$B7P83$N$J$$J}$K$b%T%C%?%j!#(B
$B5.J}$O:#$N;E;v$KBP$7$F>-MhIT0B$,$"$j$^$9$+!)(B
$B5.J}$O:#$N;E;v$KK\Ev$KK~B-$7$F$$$^$9$+!)(B
$B5.J}$OO78e$KBP$7$FIT0B$,$"$j$^$9$+!)(B
$B5.J}$O(B30$BG/8e$b%5%i%j!<%^%s([EMAIL PROTECTED];W$$$^$9$+!)(B
(YES$B$N?M$O:#$9$0$7$?$N(BURL$B$+$i;22C"-"-"-(B)
Robert> I'll look it over this weekend or early next week and
Robert> commit it then. Thanks for the patch!
Great. I'm going to wait until Monday afternoon or so for feedback on
the core part, so no hurry.
- R.
___
openib-general mailing list
On Fri, 2006-02-10 at 16:51 -0800, Roland Dreier wrote:
> Here is a series of patches that adds a new function ib_modify_qp_is_ok(),
> which low-level drivers can use to replace boilerplate logic for
> validating the parameters to the modify_qp method.
>
> In addition to getting rid of duplicated
Here is a series of patches that adds a new function ib_modify_qp_is_ok(),
which low-level drivers can use to replace boilerplate logic for
validating the parameters to the modify_qp method.
In addition to getting rid of duplicated bugs, this ends up saving
quite a lot of duplicated code across mt
Convert ehca to use ib_modify_qp_is_ok() instead of its own QP
transition table. Compile tested only because of lack of hardware.
(Well, actually IBM has kindly supplied a couple of systems, but the
rack isn't quite set up).
In any case, IBM people: please make sure I didn't break anything, and
i
Convert ipath to use ib_modify_qp_is_ok() instead of its own QP
transition table. Lightly tested.
PathScale people: please test this out and apply if it's OK.
--- infiniband/hw/ipath/ipath_verbs.c (revision 5364)
+++ infiniband/hw/ipath/ipath_verbs.c (working copy)
@@ -1003,213 +1003,6 @@ st
Convert mthca to use ib_modify_qp_is_ok() instead of its own QP
transition table.
--- infiniband/hw/mthca/mthca_cmd.c (revision 5364)
+++ infiniband/hw/mthca/mthca_cmd.c (working copy)
@@ -1566,31 +1566,56 @@ int mthca_ARM_SRQ(struct mthca_dev *dev,
CMD_TIME_CLASS_
Add ib_modify_qp_is_ok() to the IB midlayer.
--- infiniband/core/verbs.c (revision 5364)
+++ infiniband/core/verbs.c (working copy)
@@ -244,6 +244,266 @@ struct ib_qp *ib_create_qp(struct ib_pd
}
EXPORT_SYMBOL(ib_create_qp);
+static const struct {
+ int valid
Hefty, Sean wrote:
Here's an updated version of rping. I restructured to code to make it
more modular, reduce the size of some of the functions, simplify some
areas, and make it more consistent. The updated version worked for my
limited testing. Please review the changes to see if I changed an
Sean Hefty wrote:
This is yet another updated patch that releases the reference on the acquired
route.
I've committed this version of the patch. Thanks.
- Sean
___
openib-general mailing list
openib-general@openib.org
http://openib.org/mailman/list
$B7P83$N$J$$J}$K$b%T%C%?%j!#(B
$B5.J}$O:#$N;E;v$KBP$7$F>-MhIT0B$,$"$j$^$9$+!)(B
$B5.J}$O:#$N;E;v$KK\Ev$KK~B-$7$F$$$^$9$+!)(B
$B5.J}$OO78e$KBP$7$FIT0B$,$"$j$^$9$+!)(B
$B5.J}$O(B30$BG/8e$b%5%i%j!<%^%s([EMAIL PROTECTED];W$$$^$9$+!)(B
(YES$B$N?M$O:#$9$0$7$?$N(BURL$B$+$i;22C"-"-"-(B)
Sean:
Good catch. You're right. I was looking in the wrong place. It is
processed as you suggest in process_req(...).
We don't need the change to the CMA, only the mod in
addr_resolve_remote(...).
On Fri, 2006-02-10 at 17:38 -0600, Tom Tucker wrote:
> I'll try it and see what happens. I haven'
I'll try it and see what happens. I haven't tested your approach in the
latest tree, so it may work as you describe.
Thanks,
Tom
On Fri, 2006-02-10 at 14:44 -0800, Sean Hefty wrote:
> >I think what would happen in that case is that the ARP request would get
> >queued, but never resolved because
This is yet another updated patch that releases the reference on the acquired
route.
Signed-off-by: Sean Hefty <[EMAIL PROTECTED]>
---
Index: addr.c
===
--- addr.c (revision 5295)
+++ addr.c (working copy)
@@ -36,6 +36,7 @
>I think what would happen in that case is that the ARP request would get
>queued, but never resolved because the ARP reply would never hit the
>host (it's eaten by the adapter itself).
Returning success (0) from addr_resolve_remote() should mark the request as done
and queue it for a callback. T
Sean:
Sorry for the delay ...
I think what would happen in that case is that the ARP request would get
queued, but never resolved because the ARP reply would never hit the
host (it's eaten by the adapter itself).
So eventually the request would timeout, fail to find a resolved next
hop, and en
On Thu, 2006-02-09 at 23:56, Karun Beer Sharma wrote:
> Hi:
>
> I want to compile openIB with 2.6.9-22EL kernel version. It is giving me
> error with respect to iSER module.
> After browsing through openib.org website, I came to know that iSER is
> supported from 2.6.11 onwards but some work is
On Fri, 2006-02-10 at 13:56, Hal Rosenstock wrote:
> On Thu, 2006-02-09 at 10:55, Roland Dreier wrote:
> > The current code probably needs to be extended to give
> > the upper layers the port number where a MAD was received, and also to
> > make it possible to specify the port number that a directe
$B7P83$N$J$$J}$K$b%T%C%?%j!#(B
$B5.J}$O:#$N;E;v$KBP$7$F>-MhIT0B$,$"$j$^$9$+!)(B
$B5.J}$O:#$N;E;v$KK\Ev$KK~B-$7$F$$$^$9$+!)(B
$B5.J}$OO78e$KBP$7$FIT0B$,$"$j$^$9$+!)(B
$B5.J}$O(B30$BG/8e$b%5%i%j!<%^%s([EMAIL PROTECTED];W$$$^$9$+!)(B
(YES$B$N?M$O:#$9$0$7$?$N(BURL$B$+$i;22C"-"-"-(B)
$B!~"!!~(BID105246$BMMFCJL>7BT>u"!!~"!(B
$B4JC1$K$*6b$r2T$0;v$,$G$-$^$9!#K\Ev$K4JC1$G$9!#(B
$B8+$FD:$1$l$P$9$0J,$+$j$^$9!#(B
$B"->\$7$$>\:Y$O"-(B
http://www.gyakuten5.net/?okane/
$B%a!<%khttp://openib.org/mailman/listinfo/openib-general
To unsubscribe, please visit http://openib
> Arlin,
> This can be done.
>
> But I have an issue that extension call violate Transport Requirement.
> Currently, the matching semantic is well-defined since
> Recv only matches Send. Since Spec does not have any idea what
> operations are defined in extension(s) there is a problem
> with the t
Hi Tom,
> Fab:
>
> As you point out, we've been focused on the main trunk and or target
> test platforms are Linux based. That said, we actually had Beta versions
> of NDIS and Winsock Direct drivers for the AMSO adapter, so we know this
> works and we know where the dead are buried.
>
> It prob
On Thu, 2006-02-09 at 10:55, Roland Dreier wrote:
> The current code probably needs to be extended to give
> the upper layers the port number where a MAD was received, and also to
> make it possible to specify the port number that a directed route MAD
> will be sent on.
On the receive side, the po
At 09:43 AM 2/10/2006, Grant Grundler wrote:
On Fri, Feb 10, 2006 at
11:05:34AM -0500, Hal Rosenstock wrote:
> > Hi, Roland!
