On 12/16/2012 8:39 AM, Jens Domke wrote: > Hi, > > On Dec 16, 2012, at 9:32 PM, Hal Rosenstock wrote: > >> Hi, >> >> On 12/16/2012 7:03 AM, Jens Domke wrote: >>> Hello Hal, >>> >>> On Dec 15, 2012, at 5:44 AM, Hal Rosenstock wrote: >>> >>>> Hi, >>>> >>>> On 12/14/2012 3:32 PM, Jens Domke wrote: >>>>> Hello Hal, >>>>> >>>>> On Dec 15, 2012, at 3:58 AM, Hal Rosenstock wrote: >>>>> >>>>>> Hi, >>>>>> >>>>>> On 12/14/2012 1:24 PM, Jens Domke wrote: >>>>>>> Hello Hal, >>>>>>> >>>>>>> On Dec 15, 2012, at 1:42 AM, Hal Rosenstock wrote: >>>>>>> >>>>>>>> Hi again, >>>>>>>> >>>>>>>> On 12/14/2012 10:17 AM, Jens Domke wrote: >>>>>>>>> Hello Hal, >>>>>>>>> >>>>>>>>> thank you for the fast response. I will try to clarify some points. >>>>>>>>> >>>>>>>>>>> d) OpenMPI runs are executed with "--mca >>>>>>>>>>> btl_openib_ib_path_record_service_level 1" >>>>>>>>>> >>>>>>>>>> I'm not familiar with what DFSSSP does to figure out SLs exactly but >>>>>>>>>> there should be no need to set this. The proper SL for querying the >>>>>>>>>> SA >>>>>>>>>> for PathRecords, etc. is always in PortInfo.SMSL. In the case of >>>>>>>>>> DFSSSP >>>>>>>>>> (and other QoS based routing algorithms), it calculates that and the >>>>>>>>>> SM >>>>>>>>>> pushes this into each port. That should be used. It's possible that >>>>>>>>>> SL1 >>>>>>>>>> is not a valid SL for port <-> SA querying using DFSSSP. >>>>>>>>> The OpenMPI parameter btl_openib_ib_path_record_service_level does >>>>>>>>> not specify the SL for querying the PathRecords. >>>>>>>>> It just enables the functionality. And the ompi processes use the >>>>>>>>> PortInfo.SMSL to send the request. >>>>>>>>> For the request "port -> SA" every 0<=SL<=7 was used in the test, and >>>>>>>>> the SA received the requests. >>>>>>>>>> >>>>>>>>>>> e) kernel 2.6.32-220.13.1.el6.x86_64 >>>>>>>>>>> >>>>>>>>>>> As far as I understand the whole system: >>>>>>>>>>> 1. the OMPI processes are sending MAD requests >>>>>>>>>>> (SubnAdmGet:PathRecord) to the OpenSM >>>>>>>>>>> 2. the SA receives the request on QP1 >>>>>>>>>> >>>>>>>>>> There is the SL in the query itself. This should be the SMSL that >>>>>>>>>> the SM >>>>>>>>>> set for that port. >>>>>>>>> Hmm, there you might have a point. I think I saw that the query >>>>>>>>> itself had SL=0 specified. >>>>>>>>> In fact OpenMPI sets everthing to 0 except for slid and dlid. >>>>>>>>>> >>>>>>>>>>> 3. SA asks the routing algorithm (like LASH, DFSSSP or Torus_2QoS) >>>>>>>>>>> about a special service level for the slid/dlid path >>>>>>>>>> >>>>>>>>>> This is a (potentially) different SL (for MPI<->MPI port >>>>>>>>>> communication) >>>>>>>>>> than the one the query used and is the one returned inside the >>>>>>>>>> PathRecord attribute/data. >>>>>>>>> Yes, it can be different, but DFSSSP sets the same SL, because the SM >>>>>>>>> is running on a port which is also used for MPI comm. >>>>>>>> >>>>>>>> With DFSSSP are all SLs same from source port to get to any >>>>>>>> destination ? >>>>>>> No, not necessarily. In general DFSSSP does not enforce SL(LID1->LID2) >>>>>>> == SL(LID2->LID1) or SL(LID1->LID2) == SL(LID1->LID3). >>>>>> >>>>>> If SL(LID1->LID2) != SL(LID2->LID1), that's not a reversible path. >>>>> True. But i don't think that the SA asks the DFSSSP routing about the SL >>>>> for the reversible path. >>>>> So, the SA could use any SL which is a valid SL, even if the DFSSSP would >>>>> recommend another SL. >>>>> >>>>> I just read the IB Specs and it says, that "SL specified in the received >>>>> packet is used as the SL in the response packet" for MAD packets. >>>>> So, its most likely, that there is a mismatch in the way how OMPI does >>>>> the setup of the PathRequest and the way how the SA does build the >>>>> respond packet. >>>>> OMPI always specifies SL=0 (lets say SL_a) inside of the PathRequest >>>>> packet, >>>> >>>> So CompMask in the query has the SL bit on and SL is set to 0 inside the >>>> SubAdmGet of PatchRecord ? >>> >>> No, the CompMask didn't had the SL bit and the SL was set to 0. >> >> That means the SL in the request is wildcarded so the SA/SM fills in a >> valid one in the response. > Ok. >> >>> I tried to follow the path of the SL bit (IB_PR_COMPMASK_SL) and the only >>> reference I found was in osm_sa_path_record.c >>> The SA just treats the SL in the PathRequest as a "I would like to use this >>> SL" in case the SL bit is set. >>> But the routing engine can overwrite the requested SL before the reply is >>> send. >>> >>> Nevertheless, I have changed the code of OMPI so that it sets the SL bit in >>> the CompMask and sets the SL to SMSL for the PathRequest, so that SL_a == >>> SL_b. >>> Sadly, the reply send by the SA does not leave the node (for SL_b>0). Only >>> if I change the SL to 0 in the MAD right before umad_send is called by the >>> SA, the paket is able to leave the node and reaches the OMPI process. >> >> Are you sure the response doesn't leave the SA node or it's not received >> at the requester (OMPI node) ? > No, I'm not sure. Is there any possibility to check that? As far as I know, > ibdump does not show MAD pakets which leave a port, it only shows the pakets > when they are received on the other end. >> >>> >>>> >>>>> and sends the packet on SL_b (PortInfo.SMSL). >>>> >>>> Good. >>>> >>>>> The SA uses p_mad_addr->addr_type.gsi.service_level, which is SL_b, for >>>>> the response. >>>>> If SL_b is not 0, then the packet can't reach the OMPI process. Right? >>>> >>>> Depends. It may be that both SLs work but maybe not. >>>> >>>>> If I analyse this correctly, then there are two bugs. One is in OMPI, >>>>> that it does not specify the SL within the PathRequest in a appropriate >>>>> way (which would be a SL suggested by DFSSSP for the reversible path). >>>>> And the second bug is that the SA uses the SL, on which the PathRequest >>>>> packet was send, and not the SL specified within the packet. >>>>> What do you think? >>>> >>>> Yes, it might be better to wildcard the SL in the query. The only >>>> scenario that would fail with the query you are making if there's no SL >>>> 0 path between the src/dest LIDs or GIDs in the OMPI PathRecord query. >>>> If that's the case, SA should return MAD status 0xc (status code 3 - >>>> ERR_NO_RECORDS). But the response doesn't make it back to the requester >>>> OMPI node so it's not even getting that far. >>> >>> Yes, exactly. So, do you have an idea why the response hands in the SA node? >>> I have no inside of the underlying layer (kernel driver and fireware). >>> Maybe there are some implementations, which prevent the SA from sending >>> MADs back on