Re: [PATCH 0/3] drm: panfrost: Coherency support

2020-09-17 Thread Robin Murphy
On 2020-09-16 18:46, Rob Herring wrote: On Wed, Sep 16, 2020 at 11:04 AM Alyssa Rosenzweig wrote: So I get a performance regression with the dma-coherent approach, even if it's clearly the cleaner. That's bizarre -- this should really be the faster of the two. Coherency may not be free.

Re: [PATCH 0/3] drm: panfrost: Coherency support

2020-09-17 Thread Alyssa Rosenzweig
> The DDK blob has the ability to mark only certain areas of memory as > coherent for performance reasons. For simple things like kmscube I would > expect that it's basically write-only from the CPU and almost all memory the > GPU touches isn't touched by the CPU. I.e. coherency isn't helping and

Re: [PATCH 0/3] drm: panfrost: Coherency support

2020-09-17 Thread Steven Price
On 17/09/2020 11:51, Tomeu Vizoso wrote: On 9/17/20 12:38 PM, Steven Price wrote: On 16/09/2020 18:46, Rob Herring wrote: On Wed, Sep 16, 2020 at 11:04 AM Alyssa Rosenzweig wrote: So I get a performance regression with the dma-coherent approach, even if it's clearly the cleaner. That's

Re: [PATCH 0/3] drm: panfrost: Coherency support

2020-09-17 Thread Tomeu Vizoso
On 9/17/20 12:38 PM, Steven Price wrote: On 16/09/2020 18:46, Rob Herring wrote: On Wed, Sep 16, 2020 at 11:04 AM Alyssa Rosenzweig wrote: So I get a performance regression with the dma-coherent approach, even if it's clearly the cleaner. That's bizarre -- this should really be the

Re: [PATCH 0/3] drm: panfrost: Coherency support

2020-09-17 Thread Steven Price
On 16/09/2020 18:46, Rob Herring wrote: On Wed, Sep 16, 2020 at 11:04 AM Alyssa Rosenzweig wrote: So I get a performance regression with the dma-coherent approach, even if it's clearly the cleaner. That's bizarre -- this should really be the faster of the two. Coherency may not be free.

Re: [PATCH 0/3] drm: panfrost: Coherency support

2020-09-16 Thread Alyssa Rosenzweig
> So I get a performance regression with the dma-coherent approach, even if it's > clearly the cleaner. That's bizarre -- this should really be the faster of the two. signature.asc Description: PGP signature ___ iommu mailing list

Re: [PATCH 0/3] drm: panfrost: Coherency support

2020-09-16 Thread Rob Herring
On Wed, Sep 16, 2020 at 11:04 AM Alyssa Rosenzweig wrote: > > > So I get a performance regression with the dma-coherent approach, even if > > it's > > clearly the cleaner. > > That's bizarre -- this should really be the faster of the two. Coherency may not be free. CortexA9 had something like

Re: [PATCH 0/3] drm: panfrost: Coherency support

2020-09-16 Thread Neil Armstrong
Hi Robin, On 16/09/2020 01:51, Robin Murphy wrote: > Hi all, > > I polished up my original proof-of-concept a little while back, but now > that I've got my hands on my Juno again I've been able to actually test > it to my satisfaction, so here are proper patches! I tested on the Kkadas VIM3,

[PATCH 0/3] drm: panfrost: Coherency support

2020-09-15 Thread Robin Murphy
Hi all, I polished up my original proof-of-concept a little while back, but now that I've got my hands on my Juno again I've been able to actually test it to my satisfaction, so here are proper patches! It probably makes sense for patches #1 and #2 to stay together and both go via drm-misc,