Hi,
The ib_read_bw tool when used with the -V (version) option returns an
error because it is not being parsed as an option.
Regards,
Junaid
Novell, Inc.
Signed-off-by: Junaid Shariff <[EMAIL PROTECTED]>
Index: userspace/perftest/read_bw.c
===
>
> I managed to find the above missing files under the gen1 branch. I am
> not quite sure about installing packages from gen1 and gen2 on the same
> machine. Should I install IBGOLD instead of OFED ?
>
srp target sw only run on gen1 (ie. IBGD stack)
and yes you need to install/run IBGD stack
On Monday 06 November 2006 21:17, Ernst Herzberg wrote:
> On Sunday 05 November 2006 07:48, Adrian Bunk wrote:
> > ...
> > Subject: ThinkPad R50p: boot fail with (lapic && on_battery)
> > References : http://lkml.org/lkml/2006/10/31/333
> > Submitter : Ernst Herzberg <[EMAIL PROTECTED]>
> > St
On Sunday 05 November 2006 07:48, Adrian Bunk wrote:
> ...
> Subject: ThinkPad R50p: boot fail with (lapic && on_battery)
> References : http://lkml.org/lkml/2006/10/31/333
> Submitter : Ernst Herzberg <[EMAIL PROTECTED]>
> Status : problem is being debugged
Update:
2.6.19-rc4-git11 does
On Sunday 05 November 2006 10:30, Jeff Chua wrote:
> Here's results with vanilla 2.6.19-rc4 (gcc version 3.4.5) ...
>
> 1) PCI access mode (Any) ... FAILED
>
> 2) PCI access mode (MMConfig) ... FAILED
Full boot log please?
-Andi
___
openib-general m
I've been working on fixing http://openib.org/bugzilla/show_bug.cgi
which is an IPoIB crash. I figured out that the problem was a
use-after-free caused by two path record query completions being
generated, one with a successful status and one with a status of
-EINTR. In turn, this was caused by t
http://openib.org/bugzilla/show_bug.cgi?id=263
--- Comment #12 from [EMAIL PROTECTED] 2006-11-06 20:26 ---
I finally spent some time tracking this down and I believe the problem is
actually in the MAD layer. I will post more details and a patch to
openib-general.
--- You are
Hi,
I searched for the SRP target source on the openib site, and found
it under the gen1 branch. Is it available under the gen2 branch ?
I tried to install it on a machine which has OFED-1.1 (gen2) installed,
and following are the first few errors from make :-
/root/SRPTARGET/mtm_ib.h:38:29
When creating directory for db files (guid2lid) storing create it with
reasonable permissions (current 777 decimal = octal 01411) and don't do
it world writable.
Signed-off-by: Sasha Khapyorsky <[EMAIL PROTECTED]>
---
osm/opensm/osm_db_files.c |2 +-
1 files changed, 1 insertions(+), 1 delet
This removes unused (but confused) svc_id field from osm_svcr structure.
Signed-off-by: Sasha Khapyorsky <[EMAIL PROTECTED]>
---
osm/include/opensm/osm_service.h |1 -
osm/opensm/osm_service.c |1 -
2 files changed, 0 insertions(+), 2 deletions(-)
diff --git a/osm/include/opensm
Trivial indentation fixes in osm_inform.h and osm_service.h.
Signed-off-by: Sasha Khapyorsky <[EMAIL PROTECTED]>
---
osm/include/opensm/osm_inform.h | 55 ++
osm/include/opensm/osm_service.h | 11 +++
2 files changed, 37 insertions(+), 29 deletions(-)
> The point Heiko made in his comment is that with ehca_alloc_fw_ctrlblock
> returning a void*, you can (and _should_) remove the casts to other
> pointer types.
Hm, my bad. Should not happen that I totally ignored the whole sentence
from Heiko.
Roland, please throw the previous away. Here is event
> ib_mthca :02:00.0: SYS_EN DDR error: syn=4, sock=0, sladdr=0, SPD
> source=DIMM
The HCA is reporting a memory calibration error. Most likely this is
a physical hardware problem somewhere in your system.
Is it just one system failing like this, or do you have multiple
HCAs/systems with t
Fix some timeout and long disconnect delay issues discovered during scale-out
testing. Added support
to retry rdma_cm address and route resolution with configuration options.
Provide a disconnect call
when receiving the disconnect request to guarantee a disconnect reply and event
on the remote
Fix some issues supporting create qp without recv cq handle or recv qp
resources. IB verbs assume a
recv_cq handle and uDAPL dapl_ep_create assumes there is always recv_sge
resources specified.
