Re: [Xen-devel] [PATCH 06/60] xen/sched: move per-vcpu scheduler private data pointer to sched_unit

[Juergen Gross <jgross@xxxxxxxx>]

On 19.07.19 00:52, Dario Faggioli wrote:
> On Tue, 2019-05-28 at 12:32 +0200, Juergen Gross wrote:
> > This prepares making the different schedulers vcpu agnostic.
> Ok, but the scheduler private data is, actually, for all the
> schedulers, per-vcpu scheduler data such as, for instance,
> struct csched2_vcpu.
>
> After this patch we have (sticking to Credit2 as an example)
> csched2_vcpu allocated by a function called csched2_alloc_vdata() but
> stored on a per-sched_unit basis.
>
> Similarly, we have an accessor method called csched2_vcpu() which
> returns the struct csched2_vcpu which was stored in the per-unit
> private space.
>
> But shouldn't then the struct be called csched2_unit, and cited
> functions be called csched2_alloc_udata() and csched2_unit()?
> Now, I know that these transformation happen later in the series.
> And, this time, I'm not asking to consolidate the patches.
>
> However:
> - can the changelog of this patch be a little more explicit, not only
>    informing that this is a preparatory patch, but also explaining
>    briefly the temporary inconcistency
Sure.

> - could the patches that deal with this be grouped together, so that
>    they are close to each other in the series (e.g., this patch, the
>    renaming hunks of patch 10 and patches that are currently 20 to 24)?
>    Or are there dependencies that I am not considering?
There are some patches which could be reordered, but I'm not sure the
needed work is worth it. By moving the patches you named closer to each
other there will be other closely related patches ripped apart from
each other.

In the end it will make no sense to apply only some patches of the main
series. The huge amount of patches is meant only to make review easier.


Juergen

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