On Tue, Feb 14, 2006 at 06:42:36PM +0200, Michael S. Tsirkin wrote:
> All, MADV_DONTFORK patch is now part of the -mm tree.
> Everyone who's interested in fork support, please test 2.6.16-rc3-mm1 and
> publish the results here and on lkml.
>
Good news!
Should call to madvise be the part of reg_mr
涼子様からの連絡です:
いくら指名しても連絡が来ないけど、私じゃだめなの? 金額に不満がありました
ら、携帯で相談に乗ります、私の番号は090-61**0**8です。でも、わからない番号に
出るのは怖いから、
http://www.chabby-funk.com?hope
ここから登録してメールで貴方の番号と話せる時間を教えてください、こちらから連
絡します、私のIDは【112070】です。
私のこと嫌いなら
[EMAIL PROTECTED]
___
openib-general mail
Roland Dreier <[EMAIL PROTECTED]> wrote:
>
> Do we still have a chance to change this?
yes, please do.
___
openib-general mailing list
openib-general@openib.org
http://openib.org/mailman/listinfo/openib-general
To unsubscribe, please visit http://openib
Nick> May I ask, what is the rationale for ignoring the apparent
Nick> conventions of all architectures? For example parisc, you
Nick> appear to even go contrary to the comment.
Looking through include/asm-*/mman.h, I have to agree. The parisc
example seemly especially bad, as (in add
Michael S. Tsirkin wrote:
Index: linux-2.6.16-rc2/include/asm-x86_64/mman.h
===
--- linux-2.6.16-rc2.orig/include/asm-x86_64/mman.h 2006-02-14
01:22:27.0 +0200
+++ linux-2.6.16-rc2/include/asm-x86_64/mman.h 2006-02-14
$B(.(.(#(B $B!!(BAetane News$B!J(B2006$BG/HG!K(B
$B(.(.(.(B
$B(,(,(,(,(,(,(,(,(,(,(,(,(,(,(,(,(,(,(,(,(B
$B!!$3$s$K$A$O!*(B
$B!!8BDj8x3+([EMAIL PROTECTED](B
$B!!(B
$B!!2q0w8BDj$J$N$G!"B>$NJ}$K$OHkL)$K$7$F2<$5$$$M"v(B
$B!!$+$J$j2D0&$$L<$,B7$C$F$$$^$9!#(B
$B!!(B
$B!!:#$9$
I have a PCIe HCA working in the secondary x16 slot of an ASUS A8N-SLI
motherboard, but only as a x1 device.
- R.
___
openib-general mailing list
openib-general@openib.org
http://openib.org/mailman/listinfo/openib-general
To unsubscribe, please visit h
I've tried a few Asus boards with IB cards in their graphics card slots
with no luck, the boards did post but the cards weren't visable to the
drivers. Same cards in other machines were fine.
I suspect somewhere in the BIOS or chipset you can actually say that a
certain PCI-E x16 slot is for "grap
Hi All,
Does anyone have any idea if any lower end bare bones motherboards (ie the
Asus/Gigabyte type vendors) work successfully with a Mellanox PCI-E x8
card (MHGS18-XT A2 specifically)?
I'm looking for a cost-effective way to get a bunch of IB hosts for
testing, so compute performance isn't a c
On Tue, 2006-02-14 at 19:25 -0500, Hal Rosenstock wrote:
> Hi Owen,
>
> On Tue, 2006-02-14 at 14:26, Owen Stampflee wrote:
> > [EMAIL PROTECTED] ~]# opensm
> > -
> > OpenSM Rev:openib-1.1.0
> > Based on OpenIB svn 5411
> > Command Line Arguments:
>
Hi Owen,
On Tue, 2006-02-14 at 14:26, Owen Stampflee wrote:
> [EMAIL PROTECTED] ~]# opensm
> -
> OpenSM Rev:openib-1.1.0
> Based on OpenIB svn 5411
> Command Line Arguments:
> Log File: /var/log/osm.log
>
Doug> You don't have an Infiniband card in the machine (or else
Doug> the card is broken). If there was a mellanox card in the
Doug> machine there would be an entry in the lspci output similar
Doug> to this:
Exactly. There is no HCA device listed in your lspci output. Just
for c
$BFMA3$N%a!<%k$G?=$7Lu$4$6$$$^$;$s!#;[EMAIL
PROTECTED]&A49qE83+$N#3#0Be0J>e$N4{:'http://www.awg5.net/[EMAIL
PROTECTED]:w$7$F$b$i$([EMAIL PROTECTED]l9g$O!d"M([EMAIL PROTECTED]
///
___
On Tue, Feb 14, 2006 at 12:15:07PM -0800, Lamia M.Youseff wrote:
> -bash-3.00# lspci
> 00:00.0 Host bridge: Intel Corporation E7525 Memory Controller Hub (rev 0a)
> 00:02.0 PCI bridge: Intel Corporation E7525/E7520/E7320 PCI Express Port A
> (rev 0a)
> 00:03.0 PCI bridge: Intel Corporation E7525/
What does lspci -vv show?
