[PATCH V6] ARM: LPAE: Fix mapping in alloc_init_section for unaligned addresses

[Christoffer Dall]

On Sat, Mar 16, 2013 at 10:05 PM, Sricharan R <r.sricharan at ti.com> wrote:
> From: R Sricharan <r.sricharan at ti.com>
>
> With LPAE enabled, alloc_init_section() does not map the entire
> address space for unaligned addresses.
>
> The issue also reproduced with CMA + LPAE. CMA tries to map 16MB
> with page granularity mappings during boot. alloc_init_pte()
> is called and out of 16MB, only 2MB gets mapped and rest remains
> unaccessible.
>
> Because of this OMAP5 boot is broken with CMA + LPAE enabled.
> Fix the issue by ensuring that the entire addresses are
> mapped.
>
> Signed-off-by: R Sricharan <r.sricharan at ti.com>
> Cc: Catalin Marinas <catalin.marinas at arm.com>
> Cc: Christoffer Dall <chris at cloudcar.com>
> Cc: Russell King <linux at arm.linux.org.uk>
> Cc: Santosh Shilimkar <santosh.shilimkar at ti.com>
> Tested-by: Laura Abbott <lauraa at codeaurora.org>
> Acked-by: Catalin Marinas <catalin.marinas at arm.com>
> ---
>  [V2] Moved the loop to alloc_init_pte as per Russell's
>      feedback and changed the subject accordingly.
>      Using PMD_XXX instead of SECTION_XXX to avoid
>      different loop increments with/without LPAE.
>
>  [v3] Removed the dummy variable phys and updated
>       the commit log for CMA case.
>
>  [v4] Resending with updated change log and
>       updating the tags.
>
>  [v5] Renamed alloc_init_section to alloc_init_pmd
>       and moved the loop back there. Also introduced
>       map_init_section as per Catalin's comments.
>
>  [v6] Corrected tags and updated the comments for code.
>
>  arch/arm/mm/mmu.c |   73 ++++++++++++++++++++++++++++++++++-------------------
>  1 file changed, 47 insertions(+), 26 deletions(-)
>
> diff --git a/arch/arm/mm/mmu.c b/arch/arm/mm/mmu.c
> index e95a996..7897894 100644
> --- a/arch/arm/mm/mmu.c
> +++ b/arch/arm/mm/mmu.c
> @@ -598,39 +598,60 @@ static void __init alloc_init_pte(pmd_t *pmd, unsigned long addr,
>         } while (pte++, addr += PAGE_SIZE, addr != end);
>  }
>
> -static void __init alloc_init_section(pud_t *pud, unsigned long addr,
> -                                     unsigned long end, phys_addr_t phys,
> -                                     const struct mem_type *type)
> +static void __init map_init_section(pmd_t *pmd, unsigned long addr,
> +                       unsigned long end, phys_addr_t phys,
> +                       const struct mem_type *type)
>  {
> -       pmd_t *pmd = pmd_offset(pud, addr);
> -
> +#ifndef CONFIG_ARM_LPAE
>         /*
> -        * Try a section mapping - end, addr and phys must all be aligned
> -        * to a section boundary.  Note that PMDs refer to the individual
> -        * L1 entries, whereas PGDs refer to a group of L1 entries making
> -        * up one logical pointer to an L2 table.
> +        * In classic MMU format, puds and pmds are folded in to
> +        * the pgds. pmd_offset gives the PGD entry. PGDs refer to a
> +        * group of L1 entries making up one logical pointer to
> +        * an L2 table (2MB), where as PMDs refer to the individual
> +        * L1 entries (1MB). Hence increment to get the correct
> +        * offset for odd 1MB sections.
> +        * (See arch/arm/include/asm/pgtable-2level.h)
>          */
> -       if (type->prot_sect && ((addr | end | phys) & ~SECTION_MASK) == 0) {
> -               pmd_t *p = pmd;
> -
> -#ifndef CONFIG_ARM_LPAE
> -               if (addr & SECTION_SIZE)
> -                       pmd++;
> +       if (addr & SECTION_SIZE)
> +               pmd++;
>  #endif
> +       do {
> +               *pmd = __pmd(phys | type->prot_sect);
> +               phys += SECTION_SIZE;
> +       } while (pmd++, addr += SECTION_SIZE, addr != end);
>
> -               do {
> -                       *pmd = __pmd(phys | type->prot_sect);
> -                       phys += SECTION_SIZE;
> -               } while (pmd++, addr += SECTION_SIZE, addr != end);
> +       flush_pmd_entry(pmd);
> +}
>
> -               flush_pmd_entry(p);
> -       } else {
> +static void __init alloc_init_pmd(pud_t *pud, unsigned long addr,
> +                                     unsigned long end, phys_addr_t phys,
> +                                     const struct mem_type *type)
> +{
> +       pmd_t *pmd = pmd_offset(pud, addr);
> +       unsigned long next;
> +
> +       do {
>                 /*
> -                * No need to loop; pte's aren't interested in the
> -                * individual L1 entries.
> +                * With LPAE, we must loop over to map
> +                * all the pmds for the given range.
>                  */
> -               alloc_init_pte(pmd, addr, end, __phys_to_pfn(phys), type);
> -       }
> +               next = pmd_addr_end(addr, end);
> +
> +               /*
> +                * Try a section mapping - addr, next and phys must all be
> +                * aligned to a section boundary.
> +                */
> +               if (type->prot_sect &&
> +                               ((addr | next | phys) & ~SECTION_MASK) == 0) {
> +                       map_init_section(pmd, addr, next, phys, type);
> +               } else {
> +                       alloc_init_pte(pmd, addr, next,
> +                                               __phys_to_pfn(phys), type);
> +               }
> +
> +               phys += next - addr;
> +
> +       } while (pmd++, addr = next, addr != end);
>  }
>
>  static void __init alloc_init_pud(pgd_t *pgd, unsigned long addr,
> @@ -641,7 +662,7 @@ static void __init alloc_init_pud(pgd_t *pgd, unsigned long addr,
>
>         do {
>                 next = pud_addr_end(addr, end);
> -               alloc_init_section(pud, addr, next, phys, type);
> +               alloc_init_pmd(pud, addr, next, phys, type);
>                 phys += next - addr;
>         } while (pud++, addr = next, addr != end);
>  }
> --
> 1.7.9.5
>
Acked-by: Christoffer Dall <chris at cloudcar.com>