1. Patchew
  2. linux
  3. [PATCH v3 0/7] Cleanup free_pages() misuse
  4. View patch

[PATCH v3 3/7] x86: Stop calling page_address() in free_pages()

Vishal Moola (Oracle) posted 7 patches 5 months, 1 week ago
[PATCH v3 3/7] x86: Stop calling page_address() in free_pages()
Posted by Vishal Moola (Oracle) 5 months, 1 week ago 
free_pages() should be used when we only have a virtual address. We
should call __free_pages() directly on our page instead.

Signed-off-by: Vishal Moola (Oracle) <vishal.moola@gmail.com>
Acked-by: Dave Hansen <dave.hansen@linux.intel.com>
---
 arch/x86/mm/init_64.c          | 2 +-
 arch/x86/platform/efi/memmap.c | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/arch/x86/mm/init_64.c b/arch/x86/mm/init_64.c
index b9426fce5f3e..0e4270e20fad 100644
--- a/arch/x86/mm/init_64.c
+++ b/arch/x86/mm/init_64.c
@@ -1031,7 +1031,7 @@ static void __meminit free_pagetable(struct page *page, int order)
 		free_reserved_pages(page, nr_pages);
 #endif
 	} else {
-		free_pages((unsigned long)page_address(page), order);
+		__free_pages(page, order);
 	}
 }
 
diff --git a/arch/x86/platform/efi/memmap.c b/arch/x86/platform/efi/memmap.c
index 061b8ecc71a1..023697c88910 100644
--- a/arch/x86/platform/efi/memmap.c
+++ b/arch/x86/platform/efi/memmap.c
@@ -42,7 +42,7 @@ void __init __efi_memmap_free(u64 phys, unsigned long size, unsigned long flags)
 		struct page *p = pfn_to_page(PHYS_PFN(phys));
 		unsigned int order = get_order(size);
 
-		free_pages((unsigned long) page_address(p), order);
+		__free_pages(p, order);
 	}
 }
 
-- 
2.51.0
Re: [PATCH v3 3/7] x86: Stop calling page_address() in free_pages()
Posted by Mike Rapoport 5 months, 1 week ago 
On Wed, Sep 03, 2025 at 11:59:17AM -0700, Vishal Moola (Oracle) wrote:
> free_pages() should be used when we only have a virtual address. We
> should call __free_pages() directly on our page instead.
> 
> Signed-off-by: Vishal Moola (Oracle) <vishal.moola@gmail.com>
> Acked-by: Dave Hansen <dave.hansen@linux.intel.com>
> ---
>  arch/x86/mm/init_64.c          | 2 +-
>  arch/x86/platform/efi/memmap.c | 2 +-
>  2 files changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/arch/x86/mm/init_64.c b/arch/x86/mm/init_64.c
> index b9426fce5f3e..0e4270e20fad 100644
> --- a/arch/x86/mm/init_64.c
> +++ b/arch/x86/mm/init_64.c
> @@ -1031,7 +1031,7 @@ static void __meminit free_pagetable(struct page *page, int order)
>  		free_reserved_pages(page, nr_pages);
>  #endif
>  	} else {
> -		free_pages((unsigned long)page_address(page), order);
> +		__free_pages(page, order);
>  	}
>  }
>  
> diff --git a/arch/x86/platform/efi/memmap.c b/arch/x86/platform/efi/memmap.c
> index 061b8ecc71a1..023697c88910 100644
> --- a/arch/x86/platform/efi/memmap.c
> +++ b/arch/x86/platform/efi/memmap.c
> @@ -42,7 +42,7 @@ void __init __efi_memmap_free(u64 phys, unsigned long size, unsigned long flags)
>  		struct page *p = pfn_to_page(PHYS_PFN(phys));
>  		unsigned int order = get_order(size);
>  
> -		free_pages((unsigned long) page_address(p), order);

Could be just free_pages((unsigned long)phys_to_virt(phys), order), then
the page is not needed at all.