Is the ib_mthca module loaded?
BTW, when debugging problems, you should probably set
CONFIG_INFINIBAND_MTHCA_DEBUG=y.
- R.
Thanks Roland for your reply. I recompiled the kernel with infiniband
debugging. After making sure lib_mthca module is load, I tried ibstatu
[EMAIL PROTECTED] ~]# opensm
-
OpenSM Rev:openib-1.1.0
Based on OpenIB svn 5411
Command Line Arguments:
Log File: /var/log/osm.log
-
*** glibc detected *** realloc(): invalid next size: 0x1007ae
Kyle> I have come across some error codes which are not entirely
Kyle> documented (or at least very well hidden).. They are
Kyle> generated when I try to send via rdma using uverbs interface
Kyle> on mthca drivers.
They are documented in the IB spec. See section 11.6.2 of volume
I have come across some error codes which are not entirely documented
(or at least very well hidden)..
They are generated when I try to send via rdma using uverbs interface on
mthca drivers.
When polling for work completions, I get the following behavior, and
have not been able to figure out i
I will be the Cisco SQA interface.
Scott Weitzenkamp
SQA Manager
Cisco Systems
-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Matt Leininger
Sent: Tuesday, February 14, 2006 9:25 AM
To: Roland Dreier (rdreier)
Cc: Robert J Woodruff; openib-general@openi
Matt wrote,
> Limiting the "release team" to three people is an attempt to keep this
> process manageable. However, for the OpenIB code release process to
> work well we will need to take advantage of the Q&A resources at several
> companies (as well as the wider community). To first order I'd l
On Tue, 2006-02-14 at 07:00 -0800, Roland Dreier wrote:
> It's great that so many people want to help with the release, but the
> whole point of having a release team was to have a small number of
> people that can move the release rapidly. I think that the three
> members of the team is the maxim
All, MADV_DONTFORK patch is now part of the -mm tree.
Everyone who's interested in fork support, please test 2.6.16-rc3-mm1 and
publish the results here and on lkml.
Please make sure to Cc me on results: I'm not subscribed to lkml.
Thanks,
--
Michael S. Tsirkin
Staff Engineer, Mellanox Technolo
Hi!
I am trying to enable MSI-X, then disable and enable MSI.
This fails.
~>modprobe ib_mthca msi_x=1
~>rmmod ib_mthca
Works fine
~>modprobe ib_mthca msi=1
The second modprobe fails to get interrupts from the device.
An attempt to fall back to regular interrupts fails as well:
ib_mthca :04
突然のメール、大変失礼致します。
はじめまして。
私は去年の11月から在宅ビジネスをやっています。
そこで今回はネット初心者から誰にでも出来きて、堅実なビジネスを
ご紹介しようと思います。
以前私は会社を経営しておりました。
しかしこの不況で倒産してしまい、日雇いのアルバイトで生計を
立てる毎日・・・。
生活は苦しく、本当に惨めな日々が続きました。
何か副業をしなければ食べて行けなくなり、ネットで
副業を探していて、パソコンで出来る在宅ビジネスというのを
初めてそこで知りました。
パソコンの知識が無い私が出来るのか・・・?
そこには
・月収200万円!!何もしなくても必ず稼げま
It's great that so many people want to help with the release, but the
whole point of having a release team was to have a small number of
people that can move the release rapidly. I think that the three
members of the team is the maximum number already.
Please give the release team time to identif
Hi Matt,
I would be happy to join the release team as additional Voltaire
representative.
Moni Levy | +972-971-7670(o)
Project Manager, Mainstream IB host stack
Voltaire – The Grid Backbone
http://www.voltaire.com/
On 2/14/06, Tziporet Koren <[EMAIL PROTECTED]> wrote:
> Hi Matt,
>
> Go
Hi Dotan,
Thanks for your reply!On 2/14/06, Dotan Barak <
[EMAIL PROTECTED]> wrote:
Supported Linux kernels versionsLatency
---
openIB
based applications
can may better latency than IBGD based applicaitonsDid anybody get an experiment result of comparison?
Few weeks ago, Mella
Your patch got line-wrapped :( Fortunately it was trivial enough to
apply by hand.
Thanks,
Roland
___
openib-general mailing list
openib-general@openib.org
http://openib.org/mailman/listinfo/openib-general
To unsubscribe, please visit http://openib.o
Arkady> Roland, in core/cache.c
Arkady> should
device-> cache.gid_cache =
Arkady> kmalloc(sizeof *device->cache.pkey_cache *
Arkady> (end_port(device) - start_port(device) + 1), GFP_KERNEL);
Arkady> be
device-> cache.gid_cache =
Arkady> km
Lamia> There is no trace for mthca, MTHCA, pci-x, pci-express or
Lamia> hca in dmsg or /var/log/messages. I guess this would mean a
Lamia> hardware problem (possibly the card is not present), Am i
Lamia> correct in my assumption?
What does lspci -vv show?
