Re: [PATCH 2/3] mm: Convert vm_insert_pfn_prot to vmf_insert_pfn_prot

[Andy Lutomirski]

On Mon, Jan 25, 2016 at 9:25 AM, Matthew Wilcox
<matthew.r.wilcox@xxxxxxxxx> wrote:
> From: Matthew Wilcox <willy@xxxxxxxxxxxxxxx>
>
> Other than the name, the vmf_ version takes a pfn_t parameter, and
> returns a VM_FAULT_ code suitable for returning from a fault handler.
>
> This patch also prevents vm_insert_pfn() from returning -EBUSY.
> This is a good thing as several callers handled it incorrectly (and
> none intentionally treat -EBUSY as a different case from 0).
>
> Signed-off-by: Matthew Wilcox <willy@xxxxxxxxxxxxxxx>
> ---
>  arch/x86/entry/vdso/vma.c |  6 +++---
>  include/linux/mm.h        |  4 ++--
>  mm/memory.c               | 31 ++++++++++++++++++-------------
>  3 files changed, 23 insertions(+), 18 deletions(-)
>
> diff --git a/arch/x86/entry/vdso/vma.c b/arch/x86/entry/vdso/vma.c
> index 7c912fe..660bb69 100644
> --- a/arch/x86/entry/vdso/vma.c
> +++ b/arch/x86/entry/vdso/vma.c
> @@ -9,6 +9,7 @@
>  #include <linux/sched.h>
>  #include <linux/slab.h>
>  #include <linux/init.h>
> +#include <linux/pfn_t.h>
>  #include <linux/random.h>
>  #include <linux/elf.h>
>  #include <linux/cpu.h>
> @@ -131,10 +132,9 @@ static int vvar_fault(const struct vm_special_mapping *sm,
>         } else if (sym_offset == image->sym_hpet_page) {
>  #ifdef CONFIG_HPET_TIMER
>                 if (hpet_address && vclock_was_used(VCLOCK_HPET)) {
> -                       ret = vm_insert_pfn_prot(
> -                               vma,
> +                       return vmf_insert_pfn_prot(vma,
>                                 (unsigned long)vmf->virtual_address,
> -                               hpet_address >> PAGE_SHIFT,
> +                               phys_to_pfn_t(hpet_address, PFN_DEV),
>                                 pgprot_noncached(PAGE_READONLY));
>                 }

This would be even nicer if you added vmf_insert_pfn as well :)

--Andy