AMD Publishes Initial Open-Source Driver Code For Next-Gen Polaris

[darkbasic]

From agd5f's post that andrei_me linked above:

Are you seriously saying that DAL will be out of staging in 4.7? I didn't follow the review process but considering the huge amount of work needed since I saw the patches for the first time that would be an extraordinary result.

[pal666]

Are you seriously saying that DAL will be out of staging in 4.7?

staging drivers work just as good as non-staging

[rabcor]

Wow, this is AMD living up to all expectation, thats it, I'm buying polaris on release. I'm only gonna buy a upper-mid-end~ish (budget) card though, I'm gonna wait for vega and HBM2 before going for a high tier card, I just want something that's gonna noticably outclass my gtx 670.

Do someone really think DAL is going to be mainlined soon? Linux coding standards are quite a bit higher than the Catalyst code base, to have even the slightiest chance of being mainlined it requires massive amounts of work, and this will probably be enough just to enter staging...

I see this as an opportunity for AMD to update their standards for coding. They sure could use that sort of a push as both their linux and windows drivers have shown in the past for gpus, Nvidia always a step ahead of them, especially shortly after new cards are released. Nvidia's cards seem to usually perform better just after release than AMD's cards do.

Then again, AMD might not have quite predicted the changes it would bring to try to maintain a kernel side driver in linux, that they can't just write patches and update everything all in one go. Then again, I wonder about linux coding standards being high, honestly i've been pretty unimpressed by it over the past few years not maybe the code itself, just everything's performance post compilation, from the kernel itself to the guis and applications i've been running, I've rarely found something that just works and I can't find anything to complain about (to be fair the same applies to windows too though)

[boffo]

Hay, Polaris support + they are fixing DAL

[rabcor]

Would be nice to know how AMD is taking it that their dal patch was rejected from mainline, bitter? or completely willing to adapt to the environment?

[agd5f]

Would be nice to know how AMD is taking it that their dal patch was rejected from mainline, bitter? or completely willing to adapt to the environment?

I'm not sure what you are getting at. This is how open source works. You post code for review, you fix it, you repost, repeat until it's ready. The DAL team is new to working upstream so we expected there would be a good bit of commentary on the patches.

[bridgman]

See post #9... maybe half way through the changes already.

https://www.phoronix.com/forums/foru...387#post860387

Note the pushback was more about code in DAL that duplicated existing framework functionality (some that appeared in parallel with writing DAL) than about coding standards.

[riklaunim]

So as we have a AMD guy here... if you combine like 4 older 16-core Opterons (so 64 core PC ) with Radeon Pro duo and run some DX12/Vulkan/Mantle or Ashes of Singularity benchmarks - would that use all of 64 cores ?

[bridgman]

Not fully

[riklaunim]

Some Gameworks and Physx should fill in the gap... and then nuclear fusion here we come (it won't be cold fusion though... definitely not cold).