> > One issue we have with IPoIB is that IPoIB may cache a remote
node path
> > for a long time. Remote LID may get changed e.g. if the SM is
changed,
> > and IPoIB might l
On Fri, 2006-02-10 at 13:06, Michael Di Domenico wrote:
> On 10 Feb 2006 12:11:03 -0500, Hal Rosenstock <[EMAIL PROTECTED]> wrote:
> > Hi Michael,
> >
> > On Fri, 2006-02-10 at 10:11, Michael Di Domenico wrote:
> > > trying to compile libibmad, but not installing to /usr/local i'm
> > > installing
On 10 Feb 2006 12:11:03 -0500, Hal Rosenstock <[EMAIL PROTECTED]> wrote:
> Hi Michael,
>
> On Fri, 2006-02-10 at 10:11, Michael Di Domenico wrote:
> > trying to compile libibmad, but not installing to /usr/local i'm
> > installing to /opt/openib. libibcommon and libibumad compiled just
> > fine wi
Tom,
Can you tell me if this alternate patch would work? I let the ib_addr module
queue the callback. No CMA changes are necessary then.
Signed-off-by: Sean Hefty <[EMAIL PROTECTED]>
---
Index: addr.c
===
--- addr.c (revisio
On Fri, 2006-02-10 at 12:43, Grant Grundler wrote:
> On Fri, Feb 10, 2006 at 11:05:34AM -0500, Hal Rosenstock wrote:
> > > Hi, Roland!
> > > One issue we have with IPoIB is that IPoIB may cache a remote node path
> > > for a long time. Remote LID may get changed e.g. if the SM is changed,
> > > and
On Fri, Feb 10, 2006 at 11:05:34AM -0500, Hal Rosenstock wrote:
> > Hi, Roland!
> > One issue we have with IPoIB is that IPoIB may cache a remote node path
> > for a long time. Remote LID may get changed e.g. if the SM is changed,
> > and IPoIB might lose connectivity.
I wonder if this is why when
Hi Michael,
On Fri, 2006-02-10 at 10:11, Michael Di Domenico wrote:
> trying to compile libibmad, but not installing to /usr/local i'm
> installing to /opt/openib. libibcommon and libibumad compiled just
> fine with the same configure syntax
>
> (cd libibmad && ./autogen.sh && ./configure --pref
$B7P83$N$J$$J}$K$b%T%C%?%j!#(B
$B5.J}$O:#$N;E;v$KBP$7$F>-MhIT0B$,$"$j$^$9$+!)(B
$B5.J}$O:#$N;E;v$KK\Ev$KK~B-$7$F$$$^$9$+!)(B
$B5.J}$OO78e$KBP$7$FIT0B$,$"$j$^$9$+!)(B
$B5.J}$O(B30$BG/8e$b%5%i%j!<%^%s([EMAIL PROTECTED];W$$$^$9$+!)(B
(YES$B$N?M$O:#$9$0$7$?$N(BURL$B$+$i;22C"-"-"-(B)
Hal> Should we remove all the svn IDs from the OpenIB kernel tree?
It's probably worth leaving them in our svn tree. If you want to make
a patch that removes them from the upstream kernel, that would be fine.
- R.
___
openib-general mailing list
o
On Thu, 2006-02-09 at 07:10, Yael Kalka wrote:
> Hi Hal,
>
> Currently, the OpenSM sends PortInfo with ClientReRegistration bit
> turned on only during the first sweep after becoming Master.
> This doesn't cover all cases where ClientReRegistration should be
> turned on. OpenSM should turn on this
___
openib-general mailing list
openib-general@openib.org
http://openib.org/mailman/listinfo/openib-general
To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
On Wed, 2006-02-08 at 15:14, Michael S. Tsirkin wrote:
> Hi, Roland!
> One issue we have with IPoIB is that IPoIB may cache a remote node path for a
> long time. Remote LID may get changed e.g. if the SM is changed, and IPoIB
> might
> lose connectivity.
The remote LID may get changed for other r
Hi Suri,
On Fri, 2006-02-10 at 09:09, Suresh Shelvapille wrote:
> Hal:
> The way I gather port stats locally on my device is by doing a "cat" on the
> "/sys/class/Infiniband/mysw/port/0/counters" file. As Roland mentioned this
> is limited to port0 counters on a switch.
Base switch port 0 does n
On Thu, 2006-02-09 at 08:31, Sasha Khapyorsky wrote:
> Hi,
>
> This will check config dir existance and create it if neccessary. As
> previous one but for libraries and diags.