Signed-off by: Arlin Davis [EMAIL PROTECTED]
Index: dapl/common/dapl_ep_create.c
==
New series of patches with "-x -up"
Added support for new ib verbs client register event. No extra processing
required at the uDAPL
level. Shows up if opensm bounces.
Signed-off by: Arlin Davis [EMAIL PROTECTED]
Index: dapl/openib_cma/dapl_ib_util.c
=
On Mon, 2006-11-06 at 13:13, Wang, Feiyi wrote:
> Hal -
>
> Please see the output for active port 1 (although there are two ports on
> this HCA, the second one is disabled now).
>
> #smpquery portinfo 8 1
> # Port info: Lid 8 port 1
> Mkey:0x
> GidPrefi
Hi,
I have a problem in getting the IB hardware during startup.
I am running a Red Hat Linux machine with Pentium D processors.
Here is the output of the dmesg command for the ib hardware:
ib_mthca: Mellanox InfiniBand HCA driver v0.06 (June 23, 2005)
ib_mthca: Initializing :02:00.0
ib_mthca 0
On Monday 06 November 2006 22:26, Hoang-Nam Ngyuen wrote:
> - rblock = kzalloc(H_CB_ALIGNMENT, GFP_KERNEL);
> + rblock = (struct hipz_query_hca*)ehca_alloc_fw_ctrlblock();
>
> - rblock = kzalloc(H_CB_ALIGNMENT, GFP_KERNEL);
> + rblock = (struct hipz_query_port*)ehca_alloc
Hello Roland!
Below is the patch with the "ctor" bug fix and Heiko's improvement
to assure ehca_alloc_fw_ctrlblock() always returns (void*).
Hope it is in proper format.
Thanks!
Nam
This is a patch of ehca that assures 4k alignment for firmware control block
in 64k page mode, because kzalloc()'s r
On Mon, 2006-11-06 at 08:52 -0800, Roland Dreier wrote:
> > I'm surprised that something as recent as infiniband requires a
> > long-deprecated function bus_to_virt().
> >
> > What is it trying to do there that needs that call ?
>
> Don't ask -- just enjoy the fact that you don't know about t
Hal -
Please see the output for active port 1 (although there are two ports on
this HCA, the second one is disabled now).
#smpquery portinfo 8 1
# Port info: Lid 8 port 1
Mkey:0x
GidPrefix:...0xfe80
Lid:..
There is a very slight overhead since there is a test and
branch. Given modern CPU architecture, this is in the noise.
On Sun, 2006-11-05 at 09:42 +0200, Eitan Zahavi wrote:
> Hi Ralph,
>
> Is there any performance penalty for using the IB version of the DMA
> mapping functions?
>
> Thanks
>
> So instead of ib_sa_mcmember_rec_set which returned the query by pointer we
> now
> have ib_sa_join_multicast which returns the pointer. This part looks OK I
> guess, but
> I still do not understand why does the patch tinker with logic (e.g.
> setting/clearing IPOIB_MCAST_FLAG_BUSY) in the IPo
On Sun, 2006-11-05 at 17:22 +0200, Michael S. Tsirkin wrote:
> Quoting r. Ralph Campbell <[EMAIL PROTECTED]>:
> > diff -r f37bd0e41fec drivers/infiniband/ulp/ipoib/ipoib_ib.c
> > --- a/drivers/infiniband/ulp/ipoib/ipoib_ib.c Thu Oct 26 21:44:41
> > 2006 +0700
> > +++ b/drivers/infiniband/ulp
Quoting r. Roland Dreier <[EMAIL PROTECTED]>:
> Subject: Re: [PATCH 3/7] IB/ipoib - Use the new verbs DMA mapping functions
>
> > For example, ib_send_wr/ib_recv_wr could have an *optional* "void *data"
> > field. And we could have a rule that ULP must *either* pass in the void
> > *data, and
> When I unload the Infiniband driver I am getting a
> kernel Oops (consistently at the same location, with the same
> stack output).
>
> I am doing the driver unload immediately after reboot
> (the boot process loads the infiniband driver).
> I am not running opensm.
>
> Looks to me like a refe
Hi Roland!
> > As Arnd stated I need to fix this ctor issue. Do you prefer me to
resend
> > all patches in proper format (non-mangled inline) or just this one bug
fix?
