>
> Signed-off-by: Yael Kalka <[EMAIL PROTECTED]>
Thanks. Applied to both trunk and 1.0 branc
On Tue, 2006-03-21 at 06:48, Yael Kalka wrote:
> Hi Hal,
>
> The following patch includes fixes for the windows compilation:
> __msecs_to_rtv_table in osm_sa_clall_port_info is used as uint32_t
> and not regular int, and thus can be defined as such.
> Also there is some function name changing in
On Mon, 2006-03-20 at 16:12, Roland Dreier wrote:
> Hal> With the latest OpenIB svn, I got the following oops when
> Hal> attempting to remove the ib_ipoib module on x86 with 2.6.15.
>
> Hmm, I can't understand what's going on here, nor can I reproduce it.
> Can you try to get it to happen
On Mon, 2006-03-20 at 16:21, Sean Hefty wrote:
> Hal Rosenstock wrote:
> > mad: Add RMPP support for additional management classes which support it
> > Also, validate RMPP is consistent with management class specified
>
> This looks good to me.
I'll check it in aft
mad: Add RMPP support for additional management classes which support it
Also, validate RMPP is consistent with management class specified
Signed-off-by: Hal Rosenstock <[EMAIL PROTECTED]>
Index: include/rdma/ib
Hi Sean,
On Fri, 2006-03-17 at 13:03, Hal Rosenstock wrote:
> On Fri, 2006-03-17 at 13:00, Sean Hefty wrote:
> > >-static int data_offset(u8 mgmt_class)
> > >+int ib_get_rmpp_data_offset(u8 mgmt_class)
> > > {
> > > if (mgmt_class == IB_MGMT_CLASS_SUBN_ADM
Hi,
With the latest OpenIB svn, I got the following oops when attempting to
remove the ib_ipoib module on x86 with 2.6.15.
-- Hal
Unable to handle kernel NULL pointer dereference at virtual address 0004
printing eip:
d1b8299f
*pde = 07114067
*pte =
Oops: 0002 [#1]
DEBUG_PAGEALLOC
M
Hi Ofer,
On Mon, 2006-03-20 at 07:53, Ofer Gigi wrote:
> Hi Hal,
>
> In osm_mcast_mgr.c the line
>
> osm_mgrp_send_delete_notice( p_mgr->p_subn, p_mgr->p_log, p_mgrp );
>
> caused segmentation fault since the handle p_mgrp was already deleted while
> the
> function was called.
>
> The fix i
On Mon, 2006-03-20 at 11:00, Hal Rosenstock wrote:
> On Sun, 2006-03-19 at 15:15, Sasha Khapyorsky wrote:
> > Hello,
> >
> > There is simplification of mad_dump_sltovl() output (and function). This
> > is helpful when multiple tables are queried for gracefull results
&
On Sun, 2006-03-19 at 15:21, Sasha Khapyorsky wrote:
> Hello,
>
> There is support for querying SL2VLMapping and VLArbitration tables -
> 'sl2vl' and 'vlarb' operations are added to smpquery.
>
> Sasha.
>
>
> Support for SL2VLMapping and VLArbitration tables querying.
>
> Signed-off-by: Sasha
On Sun, 2006-03-19 at 15:15, Sasha Khapyorsky wrote:
> Hello,
>
> There is simplification of mad_dump_sltovl() output (and function). This
> is helpful when multiple tables are queried for gracefull results
> printing.
>
> Sasha.
>
>
> This simplifies mad_dump_sltovl() output and makes it possi
Hi Yael,
On Sun, 2006-03-19 at 07:55, Yael Kalka wrote:
> Hi Hal,
>
> We noticed that when sending MCMemberRecord query, there is a check if
> the requestor is trusted (if not then PortGid, join_state and
> proxy_join are returned as zero). But, currently the trusted check is
> done on the sm_key
Hi Ofer,
On Thu, 2006-03-16 at 12:14, Ofer Gigi wrote:
> Hi Hal,
>
> RMPP mad can only sent in 4 kinds of mad:
> 1. IB_MCLASS_SUBN_ADM
> 2. IB_MCLASS_DEV_MGMT
> 3. BIS
> 4. DevAdm
>
> If the packet is not 1 or 2 - the mad is not checked and returned in advance
> as not a
> rmpp packet.
> Si
a proper patch for this. Several cuts at this were
sent out after you posted the issue and an initial patch.
-- Hal
> Thanks!
> Ofer
>
>
> -Original Message-
> From: Hal Rosenstock [mailto:[EMAIL PROTECTED]
> Sent: Friday, March 17, 2006 2:07 AM
> To: Ofer Gig
On Fri, 2006-03-17 at 13:00, Sean Hefty wrote:
> >-static int data_offset(u8 mgmt_class)
> >+int ib_get_rmpp_data_offset(u8 mgmt_class)
> > {
> > if (mgmt_class == IB_MGMT_CLASS_SUBN_ADM)
> > return IB_MGMT_SA_HDR;
> >+else if ((mgmt_class == IB_MGMT_CLASS_DEVICE_MGMT) ||
> >+
mad: Add RMPP support for additional management classes which support it
Signed-off-by: Hal Rosenstock <[EMAIL PROTECTED]>
Index: include/rdma/ib_mad.h
===
--- include/rdma/ib_mad.h (revision 5874)
+++ include/rdma/ib
On Thu, 2006-03-16 at 23:09, Sean Hefty wrote:
> > /**
> >+ * ib_get_rmpp_data_offset - returns the data offset for a given
> >+ * management class.
> >+ * @mgmt_class: management class
> >+ *
> >+ * This routine returns the data offset in the MAD for the management
> >+ * class requested.
> >+ */
On Thu, 2006-03-16 at 23:09, Sean Hefty wrote:
> > /**
> >+ * ib_get_rmpp_data_offset - returns the data offset for a given
> >+ * management class.
> >+ * @mgmt_class: management class
> >+ *
> >+ * This routine returns the data offset in the MAD for the management
> >+ * class requested.
> >+ */
901 - 1000 of 3498 matches
Mail list logo