Thanks. Applied.
> Sasha.
>
>
> This will check config dir existance and create it if neccessary.
>
> Signed-off-by:
Hi Suri,
You can send a PMA PortCounters MAD either using "process local" or to
the device using its LID.
Eitan
> I am not using user space verbs library, is there any other way to get
at
> the different port#N stats?
>
> Thanks,
> Suri
___
openib-gene
This patch adds support for devices that do ARP internally. The
patch checks the flags in the netdevice during ARP resolution
processing to see if it supports ARP. If it does not, it will
return 'done' up the stack until it hits the CMA where a callback
will be generated on the CMA workqueue thre
trying to compile libibmad, but not installing to /usr/local i'm
installing to /opt/openib. libibcommon and libibumad compiled just
fine with the same configure syntax
(cd libibmad && ./autogen.sh && ./configure --prefix=/opt/openib
LDFLAGS=-L/opt/openib/lib CFLAGS=-I/opt/openib/include && make &
In message <[EMAIL PROTECTED]>,Roland Dreier writes:
>Yes, it's exactly because we know that work queues run in process
>context with interrupts enabled which lets us use spin_lock_irq.
thanks for the reply. you are quite right. i dont know what i was
thinking.
>There's no limitation on number
Hi,
In most all the OpenIB code, we have been putting in svn IDs. In the
kernel submission of the ipath code, the following comment came back
from lkml from Christoph.
Should we remove all the svn IDs from the OpenIB kernel tree ?
-- Hal
-Forwarded Message-
From: Christoph Hellwig <[EM
Hal:
The way I gather port stats locally on my device is by doing a "cat" on the
"/sys/class/Infiniband/mysw/port/0/counters" file. As Roland mentioned this
is limited to port0 counters on a switch.
If I need stats for port#N I need to expand this by providing more class
attributes and getting to
On Wed, 2006-02-08 at 03:50, Yael Kalka wrote:
> Hi Hal,
>
> There was a problem with some of the types defined when compiling on
> 64bit windows machines. The following patch adds support for these as
> well.
Thanks. Applied.
-- Hal
> Thanks,
> Yael
>
> Signed-off-by: Yael Kalka <[EMAIL PROT
On Wed, 2006-02-08 at 03:30, Yael Kalka wrote:
> Hi Hal,
>
> The following patch adds a missing type casting in the return value of
> the function osm_mcast_mgr_compute_max_hops.
Thanks. Applied.
-- Hal
> Thanks,
> Yael
___
openib-general mailing li
On Tue, 2006-02-07 at 03:28, Yael Kalka wrote:
> Hi Hal,
>
> The following patch includes some fixes for the windows stack:
> 1. Add needed __cdecl.
> 2. Change the default directories/files names.
Thanks. Applied.
-- Hal
> Thanks,
> Yael
>
> Signed-off-by: Yael Kalka <[EMAIL PROTECTED]>
__
When it runs I get very good results about 228 MB/s
with the same sample program without sdp with ipoib i get about 130 MB/s
But if I try to connect to a non running server I get this :
iommu_free: invalid entry
entry = 0x0
dma_addr = 0x0
Table = 0xc000edf3448
*☆*〜*☆*出逢いを求める人へ*☆*〜*☆*
◆男女登録割合☆53%・47%
◆女性スタッフもいるので安心して登録できます
覗いてみよう【http://www.deai-style.net/casanova/?1927】
◆◇◆遊び半分の出逢い〜本当の出逢いまで提供します◆◇◆
出逢いはここから
http://www.deai-style.net/casanova/?1927
★☆★今回は体験談を紹介しま〜す★☆★
☆‥…━☆ある日‥…☆‥…━☆‥…━
あきら 28歳 男 埼玉県
これは、9月の初めの出来
Le jeudi 09 février 2006 à 09:09 -0800, Ralph Campbell a écrit :
> My guess is there is a bug in the zero-copy code.
> Try "echo 100 > /sys/module/ib_sdp/parameters/sdp_zcopy_thrsh_src_default"
> and see if the problem still exists.
> This raises the zero-copy threshold.
I think you are right,
49 matches
Mail list logo