Re: [Xen-devel] [PATCH 0/7] get rid of the SEDF scheduler

[George Dunlap <george.dunlap@xxxxxxxxxxxxx>]

On 06/26/2015 05:18 PM, Dario Faggioli wrote:
> Hi everyone,
> 
> This series is the follow up of this proposal and conversation:
> http://lists.xen.org/archives/html/xen-devel/2015-05/msg02874.html
> 
> Let me quote this (again), from 2006:
>   git show db51cd09d37ea44b126bb259f9392248afd768e6
>   ...
>   diff --git a/docs/src/interface.tex b/docs/src/interface.tex
>   index c9017c7..9a59840 100644
>   --- a/docs/src/interface.tex
>   +++ b/docs/src/interface.tex
>   @@ -209,8 +209,8 @@ implement timeout values when they block.
> 
>    Xen offers a uniform API for CPU schedulers.  It is possible to choose
>    from a number of schedulers at boot and it should be easy to add more.
>   -The SEDF, BVT, and Credit schedulers are part of the normal Xen
>   -distribution.  BVT and SEDF will be going away and their use should be
>   +The SEDF and Credit schedulers are part of the normal Xen
>   +distribution.  SEDF will be going away and its use should be
>    avoided once the credit scheduler has stabilized and become the default.
>    The Credit scheduler provides proportional fair shares of the
>    host's CPUs to the running domains. It does this while transparently
> 
> So, after being "going away" for 9 years, here we are, getting rid of it!
> Hopefully, I've killed every sign of it, and did not overlook anything.
> 
> I'm not sure whether the python bindings are an interface that we consider
> stable. I've killed everything about SEDF from there, but I don't know sure 
> how
> to test this. If I should do it differently, let me know.
> 
> In libxl, I left the symbols and the data types in place, for the sake of
> build-time backward compatibility, but any attempt of calling one of the
> scheduler API function with LIBXL_SCHEDULED_SEDF as the scheduler id, results
> in an error.
> 
> As far as testing goes, I have verified that:
>  - the repo builds, both xen and tools, at each stage of the series
>  - booting with sched=sedf no longer works
>  - with all the series applied, `xl sched-sedf' no longer exists
>  - with the libxl+libxc patches patch applied (so, with SEDF still available 
> in Xen
>    and xl), `xl sched-sedf' exists, but produces an error
> 
> There is an OSSTest patch that shuts down SEDF testing for xen-unstable
> (actually, for everything greater than xen-4.5) here:
> http://lists.xen.org/archives/html/xen-devel/2015-06/msg04399.html
> 
> Regards,
> Dario
> ---
> Dario Faggioli (7):
>       libxl: get rid of the SEDF scheduler
>       tools: python: get rid of the SEDF scheduler bindings
>       libxc: get rid of the SEDF scheduler
>       xen: get rid of the SEDF scheduler
>       xen: kill sched_sedf.c
>       xl: get rid of the SEDF scheduler
>       docs: get rid of the SEDF scheduler

The entire series:

Reviewed-by: George Dunlap <george.dunlap@xxxxxxxxxxxxx>

Good to see it finally go. :-)



_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel