Re: [RFC PATCH] sched/eevdf: Return leftmost entity in pick_eevdf() if no eligible entity is found

[Tianchen Ding]

On 2024/4/18 13:52, Chen Yu wrote:
> On 2024-04-18 at 11:37:31 +0800, Tianchen Ding wrote:
> > Hi, we've found a bug about reweight_eevdf(). Would you also please try this
> > patchset?
> >
> > https://lore.kernel.org/all/20240306022133.81008-1-dtcccc@xxxxxxxxxxxxxxxxx/
>
> Sure, I'll have a try after I test Xuewen's patch. I assume your fix is mainly in
> 2/2 which adjusts the se->vruntime in reweight_eevdf() by using the correct
> avg_vruntime(), although it might not be directly related to the s64 overflow issue
> here, it is worthy a fix.

Yes. I'm hoping to reduce "noises" about the unexpected avg_vruntime() varying 
during your tests.

Thanks.