Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1637795 > unrolled thread

ION changes in staging for 4.12

Back to article view | Back to linux.kernel

Contents

  ION changes in staging for 4.12 John Stultz <john.stultz@linaro.org> - 2017-05-09 02:20 +0200
    Re: ION changes in staging for 4.12 Laura Abbott <labbott@redhat.com> - 2017-05-09 17:10 +0200

#1637795 — ION changes in staging for 4.12

Hey Laura,
   Wanted to check in on ION changes that landed in the staging merge
for 4.12...

So with these changes, I can no longer get HiKey to boot to graphics.
We had previously been using the dummy ion driver, and I've added the
new config entries for the various standard heaps, but it seems like
things still aren't working.

Was curious if this is an expected hard ABI break, or if I'm missing
something, and if so, maybe how the resulting patches were tested?  Is
there any sort of transition document to help folks figure out what
changes they might need to make to support the new ABI?

thanks
-john

[toc] | [next] | [standalone]

#1638181

On 05/08/2017 05:18 PM, John Stultz wrote:
> Hey Laura,
>    Wanted to check in on ION changes that landed in the staging merge
> for 4.12...
> 
> So with these changes, I can no longer get HiKey to boot to graphics.
> We had previously been using the dummy ion driver, and I've added the
> new config entries for the various standard heaps, but it seems like
> things still aren't working.
> 
> Was curious if this is an expected hard ABI break, or if I'm missing
> something, and if so, maybe how the resulting patches were tested?  Is
> there any sort of transition document to help folks figure out what
> changes they might need to make to support the new ABI?
> 
> thanks
> -john
> 

Yes, this was intentionally decided to be ABI breaking. It turned out
to be too much of a pain to be backwards compatible and I'm not too
surprised things have broken (sorry...). What I suspect is happening
is that the heap ids no longer match so userspace needs to call the
query ioctl to get the heap information. This is also lacking in
documentation at the moment (sorry again) but should be coming soon.

Thanks,
Laura

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web