pte_page

pte_page

[mdaljeet]

I use the 'pgt_offset', 'pmd_offset', 'pte_offset' and 'pte_page' inside a
module to get the physical address of a user space virtual address. The
physical address returned by 'pte_page' is not page aligned whereas the
virtual address was page aligned. Can somebody tell me the reason?

Also, can i use these functions to get the physical address of a kernel
virtual address using init_mm?

regards,
Daljeet.

Re: pte_page

[Brian Gerst]

mdaljeet@in.ibm.com wrote:
> 
> I use the 'pgt_offset', 'pmd_offset', 'pte_offset' and 'pte_page' inside a
> module to get the physical address of a user space virtual address. The
> physical address returned by 'pte_page' is not page aligned whereas the
> virtual address was page aligned. Can somebody tell me the reason?
> 
> Also, can i use these functions to get the physical address of a kernel
> virtual address using init_mm?

pte_page() returns the struct page * for the page, not the page
address.  To get the physical address of a page use __pa(virtaddr), but
this works if and only if the page is not highmem and not vmalloced.

-- 

						Brian Gerst

Re: pte_page

[Brian Gerst]

Brian Gerst wrote:
> 
> mdaljeet@in.ibm.com wrote:
> >
> > I use the 'pgt_offset', 'pmd_offset', 'pte_offset' and 'pte_page' inside a
> > module to get the physical address of a user space virtual address. The
> > physical address returned by 'pte_page' is not page aligned whereas the
> > virtual address was page aligned. Can somebody tell me the reason?
> >
> > Also, can i use these functions to get the physical address of a kernel
> > virtual address using init_mm?
> 
> pte_page() returns the struct page * for the page, not the page
> address.  To get the physical address of a page use __pa(virtaddr), but
> this works if and only if the page is not highmem and not vmalloced.

Oops, I forgot that you said user page...  __pa() only works for kernel
addresses.  I don't see any other macros that can help you.  Maybe one
of the VM gurus can add something here?

-- 

						Brian Gerst

Re: pte_page

[Ingo Molnar]

On Wed, 30 May 2001 mdaljeet@in.ibm.com wrote:

> I use the 'pgt_offset', 'pmd_offset', 'pte_offset' and 'pte_page'
> inside a module to get the physical address of a user space virtual
> address. The physical address returned by 'pte_page' is not page
> aligned whereas the virtual address was page aligned. Can somebody
> tell me the reason?

__pa(page_address(pte_page(pte))) is the address you want. [or
pte_val(*pte) & (PAGE_SIZE-1) on x86 but this is platform-dependent.]

> Also, can i use these functions to get the physical address of a
> kernel virtual address using init_mm?

nope. Eg. on x86 these functions only walk normal 4K page pagetables, they
do not walk 4MB pages correctly. (which are set up on pentiums and better
CPUs, unless mem=nopentium is specified.)

a kernel virtual address can be decoded by simply doing __pa(kaddr). If
the page is a highmem page [and you have the struct page pointer] then you
can do [(page-mem_map) << PAGE_SHIFT] to get the physical address, but
only on systems where mem_map[] starts at physical address 0.

	Ingo

Re: pte_page

[Pete Wyckoff]

mingo@elte.hu said:
> On Wed, 30 May 2001 mdaljeet@in.ibm.com wrote:
> > I use the 'pgt_offset', 'pmd_offset', 'pte_offset' and 'pte_page'
> > inside a module to get the physical address of a user space virtual
> > address. The physical address returned by 'pte_page' is not page
> > aligned whereas the virtual address was page aligned. Can somebody
> > tell me the reason?
> 
> __pa(page_address(pte_page(pte))) is the address you want. [or
> pte_val(*pte) & (PAGE_SIZE-1) on x86 but this is platform-dependent.]

Does this work on x86 non-kmapped highmem user pages too?  (i.e. is
page->virtual valid for every potential user page.)

		-- Pete

> > Also, can i use these functions to get the physical address of a
> > kernel virtual address using init_mm?
> 
> nope. Eg. on x86 these functions only walk normal 4K page pagetables, they
> do not walk 4MB pages correctly. (which are set up on pentiums and better
> CPUs, unless mem=nopentium is specified.)
> 
> a kernel virtual address can be decoded by simply doing __pa(kaddr). If
> the page is a highmem page [and you have the struct page pointer] then you
> can do [(page-mem_map) << PAGE_SHIFT] to get the physical address, but
> only on systems where mem_map[] starts at physical address 0.

Re: pte_page

[Ingo Molnar]

On Wed, 30 May 2001, Pete Wyckoff wrote:

> > __pa(page_address(pte_page(pte))) is the address you want. [or
> > pte_val(*pte) & (PAGE_SIZE-1) on x86 but this is platform-dependent.]
>
> Does this work on x86 non-kmapped highmem user pages too?  (i.e. is
> page->virtual valid for every potential user page.)

you are right, the highmem-compatible solution is to use page-mem_map as
the physical page index.

	Ingo

Re: pte_page

[mdaljeet]

I am doing a DMA from a card to system memory. The system memory "physical
address" is '0x104000'. I am doing this on x86 with kernel version 2.4.2.
Can this address be the address of a user space buffer?

Regards,
Daljeet.


|--------+----------------------->
|        |          Ingo Molnar  |
|        |          <mingo@elte.h|
|        |          u>           |
|        |                       |
|        |          05/30/01     |
|        |          11:09 PM     |
|        |          Please       |
|        |          respond to   |
|        |          mingo        |
|        |                       |
|--------+----------------------->
  >--------------------------------------------------------|
  |                                                        |
  |       To:     Pete Wyckoff <pw@osc.edu>                |
  |       cc:     Daljeet Maini/India/IBM@IBMIN,           |
  |       linux-kernel@vger.kernel.org                     |
  |       Subject:     Re: pte_page                        |
  >--------------------------------------------------------|






On Wed, 30 May 2001, Pete Wyckoff wrote:

> > __pa(page_address(pte_page(pte))) is the address you want. [or
> > pte_val(*pte) & (PAGE_SIZE-1) on x86 but this is platform-dependent.]
>
> Does this work on x86 non-kmapped highmem user pages too?  (i.e. is
> page->virtual valid for every potential user page.)

you are right, the highmem-compatible solution is to use page-mem_map as
the physical page index.

     Ingo