On Sun, Jul 26, 2020 at 11:27:03PM -0700, Bjorn Andersson wrote:
> On Mon 20 Jul 08:40 PDT 2020, Jordan Crouse wrote:
> > diff --git a/drivers/iommu/arm-smmu-qcom.c b/drivers/iommu/arm-smmu-qcom.c
> [..]
> > +static int qcom_adreno_smmu_alloc_context_bank(struct arm_smmu_domain *smmu_domain,
> > + struct device *dev, int start, int count)
> > +{
> > + struct arm_smmu_device *smmu = smmu_domain->smmu;
> > +
> > + /*
> > + * Assign context bank 0 to the GPU device so the GPU hardware can
> > + * switch pagetables
> > + */
> > + if (qcom_adreno_smmu_is_gpu_device(dev)) {
> > + if (start > 0 || test_bit(0, smmu->context_map))
> > + return -ENOSPC;
> > +
> > + set_bit(0, smmu->context_map);
> > + return 0;
> > + }
> > +
> > + return __arm_smmu_alloc_bitmap(smmu->context_map, start, count);
>
> If we end up here before the GPU device shows up this is going to
> steal the first context bank, causing the subsequent allocation for the
> GPU to always fail.
>
> As such I think it would be appropriate for you to adjust "start" to
> never be 0 here. And I think it would be appropriate to write this
> function as:
>
> if (gpu) {
> start = 0;
> count = 1;
> } else {
> if (start == 0)
> start = 1;
> }
>
> return __arm_smmu_alloc_bitmap(smmu->context_map, start, count);
Excellent suggestions. Thanks.
Jordan
> Regards,
> Bjorn
--
The Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
a Linux Foundation Collaborative Project