SL>0? >> >> If you're sure this response doesn't get out of the SA node, please >> contact Mellanox support with the details. > Ok, I can do this, if it turns out to be true. >> >>>> >>>>> I can try to change the PathRequest of OMPI tomorrow, so that it matches >>>>> addr_type.gsi.service_level. >>>>> Maybe, with this change the packets of the SA will reach the OMPI process >>>>> on a SL>0. >>>>>> >>>>>>>> >>>>>>>>>> >>>>>>>>>>> 4. SA sends the PathRecord back to the OMPI process via umad_send >>>>>>>>>>> in libvendor/osm_vendor_ibumad.c >>>>>>>>>> >>>>>>>>>> By the response reversibility rule, I think this is returned on the >>>>>>>>>> SL >>>>>>>>>> of the original query but haven't verified this in the code base yet. >>>>>>>>> Ok, I was not aware of that rule. But if this is true, then the SA >>>>>>>>> should also be able to send via SL>0. >>>>>>>> >>>>>>>> I doubled checked and indeed the SA response does use the SL that the >>>>>>>> incoming request was received on. >>>>>>>> >>>>>>>>>> >>>>>>>>>>> The osm_vendor_send() function builds the MAD packet with the >>>>>>>>>>> following attributes: >>>>>>>>>>> /* GS classes */ >>>>>>>>>>> umad_set_addr_net(p_vw->umad, p_mad_addr->dest_lid, >>>>>>>>>>> p_mad_addr->addr_type.gsi.remote_qp, >>>>>>>>>>> p_mad_addr->addr_type.gsi.service_level, >>>>>>>>>>> IB_QP1_WELL_KNOWN_Q_KEY); >>>>>>>>>>> So, the SL is the same like the one which was used by the OMPI >>>>>>>>>>> process. The Q_Key matches the Q_key on the OMPI process, and >>>>>>>>>>> remote_qp and dest_lid is correct, too. >>>>>>>>>>> Afterwards umad_send(…) is used to send the reply with the >>>>>>>>>>> PathRecord, and this send does not work (except for SL=0). >>>>>>>>>> >>>>>>>>>> By not working, what do you mean ? Do you mean it's not received at >>>>>>>>>> the >>>>>>>>>> requester with no message in the OpenSM log or not received at the >>>>>>>>>> OpenSM or something else ? It could be due to the wrong SL being >>>>>>>>>> used in >>>>>>>>>> the original request (forcing it to SL 1). That could cause it not >>>>>>>>>> to be >>>>>>>>>> received at the SM or the response not to make it back to the >>>>>>>>>> requester >>>>>>>>>> from the SA if the SL used is not "reversible". >>>>>>>>> By "not working" I mean, that the MPI process does not receive any >>>>>>>>> response from the SA. >>>>>>>>> I get messages from the MPI process like the following: >>>>>>>>> [rc011][[14851,1],1][connect/btl_openib_connect_sl.c:301:get_pathrecord_info] >>>>>>>>> No response from SA after 20 retries >>>>>>>>> The log of OpenSM shows that the SA received the PathRequest query, >>>>>>>>> dumps the query into the log, and sends the reply back. >>>>>>>>> And I think I was some messages in the log about "…1 outstanding >>>>>>>>> MAD…". >>>>>>>>>> >>>>>>>>>>> If I look into the MAD before it is send, then it looks like this: >>>>>>>>>>> Breakpoint 2, umad_send (fd=9, agentid=2, umad=0x7fffe8012530, >>>>>>>>>>> length=120, timeout_ms=0, retries=3) >>>>>>>>>>> at src/umad.c:791 >>>>>>>>>>> 791 if (umaddebug > 1) >>>>>>>>>>> (gdb) p *mad >>>>>>>>>>> $1 = {agent_id = 2, status = 0, timeout_ms = 0, retries = 3, length >>>>>>>>>>> = 0, addr = {qpn = 1325427712, qkey = 384, >>>>>>>>>>> lid = 4096, sl = 6 '\006', path_bits = 0 '\000', grh_present = 0 >>>>>>>>>>> '\000', gid_index = 0 '\000', >>>>>>>>>>> hop_limit = 0 '\000', traffic_class = 0 '\000', gid = '\000' >>>>>>>>>>> <repeats 15 times>, flow_label = 0, >>>>>>>>>>> pkey_index = 0, reserved = "\000\000\000\000\000"}, data = >>>>>>>>>>> 0x7fffe8012530 "\002"} >>>>>>>>>> >>>>>>>>>> Is this the PathRecord query on the OpenMPI side or the response on >>>>>>>>>> the >>>>>>>>>> OpenSM side ? SL is 6 rather than 1 here. >>>>>>>>> This is the response on the OpenSM side (inside the umad_send >>>>>>>>> function, right before it is written to the device with write(fd, …). >>>>>>>>> SL=6 indicates, that the MPI process was sending the request on SL 6. >>>>>>>> >>>>>>>> What is SMSL for the requester ? Was it SL 6 ? >>>>>>> Yes, it was SL 6. >>>>>>> Here is a content of a similar packet which was received by the SA. I >>>>>>> have used ibdump on the port where the OpenSM was running: >>>>>>> ====================================================================================== >>>>>>> No. Time Source Destination >>>>>>> Protocol Length Info >>>>>>> 785 14.352168 LID: 384 LID: 4140 InfiniBand >>>>>>> 290 UD Send Only SubnAdmGet(PathRecord) >>>>>>> >>>>>>> Frame 785: 290 bytes on wire (2320 bits), 290 bytes captured (2320 bits) >>>>>>> Arrival Time: Dec 13, 2012 18:09:44.437633332 JST >>>>>>> Epoch Time: 1355389784.437633332 seconds >>>>>>> [Time delta from previous captured frame: 4.332020528 seconds] >>>>>>> [Time delta from previous displayed frame: 4.332020528 seconds] >>>>>>> [Time since reference or first frame: 14.352168681 seconds] >>>>>>> Frame Number: 785 >>>>>>> Frame Length: 290 bytes (2320 bits) >>>>>>> Capture Length: 290 bytes (2320 bits) >>>>>>> [Frame is marked: False] >>>>>>> [Frame is ignored: False] >>>>>>> [Protocols in frame: erf:infiniband] >>>>>>> Extensible Record Format >>>>>>> [ERF Header] >>>>>>> Timestamp: 0x50c99b587008bcf2 >>>>>>> [Header type] >>>>>>> .001 0101 = type: INFINIBAND (21) >>>>>>> 0... .... = Extension header present: 0 >>>>>>> 0000 0100 = flags: 4 >>>>>>> .... ..00 = capture interface: 0 >>>>>>> .... .1.. = varying record length: 1 >>>>>>> .... 0... = truncated: 0 >>>>>>> ...0 .... = rx error: 0 >>>>>>> ..0. .... = ds error: 0 >>>>>>> 00.. .... = reserved: 0 >>>>>>> record length: 306 >>>>>>> loss counter: 0 >>>>>>> wire length: 290 >>>>>>> InfiniBand >>>>>>> Local Route Header >>>>>>> 0110 .... = Virtual Lane: 0x06 >>>>>>> .... 0000 = Link Version: 0 >>>>>>> 0110 .... = Service Level: 6 >>>>>>> .... 00.. = Reserved (2 bits): 0 >>>>>>> .... ..10 = Link Next Header: 0x02 >>>>>>> Destination Local ID: 19 >>>>>>> 0000 0... .... .... = Reserved (5 bits): 0 >>>>>>> .... .000 0100 1000 = Packet Length: 72 >>>>>>> Source Local ID: 16 >>>>>>> Base Transport Header >>>>>>> Opcode: 100 >>>>>>> 1... .... = Solicited Event: True >>>>>>> .1.. .... = MigReq: True >>>>>>> ..00 .... = Pad Count: 0 >>>>>>> .... 