> I have the rest of the patches, so you just need to resend a fixed
> version of this one. BTW see my previous response about
>
On Mon, 6 Nov 2006, Linev Sergei wrote:
> When I cahnge limits in limits.conf file, I can run most of my uDAPL
> code except disconnection of nodes (which not principal for me). I
> get error message:
>
> dapl/common/dapl_ep_free.c:114: dapl_ep_free: Assertion
> `ep_ptr->param.ep_state == DA
Hi,
>
> Did you follow these directions from the libibverbs README?
>
> --
> https://openfabrics.org/svn/gen2/trunk/src/userspace/libibverbs/README
>
>
> To use IB verbs from userspace, a process must also have permission to
> tell the kernel to lock sufficient memory for all of your regist
> From: Todd Rimmer
> Sent: Thursday, November 02, 2006 7:15 PM
> To: 'Michael S. Tsirkin'; Hal Rosenstock
> Cc: Or Gerlitz; openib-general; Arlin R Davis
> Subject: RE: [openib-general] scaling issues, was: uDAPL cma: add
support
> for address and route retries, call disconnect when recving dreq
>
> -Original Message-
> From: Todd Rimmer
> Sent: Thursday, November 02, 2006 7:12 PM
> To: 'Michael S. Tsirkin'; Arlin Davis
> Cc: Or Gerlitz; openib-general; Arlin Davis
> Subject: RE: [openib-general] scaling issues, was: uDAPL cma: add
support
> for address and route retries, call disco
> Hmm, since ib_dma_unmap_single calls a function through a pointer,
> this seems to introduce overhead on data path operations in ipoib.
> For apps like ipoib always working with low memory, I think it is important
> to avoid this
> overhead of extra indirect function calls at least on syste
> I'm surprised that something as recent as infiniband requires a
> long-deprecated function bus_to_virt().
>
> What is it trying to do there that needs that call ?
Don't ask -- just enjoy the fact that you don't know about this...
___
openib-gener
> As Arnd stated I need to fix this ctor issue. Do you prefer me to resend
> all patches in proper format (non-mangled inline) or just this one bug fix?
I have the rest of the patches, so you just need to resend a fixed
version of this one. BTW see my previous response about
kmem_cache_zalloc()
> I'd simply move the memset into the alloc function and get rid of the
> constructor here.
Slightly better still would be to use kmem_cache_zalloc() (save a tiny
bit of text by getting rid of the call to memset).
- R.
___
openib-general mailing lis
Linev Sergei wrote:
> Hi
>
> I found, that only SuSE Enterprise Linux 10 is supported.
> Can I try normal SuSE Linux 10.1 or it should be only Enterprise edition?
>
>
We have run here testing on SuSE Linux 10, but not full QA. Another
option is to install SuSE Linux 10 and then replace kernel t
I tried using ifconfig to limit the ipoib mtu.
Once I do this on *either* both server and client, or only on the client side,
UDP seems to stop working:
#ifconfig ib0 mtu 512
#netperf -c -C -H 11.4.3.68 -f M -t UDP_STREAM
UDP UNIDIRECTIONAL SEND TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to 11.4.3
As I explained in my mail, no one had replied to any of the posts
containing my very directed and specific questions (not even you --
and you still haven't), so I figured that no one cared. That's not
an unreasonable assumption given that I posted the same questions 3
times and got silence
[EMAIL PROTECTED] wrote on 06.11.2006 15:58:39:
> Maybe you want to make sure that ehca_alloc_fw_ctrlblock() always returns
a
> void pointer, so you can avoid all the casts in your code?
> static inline void *ehca_alloc_fw_ctrlblock(void)
> {
>return (void *)get_zeroed_page(GFP_KERNEL);
> }
Yes
On Sun, 5 Nov 2006, Michael S. Tsirkin wrote:
> Quoting r. Roland Dreier <[EMAIL PROTECTED]>:
> > Subject: Re: static ARP entries for IPoIB?
> >
> > > 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_INFINI
Did you follow these directions from the libibverbs README?
--
https://openfabrics.org/svn/gen2/trunk/src/userspace/libibverbs/README
To use IB verbs from userspace, a process must also have permission to
tell the kernel to lock sufficient memory for all of your registered
memory regions as wel
Hi
I found, that only SuSE Enterprise Linux 10 is supported.
Can I try normal SuSE Linux 10.1 or it should be only Enterprise edition?
