[openib-general] Re: when executing sminfo with a port in down state, there is a retur n value 0

[Hal Rosenstock]

Hi Dotan,

On Tue, 2005-08-23 at 02:33, Dotan Barak wrote:
> I'm working with gen2 svn rev. 3155 with 2 Mellanox HCAs (23108) (1 on
> each host; they are connected b2b: port 1 to port 1).
> 
> I executed opensm on host 1, port 1.
> When i executed sminfo on host 2 port 1 everything was as expected
> (return value = 0).
> 
> I killed the opensm
> When i executed sminfo on host 2 port 1 everything was as expected
> (return value = 255).
> 
> When i executed sminfo on host 2 port 2 everything i got 0 (i expected
> to get return value = 255).
> 
> Port 2 in host 2 was down, so i don't know why i got the return value
> 0.

I just tried this and got 255. Can you try this again ?

> here is the output:
> 
> host2:~ # /usr/local/bin/sminfo -C mthca0 -P 2
> sminfo: sm lid 0x0 sm guid 0x8200000000, activity count 0 priority 0
> state SMINFO_NOTACT 0
> host2:~ # echo $?
> 0
> host2:~ # /usr/local/bin/sminfo -C mthca0 -P 2
> sminfo: sm lid 0x0 sm guid 0x0, activity count 0 priority 0 state
> SMINFO_STANDBY 2
> host2:~ # echo $?
> 0

It looks like 0 is set because there is some SMInfo response but I don't
understand how that would be the case. sminfo can use either DR or LR.
This form (without -D) uses LR so that shouldn't work if the port is
down. In either case, the other end wouldn't respond if there is no SM
there. Also, the GUID looks suspicious.

> host2:~ #  vstat
> hca_id: mthca0
>         phys_port_cnt:                  2
>                 port:   1
>                         state:                  PORT_ACTIVE (4)
>                         max_mtu:                0 (0)
>                         active_mtu:             0 (0)
>                         sm_lid:                 1
>                         port_lid:               2
>                         port_lmc:               0x00
> 
>                 port:   2
>                         state:                  PORT_DOWN (1)
>                         max_mtu:                0 (0)
>                         active_mtu:             0 (0)
>                         sm_lid:                 0
>                         port_lid:               0
>                         port_lmc:               0x00

How about ibstat or ibstatus ?

> can you please help me with this issue?

If you can reproduce this, not sure what is different about your setup.
Is port 2 on host 2 cabled to anything ?

-- Hal