0000 = Header Version: 0 >>>>>>> Partition Key: 65535 >>>>>>> Reserved (8 bits): 0 >>>>>>> Destination Queue Pair: 0x000001 >>>>>>> 0... .... = Acknowledge Request: False >>>>>>> .000 0000 = Reserved (7 bits): 0 >>>>>>> Packet Sequence Number: 0 >>>>>>> DETH - Datagram Extended Transport Header >>>>>>> Queue Key: 2147549184 >>>>>>> Reserved (8 bits): 0 >>>>>>> Source Queue Pair: 0x00380050 >>>>>>> MAD Header - Common Management Datagram >>>>>>> Base Version: 0x01 >>>>>>> Management Class: 0x03 >>>>>>> Class Version: 0x02 >>>>>>> Method: Get() (0x01) >>>>>>> Status: 0x0000 >>>>>>> Class Specific: 0x0000 >>>>>>> Transaction ID: 0x0010000f38005000 >>>>>>> Attribute ID: 0x0035 >>>>>>> Reserved: 0x0000 >>>>>>> Attribute Modifier: 0x00000000 >>>>>>> MAD Data Payload: >>>>>>> 000000000000000000000000000000000000000000000000... >>>>>>> Illegal RMPP Type (0)! >>>>>>> RMPP Type: 0x00 >>>>>>> RMPP Type: 0x00 >>>>>>> 0000 .... = R Resp Time: 0x00 >>>>>>> .... 0000 = RMPP Flags: Unknown (0x00) >>>>>>> RMPP Status: (Normal) (0x00) >>>>>>> RMPP Data 1: 0x00000000 >>>>>>> RMPP Data 2: 0x00000000 >>>>>>> SMASubnAdmGet(PathRecord) >>>>>>> SM_Key (Verification Key): 0x0000000000000000 >>>>>>> Attribute Offset: 0x0000 >>>>>>> Reserved: 0x0000 >>>>>>> Component Mask: 0x0000003000000000 >>>>>>> Attribute (PathRecord) >>>>>>> PathRecord >>>>>>> DGID: :: (::) >>>>>>> SGID: ::0.15.0.16 (::0.15.0.16) >>>>>>> DLID: 0x0000 >>>>>>> SLID: 0x0000 >>>>>>> 0... .... = RawTraffic: 0x00 >>>>>>> .... 0000 0000 0000 0000 0000 = FlowLabel: 0x000000 >>>>>>> HopLimit: 0x00 >>>>>>> TClass: 0x00 >>>>>>> 0... .... = Reversible: 0x00 >>>>>>> .000 0000 = NumbPath: 0x00 >>>>>>> P_Key: 0x0000 >>>>>>> .... .... .... 0000 = SL: 0x0000 >>>>>>> 00.. .... = MTUSelector: 0x00 >>>>>>> ..00 0000 = MTU: 0x00 >>>>>>> 00.. .... = RateSelector: 0x00 >>>>>>> ..00 0000 = Rate: 0x00 >>>>>>> 00.. .... = PacketLifeTimeSelector: 0x00 >>>>>>> ..00 0000 = PacketLifeTime: 0x00 >>>>>>> Preference: 0x00 >>>>>>> Variant CRC: 0xad4e >>>>>>> ====================================================================================== >>>>>> >>>>>> And the SubnAdmGetResp(PathRecord) is not seen ? If not, it doesn't get >>>>>> out that machine and the issue is internal to that machine. It could be >>>>>> because of the underlying issue which hangs OpenSM when some IB program >>>>>> tried to unregister from the MAD layer but there were outstanding work >>>>>> completions. That's based on your original email earlier this AM. >>>>> No, the SubnAdmGetResp does not show up, if I use ibdump on the OMPI side >>>>> and the SA uses a SL>0. >>>> >>>> Can ibdump be used to capture output on the SM port ? >>> >>> Yes, that works quite well, despite the warning in the ibdump manual. >>> But I have started ibdump before opensm, maybe that makes a difference, not >>> sure. >>> >>> Regards, >>> Jens >>> >>> PS: I have seen a small bug. Not sure if its a bug in wireshark or ibdump, >>> but the response received by the OMPI node isn't shown correctly. The >>> PathRecord contains an offset which is either missing in the dump or is not >>> treated correctly be wireshark. But it causes wireshark to show the >>> PathRecord data with wrong values. >>> Maybe you could redirect this to the developer of ibdump, so that he can >>> check/fix it. >> >> Are you referring to the fields after the SA AttributeOffset or >> something else ? > Yes, after the SMASubnAdmGet Attribute Offset. Here an example: > I get on the OMPI side: > SMASubnAdmGetResp(PathRecord) > SM_Key (Verification Key): 0x0000000000000000 > Attribute Offset: 0x0008 > Reserved: 0x0000 > Component Mask: 0x0000803000000000 > Attribute (PathRecord) > PathRecord > DGID: ::8:f104:399:ebb5:fe80:0 (::8:f104:399:ebb5:fe80:0) > SGID: ::8:f104:399:ecd5:4:8 (::8:f104:399:ecd5:4:8) > DLID: 0x0000 > SLID: 0x0000 > 0... .... = RawTraffic: 0x00 > .... 