And another question - why in Mellanox release notes for OFED 1.1 stated, that
uDAPL is not supported -
is it not working or just has some bugs/unattended featu
> +#ifdef CONFIG_PPC_64K_PAGES
> +void *ehca_alloc_fw_ctrlblock(void);
> +void ehca_free_fw_ctrlblock(void *ptr);
> +#else
> +#define ehca_alloc_fw_ctrlblock() get_zeroed_page(GFP_KERNEL)
> +#define ehca_free_fw_ctrlblock(ptr) free_page((unsigned long)(ptr))
> +#endif
Maybe you want to make sure t
Jeff:
May I know on with what `right' you are making this motion to remove
the code.
To have the code there was decided by the OpenIB community and the
organizers. It needs to be decided by the community, not by an
individual person.
Let me suggest that we we discuss this at the Developers Summi
Ralph Campbell wrote:
> IB/iser - Use the new verbs DMA mapping functions
>
> This patch converts iser to use the new verbs DMA mapping functions
> for kernel verbs consumers.
>
>
I tested it and ran some sanity checks and it looks ok.
--
__
Linev Sergei wrote:
> Hi
>
> I was trying to install OFED 1.1 on SuSE 9.3 Linux (2.6.11.4-20a-smp).
> We are using Opterons with Mellanox MHES18-XT PCIe host adapters.
> Previousely we were using IB Gold 1.8.0 and mostly working with uDAPL.
>
OFED 1.1 does not support SuSE 9.3.
Best if you move
On Monday 06 November 2006 14:40, Or Gerlitz wrote:
> By "unload the driver" you mean modprobe -r to which module?
>
As indicated in my mail, the active modprobe was to module
ib_ipoib (as part of the "/etc/init.d/openibd stop" script).
>By "our driver" do you mean the OFED 1.1 IB kernel drivers
Having received no replies for 2 weeks as to why it is useful to have
MVAPICH in the OpenFabrics SVN, I can only conclude that no one
cares. If someone does care, please respond to my original questions
included below ASAP (originally posted 23 Oct, 27 Oct, 1 Nov).
I therefore make the moti
Hi
I was trying to install OFED 1.1 on SuSE 9.3 Linux (2.6.11.4-20a-smp).
We are using Opterons with Mellanox MHES18-XT PCIe host adapters.
Previousely we were using IB Gold 1.8.0 and mostly working with uDAPL.
Now I trying uDAPL with OFED and find out, that it is not working for me.
Actually, I
Jack Morgenstein wrote:
> On Wednesday 25 October 2006 00:25, Sean Hefty 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
>>
Hi Roland!
> Arnd wrote:
> This seems broken. You have a constructor for newly allocated objects,
but
> there is no destructor and it seems that objects passed to
> ehca_free_fw_ctrlblock are not guaranteed to be initialized either.
> I'd simply move the memset into the alloc function and get rid o
Quoting Sean Hefty <[EMAIL PROTECTED]>:
> diff --git a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c
> b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c
> index 3faa182..d90f804 100644
> --- a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c
> +++ b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c
> @@
On Wednesday 25 October 2006 00:25, Sean Hefty 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
Sean Hefty wrote:
>>> Are you wanting the rdma cm to join the same multicast groups that ipoib
>> does?
>>> (This is simple to change, but it does not join the same groups today.)
>> Actually, yes, we strogly prefer that you will remove the rdma cm
>> signature byte from MGIDs generated by the cma.
Hi Arnd,
> This seems broken. You have a constructor for newly allocated objects,
but
> there is no destructor and it seems that objects passed to
> ehca_free_fw_ctrlblock are not guaranteed to be initialized either.
> I'd simply move the memset into the alloc function and get rid of the
> construc
On Sun, 2006-11-05 at 23:42 -0800, Robert Walsh wrote:
> Michael Ellerman wrote:
> > Hi,
> >
> > Just a heads-up, drivers/infiniband/hw/ipath/ipath_keys.c doesn't build
> > on powerpc because we have don't have bus_to_virt().
> >
> > (Actually you can't select CONFIG_INFINIBAND_IPATH on mainline
On Mon, 2006-11-06 at 18:33 +1100, Michael Ellerman wrote:
> Hi,
>
> Just a heads-up, drivers/infiniband/hw/ipath/ipath_keys.c doesn't build
> on powerpc because we have don't have bus_to_virt().
>
> (Actually you can't select CONFIG_INFINIBAND_IPATH on mainline because
> powerpc doesn't enable P
>> Are you wanting the rdma cm to join the same multicast groups that ipoib
>does?
>> (This is simple to change, but it does not join the same groups today.)
>
>Actually, yes, we strogly prefer that you will remove the rdma cm
>signature byte from MGIDs generated by the cma.
Would there be any iss
59 matches
Mail list logo