<br><br><div class="gmail_quote">On Fri, Jan 9, 2009 at 10:49 PM, Fab Tillier <span dir="ltr"><<a href="mailto:ftillier@windows.microsoft.com">ftillier@windows.microsoft.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
<div link="blue" vlink="purple" lang="EN-US">
<div>
<p><span style="font-size: 11pt; color: rgb(31, 73, 125);">I had suggested allowing devices to have a 'parent' device, in
order to let the ibbus driver create a hierarchy of devices. This would
eliminate the need for the ibiou driver altogether. </span></p></div></div></blockquote><div><br>So which driver will act as function driver for IOU PDO's created.<br><br>OR there will be no PDO's for IOU's and instead ibbus will create child PDO's representing IOC's.<br>
<br><br>Regards<br>Deepak<br></div><div><br></div><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;"><div link="blue" vlink="purple" lang="EN-US">
<div><br><p><span style="font-size: 11pt; color: rgb(31, 73, 125);"> </span></p>
<p><span style="font-size: 11pt; color: rgb(31, 73, 125);">Personally, I think it would make sense to start a new bus
driver, based on the KMDF, to implement this. It makes (in theory at least) handling
the child device lists much easier.</span></p>
<p><span style="font-size: 11pt; color: rgb(31, 73, 125);"> </span></p>
<p><span style="font-size: 11pt; color: rgb(31, 73, 125);">-Fab</span></p>
<p><span style="font-size: 11pt; color: rgb(31, 73, 125);"> </span></p>
<div style="border-style: none none none solid; border-color: -moz-use-text-color -moz-use-text-color -moz-use-text-color blue; border-width: medium medium medium 1.5pt; padding: 0in 0in 0in 4pt;">
<div>
<div style="border-style: solid none none; border-color: rgb(181, 196, 223) -moz-use-text-color -moz-use-text-color; border-width: 1pt medium medium; padding: 3pt 0in 0in;">
<p><b><span style="font-size: 10pt;">From:</span></b><span style="font-size: 10pt;">
<a href="mailto:ofw-bounces@lists.openfabrics.org" target="_blank">ofw-bounces@lists.openfabrics.org</a> [mailto:<a href="mailto:ofw-bounces@lists.openfabrics.org" target="_blank">ofw-bounces@lists.openfabrics.org</a>] <b>On
Behalf Of </b>Deepak Gupta<br>
<b>Sent:</b> Thursday, January 08, 2009 5:08 AM<br>
<b>To:</b> James Yang<br>
<b>Cc:</b> <a href="mailto:ofw@lists.openfabrics.org" target="_blank">ofw@lists.openfabrics.org</a>; Fab Tillier<br>
<b>Subject:</b> ***SPAM*** Re: [ofw] Expose a vendor defined device in ibbus?</span></p>
</div>
</div><div><div></div><div class="Wj3C7c">
<p> </p>
<p style="margin-bottom: 12pt;">James,<br>
<br>
Although I didn't get the time to test it but I had a look at the patch,<br>
<br>
Going by a dry run on the patch it looks to me that it is for creating vendor
defined child devices on "IBA".<br>
<br>
It looks ok to me, I will do a test on it and will reply.<br>
<br>
<br>
We need multiple IOC device interfaces (more than one child device mapping to
same IOC).<br>
If we go on the similar lines, it seems to me that we need to put the same
device creation logic in "ibiou" driver.<br>
<br>
But as specified above by Fab, If we are removing the IOC PnP Manager code from
IBAL, then there will be changes<br>
required in "ibiou" driver because currently "ibiou"
registers a PnP callback for IOC events. <br>
<br>
If IOC PnP manager code is removed from IBAL then we will need SA queries to be
done specifically from "ibiou" driver.<br>
<br>
I want to implement child device creations from "ibiou" too.<br>
Should I remove IOC PnP callback mechanisms in "ibiou" and make it
dependent on user configurations for creating the child devices?<br>
<br>
I am new to IBAL's code, So please let me know if I am getting anything wrong.<br>
<br>
Regards<br>
Deepak</p>
<div>
<p>On Thu, Jan 8, 2009 at 12:35 AM, James Yang <<a href="mailto:jyang@xsigo.com" target="_blank">jyang@xsigo.com</a>> wrote:</p>
<div>
<div>