0000 1000 0000 1111 1111 = FlowLabel: 0x0080ff > HopLimit: 0xff > TClass: 0x00 > 0... .... = Reversible: 0x00 > .000 0011 = NumbPath: 0x03 > P_Key: 0x8486 > .... .... .... 0000 = SL: 0x0000 > 00.. .... = MTUSelector: 0x00 > ..00 0000 = MTU: 0x00 > 00.. .... = RateSelector: 0x00 > ..00 0000 = Rate: 0x00 > 00.. .... = PacketLifeTimeSelector: 0x00 > ..00 0000 = PacketLifeTime: 0x00 > Preference: 0x00 > > But it should show (see the difference in SLID, DLID, SL which are now > correct): > SMASubnAdmGetResp(PathRecord) > SM_Key (Verification Key): 0x0000000000000000 > Attribute Offset: 0x0008 > Reserved: 0x0000 > Component Mask: 0x0000803000000000 > Attribute (PathRecord) > PathRecord > DGID: ::8:f104:399:ebb5 (::8:f104:399:ebb5) > SGID: fe80::8:f104:399:ecd5 (fe80::8:f104:399:ecd5) > DLID: 0x0004 > SLID: 0x0008 > 0... .... = RawTraffic: 0x00 > .... 0000 0000 0000 0000 0000 = FlowLabel: 0x000000 > HopLimit: 0x00 > TClass: 0x00 > 1... .... = Reversible: 0x01 > .000 0000 = NumbPath: 0x00 > P_Key: 0xffff > .... .... .... 0011 = SL: 0x0003 > 10.. .... = MTUSelector: 0x02 > ..00 0100 = MTU: 0x04 > 10.. .... = RateSelector: 0x02 > ..00 0110 = Rate: 0x06 > 10.. .... = PacketLifeTimeSelector: 0x02 > ..01 0010 = PacketLifeTime: 0x12 > Preference: 0x00
I think everything after AttributeOffset is off by 2 bytes. DGID doesn't look right to me (no subnet prefix fe80:: in front of GUID). -- Hal > > Regards, > Jens > >> >> -- Hal >> >>>> >>>> -- Hal >>>> >>>>>> >>>>>>>> >>>>>>>> One would need to walk the SLToVLMappingTables from requester (OMPI >>>>>>>> port) to SA and back to see whether SL6 would even have a chance of >>>>>>>> working (not dropping) aside from whether it's really the correct SL >>>>>>>> to use. >>>>>>> All SL2VL tables look the same. I checked the output of OpenSM. >>>>>>> SL: | 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | >>>>>>> 11 | 12 | 13 | 14 | 15 | >>>>>>> VL: | 0x0 |0x1 |0x2 |0x3 |0x4 |0x5 |0x6 |0x7 |0x0 |0x1 |0x2 >>>>>>> |0x3 |0x4 |0x5 |0x6 |0x7 | >>>>>>> But this is also as expected, because I have set the QoS in the opensm >>>>>>> config as follows: >>>>>>> qos_sl2vl 0,1,2,3,4,5,6,7,0,1,2,3,4,5,6,7 >>>>>>> This was set for "default", "CA" and "Switch external ports". I have >>>>>>> not touched the config for "Switch Port 0" and "Router ports", they >>>>>>> remained: qos_[sw0 | rtr]_sl2vl (null) >>>>>> >>>>>> That works as long as all links have (at least) 8 data VLs (VLCap 4). >>>>> Yes, all VL_CAP show 4 in the OpenSM log file. >>>>> >>>>> Regards >>>>> Jens >>>>> >>>>> >>>>> >>>>>> >>>>>> -- Hal >>>>>> >>>>>>> Regards >>>>>>> Jens >>>>>>> >>>>>>>> >>>>>>>> -- Hal >>>>>>>> >>>>>>>>>> >>>>>>>>>>> The output of OpenMPI or OpenSM's log file don't show any useful >>>>>>>>>>> information for this problem, even with higher debug levels. >>>>>>>>>> >>>>>>>>>> So nothing interesting logged relative to the PathRecord queries ? >>>>>>>>> In the OpenSM log, only that it was received, how the request looks >>>>>>>>> like, and that it was send back. >>>>>>>>> And a few "outstanding MADs" a few lines later in the log. >>>>>>>>>> >>>>>>>>>>> So, right now I'm stuck, and have no idea if there is an error in >>>>>>>>>>> the kernel