Thanks, I applied cleaned-up versions of all three patches for 2.6.24. One thing I changed was to just pass an error back to the caller rather than doing BUG_ON() anywhere. It's very unfriendly to the user to crash the whole machine just because of a driver bug -- much better to try and continue so that the user sees the error and can report it. _______________________________________________ general mailing list general@lists.openfabrics.org [ message continues ]

> On one run we got this in syslog (ib_mthca's debug_level set to 1): > > 15:34:34 ib_mthca 0012:01:00.0: Command 21 completed with status 09 > 15:35:34 ib_mthca 0012:01:00.0: HW2SW_MPT failed (-16) > .... > (status 0x9==MTHCA_CMD_STAT_BAD_RES_STATE => problem with mpi?) > > or on another run: > > 13:57:15 ib_mthca 0005:01:00.0: Command 1a completed with status 01 > 13:57:15 ib_mthca 0005:01:00.0: modify QP 1->2 returnedstatus 01. > ....

Genius! I have completed over 275 runs with the patch below, so we can be very confident that this has fixed things. Roland, should I submit a proper patch, or do you want to take care of this? (And thanks alot, too!) diff -rup ofa_kernel-1.2.orig/drivers/infiniband/hw/mthca/mthca_cmd.c ofa_kernel-1.2/drivers/infiniband/hw/mthca/mthca_cmd.c --- ofa_kernel-1.2.orig/drivers/infiniband/hw/mthca/mthca_cmd.c 2007-06-21 07:38:47.000000000 -0700 +++ ofa_kernel-1.2/drivers/infiniband/hw/mthca/mthc...

OFA Admins: It would be nice to put the man pages on-line... If we installed the man pages, then used man2html or something we could point folks at that for on-line docs... Zulfi, if you build/install ofed-1.2.5, you can then get man pages for the verbs and rdmacm APIs. Also there are header files and examples that get build/installed. Steve. _______________________________________________ general mailing list general@lists.openfabrics.org [ message continues ]

Rick, have you had a chance to test out this patch? _______________________________________________ general mailing list general@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general

FRLE begins to deliver promised returns, Shares up over 31%. Fearless International Inc. (F R L E) $0.25 UP 31.76 % Hard climb for the hottest new yacht on the market, shares jumped nearly 32% today. You cant ignore these kind of numbers, this is going to be huge. There is a time and place for everything, and Friday more is yours, grab this one early. _______________________________________________ general mailing list general@lists.openfabrics.org [ message continues ]

----boundary_2054012_0b00d57b-fba4-47c8-849c-606cdcdd600d content-type: text/plain; charset=iso-8859-1 content-transfer-encoding: quoted-printable I just joined Shelfari to connect with other book lovers. Come see the books= I love and see if we have any in common. Then pick my next book so I can= keep on reading.=0A=0AClick below to join my group of friends on Shelfari!= =0A=0Ahttp://www.shelfari.com/Register.aspx?ActivityId=3D22801633&InvitationCode=3Dc70c284d-a0dd-4a69-a023-3022d4752243= ...

I think it would help usability a lot to put the PARTITION CONFIGURATION section in a separate 'opensm-partitions.conf' man page with the values for rate, mtu and scope listed directly. _______________________________________________ general mailing list general@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general

> I tested this by simulating a slow passive side responder, and it worked as > expected for those tests. Using an MRA does add another MAD to the CM exchange, > which is why it is sent only after seeing a duplicate request. Alternatively, > we can take the OFED module parameter patch. What the heck, I added this for 2.6.24. If it doesn't work out we can back it out. - R. _______________________________________________ general mailing list general@lists.openfabrics.org http...

------=_Part_13547_30004409.1191621670073 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline Hi Dotan, ifconfig shows up ib0 Link encap:InfiniBand HWaddr 80:00:00:03:FE:80:00:00:00:00:00:00:00:00:00:00:00:00:00:00 inet addr: 140.221.37.32 Bcast: 140.221.37.255 Mask: 255.255.255.0 inet6 addr: fe80::211:7500:ff:d802/64 Scope:Link UP BROADCAST RUNNING MULTICAST MTU:65520 Metric:1 RX packet...

------=_Part_13764_7924086.1191626722839 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline I restarted openibd and now my interfaces are up. mach#1 ib0 Link encap:InfiniBand HWaddr 80:00:00:02:FE:80:00:00:00:00:00:00:00:00:00:00:00:00:00:00 inet addr:140.221.37.46 Bcast:140.221.37.255 Mask:255.255.255.0 inet6 addr: fe80::211:7500:ff:d7f2/64 Scope:Link mach#2 ib0 Link encap:InfiniBand HWaddr 80:00:00:...

------=_Part_13612_8676690.1191623167533 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline For machine#1 my IPoIB interface is ib0 Link encap:InfiniBand HWaddr 80:00:00:03:FE:80:00:00:00:00:00:00:00:00:00:00:00:00:00:00 inet addr:140.221.37.32 Bcast:140.221.37.255 Mask:255.255.255.0 inet6 addr: fe80::211:7500:ff:d802/64 Scope:Link UP BROADCAST RUNNING MULTICAST MTU:65520 Metric:1 RX packe...

> I understand this desire... just need a little clarification from you > re hotplug. First, as for OFED, looking on the openibd service script > (excerpts below) installed by OFED 1.3 I see that mode and mtu are set > "manually", that is the user sets/provides the mode and mtu params for > the script and the script uses sysfs to configure the device. This > does not address devices created after the service has started nor > seem a very elegant way to do so. I don't kno...

Add performance/debug counters to track sent/received messages, retries, and duplicates. Counters are tracked per CM message type, per port. The counters are always enabled, so intrusive state tracking is not done. Counters are exported as: /sys/class/infiniband_cm/device/port/counter_description/cm_attribute for example: /sys/class/infiniband_cm/mthca0/1/cm_tx_msgs/req /sys/class/infiniband_cm/mthca0/1/cm_tx_retries/rep Signed-off-by: Sean Hefty <sean.hefty@intel.com> This m...