<p><span style="font-size: 10pt; color: navy;">Please
use this new patch which fixed a few compile errors in free build.</span></p>
<p><span style="font-size: 10pt; color: navy;"> </span></p>
<p><span style="font-size: 10pt; color: navy;">Thanks,</span></p>
<p><span style="font-size: 10pt; color: navy;">James</span></p>
<p><span style="font-size: 10pt; color: navy;"> </span></p>
<p><span style="font-size: 10pt; color: navy;"> </span></p>
<div>
<div style="text-align: center;" align="center">
<hr align="center" size="2" width="100%">
</div>
<p><b><span style="font-size: 10pt;">From:</span></b><span style="font-size: 10pt;"> <a href="mailto:ofw-bounces@lists.openfabrics.org" target="_blank">ofw-bounces@lists.openfabrics.org</a>
[mailto:<a href="mailto:ofw-bounces@lists.openfabrics.org" target="_blank">ofw-bounces@lists.openfabrics.org</a>]
<b>On Behalf Of </b>James Yang<br>
<b>Sent:</b> Tuesday, January 06, 2009 12:11 PM<br>
<b>To:</b> Deepak Gupta; Fab Tillier<br>
<b>Cc:</b> <a href="mailto:ofw@lists.openfabrics.org" target="_blank">ofw@lists.openfabrics.org</a><br>
<b>Subject:</b> RE: [ofw] Expose a vendor defined device in ibbus?</span></p>
</div>
<div>
<div>
<p> </p>
<p><span style="font-size: 10pt; color: navy;">Hi,</span></p>
<p><span style="font-size: 10pt; color: navy;"> </span></p>
<p><span style="font-size: 10pt; color: navy;">Please
review the patch to create user defined devices by reading from registry. By
default there is only one IpoIB device enabled in mlx4_hca.inx file. This patch
will only work for ConnectX.</span></p>
<p><span style="font-size: 10pt; color: navy;"> </span></p>
<p><span style="font-size: 10pt; color: navy;">The
paritition key if set to default to FFFF, I didn't test on other value. And the
Ioctl part to add partition key may also need to be verified.</span></p>
<p><span style="font-size: 10pt; color: navy;"> </span></p>
<p><span style="font-size: 10pt; color: navy;">Thanks,</span></p>
<p><span style="font-size: 10pt; color: navy;">James</span></p>
<p><span style="font-size: 10pt; color: navy;"> </span></p>
<p><span style="font-size: 10pt; color: navy;"> </span></p>
<div>
<div style="text-align: center;" align="center">
<hr align="center" size="2" width="100%">
</div>
<p><b><span style="font-size: 10pt;">From:</span></b><span style="font-size: 10pt;"> <a href="mailto:mailmeatdkg@gmail.com" target="_blank">mailmeatdkg@gmail.com</a>
[mailto:<a href="mailto:mailmeatdkg@gmail.com" target="_blank">mailmeatdkg@gmail.com</a>]
<b>On Behalf Of </b>Deepak Gupta<br>
<b>Sent:</b> Monday, January 05, 2009 10:51 PM<br>
<b>To:</b> Fab Tillier<br>
<b>Cc:</b> Tzachi Dar; James Yang; <a href="mailto:ofw@lists.openfabrics.org" target="_blank">ofw@lists.openfabrics.org</a><br>
<b>Subject:</b> Re: [ofw] Expose a vendor defined device in ibbus?</span></p>
</div>
<p> </p>
<p style="margin-bottom: 12pt;">Have a gr8 New Year to all the members!!!.<br>
<br>
Do you we any updates on "vendor defined device in ibbus"?<br>
<br>
I wanted to create multiple vnic interfaces irrespective of number of reachable
IOCs.<br>
<br>
Currently I am creating vnic child devices on "root" bus.<br>
Inside VNIC driver, I am looking for arrival GUID_IB_AL_INTERFACE and hence
contacting the IBAL.<br>
But since vnic devices are root enumerated, VNIC driver gets loaded very
earlier in boot phase (Before "Extended Base" group to which IB Stack
drivers belong) and hence VNIC device interfaces are not getting initialized
properly.<br>
<br>
If we are providing a vendor defined device functionality in ibbus in near
future, then it would be worth for me to wait for it.<br>
<br>
Can any one please comment on this.<br>
<br>
Regards<br>
Deepak</p>
<div>
<p>On Mon, Dec 22, 2008 at 1:12 PM, Deepak Gupta <<a href="mailto:deepak.gupta@qlogic.com" target="_blank">deepak.gupta@qlogic.com</a>>
wrote:</p>
<p>All,<br>
<br>