driver, the HCA firmware or something completely >>>>>>>>>>> different. Or if umad_send basically does not support SL>0. >>>>>>>>>>> A workaround for the moment is to set the SL in the >>>>>>>>>>> umad_set_addr_net(...) call to 0. >>>>>>>>>> >>>>>>>>>> So SL 0 works between all nodes and SA for querying/responses. >>>>>>>>>> Wonder if >>>>>>>>>> that's how SMSL is set by DFSSSP. >>>>>>>>> No, the SMSL set by DFSSSP is different from 0, I have checked this. >>>>>>>>> In our case (OpenSM running on a compute node), it sets the same SL, >>>>>>>>> which is used >>>>>>>> for MPI<->MPI traffic, to ensure deadlock freedom. >>>>>>>>> >>>>>>>>> Regards >>>>>>>>> Jens >>>>>>>>> >>>>>>>>> -------------------------------- >>>>>>>>> Dipl.-Math. Jens Domke >>>>>>>>> Researcher - Tokyo Institute of Technology >>>>>>>>> Satoshi MATSUOKA Laboratory >>>>>>>>> Global Scientific Information and Computing Center >>>>>>>>> 2-12-1-E2-7 Ookayama, Meguro-ku, >>>>>>>>> Tokyo, 152-8550, JAPAN >>>>>>>>> Tel/Fax: +81-3-5734-3876 >>>>>>>>> E-Mail: domke.j...@m.titech.ac.jp >>>>>>>>> -------------------------------- >>>>>>>>> >>>>>>>>> >>>>>>>> >>>>>>>> -- >>>>>>>> To unsubscribe from this list: send the line "unsubscribe linux-rdma" >>>>>>>> in >>>>>>>> the body of a message to majord...@vger.kernel.org >>>>>>>> More majordomo info at http://vger.kernel.org/majordomo-info.html >>>>>>> >>>>>>> -------------------------------- >>>>>>> Dipl.-Math. Jens Domke >>>>>>> Researcher - Tokyo Institute of Technology >>>>>>> Satoshi MATSUOKA Laboratory >>>>>>> Global Scientific Information and Computing Center >>>>>>> 2-12-1-E2-7 Ookayama, Meguro-ku, >>>>>>> Tokyo, 152-8550, JAPAN >>>>>>> Tel/Fax: +81-3-5734-3876 >>>>>>> E-Mail: domke.j...@m.titech.ac.jp >>>>>>> -------------------------------- >>>>>>> >>>>>>> >>>>>> >>>>>> -- >>>>>> To unsubscribe from this list: send the line "unsubscribe linux-rdma" in >>>>>> the body of a message to majord...@vger.kernel.org >>>>>> More majordomo info at http://vger.kernel.org/majordomo-info.html >>>>> >>>>> -------------------------------- >>>>> Dipl.-Math. Jens Domke >>>>> Researcher - Tokyo Institute of Technology >>>>> Satoshi MATSUOKA Laboratory >>>>> Global Scientific Information and Computing Center >>>>> 2-12-1-E2-7 Ookayama, Meguro-ku, >>>>> Tokyo, 152-8550, JAPAN >>>>> Tel/Fax: +81-3-5734-3876 >>>>> E-Mail: domke.j...@m.titech.ac.jp >>>>> -------------------------------- >>>>> >>>>> >>>> >>>> -- >>>> To unsubscribe from this list: send the line "unsubscribe linux-rdma" in >>>> the body of a message to majord...@vger.kernel.org >>>> More majordomo info at http://vger.kernel.org/majordomo-info.html >>> >>> -------------------------------- >>> Dipl.-Math. Jens Domke >>> Researcher - Tokyo Institute of Technology >>> Satoshi MATSUOKA Laboratory >>> Global Scientific Information and Computing Center >>> 2-12-1-E2-7 Ookayama, Meguro-ku, >>> Tokyo, 152-8550, JAPAN >>> Tel/Fax: +81-3-5734-3876 >>> E-Mail: domke.j...@m.titech.ac.jp >>> -------------------------------- >>> >>> >> >> -- >> To unsubscribe from this list: send the line "unsubscribe linux-rdma" in >> the body of a message to majord...@vger.kernel.org >> More majordomo info at http://vger.kernel.org/majordomo-info.html > > > > -- To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html