Re: [PATCH 5/7] iommu/arm-smmu-v3: Move the DMA API comment to flush_iotlb_all

[Mostafa Saleh <smostafa@google.com>]

On Mon, Jul 06, 2026 at 01:29:11PM -0300, Jason Gunthorpe wrote:
> arm_smmu_tlb_inv_context() is the wrong flush_all for this comment,
> it is only called during io-pgtable destruction not during the dma-iommu
> operation. Move it to arm_smmu_flush_iotlb_all() which is the flush
> that is triggered by the dma-iommu lazy flush thread.
> 
> Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Reviewed-by: Mostafa Saleh <smostafa@google.com>

Thanks,
Mostafa
> ---
>  drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c | 31 +++++++++++----------
>  1 file changed, 16 insertions(+), 15 deletions(-)
> 
> diff --git a/drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c b/drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c
> index 5087603ea18e62..dd22c0f881bfba 100644
> --- a/drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c
> +++ b/drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c
> @@ -2392,21 +2392,6 @@ static void arm_smmu_tlb_inv_context(void *cookie)
>  {
>  	struct arm_smmu_domain *smmu_domain = cookie;
>  
> -	/*
> -	 * If the DMA API is running in non-strict mode then another CPU could
> -	 * have changed the page table and not invoked any flush op. Instead the
> -	 * other CPU will do an atomic_read() and this CPU will have done an
> -	 * atomic_write(). That handshake is enough to acquire the page table
> -	 * writes from the other CPU.
> -	 *
> -	 * All command execution has a dma_wmb() to release all the in-memory
> -	 * structures written by this CPU, that barrier must also release the
> -	 * writes acquired from all the other CPUs too.
> -	 *
> -	 * There are other barriers and atomics on this path, but the above is
> -	 * the essential mechanism for ensuring that HW sees the page table
> -	 * writes from another CPU before it executes the IOTLB invalidation.
> -	 */
>  	arm_smmu_domain_inv(smmu_domain);
>  }
>  
> @@ -4078,6 +4063,22 @@ static void arm_smmu_flush_iotlb_all(struct iommu_domain *domain)
>  {
>  	struct arm_smmu_domain *smmu_domain = to_smmu_domain(domain);
>  
> +	/*
> +	 * If the DMA API is running in non-strict mode then another CPU could
> +	 * have changed the page table and not invoked any flush op. Instead the
> +	 * other CPU will do an atomic_read() and this CPU will have done an
> +	 * atomic_write(). That handshake is enough to acquire the page table
> +	 * writes from the other CPU.
> +	 *
> +	 * All command execution has a dma_wmb() to release all the in-memory
> +	 * structures written by this CPU, that barrier must also release the
> +	 * writes acquired from all the other CPUs too.
> +	 *
> +	 * There are other barriers and atomics on this path, but the above is
> +	 * the essential mechanism for ensuring that HW sees the page table
> +	 * writes from another CPU before it executes the IOTLB invalidation.
> +	 */
> +
>  	if (smmu_domain->smmu)
>  		arm_smmu_tlb_inv_context(smmu_domain);
>  }
> -- 
> 2.43.0
>