I came across one more question in my mind which are I think is not clear to me
after reading the whole thread.<br>
<br>
In new design being discussed:- <br>
Are we making sure that we can have more than one child devices configured for
the same IOC.<br>
Currently, there is one child device created per IOC discovered.<br>
<br>
Having more than one child device configured for same IOC is a requirement if a
user wants two different ULP interfaces to be created on host side.<br>
Consider a case in which a host is connected to a single IOC and IOC is
connected to a ethernet network via switch.<br>
If there are two different IP subnets then there is a requirement of two
different Ethernet interfaces on the host side too.<br>
<br>
Please let me know if you need more clarification of my question.<br>
<br>
Regards<br>
<span style="color: rgb(136, 136, 136);">Deepak</span></p>
<div>
<div>
<p style="margin-bottom: 12pt;"> </p>
<div>
<p>On Sat, Dec 20, 2008 at 2:19 PM, Deepak Gupta <<a href="mailto:deepak.gupta@qlogic.com" target="_blank">deepak.gupta@qlogic.com</a>>
wrote:</p>
<p style="margin-bottom: 12pt;">Please see below.<br>
<br>
Regards<br>
Deepak</p>
<div>
<div>
<div>
<p>On Sat, Dec 20, 2008 at 12:42 AM, Fab Tillier <<a href="mailto:ftillier@windows.microsoft.com" target="_blank">ftillier@windows.microsoft.com</a>>
wrote:</p>
<div>
<p>> On Wed, Dec 17, 2008 at 11:43 PM, Fab Tillier<br>
> <<a href="mailto:ftillier@windows.microsoft.com" target="_blank">ftillier@windows.microsoft.com</a>>
wrote:</p>
</div>
<div>
<p style="margin-bottom: 12pt;">>> Are there other properties that I have
missed that are needed?<br>
><br>
> We need a way in which devices created should be configured for<br>
> failovers ( ULPs like VNIC, SRP need more configurable failovers).<br>
> Looking at IBAL's code it create the devices based on the reachable<br>
> IOC's and thus failover's are possible across the HCA/ports and not<br>
> across two different IOCs.<br>
> Users can have a case in which two different IOCs connected to same<br>
> physical network/storage (redundancy is provided for high availability)<br>
> and want a failover across the IOCs.</p>
</div>
<p>This would be done via LBFO for network devices, and MPIO for storage
devices. I think having the bus driver report a single IOC that really
maps to two IOCs on the fabric is asking for management problems.
Further, LBFO/MPIO can provide failover between different device types,
so the failover devices don't have to be identical HW.</p>
</div>
</div>
<div>
<p><br>
I don't know about how MPIO works. But for LBFO, BundleID param will have
to be included in extended params then so that user gets the freedom of bundling
different failover configurations.<br>
</p>
</div>
<blockquote style="border-style: none none none solid; border-color: -moz-use-text-color -moz-use-text-color -moz-use-text-color rgb(204, 204, 204); border-width: medium medium medium 1pt; margin: 5pt 0in 5pt 4.8pt; padding: 0in 0in 0in 6pt;">
<p><br>
<span style="color: rgb(136, 136, 136);"><br>
-Fab</span></p>
</blockquote>
</div>
<p> </p>
</div>
<p> </p>
</div>
</div>
</div>
<p> </p>
</div>
</div>
</div>
</div>
<p><br>
_______________________________________________<br>
ofw mailing list<br>
<a href="mailto:ofw@lists.openfabrics.org" target="_blank">ofw@lists.openfabrics.org</a><br>
<a href="http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ofw" target="_blank">http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ofw</a></p>
</div>
<p> </p>
</div></div></div>
</div>
</div>
<br>_______________________________________________<br>
ofw mailing list<br>
<a href="mailto:ofw@lists.openfabrics.org">ofw@lists.openfabrics.org</a><br>
<a href="http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ofw" target="_blank">http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ofw</a><br></blockquote></div><br>