Re: [PATCH v3 3/4] arm64: use both ZONE_DMA and ZONE_DMA32

From: Nicolas Saenz Julienne
Date: Thu Sep 05 2019 - 13:28:56 EST


On Thu, 2019-09-05 at 18:19 +0100, Catalin Marinas wrote:
> On Mon, Sep 02, 2019 at 04:10:41PM +0200, Nicolas Saenz Julienne wrote:
> > diff --git a/arch/arm64/mm/init.c b/arch/arm64/mm/init.c
> > index 8956c22634dd..f02a4945aeac 100644
> > --- a/arch/arm64/mm/init.c
> > +++ b/arch/arm64/mm/init.c
> > @@ -50,6 +50,13 @@
> > s64 memstart_addr __ro_after_init = -1;
> > EXPORT_SYMBOL(memstart_addr);
> >
> > +/*
> > + * We create both ZONE_DMA and ZONE_DMA32. ZONE_DMA covers the first 1G of
> > + * memory as some devices, namely the Raspberry Pi 4, have peripherals with
> > + * this limited view of the memory. ZONE_DMA32 will cover the rest of the
> > 32
> > + * bit addressable memory area.
> > + */
> > +phys_addr_t arm64_dma_phys_limit __ro_after_init;
> > phys_addr_t arm64_dma32_phys_limit __ro_after_init;
> >
> > #ifdef CONFIG_KEXEC_CORE
> > @@ -164,9 +171,9 @@ static void __init reserve_elfcorehdr(void)
> > }
> > #endif /* CONFIG_CRASH_DUMP */
> > /*
> > - * Return the maximum physical address for ZONE_DMA32 (DMA_BIT_MASK(32)).
> > It
> > - * currently assumes that for memory starting above 4G, 32-bit devices will
> > - * use a DMA offset.
> > + * Return the maximum physical address for ZONE_DMA32 (DMA_BIT_MASK(32))
> > and
> > + * ZONE_DMA (DMA_BIT_MASK(30)) respectively. It currently assumes that for
> > + * memory starting above 4G, 32-bit devices will use a DMA offset.
> > */
> > static phys_addr_t __init max_zone_dma32_phys(void)
> > {
> > @@ -174,12 +181,23 @@ static phys_addr_t __init max_zone_dma32_phys(void)
> > return min(offset + (1ULL << 32), memblock_end_of_DRAM());
> > }
> >
> > +static phys_addr_t __init max_zone_dma_phys(void)
> > +{
> > + phys_addr_t offset = memblock_start_of_DRAM() & GENMASK_ULL(63, 32);
> > +
> > + return min(offset + (1ULL << ARCH_ZONE_DMA_BITS),
> > + memblock_end_of_DRAM());
> > +}
>
> I think we could squash these two functions into a single one with a
> "bits" argument that is either 32 or ARCH_ZONE_DMA_BITS.

Hi Catalin, thanks for the review.

Agree, it'll look nicer.

> > +
> > #ifdef CONFIG_NUMA
> >
> > static void __init zone_sizes_init(unsigned long min, unsigned long max)
> > {
> > unsigned long max_zone_pfns[MAX_NR_ZONES] = {0};
> >
> > +#ifdef CONFIG_ZONE_DMA
> > + max_zone_pfns[ZONE_DMA] = PFN_DOWN(arm64_dma_phys_limit);
> > +#endif
> > #ifdef CONFIG_ZONE_DMA32
> > max_zone_pfns[ZONE_DMA32] = PFN_DOWN(arm64_dma32_phys_limit);
> > #endif
> > @@ -195,13 +213,17 @@ static void __init zone_sizes_init(unsigned long min,
> > unsigned long max)
> > struct memblock_region *reg;
> > unsigned long zone_size[MAX_NR_ZONES], zhole_size[MAX_NR_ZONES];
> > unsigned long max_dma32 = min;
> > + unsigned long max_dma = min;
> >
> > memset(zone_size, 0, sizeof(zone_size));
> >
> > - /* 4GB maximum for 32-bit only capable devices */
> > +#ifdef CONFIG_ZONE_DMA
> > + max_dma = PFN_DOWN(arm64_dma_phys_limit);
> > + zone_size[ZONE_DMA] = max_dma - min;
> > +#endifmax_dma32
> > #ifdef CONFIG_ZONE_DMA32
> > max_dma32 = PFN_DOWN(arm64_dma32_phys_limit);
> > - zone_size[ZONE_DMA32] = max_dma32 - min;
> > + zone_size[ZONE_DMA32] = max_dma32 - max_dma;
> > #endif
> > zone_size[ZONE_NORMAL] = max - max_dma32;
>
> Does this still work if we have ZONE_DMA32 disabled but ZONE_DMA
> enabled? You could use a max(max_dma32, max_dma) or just update
> max_dma32 to max_dma in the CONFIG_ZONE_DMA block.

You're right, I missed that scenario. I'll fix it and give it a test for the
next series.

Regards,
Nicolas

Attachment: signature.asc
Description: This is a digitally signed message part