See inline..<br><br><div><span class="gmail_quote">On 02 Sep 2005 17:04:42 -0400, <b class="gmail_sendername">Hal Rosenstock</b> <<a href="mailto:halr@voltaire.com">halr@voltaire.com</a>> wrote:</span><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
On Fri, 2005-09-02 at 16:59, Viswanath Krishnamurthy wrote:<br>> Here is the setup..<br><br>Thanks. A couple more questions:<br><br>> #svn info<br>> Path: .<br>><br>> URL: <a href="https://openib.org/svn/gen2/trunk">
https://openib.org/svn/gen2/trunk</a><br>> Repository UUID: 21a7a0b7-18d7-0310-8e21-e8b31bdbf5cd<br>> Revision: 3295<br>> Node Kind: directory<br>> Schedule: normal<br>> Last Changed Author: halr<br>> Last Changed Rev: 3295
<br>> Last Changed Date: 2005-09-01 12:07:54 -0700 (Thu, 01 Sep 2005)<br>><br>><br>> Patch applied to core/at.c and kernel 2.6.13 recompiled.<br>><br>><br>> Machine A<br>> =========<br>> Running opensm
<br>><br>> Run ucmpost<br>><br>> machine B<br>> =========<br>> ./ucmpost <ipaddr_of_machineA><br><br>Are these back to back HCAs or is there a switch in between ?</blockquote><div><br>
<br>
There is a switch in between. A simple setup with 2 machines and a switch. The machines are running<br>
2.6.13. One of them is running opensm.<br>
</div><br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">> The problem is reproducible when you *cannot* ping each other<br><br>over IPoIB ?
</blockquote><div><br>
<br>
Yes.. <br>
</div><br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">> [root@subnetmgr4 ~]# ibv_devinfo<br>> hca_id: mthca0<br>>
fw_ver:
1.0.1<br>>
node_guid: 0002:c902:0040:0d00<br>>
sys_image_guid:
0002:c902:0040:0d03<br>>
max_mr_size: 0xffffffffffffffff<br>>
page_size_cap: 0x0<br>>
vendor_id: 0x02c9<br>>
vendor_part_id:
25204<br>>
hw_ver:
0x0<br>>
phys_port_cnt: 1<br>> port: 1<br>>
state: PORT_ACTIVE
(4)<br>>
max_mtu: invalid
MTU (0) <<br>> What is this ??><br>>
active_mtu:
invalid MTU (0)<br><br>If the program is right and those are the real values, somehow max_mtu<br>is trashed which causes active_mtu to be invalid which could break all<br>sorts of things...</blockquote><div><br>
Is there some issue with the HCA ? <br>
</div><br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">>
sm_lid:
1<br>>
port_lid:
3<br>>
port_lmc:
0x00<br><br>That's on the remote (from the SM) machine.<br><br>-- Hal<br><br></blockquote></div><br>