On 01/03, Srinivas Kandagatla wrote:
> Thanks for your review comments,
>
> On 03/01/18 17:20, Stephen Boyd wrote:
> >>+
> >>+ return ret;
> >>+}
> >>+
> >>+static int msm_snd_apq8096_probe(struct platform_device *pdev)
> >>+{
> >>+ int ret;
> >>+ struct snd_soc_card *card;
> >>+
> >>+ card = devm_kzalloc(&pdev->dev, sizeof(*card), GFP_KERNEL);
> >>+ if (!card)
> >>+ return -ENOMEM;
> >>+
> >>+ card->dev = &pdev->dev;
> >>+
> >>+ ret = dma_coerce_mask_and_coherent(card->dev, DMA_BIT_MASK(32));
> >
> >Why do we need to do this? Can you add some sort of comment in the code
> >about why?
>
> Even though dsp supports 64 bit addresses, but the sid sits at
> offset of 32, which brings this restriction of supporting only 32
> bit iova.
>
Doesn't the dsp have an iommu in place to make the address
translation from 64 to 32 bits transparent? I thought this was
what dma-ranges and iommu binding was for, but I'm not well
versed on all the details here.
--
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
a Linux Foundation Collaborative Project
