AMD GPU new API for new module

[Oded Gabbay]

On 09/10/14 11:02, Jerome Glisse wrote:
> On Thu, Oct 09, 2014 at 09:54:14AM +0300, Oded Gabbay wrote:
>> Hi Jerome,
>>
>> Do you think your proposed change should also be applied to amdkfd's
>> IOCTLs ?
> 
> It might make sense it really depends on the lifespan you expect for
> amdkfd, do you expect that you will need substential API evolution
> during that timespan or do you foresee that the currently propose API
> is good enough.

Well, I don't expect to reach 100 ioctls anytime soon, but I can tell
you that for the features we have in the pipeline, I can see the IOCTL
number go up to 20-30, just for the current H/W generation.
> 
> The thing i am proposing here is really to address the issue of IOCTL
> needing fixes because new generation of GPU requires new informations
> or do not care about old and now meaningless informations. The radeon
> ioctl sets is a testimony to the fact that we can not predict what kind
> of broken API we expose and have to leave with.
Understood, and its really hard for me to predict now if we would
absolutely need fixing IOCTLS for newer hardware generations. However,
better be safe than sorry.
> 
> My proposal is really about allowing to evolve each ioctl in what i
> believe to be the simplest and cleanest way to do it while still
> preserving backward compatibility ie i expect that old version of
> an ioctl will continue to live in there own functions untouch and
> unlove but still working. While new development can move to updated
> ioctl.
Yeah, and I think this is a very good forward thinking and I would like
to adopt it as well. This is the perfect time to do it as we still don't
have any history to maintain. I will run it by our internal team and
hopefully publish a new set with this fix for you to review :)
> 
> Other solution would be to just keep adding new ioctl but i fear we
> might run out of ioctl number.

> 
> Cheers,
> Jérôme
> 
>>
>> 	Oded
>>
>> On 08/10/14 19:00, Jerome Glisse wrote:
>>> Hi,
>>>
>>> So if i do not start the discussion now it might be already too late. Given
>>> plan to converge open source driver and closed source driver to use a single
>>> common kernel driver and that this would be a new kernel driver. This is an
>>> opportunity to fix some of the radeon design issues (at least things that i
>>> would have done differently if only i could get some gas for my DeLorean).
>>>
>>> Among the thing that i will not do is the chunk stuff associated with cs
>>> ioctl. I find it ugly, if my memory serve me well i was trying to be future
>>> proof and allow the cs ioctl to be extended. While this original aim have
>>> been somewhat successfully, i think it was the wrong day to do it.
>>>
>>> My lastest (and what i still believe to be a good idea until proven wrong),
>>> is to change the way we do ioctl and use a little trick. This idea was also
>>> spark by the continuous additions we do to info ioctl which is getting ugly.
>>>
>>> So idea is simple, each ioctl would use some struct like :
>>>
>>> struct radeon_ioctl {
>>> 	u32	version;
>>> 	u32	size;
>>> };
>>>
>>> The version field is the key here, think of it as an index into an array of
>>> ioctl dispatch functions. So something like :
>>>
>>> struct radeon_ioctls {
>>> 	int (*iotcl)[MAX_IOCTL_NUM](void *data, ...);
>>> };
>>>
>>> struct radeon_ioctls rdispatch_ioctls[N];
>>>
>>> And now all ioctl go through this single entry point :
>>>
>>> int radeon_ioctl_stub(int ioctl, void *data, ...)
>>> {
>>> 	struct radeon_ioctl *rio = data;
>>>
>>> 	return rdispatch_ioctls[rio->version][ioctl](data, ...);
>>> }
>>>
>>> So this is rough idea but the point is that we can do proper ioctl versioning
>>> and have separate functions for each new versions and avoid ioctl cruft, or
>>> at least this is the theory.
>>>
>>> The two things this gave us, is feedback from userspace as we the version the
>>> kernel will know which version of userspace it is dealing with. The others one
>>> is that it does allow you to introduce a completely new API either for new
>>> generation of hardware or just as an evolution. And small bonus is that it
>>> does allow to slowly phase out API that we consider broken (ioctl per ioctl).
>>>
>>> So this is the main design change that i would do. I should probably rought
>>> up something that goes deeper for the cs ioctl.
>>>
>>> Cheers,
>>> Jérôme
>>> _______________________________________________
>>> dri-devel mailing list
>>> dri-devel at lists.freedesktop.org
>>> http://lists.freedesktop.org/mailman/listinfo/dri-devel
>>>