AMD/IOMMU: assorted corrections (leftover)

[PATCH v3 for-4.20 0/4] AMD/IOMMU: assorted corrections (leftover)

Posted by Jan Beulich 1 month, 1 week ago

What previously was the main bug fix in this series was dropped for v3. It
turns out what is now patch 2 is sufficient to address the issue, while
doing the requested tidying. The latter two patches are for 4.21 only, with
the final one being up for debate altogether.

1: radix-tree: purge node allocation override hooks
2: radix-tree: introduce RADIX_TREE{,_INIT}()
3: radix-tree: drop "root" parameters from radix_tree_node_{alloc,free}()
4: PCI: drop pci_segments_init()

Jan

Re: [PATCH v3 for-4.20 0/4] AMD/IOMMU: assorted corrections (leftover)

Posted by Jan Beulich 1 month ago

Oleksii,

On 04.02.2025 14:00, Jan Beulich wrote:
> What previously was the main bug fix in this series was dropped for v3. It
> turns out what is now patch 2 is sufficient to address the issue, while
> doing the requested tidying. The latter two patches are for 4.21 only, with
> the final one being up for debate altogether.
> 
> 1: radix-tree: purge node allocation override hooks
> 2: radix-tree: introduce RADIX_TREE{,_INIT}()
> 3: radix-tree: drop "root" parameters from radix_tree_node_{alloc,free}()
> 4: PCI: drop pci_segments_init()

any verdict one way or the other for the first two patches?

Thanks, Jan

Re: [PATCH v3 for-4.20 0/4] AMD/IOMMU: assorted corrections (leftover)

Posted by Oleksii Kurochko 1 month ago

On 2/6/25 1:35 PM, Jan Beulich wrote:
> Oleksii,
>
> On 04.02.2025 14:00, Jan Beulich wrote:
>> What previously was the main bug fix in this series was dropped for v3. It
>> turns out what is now patch 2 is sufficient to address the issue, while
>> doing the requested tidying. The latter two patches are for 4.21 only, with
>> the final one being up for debate altogether.
>>
>> 1: radix-tree: purge node allocation override hooks
>> 2: radix-tree: introduce RADIX_TREE{,_INIT}()
>> 3: radix-tree: drop "root" parameters from radix_tree_node_{alloc,free}()
>> 4: PCI: drop pci_segments_init()
> any verdict one way or the other for the first two patches?

Sorry, I missed the fixes tag so I thought that they are just refactoring.

For first two patches:
R-Acked-by: Oleksii Kurochko<oleksii.kurochko@gmail.com>

~ Oleksii