Re: [PATCH v3 1/6] xen/arm: Skip initializing the BSS section when it is empty

[Julien Grall <julien@xxxxxxx>]

Hi Ayan,

On 21/10/2024 12:12, Ayan Kumar Halder wrote:
> On 18/10/2024 14:41, Julien Grall wrote:
> > Hi Ayan,
> Hi Julien,
> > On 10/10/2024 15:03, Ayan Kumar Halder wrote:
> > > If the BSS section is empty, then the function can just return.
> > This is more than "can", right? If we don't do it, we will end up to 
> > zero outside of BSS. This could be critical data...
> s/can just/should
> > Also, I am tempted to suggest to add a Fixes tag because even if it is 
> > unlikely BSS will be zero in the current Xen, it is also not unlikely.
> > The tag would be:
> >
> > Fixes: dac84b66cc9a ("xen: arm64: initial build + config changes, 
> > start of day code")
> Ack.
> > > Signed-off-by: Ayan Kumar Halder <ayan.kumar.halder@xxxxxxx>
> > I saw the discussion. I don't have a strong opinion on the exact 
> > approach choosen for zeroing. With the commit message updated:
> > Acked-by: Julien Grall <jgrall@xxxxxxxxxx>
> I propose that this patch to be committed. The changes to the commit 
> message can be done on commit.
> > > ---
> > > Changes from :-
> > >
> > > v1..v2 - New patch introduced in v3.
> > >
> > >   xen/arch/arm/arm64/head.S | 2 ++
> > Don't we need a similar change on the arm32 code?
> I haven't looked at the arm32 code. My idea is to get the earlyboot (ie 
> the asm part) of Xen working on R82 and then do the similar changes for 
> R52 (ie arm32).
AFAIU, this change is not related to the MPU. So I would rather prefer 
if we keep this change in sync. I am happy to send a patch for it if you 
don't have time.
Cheers,

--
Julien Grall