[PATCH v2 33/34] ARM: mmu32: Skip reserved ranges during initialization

[Sascha Hauer]

On Wed, May 17, 2023 at 04:43:39PM +0200, Ahmad Fatoum wrote:
> On 17.05.23 11:03, Sascha Hauer wrote:
> > The early MMU code now uses pages to map the OP-TEE area non executable.
> > This mapping is overwritten with sections in barebox proper. Refrain
> > from doing so by using arch_remap_range() and bypassing reserved areas.
> > 
> > Signed-off-by: Sascha Hauer <s.hauer at pengutronix.de>
> > ---
> >  arch/arm/cpu/mmu_32.c | 14 +++++++++++---
> >  1 file changed, 11 insertions(+), 3 deletions(-)
> > 
> > diff --git a/arch/arm/cpu/mmu_32.c b/arch/arm/cpu/mmu_32.c
> > index 705d27a045..47711bed35 100644
> > --- a/arch/arm/cpu/mmu_32.c
> > +++ b/arch/arm/cpu/mmu_32.c
> > @@ -522,9 +522,17 @@ void __mmu_init(bool mmu_on)
> >  	vectors_init();
> 
> Took me a while to parse the below code, so I think a comment may be apt:

That's just fair, it likely tool me even longer to write it ;)

> 
> /*
>  * Early mmu init will have mapped everything but the initial memory area
>  * (excluding final OPTEE_SIZE bytes) uncached. We have now discovered
>  * all memory banks, so let's map all pages, excluding reserved memory areas,
>  * cacheable and executable.
>  */
> 
> >  
> >  	for_each_memory_bank(bank) {
> > -		create_sections(bank->start, bank->start + bank->size - 1,
> > -				PMD_SECT_DEF_CACHED);
> > -		__mmu_cache_flush();
> > +		struct resource *rsv;
> > +		resource_size_t pos;
> > +
> > +		pos = bank->start;
> > +
> > +		for_each_reserved_region(bank, rsv) {
> > +			arch_remap_range((void *)pos, rsv->start - pos, MAP_CACHED);
> > +			pos = rsv->end + 1;
> > +		}
> > +
> > +		arch_remap_range((void *)pos, bank->start + bank->size - pos, MAP_CACHED);
> 
> I am a bit bothered by the asymmetry here: Reserved regions in the extra memory banks
> will be initially uncached (because outside initmem), so this loop does the correct thing.
> 
> For reserved regions within the initial memory, only the OP-TEE region would be uncached,
> everything else would just be requested, but is still mapped cacheable.
> 
> IMO, that's surprising behavior.

I agree. What would you suggest. Map all reserved regions uncached?

Sascha

-- 
Pengutronix e.K.                           |                             |
Steuerwalder Str. 21                       | http://www.pengutronix.de/  |
31137 Hildesheim, Germany                  | Phone: +49-5121-206917-0    |
Amtsgericht Hildesheim, HRA 2686           | Fax:   +49-5121-206917-5555 |