Re: [Xen-devel] [PATCH 0/2] Dangling fixes for ARM iommu

[Andrii Anisov <andrii.anisov@xxxxxxxxx>]

Hello Julien,

On 21.01.19 19:29, Julien Grall wrote:
> (+ Juergen)
> All patches candidate for Xen 4.12 should have the release manager CCed and 
> explain the pros/cons to have those patches for this release. It is also useful 
> if you add for-4.12 (or similar) in the [...] so we can prioritize it.
I've got the point. Thank you for adding Juergen.

> Indeed code base evolved, for instance iommu_use_hap_pt has a slight different 
> behavior now.
Actually you properly pointed the problem. I'm not sure how it happened, 
because I did test build and run for those patches yesterday. But now, 
progressing with our stuff porting, I realized that the patch 1 is missing 
dependency on `xen/sched.h` required by `iommu_use_hap_pt`.

> So please stripped my RB
OK.

> The AB is arguable as it is quite old.
I could drop it for v2.

> Secondly, it often happens that some patches have the required acked-by and reviewed-by but are not merged because they makes little sense without the rest of the series. 
Yes, sure, but fixes are still relevant and are quite autonomous. The only 
relation to the series is that they were discovered during its implementation.

> In that case, the current code base does not support the case where the P2M is 
> not shared with the IOMMU and does not support new IOMMU bindings in the 
> current setup.
It is the intention of the rest of the series.

> So I am not convinced they should be included in Xen 4.12 or even without the 
> rest of the series.
IMO, the pure fixes, like patch 2, and the first hunk of patch 1 should be OK 
for 4.12.

--
Sincerely,
Andrii Anisov.

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxx
https://lists.xenproject.org/mailman/listinfo/xen-devel