On Apr 28, 2010 04:55 PM, Amir Vadai <am...@mellanox.co.il> wrote: > comment it out in sdp_cma.c at: > sdp_remove_device() and sdp_add_device() > > If I am right, the problem that you see, is that ib_create_fmr_pool() > is > failing (you should see an error in dmesg), and because of that > ib_set_client_data() is not called. And this is the reason for the > error > that you saw in the logs. > > Because you won't be using ZCopy, no need to call ib_create_fmr_pool() > - > ofcourse, make sure you comment out the calls to ib_destroy_fmr_pool > and > ib_flush_fmr_pool() > > - Amir
Hi Amir, I put in off status the ZCopy with command: # echo 0 > /sys/module/ib_sdp/parameters/sdp_zcopy_thresh I comment out some lines in file sdp_main.c, but now the programs don't compile with a lot of errors. I list out the lines that I have commented out: file sdp_main.c line 2625: struct ib_fmr_pool_param fmr_param lines from 2645 to 2661 lines 2665 2666 lines from 2739 to 2745 All these lines contain a word "fmr". I don't find word "fmr" in file sdp_cma.c; sdp_main.c calls objects of sdp_cma.c SDP doesn't work. The output message of dmesg command is: sdp_init_qp:105 sdp_sock SDP not available on device nes0. I would like identify the problem: if we are almost sure that Fast Memory Region FMR is responsible for that, the bug 2028 solution in July will remove the error and I hope that SDP will work fine. On the other side, if FMR is not involved, there is an unknown situation and I am not happy. Could you evaluate the written error? Could you give me your opinion? Amir and Steve, thank you very much for help. Regards, Andrea Andrea Gozzelino INFN - Laboratori Nazionali di Legnaro (LNL) Viale dell'Universita' 2 I-35020 - Legnaro (PD)- ITALIA Tel: +39 049 8068346 Fax: +39 049 641925 Mail: andrea.gozzel...@lnl.infn.it -- 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