linux-kernel - Re: [PATCH 2/2] drivers/arm-smmu-v3: Implement .iotlb_sync

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <aNp60ak6OASbPd_J@google.com>
Date: Mon, 29 Sep 2025 12:25:53 +0000
From: Mostafa Saleh <smostafa@...gle.com>
To: Daniel Mentz <danielmentz@...gle.com>
Cc: iommu@...ts.linux.dev, linux-kernel@...r.kernel.org,
	Will Deacon <will@...nel.org>,
	Pranjal Shrivastava <praan@...gle.com>,
	Liviu Dudau <liviu.dudau@....com>, Jason Gunthorpe <jgg@...dia.com>,
	Rob Clark <robin.clark@....qualcomm.com>
Subject: Re: [PATCH 2/2] drivers/arm-smmu-v3: Implement .iotlb_sync_map
 callback

On Sat, Sep 27, 2025 at 10:39:53PM +0000, Daniel Mentz wrote:
> Implement .iotlb_sync_map callback based on the new callback in the
> io-pgtable code.
> 
> Signed-off-by: Daniel Mentz <danielmentz@...gle.com>
> ---
>  drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c | 14 ++++++++++++++
>  1 file changed, 14 insertions(+)
> 
> 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 2a8b46b948f0..0ffcc6c8e4bf 100644
> --- a/drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c
> +++ b/drivers/iommu/arm/arm-smmu-v3/arm-smmu-v3.c
> @@ -2541,6 +2541,7 @@ static int arm_smmu_domain_finalise(struct arm_smmu_domain *smmu_domain,
>  	pgtbl_cfg = (struct io_pgtable_cfg) {
>  		.pgsize_bitmap	= smmu->pgsize_bitmap,
>  		.coherent_walk	= smmu->features & ARM_SMMU_FEAT_COHERENCY,
> +		.defer_sync_pte	= true,
>  		.tlb		= &arm_smmu_flush_ops,
>  		.iommu_dev	= smmu->dev,
>  	};
> @@ -3366,6 +3367,18 @@ static int arm_smmu_map_pages(struct iommu_domain *domain, unsigned long iova,
>  	return ops->map_pages(ops, iova, paddr, pgsize, pgcount, prot, gfp, mapped);
>  }
>  
> +static int arm_smmu_iotlb_sync_map(struct iommu_domain *domain,
> +				    unsigned long iova, size_t size)
> +{
> +	struct io_pgtable_ops *ops = to_smmu_domain(domain)->pgtbl_ops;
> +
> +	if (!ops || !ops->iotlb_sync_map)
> +		return 0;
> +
> +	ops->iotlb_sync_map(ops, iova, size);
> +	return 0;

Shouldn't that be
	return ops->iotlb_sync_map(ops, iova, size);

As now we ignore the return from the ops and mask it to 0.

Thanks,
Mostafa

> +}
> +
>  static size_t arm_smmu_unmap_pages(struct iommu_domain *domain, unsigned long iova,
>  				   size_t pgsize, size_t pgcount,
>  				   struct iommu_iotlb_gather *gather)
> @@ -3700,6 +3713,7 @@ static const struct iommu_ops arm_smmu_ops = {
>  		.map_pages		= arm_smmu_map_pages,
>  		.unmap_pages		= arm_smmu_unmap_pages,
>  		.flush_iotlb_all	= arm_smmu_flush_iotlb_all,
> +		.iotlb_sync_map		= arm_smmu_iotlb_sync_map,
>  		.iotlb_sync		= arm_smmu_iotlb_sync,
>  		.iova_to_phys		= arm_smmu_iova_to_phys,
>  		.free			= arm_smmu_domain_free_paging,
> -- 
> 2.51.0.570.gb178f27e6d-goog
>