[openib-general] Re: [PATCHv3][RFC] kDAPL: use cm timers instead of own

[James Lentini]

On Thu, 2 Jun 2005, Tom Duffy wrote:

> On Thu, 2005-06-02 at 16:18 -0400, James Lentini wrote:
>> I'd recommend these changes (see attached):
>>
>> - use the new microsecond conversion function you sent
>
> great
>
>> - keep a default DAT timeout constant (I propose DAT_TIMEOUT_MAX)
>
> fair enough.
>
>> - set the timeout a little differently. Instead of storing it in the
>>    ep, store it directly in the CM param
>
> that is better
>
>> - add a kdat.h file to the patches directory. We can start using this
>>    to place symbols that are in the official DAT spec (like
>>    DAT_TIMEOUT_INFINITE), but not our linux variant.
>
> alright, but this could get ugly.

I expect that it will. This kdat.h file will not be something we would 
submit for inclusion in the Linux kernel. It's the best way I've come 
up with to support applications written against the official DAT 
headers.

>> Did I call dapl_convert_us_to_kookyib() correctly? You were passing a
>> value divided by 1000 to dapl_convert_ms_to_kookyib(), so I assumed
>> that I could get rid of that.
>
> Yup.
>
>> What do you think?
>
> Looks good, James.  I didn't test it, but it looks good.  I'd say go for
> it.

I tested it with dapltest and it worked. (i.e. it doesn't break things 
that used to work). I'll go ahead and commit it. Thanks for working on 
this Tom.

>
> -tduffy
>