> I'd like to create static ARP entries for some IPoIB
> devices. The arp (8) command that I'm using doesn't
> know about ARPHRD_INFINIBAND (this is arp version 1.88
> from net-tools-1.60-583.4.src.rpm, in SLES10.) Is there
> a version of arp (8) that works with IB? Or some other
> utility or
I'd like to create static ARP entries for some IPoIB
devices. The arp (8) command that I'm using doesn't
know about ARPHRD_INFINIBAND (this is arp version 1.88
from net-tools-1.60-583.4.src.rpm, in SLES10.) Is there
a version of arp (8) that works with IB? Or some other
utility or means to make st
On Fri, 2006-11-03 at 15:56, Wang, Feiyi wrote:
> 255
>
> I think I tested with default 0 before, that is send at most one packet
> before give low priority table the chance according to IBA. It doesn't
> seem to make a difference though.
I was hoping you would say 0 as that means 1 packet befor
255
I think I tested with default 0 before, that is send at most one packet
before give low priority table the chance according to IBA. It doesn't
seem to make a difference though.
Feiyi
-Original Message-
From: Hal Rosenstock [mailto:[EMAIL PROTECTED]
Sent: Friday, November 03, 2006
On Fri, 2006-11-03 at 15:12, Feiyi Wang wrote:
> In our test at the ORNL - it appears you can "turn off" the traffic by
> giving every VL weight 0.
A weight of 0 indicates to skip that entry.
> As soon as you assign non-zero VL weight,
> the traffic starts to flow, however, VL with more weight d
In our test at the ORNL - it appears you can "turn off" the traffic by
giving every VL weight 0. As soon as you assign non-zero VL weight,
the traffic starts to flow, however, VL with more weight doesn't have
expected preference treatment. In other words, traffic shaping didn't
take place. smpquery
We were able to get some more test time on the cluster. Our latest findings
are
below.
> The main issue that we saw was that the SA simply doesn't scale.
From what we could see, it didn't appear that _any_ path record queries were
ever lost, even when scaling up to 500,000+ requests. As lon
On Fri, 3 Nov 2006, Ralph Campbell wrote:
> On Fri, 2006-11-03 at 12:54 -0500, James Lentini wrote:
> > Ralph,
> >
> > Could you add documentation to Documentation/infiniband/ explainin
> > when a ULP needs to use the new verbs DMA mapping functions (what
> > structures/function parameters mu
On Fri, 3 Nov 2006, Steve Wise wrote:
> > >
> > > But how?
> > > Wait, work request posting functions actually get a virtual
> > > address and a key, not a dma address.
> >
> > Work requests posted using ib_post_send/recv are specified using a dma
> > address obtained using the appropriate L
On Fri, 2006-11-03 at 12:08 -0600, Steve Wise wrote:
> > >
> > > But how?
> > > Wait, work request posting functions actually get a virtual
> > > address and a key, not a dma address.
> >
> > Work requests posted using ib_post_send/recv are specified using a dma
> > address obtained using the a
On Fri, 2006-11-03 at 12:54 -0500, James Lentini wrote:
> Ralph,
>
> Could you add documentation to Documentation/infiniband/ explainin
> when a ULP needs to use the new verbs DMA mapping functions (what
> structures/function parameters must be specified using this API)?
Good idea. I will.
>
> >
> > But how?
> > Wait, work request posting functions actually get a virtual
> > address and a key, not a dma address.
>
> Work requests posted using ib_post_send/recv are specified using a dma
> address obtained using the appropriate Linux DMA mapping API function.
> They are not virtual
Ralph,
Could you add documentation to Documentation/infiniband/ explainin
when a ULP needs to use the new verbs DMA mapping functions (what
structures/function parameters must be specified using this API)?
Did you consider creating a new type, ib_dma_addr_t, and using this
type for the verb
On Fri, 3 Nov 2006, Michael S. Tsirkin wrote:
> Quoting r. Roland Dreier <[EMAIL PROTECTED]>:
> > Subject: Re: [PATCH 1/7] IB/core - Add DMA mapping functions to allow
> > device drivers to interpose
> >
> > > However, this means that the API must give the HCA the choice of
> > > what to kee
On Thu, 2006-11-02 at 05:53, Sasha Khapyorsky wrote:
> This adds gcc attribute to osm_log() which causes the compiler to check
> argument types against a format string. And also there are related fixes
> in osm_log() usage in opensm and osmtest.
>
> Signed-off-by: Sasha Khapyorsky <[EMAIL PROTECTE
On Fri, 2006-11-03 at 10:59, Sasha Khapyorsky wrote:
> x86_64 related osm_log() format fixes. Found with strict format checking
> attribute.
>
> Signed-off-by: Sasha Khapyorsky <[EMAIL PROTECTED]>
Thanks. Applied.
-- Hal
___
openib-general mailing li
x86_64 related osm_log() format fixes. Found with strict format checking
attribute.
Signed-off-by: Sasha Khapyorsky <[EMAIL PROTECTED]>
---
osm/libvendor/osm_vendor_ibumad_sa.c |2 +-
osm/opensm/osm_mcast_mgr.c |2 +-
osm/opensm/osm_sa_link_record.c |4 ++--
osm/opensm
Sean posted 7 patches that include the ucma support.
You'll need those + the one librdmacm patch he posted.
Steve.
On Fri, 2006-11-03 at 13:59 +0530, Krishna Kumar2 wrote:
> Hi,
>
> I installed the 2.6.19-rc3 bits, and when I try to run
> perftest/rdma_bw (with '-c' option), I get the error :
Hi,
I havent tried this but want to hear from IB experts whether IB broadcast (where switch does the real broadcasting task and where all the hosts are connected to a single switch) is as fast as point-to-point IB data transfer and secondly if data transfer over UD QP is faster then over RC QP ??
On Wed, 2006-11-01 at 13:45, Sasha Khapyorsky wrote:
> This removes obsolete now shared sm->p_report_buf buffer and cleans
> up related code. And also introduces new log function osm_log_printf()
> which currently trivially sends formatted output to stdout.
>
> Signed-off-by: Sasha Khapyorsky <[EM
On Wed, 2006-11-01 at 15:39, Sasha Khapyorsky wrote:
> When partition is specified in partition configuration file without
> desired PKey value OpenSM will generate one dynamically.
>
> The problem is that when the list of such "dynamic" partitions is
> edited (some partitions are removed and/or s
Hi,
I installed the 2.6.19-rc3 bits, and when I try to run
perftest/rdma_bw (with '-c' option), I get the error :
"librdmacm: Couldnt open rdma_cm ABI version".
I found that this is due to ucma not being present in
mainline kernel bits (which creates /sys/class/misc/rdma_cm).
So how can I resolve
22 matches
Mail list logo
