Re: [RESEND][PATCH v8 0/5] DMA-BUF Heaps (destaging ION)

[Ayan Halder]

On Fri, Oct 18, 2019 at 09:55:17AM +0000, Brian Starkey wrote:
> On Thu, Oct 17, 2019 at 01:57:45PM -0700, John Stultz wrote:
> > On Thu, Oct 17, 2019 at 12:29 PM Andrew F. Davis <afd@xxxxxx> wrote:
> > > On 10/17/19 3:14 PM, John Stultz wrote:
> > > > But if the objection stands, do you have a proposal for an alternative
> > > > way to enumerate a subset of CMA heaps?
> > > >
> > > When in staging ION had to reach into the CMA framework as the other
> > > direction would not be allowed, so cma_for_each_area() was added. If
> > > DMA-BUF heaps is not in staging then we can do the opposite, and have
> > > the CMA framework register heaps itself using our framework. That way
> > > the CMA system could decide what areas to export or not (maybe based on
> > > a DT property or similar).
> > 
> > Ok. Though the CMA core doesn't have much sense of DT details either,
> > so it would probably have to be done in the reserved_mem logic, which
> > doesn't feel right to me.
> > 
> > I'd probably guess we should have some sort of dt binding to describe
> > a dmabuf cma heap and from that node link to a CMA node via a
> > memory-region phandle. Along with maybe the default heap as well? Not
> > eager to get into another binding review cycle, and I'm not sure what
> > non-DT systems will do yet, but I'll take a shot at it and iterate.
> > 
> > > The end result is the same so we can make this change later (it has to
> > > come after DMA-BUF heaps is in anyway).
> > 
> > Well, I'm hesitant to merge code that exposes all the CMA heaps and
> > then add patches that becomes more selective, should anyone depend on
> > the initial behavior. :/
> 
> How about only auto-adding the system default CMA region (cma->name ==
> "reserved")?
> 
> And/or the CMA auto-add could be behind a config option? It seems a
> shame to further delay this, and the CMA heap itself really is useful.
>
A bit of a detour, comming back to the issue why the following node
was not getting detected by the dma-buf heaps framework.

        reserved-memory {
                #address-cells = <2>;
                #size-cells = <2>;
                ranges;

                display_reserved: framebuffer@60000000 {
                        compatible = "shared-dma-pool";
                        linux,cma-default;
                        reusable; <<<<<<<<<<<<-----------This was missing in our
earlier node
                        reg = <0 0x60000000 0 0x08000000>;
                };
 
Quoting reserved-memory.txt :-
"The operating system can use the memory in this region with the limitation that
 the device driver(s) owning the region need to be able to reclaim it back"

Thus as per my observation, without 'reusable', rmem_cma_setup()
returns -EINVAL and the reserved-memory is not added as a cma region.

With 'reusable', rmem_cma_setup() succeeds , but the kernel crashes as follows :-

[    0.450562] WARNING: CPU: 2 PID: 1 at mm/cma.c:110 cma_init_reserved_areas+0xec/0x22c
[    0.458415] Modules linked in:                                                                                                             
[    0.461470] CPU: 2 PID: 1 Comm: swapper/0 Not tainted 5.3.0-rc4-01377-g51dbcf03884c-dirty #15                                              
[    0.470017] Hardware name: ARM Juno development board (r0) (DT)                                                                            
[    0.475953] pstate: 80000005 (Nzcv daif -PAN -UAO)                                                                                         
[    0.480755] pc : cma_init_reserved_areas+0xec/0x22c  
[    0.485643] lr : cma_init_reserved_areas+0xe8/0x22c 
<----snip register dump --->

[    0.600646] Unable to handle kernel paging request at virtual address ffff7dffff800000
[    0.608591] Mem abort info:
[    0.611386]   ESR = 0x96000006
<---snip uninteresting bits --->
[    0.681069] pc : cma_init_reserved_areas+0x114/0x22c
[    0.686043] lr : cma_init_reserved_areas+0xe8/0x22c


I am looking into this now. My final objective is to get "/dev/dma_heap/framebuffer"
(as a cma heap).
Any leads?

> Cheers,
> -Brian
> 
> > 
> > So, <sigh>, I'll start on the rework for the CMA bits.
> > 
> > That said, I'm definitely wanting to make some progress on this patch
> > series, so maybe we can still merge the core/helpers/system heap and
> > just hold the cma heap for a rework on the enumeration bits. That way
> > we can at least get other folks working on switching their vendor
> > heaps from ION.
> > 
> > Sumit: Does that sound ok? Assuming no other objections, can you take
> > the v11 set minus the CMA heap patch?
> > 
> > thanks
> > -john