2020-09-14 07:26:46

by Srinath Mannam

[permalink] [raw]
Subject: [PATCH v3] iommu/dma: Fix IOVA reserve dma ranges

Fix IOVA reserve failure in the case when address of first memory region
listed in dma-ranges is equal to 0x0.

Fixes: aadad097cd46f ("iommu/dma: Reserve IOVA for PCIe inaccessible DMA address")
Signed-off-by: Srinath Mannam <[email protected]>
---
Changes from v2:
Modify error message with useful information based on Bjorn's comments.

Changes from v1:
Removed unnecessary changes based on Robin's review comments.

drivers/iommu/dma-iommu.c | 6 ++++--
1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/drivers/iommu/dma-iommu.c b/drivers/iommu/dma-iommu.c
index 5141d49a046b..5b9791f35c5e 100644
--- a/drivers/iommu/dma-iommu.c
+++ b/drivers/iommu/dma-iommu.c
@@ -217,9 +217,11 @@ static int iova_reserve_pci_windows(struct pci_dev *dev,
lo = iova_pfn(iovad, start);
hi = iova_pfn(iovad, end);
reserve_iova(iovad, lo, hi);
- } else {
+ } else if (end < start) {
/* dma_ranges list should be sorted */
- dev_err(&dev->dev, "Failed to reserve IOVA\n");
+ dev_err(&dev->dev,
+ "Failed to reserve IOVA [%#010llx-%#010llx]\n",
+ start, end);
return -EINVAL;
}

--
2.17.1


2021-06-02 20:22:25

by Sven Peter

[permalink] [raw]
Subject: Re: [PATCH v3] iommu/dma: Fix IOVA reserve dma ranges

Hi,

I just ran into the exact same issue while working on the M1 DART IOMMU driver
and it was fixed by this commit. Thanks!

Would be great if this could be picked up.

Tested-by: Sven Peter <[email protected]>


Best,


Sven


On Mon, Sep 14, 2020, at 09:23, Srinath Mannam via iommu wrote:
> Fix IOVA reserve failure in the case when address of first memory region
> listed in dma-ranges is equal to 0x0.
>
> Fixes: aadad097cd46f ("iommu/dma: Reserve IOVA for PCIe inaccessible
> DMA address")
> Signed-off-by: Srinath Mannam <[email protected]>
> ---
> Changes from v2:
> Modify error message with useful information based on Bjorn's
> comments.
>
> Changes from v1:
> Removed unnecessary changes based on Robin's review comments.
>
> drivers/iommu/dma-iommu.c | 6 ++++--
> 1 file changed, 4 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/iommu/dma-iommu.c b/drivers/iommu/dma-iommu.c
> index 5141d49a046b..5b9791f35c5e 100644
> --- a/drivers/iommu/dma-iommu.c
> +++ b/drivers/iommu/dma-iommu.c
> @@ -217,9 +217,11 @@ static int iova_reserve_pci_windows(struct pci_dev *dev,
> lo = iova_pfn(iovad, start);
> hi = iova_pfn(iovad, end);
> reserve_iova(iovad, lo, hi);
> - } else {
> + } else if (end < start) {
> /* dma_ranges list should be sorted */
> - dev_err(&dev->dev, "Failed to reserve IOVA\n");
> + dev_err(&dev->dev,
> + "Failed to reserve IOVA [%#010llx-%#010llx]\n",
> + start, end);
> return -EINVAL;
> }
>
> --
> 2.17.1
>
> _______________________________________________
> iommu mailing list
> [email protected]
> https://lists.linuxfoundation.org/mailman/listinfo/iommu
>
>

2021-06-03 11:02:49

by Robin Murphy

[permalink] [raw]
Subject: Re: [PATCH v3] iommu/dma: Fix IOVA reserve dma ranges

On 2021-06-02 21:18, Sven Peter wrote:
> Hi,
>
> I just ran into the exact same issue while working on the M1 DART IOMMU driver
> and it was fixed by this commit. Thanks!
>
> Would be great if this could be picked up.

Oops, apparently I was happy enough with this 9 months ago to forget
about it, so if it helps,

Reviewed-by: Robin Murphy <[email protected]>

> Tested-by: Sven Peter <[email protected]>
>
>
> Best,
>
>
> Sven
>
>
> On Mon, Sep 14, 2020, at 09:23, Srinath Mannam via iommu wrote:
>> Fix IOVA reserve failure in the case when address of first memory region
>> listed in dma-ranges is equal to 0x0.
>>
>> Fixes: aadad097cd46f ("iommu/dma: Reserve IOVA for PCIe inaccessible
>> DMA address")
>> Signed-off-by: Srinath Mannam <[email protected]>
>> ---
>> Changes from v2:
>> Modify error message with useful information based on Bjorn's
>> comments.
>>
>> Changes from v1:
>> Removed unnecessary changes based on Robin's review comments.
>>
>> drivers/iommu/dma-iommu.c | 6 ++++--
>> 1 file changed, 4 insertions(+), 2 deletions(-)
>>
>> diff --git a/drivers/iommu/dma-iommu.c b/drivers/iommu/dma-iommu.c
>> index 5141d49a046b..5b9791f35c5e 100644
>> --- a/drivers/iommu/dma-iommu.c
>> +++ b/drivers/iommu/dma-iommu.c
>> @@ -217,9 +217,11 @@ static int iova_reserve_pci_windows(struct pci_dev *dev,
>> lo = iova_pfn(iovad, start);
>> hi = iova_pfn(iovad, end);
>> reserve_iova(iovad, lo, hi);
>> - } else {
>> + } else if (end < start) {
>> /* dma_ranges list should be sorted */
>> - dev_err(&dev->dev, "Failed to reserve IOVA\n");
>> + dev_err(&dev->dev,
>> + "Failed to reserve IOVA [%#010llx-%#010llx]\n",
>> + start, end);
>> return -EINVAL;
>> }
>>
>> --
>> 2.17.1
>>
>> _______________________________________________
>> iommu mailing list
>> [email protected]
>> https://lists.linuxfoundation.org/mailman/listinfo/iommu
>>
>>

2021-06-04 15:36:10

by Joerg Roedel

[permalink] [raw]
Subject: Re: [PATCH v3] iommu/dma: Fix IOVA reserve dma ranges

On Mon, Sep 14, 2020 at 12:53:19PM +0530, Srinath Mannam wrote:
> Fix IOVA reserve failure in the case when address of first memory region
> listed in dma-ranges is equal to 0x0.
>
> Fixes: aadad097cd46f ("iommu/dma: Reserve IOVA for PCIe inaccessible DMA address")
> Signed-off-by: Srinath Mannam <[email protected]>
> ---
> Changes from v2:
> Modify error message with useful information based on Bjorn's comments.
>
> Changes from v1:
> Removed unnecessary changes based on Robin's review comments.
>
> drivers/iommu/dma-iommu.c | 6 ++++--
> 1 file changed, 4 insertions(+), 2 deletions(-)

Applied, thanks.