[PATCH] iommu/arm-smmu: fix bug in pmd construction

[PATCH] iommu/arm-smmu: fix bug in pmd construction

[Mitchel Humpherys]

We are using the same pfn for every pte we create while constructing the
pmd. Fix this by actually updating the pfn on each iteration of the pmd
construction loop.

It's not clear if we can actually hit this bug right now since iommu_map
splits up the calls to .map based on the page size, so we only ever seem to
iterate this loop once. However, things might change in the future that
might cause us to hit this.

Signed-off-by: Mitchel Humpherys <mitchelh@codeaurora.org>
---
Will, I was unable to come up with a test case to hit this bug based on
what I said in the commit message above. Not sure if my analysis is
completely off base, my head is still spinning from all these page tables
:).
---
 drivers/iommu/arm-smmu.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/iommu/arm-smmu.c b/drivers/iommu/arm-smmu.c
index ca18d6d42a..eba4cb390c 100644
--- a/drivers/iommu/arm-smmu.c
+++ b/drivers/iommu/arm-smmu.c
@@ -1368,6 +1368,7 @@ static int arm_smmu_alloc_init_pmd(struct arm_smmu_device *smmu, pud_t *pud,
 		ret = arm_smmu_alloc_init_pte(smmu, pmd, addr, next, pfn,
 					      prot, stage);
 		phys += next - addr;
+		pfn = __phys_to_pfn(phys);
 	} while (pmd++, addr = next, addr < end);
 
 	return ret;
-- 
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum,
hosted by The Linux Foundation

[PATCH] iommu/arm-smmu: fix bug in pmd construction

[Varun Sethi]

Hi Mitchel,
I have made changes to the arm smmu driver paging support, as a part of the huge page support patch. Will is currently reviewing changes made by me. I should be able to post my patch in the next couple of weeks.

Regards
Vaun

> -----Original Message-----
> From: iommu-bounces at lists.linux-foundation.org [mailto:iommu-
> bounces at lists.linux-foundation.org] On Behalf Of Mitchel Humpherys
> Sent: Saturday, September 20, 2014 3:29 AM
> To: linux-arm-kernel at lists.infradead.org; iommu at lists.linux-foundation.org;
> Will Deacon
> Subject: [PATCH] iommu/arm-smmu: fix bug in pmd construction
> 
> We are using the same pfn for every pte we create while constructing the pmd.
> Fix this by actually updating the pfn on each iteration of the pmd construction
> loop.
> 
> It's not clear if we can actually hit this bug right now since iommu_map splits up
> the calls to .map based on the page size, so we only ever seem to iterate this
> loop once. However, things might change in the future that might cause us to
> hit this.
> 
> Signed-off-by: Mitchel Humpherys <mitchelh@codeaurora.org>
> ---
> Will, I was unable to come up with a test case to hit this bug based on what I
> said in the commit message above. Not sure if my analysis is completely off
> base, my head is still spinning from all these page tables :).
> ---
>  drivers/iommu/arm-smmu.c | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/drivers/iommu/arm-smmu.c b/drivers/iommu/arm-smmu.c index
> ca18d6d42a..eba4cb390c 100644
> --- a/drivers/iommu/arm-smmu.c
> +++ b/drivers/iommu/arm-smmu.c
> @@ -1368,6 +1368,7 @@ static int arm_smmu_alloc_init_pmd(struct
> arm_smmu_device *smmu, pud_t *pud,
>  		ret = arm_smmu_alloc_init_pte(smmu, pmd, addr, next, pfn,
>  					      prot, stage);
>  		phys += next - addr;
> +		pfn = __phys_to_pfn(phys);
>  	} while (pmd++, addr = next, addr < end);
> 
>  	return ret;
> --
> The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum,
> hosted by The Linux Foundation
> 
> _______________________________________________
> iommu mailing list
> iommu at lists.linux-foundation.org
> https://lists.linuxfoundation.org/mailman/listinfo/iommu

[PATCH] iommu/arm-smmu: fix bug in pmd construction

[Will Deacon]

On Sun, Sep 21, 2014 at 08:56:28AM +0100, Varun Sethi wrote:
> I have made changes to the arm smmu driver paging support, as a part of
> the huge page support patch. Will is currently reviewing changes made by
> me. I should be able to post my patch in the next couple of weeks.

Yup, but this patch looks like a sensible change for now. I agree that we
can't currently hit the bug, but that code is intended to be usable for use
across multiple PMDs, so I'll apply this.

Will