Re: [PATCH 4/9] xen/arm: static memory initialization

[Julien Grall <julien@xxxxxxx>]

On 06/07/2021 10:20, Penny Zheng wrote:
> Hi Julien
Hi Penny,

> > -----Original Message-----
> > From: Julien Grall <julien@xxxxxxx>
> > Sent: Tuesday, July 6, 2021 5:10 PM
> > To: Penny Zheng <Penny.Zheng@xxxxxxx>; xen-devel@xxxxxxxxxxxxxxxxxxxx;
> > sstabellini@xxxxxxxxxx; jbeulich@xxxxxxxx
> > Cc: Bertrand Marquis <Bertrand.Marquis@xxxxxxx>; Wei Chen
> > <Wei.Chen@xxxxxxx>
> > Subject: Re: [PATCH 4/9] xen/arm: static memory initialization
> >
> >
> >
> > On 05/07/2021 08:28, Penny Zheng wrote:
> > > Hi Julien
> > Hi Penny,
> >
> > > > -----Original Message-----
> > > > From: Julien Grall <julien@xxxxxxx>
> > > > Sent: Thursday, July 1, 2021 2:10 AM
> > > > To: Penny Zheng <Penny.Zheng@xxxxxxx>;
> > > > xen-devel@xxxxxxxxxxxxxxxxxxxx; sstabellini@xxxxxxxxxx;
> > > > jbeulich@xxxxxxxx
> > > > Cc: Bertrand Marquis <Bertrand.Marquis@xxxxxxx>; Wei Chen
> > > > <Wei.Chen@xxxxxxx>
> > > > Subject: Re: [PATCH 4/9] xen/arm: static memory initialization
> > > >
> > > > Hi Penny,
> > > >
> > > > On 07/06/2021 03:43, Penny Zheng wrote:
> > > > > This patch introduces static memory initialization, during system
> > > > > RAM boot
> > > > up.
> > > >
> > > > The word "RAM" looks spurious.
> > > Thx. I check the "spurious" in dictionary, it means fake? So I will leave
> > "during system boot up"
> > > here.
> > Yes, this reads better.
> >
> > > > > 2. scrub the page in need synchronously.
> > > > Can you explain why this is necessary?
> > > Since I'm borrowing the logic in free_heap_pages, I'm also trying to cover all
> > the scenarios here like it does.
> > > So I assume that free_staticmem_page will not only be used on initialization,
> > but also when destroying/rebooting the domain.
> > > On these cases, it is necessary to scrub the page, ig.
> > I wasn't asking about scrubbing specifically but instead why it is synchronous.
> > Sorry for the confusion.
> I've read asynchronous scrubbing in buddy allocator, pages that need a scrub 
> are added to tail,
> and specific working thread is working on it, hmm, imo, I don't think it could 
> be simply applied into static
> memory. :/
I am not sure to understand why the same concept can't be applied. 
Anyway, what I am asking is to clarify in the commit message why this is 
synchronous (e.g. this just temporary).
>  > So I put asynchronous scrubbing in #TODO.
We seem to have quite a few TODOs (some of them are implicit). Can we 
start to gather a list of what's missing? A first good place would be 
the cover lette.r
Cheers,

--
Julien Grall