Re: device attr cleanup (was: Handle mlx4 max_sge_rd correctly)

[Doug Ledford <dledford-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>]

[-- Attachment #1: Type: text/plain, Size: 1871 bytes --]

On 12/08/2015 04:00 PM, Or Gerlitz wrote:
> On Tue, Dec 8, 2015 at 8:17 PM, Doug Ledford <dledford-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org> wrote:
>>
>> On 12/08/2015 01:13 PM, Sagi Grimberg wrote:
> 
> 
>>
>>> I mentioned this in v1. This patch set is applied over Christoph's
>>> device attributes patch. Will it go in as well?
>>
>> No, that's too big and not the right sort of fix for 4.4-rc.  I had it
>> on my radar for getting into for-next.
> 
> 
> 
> Doug, re the device attribute patch, I have expressed my opinion that
> --- we should be going in a slightly different direction of stashing a
> struct ib_device_attr as a direct or pointer field of struct ib_device
> and have the device or the core to fill it up just before the device
> instance creation with the core is to be complete. This way we can
> remove all the annoying calls from ULPs to query device and avoid
> adding so many fields to the device structure itself.
> 
> So there are two suggestions on the table here and we need to hear the
> maintainer thinking. I do expect that your response will not be "I
> applied X" but rather you'll allow the time to try and convince you on
> whatever direction, this is pending for long time and we're on
> 4.4-rc4, there should be no rush to cut this over night and push to
> k.o the code for this small cleanup.
> 
> I didn't submit a patch that follows my suggestion, but Ira did so in
> one of the OPA submission rounds, if you're OK with that proposal or
> if you even just want to see the patch to cast your vote/decision, I
> will rebase it and complete.
> 
> Makes sense?

Makes sense.  Show me what you are talking about (either a link to Ira's
patch you are referring to or your own patch).


-- 
Doug Ledford <dledford-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
              GPG KeyID: 0E572FDD



[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 884 bytes --]