> +		__free_pages(p, order);
>  	}
>  }
>  
> -- 
> 2.51.0
> 
> 

-- 
Sincerely yours,
Mike.
Re: [PATCH v3 3/7] x86: Stop calling page_address() in free_pages()
Posted by Mike Rapoport 5 months, 1 week ago 
On Thu, Sep 04, 2025 at 02:51:14PM +0300, Mike Rapoport wrote:
> On Wed, Sep 03, 2025 at 11:59:17AM -0700, Vishal Moola (Oracle) wrote:
> > free_pages() should be used when we only have a virtual address. We
> > should call __free_pages() directly on our page instead.
> > 
> > Signed-off-by: Vishal Moola (Oracle) <vishal.moola@gmail.com>
> > Acked-by: Dave Hansen <dave.hansen@linux.intel.com>
> > ---
> >  arch/x86/mm/init_64.c          | 2 +-
> >  arch/x86/platform/efi/memmap.c | 2 +-
> >  2 files changed, 2 insertions(+), 2 deletions(-)
> > 
> > diff --git a/arch/x86/mm/init_64.c b/arch/x86/mm/init_64.c
> > index b9426fce5f3e..0e4270e20fad 100644
> > --- a/arch/x86/mm/init_64.c
> > +++ b/arch/x86/mm/init_64.c
> > @@ -1031,7 +1031,7 @@ static void __meminit free_pagetable(struct page *page, int order)
> >  		free_reserved_pages(page, nr_pages);
> >  #endif
> >  	} else {
> > -		free_pages((unsigned long)page_address(page), order);
> > +		__free_pages(page, order);
> >  	}
> >  }
> >  
> > diff --git a/arch/x86/platform/efi/memmap.c b/arch/x86/platform/efi/memmap.c
> > index 061b8ecc71a1..023697c88910 100644
> > --- a/arch/x86/platform/efi/memmap.c
> > +++ b/arch/x86/platform/efi/memmap.c
> > @@ -42,7 +42,7 @@ void __init __efi_memmap_free(u64 phys, unsigned long size, unsigned long flags)
> >  		struct page *p = pfn_to_page(PHYS_PFN(phys));
> >  		unsigned int order = get_order(size);
> >  
> > -		free_pages((unsigned long) page_address(p), order);
> 
> Could be just free_pages((unsigned long)phys_to_virt(phys), order), then
> the page is not needed at all.

Or even __free_pages(phys_to_page(phys), order);
 
> > +		__free_pages(p, order);
> >  	}
> >  }
> >  
> > -- 
> > 2.51.0
> > 
> > 
> 
> -- 
> Sincerely yours,
> Mike.

-- 
Sincerely yours,
Mike.
Re: [PATCH v3 3/7] x86: Stop calling page_address() in free_pages()
Posted by Vishal Moola (Oracle) 5 months ago 
On Thu, Sep 04, 2025 at 02:54:24PM +0300, Mike Rapoport wrote:
> On Thu, Sep 04, 2025 at 02:51:14PM +0300, Mike Rapoport wrote:
> > On Wed, Sep 03, 2025 at 11:59:17AM -0700, Vishal Moola (Oracle) wrote:
> > > free_pages() should be used when we only have a virtual address. We
> > > should call __free_pages() directly on our page instead.
> > > 
> > > Signed-off-by: Vishal Moola (Oracle) <vishal.moola@gmail.com>
> > > Acked-by: Dave Hansen <dave.hansen@linux.intel.com>
> > > ---
> > >  arch/x86/mm/init_64.c          | 2 +-
> > >  arch/x86/platform/efi/memmap.c | 2 +-
> > >  2 files changed, 2 insertions(+), 2 deletions(-)
> > > 
> > > diff --git a/arch/x86/mm/init_64.c b/arch/x86/mm/init_64.c
> > > index b9426fce5f3e..0e4270e20fad 100644
> > > --- a/arch/x86/mm/init_64.c
> > > +++ b/arch/x86/mm/init_64.c
> > > @@ -1031,7 +1031,7 @@ static void __meminit free_pagetable(struct page *page, int order)
> > >  		free_reserved_pages(page, nr_pages);
> > >  #endif
> > >  	} else {
> > > -		free_pages((unsigned long)page_address(page), order);
> > > +		__free_pages(page, order);
> > >  	}
> > >  }
> > >  
> > > diff --git a/arch/x86/platform/efi/memmap.c b/arch/x86/platform/efi/memmap.c
> > > index 061b8ecc71a1..023697c88910 100644
> > > --- a/arch/x86/platform/efi/memmap.c
> > > +++ b/arch/x86/platform/efi/memmap.c
> > > @@ -42,7 +42,7 @@ void __init __efi_memmap_free(u64 phys, unsigned long size, unsigned long flags)
> > >  		struct page *p = pfn_to_page(PHYS_PFN(phys));
> > >  		unsigned int order = get_order(size);
> > >  
> > > -		free_pages((unsigned long) page_address(p), order);
> > 
> > Could be just free_pages((unsigned long)phys_to_virt(phys), order), then
> > the page is not needed at all.
> 
> Or even __free_pages(phys_to_page(phys), order);

Right. It actually looks like we could inline this whole block if we
really wanted to...

__free_pages(phys_to_page(phys), get_order(size));

Should I send a fixup (or v4) with this change?

Patchew 2.1-devel-b67e4c2

© 2016 - 2026 Red Hat, Inc.