Is the ib_mthca module lo
Roland,
in
core/cache.c
should
device->cache.gid_cache =
kmalloc(sizeof *device->cache.pkey_cache *
(end_port(device) - start_port(device) + 1),
GFP_KERNEL);
be
device->cache.gid_cache =
kmalloc(sizeof *device->cache.gid_cache *
Hi Ian and welcome to the IB scene …
There are several things you should
consider when you choose the stack to work with:
Life
-
The IBGD is based on the VAPI driver (which won’t be
developed much in the future)
The openIB is under development
Stability
Hi all,
We are going to develop some applications over the InfiniBand network.
And I would like to get some advice on choosing the IB driver: IBGD or
OpenIB.
These applications are of certain scale and the developing will last
for a certain time. So a stable and reliable environment is needed.
As
Hi Matt,
Good that we start the release effort. I would like to join the release
team as Mellanox representative.
Tziporet
___
openib-general mailing list
openib-general@openib.org
http://openib.org/mailman/listinfo/openib-general
To unsubscribe, pl
DATE: 14th OF FEBRUARY 2006.
FROM: THE DESK OF THE VICE PRESIDENT.
PACIFIC PROMOTIONS ORGANIZATION PRIZE/AWARD
BANGKOK THAILAND.
BATCH: EGS/ 33504992/03: REFERENCE: 17/9918/IPD
ATTENTION: CONGRATULATIONS!
This is to inform you of the release of PACIFIC PROMOTIONS ORGANIZATION PRIZE
DRAW, held
In openib workshop Mellanox decided to open source our gen1 SRP target.
This implements an SRP target in software on top of a local SCSI device
supported by Linux.
Code is posted under:
https://openib.org/svn/trunk/contrib/mellanox/gen1/ib_srpt
This SRP target passed basic tests with gen2 SRP
$B$*5RMM$K3Ne$N%a!<%k%^!<%/$N%"[EMAIL
PROTECTED]>%"%I$K%a%C%;!<%8$rAw?.$7$F$/[EMAIL PROTECTED]AwJ8$H$J$j$^$9!#(B
$B%-%(!j!J(B33$B!K$O$8$a$^$7$F!"[EMAIL PROTECTED]>/[EMAIL
PROTECTED]($?$/[EMAIL PROTECTED])$rP$$$,@d$($J$$AGE($J4X78$r4|BT$7$F$$$^$9!#59$7$/$*4j$$$7$^$9!#(B
$B$7$g$&$3!j!J(
Correct, I missed that info in your post: the fact that you dont
have the devices under /sys/infiniband implies some hardware problem.
Do you see any messages from MTHCA in dmesg or /var/log/message
There is no trace for mthca, MTHCA, pci-x, pci-express or hca in dmsg or
/var/log/messages.
Quoting Lamia M.Youseff <[EMAIL PROTECTED]>:
> I am wondering if this error may be caused by absence/malfunction of the
> physical card, since this is a remote machine and I can not see it. If
> so, How can I check that card is pluged-in and working properly.
> I have included strace of devinfo a
This means that the problem is with the udev: it didnt create the devices
properly. I dont know how to debug this. Maybe updating to latest udev will
help.
It does not seem it is a udev problem. I have upgraded from udev-039-10.10.EL4 to udev-071-0.FC4.2, unloaded, reloaded the modules and then
Hi Roland.
The following patch wasn't committed:
Index: last_stable/drivers/infiniband/core/uverbs_main.c
===
--- last_stable.orig/drivers/infiniband/core/uverbs_main.c
2006-02-07 21:13:38.0 +0200
+++ last_stable/drivers/inf
Quoting r. Lamia M.Youseff <[EMAIL PROTECTED]>:
> Subject: Re: newbie to openib
>
>
> >>Hi Michael,
> >>Thank you for reply. The way I created the character device was by
> >>adding /etc/udev/rules.d/90-ib.rules, unloading the ib kernel modules,
> >>reloading them again, and restarting the udev
Hi Michael,
Thank you for reply. The way I created the character device was by
adding /etc/udev/rules.d/90-ib.rules, unloading the ib kernel modules,
reloading them again, and restarting the udev, which is the same way
described in the wiki (detailed sequence of commands is shown below). I
a
Quoting r. Lamia M.Youseff <[EMAIL PROTECTED]>:
> Subject: Re: newbie to openib
>
> Hi Michael,
> Thank you for reply. The way I created the character device was by
> adding /etc/udev/rules.d/90-ib.rules, unloading the ib kernel modules,
> reloading them again, and restarting the udev, which is
On Mon, 2006-02-13 at 22:30 -0800, Roland Dreier wrote:
> Or is the command doorbell stuff inside the UAR, which we already
> request in mthca_request_regions()?
Yes. These are offsets into UAR0 so there is no need to request a
region.
___
openib-genera
Hi Michael,
Thank you for reply. The way I created the character device was by
adding /etc/udev/rules.d/90-ib.rules, unloading the ib kernel modules,
reloading them again, and restarting the udev, which is the same way
described in the wiki (detailed sequence of commands is shown below). I
am
46 matches
Mail list logo