Re: [PATCH] opensm: Fix sl2vl configuration

2010-08-25 Thread Eli Dorfman (Voltaire)
Hal Rosenstock wrote: Hi Eli, On Mon, Aug 23, 2010 at 3:37 PM, Eli Dorfman dorfman@gmail.com wrote: Hi Hal, On Mon, Aug 23, 2010 at 5:25 PM, Hal Rosenstock hal.rosenst...@gmail.com wrote: On Wed, Jul 28, 2010 at 12:26 PM, Eli Dorfman (Voltaire) dorfman@gmail.com wrote: Subject:

srp target - write to raw block device like /dev/sda ist fragmented into very small requests

2010-08-25 Thread Billich Heiner
Hello, we employ Infiniband SRP storage. For basic tests I just did write/read with dd to the raw block device like /dev/sda. I tell dd to use 1M sized blocks, but I see very small few KB sized requests only. To make this tests of real value I would need to read and write with large 1MB sized

Re: IPoIB CM connection establishment protocol question

2010-08-25 Thread Eli Cohen
Hi stan, Here's a short summary: An IPoIB interface encodes it capability of supporting CM mode in the hardware adress it publishes. Once the stack hands an SKB to IPoIB for transmission (unicast), IPoIB needs to resolve the hardware address to IB adress information (LID, SL etc.). Once

RE: IPoIB: Broken IGMP processing

2010-08-25 Thread Christoph Lameter
On Mon, 23 Aug 2010, Yossi Etigin wrote: Simplest then is to check if byte 24 of the packet is 0xff. (ie IN6_IS_ADDR_MULTICAST) No need to worry about if it is properly formed or anything, if it is a multicast DGID then it is a multicast packet at the link level. Sounds good to me

Re: [PATCH] complib/cl_timer.c: fixing cl_timer calculation

2010-08-25 Thread Sasha Khapyorsky
On 15:01 Tue 24 Aug , Yevgeny Kliteynik wrote: BTW, any idea what are the (obviously historical) reasons for these lines in the code? 323: /* do not do 0 wait ! */ 324: /* if (delta_time 1000.0) {delta_time = 1000;} */ Have no idea. This is from day zero of git history. Sasha --

Re: [PATCH v2] complib/cl_timer.c: fixing cl_timer calculation

2010-08-25 Thread Sasha Khapyorsky
On 16:21 Tue 24 Aug , Yevgeny Kliteynik wrote: When calculating p_timer-timeout.tv_sec and p_timer-timeout.tv_nsec, the carry was ignored, resulting in wrong value in p_timer-timeout.tv_sec, and value 10^9 in p_timer-timeout.tv_nsec (illegal value). Signed-off-by: Yevgeny Kliteynik

Re: opensm/osm_opensm.c: no report when SM is exiting

2010-08-25 Thread Sasha Khapyorsky
On 16:38 Wed 18 Aug , Yevgeny Kliteynik wrote: Don't bother reporting events to plug-ins while SM is exiting. Signed-off-by: Yevgeny Kliteynik klit...@dev.mellanox.co.il Applied. Thanks. Sasha -- To unsubscribe from this list: send the line unsubscribe linux-rdma in the body of a message

Re: [PATCH 2/5 v6] IB/srp: IB/srp: Implement SRP_CRED_REQ and SRP_AER_REQ

2010-08-25 Thread Bart Van Assche
On Wed, Aug 25, 2010 at 4:48 AM, David Dillow d...@thedillows.org wrote: On Sun, 2010-08-22 at 18:56 +0200, Bart Van Assche wrote: [ ... ]  static struct srp_iu *__srp_get_tx_iu(struct srp_target_port *target, -                                     enum srp_request_type req_type) +        

Re: [PATCH] ibnetdiscover: add '-f' flag to show full information (ports' speed and width).

2010-08-25 Thread Sasha Khapyorsky
On 09:45 Tue 24 Aug , Hal Rosenstock wrote: What about the flag? do we still need it if we pass the output after the comment? I wouldn't think so. I also think we've made commentary changes to the ibnetdiscover output format like this before. If we wanted to be absolutely sure it

Re: [ofw] [Patch] [Tools][infiniband-diags]

2010-08-25 Thread Sasha Khapyorsky
On 20:19 Sun 01 Aug , Sasha Khapyorsky wrote: On 14:36 Wed 28 Jul , Irena Kruchkovsky wrote: A patch that fixes the GUID output in ibstat to work correctly in windows 2003. Index: D:/Windows/MLNX_VPI/tools/infiniband-diags/src/ibstat.c

Re: [PATCH] opensm: Fix sl2vl configuration

2010-08-25 Thread Hal Rosenstock
On Wed, Aug 25, 2010 at 4:43 AM, Eli Dorfman (Voltaire) dorfman@gmail.com wrote: Hal Rosenstock wrote: Hi Eli, On Mon, Aug 23, 2010 at 3:37 PM, Eli Dorfman dorfman@gmail.com wrote: Hi Hal, On Mon, Aug 23, 2010 at 5:25 PM, Hal Rosenstock hal.rosenst...@gmail.com wrote: On Wed, Jul

ibnetdiscover issue with multiported CA (or router) with multiple ports on same subnet

2010-08-25 Thread Hal Rosenstock
Sasha, I'm seeing an issue with ibnetdiscover from a CA port where it appears to extend a path at a remote CA port (it's actually another port on the same CA) to query NodeInfo of the next hop beyond it. I get the following error message: src/query_smp.c:188; umad (DR path slid 0; dlid 0;

[PATCH] infiniband-diags/ibnetdiscover: Fix handling of CA ports in recv_port_info

2010-08-25 Thread Hal Rosenstock
When multiple ports on the same CA are connected to the same subnet and an ibnetdiscover is initiated from one of them, the discovery continues past the other CA port and an error occurs. The error is: src/query_smp.c:188; umad (DR path slid 0; dlid 0; 0,1,20,2 Attr0x11:0) bad status 110;

Re: [PATCH] complib/cl_timer.c: fixing cl_timer calculation

2010-08-25 Thread Yevgeny Kliteynik
On 25-Aug-10 7:07 PM, Sasha Khapyorsky wrote: On 15:01 Tue 24 Aug , Yevgeny Kliteynik wrote: BTW, any idea what are the (obviously historical) reasons for these lines in the code? 323: /* do not do 0 wait ! */ 324: /* if (delta_time 1000.0) {delta_time = 1000;} */ Have no idea. This

RE: IPoIB CM connection establishment protocol question

2010-08-25 Thread Smith, Stan
Eli Cohen wrote: Hi stan, Here's a short summary: An IPoIB interface encodes it capability of supporting CM mode in the hardware address it publishes. Once the stack hands an SKB to IPoIB for transmission (unicast), IPoIB needs to resolve the hardware address to IB address information (LID,