linux.git/include/asm-mips/page.h, branch v2.6.26

CONFIG_HIGHPTE vs. sub-page page tables.

2008-02-08T17:22:42+00:00

Background: I've implemented 1K/2K page tables for s390.  These sub-page
page tables are required to properly support the s390 virtualization
instruction with KVM.  The SIE instruction requires that the page tables
have 256 page table entries (pte) followed by 256 page status table entries
(pgste).  The pgstes are only required if the process is using the SIE
instruction.  The pgstes are updated by the hardware and by the hypervisor
for a number of reasons, one of them is dirty and reference bit tracking.
To avoid wasting memory the standard pte table allocation should return
1K/2K (31/64 bit) and 2K/4K if the process is using SIE.

Problem: Page size on s390 is 4K, page table size is 1K or 2K.  That means
the s390 version for pte_alloc_one cannot return a pointer to a struct
page.  Trouble is that with the CONFIG_HIGHPTE feature on x86 pte_alloc_one
cannot return a pointer to a pte either, since that would require more than
32 bit for the return value of pte_alloc_one (and the pte * would not be
accessible since its not kmapped).

Solution: The only solution I found to this dilemma is a new typedef: a
pgtable_t.  For s390 pgtable_t will be a (pte *) - to be introduced with a
later patch.  For everybody else it will be a (struct page *).  The
additional problem with the initialization of the ptl lock and the
NR_PAGETABLE accounting is solved with a constructor pgtable_page_ctor and
a destructor pgtable_page_dtor.  The page table allocation and free
functions need to call these two whenever a page table page is allocated or
freed.  pmd_populate will get a pgtable_t instead of a struct page pointer.
 To get the pgtable_t back from a pmd entry that has been installed with
pmd_populate a new function pmd_pgtable is added.  It replaces the pmd_page
call in free_pte_range and apply_to_pte_range.

Signed-off-by: Martin Schwidefsky 
Cc: 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

Cleanup asm/{elf,page,user}.h: #ifdef KERNEL is no longer needed

2008-02-07T16:42:30+00:00

asm/elf.h, asm/page.h and asm/user.h don't export to userspace now, so we can
drop #ifdef __KERNEL__ for them.

[k.shutemov@gmail.com: remove #ifdef __KERNEL_]
Signed-off-by: Kirill A. Shutemov 
Reviewed-by: David Woodhouse 
Cc: 
Signed-off-by: Kirill A. Shutemov 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

[MIPS] checkfiles: Fix "need space after that ','" errors.

2007-10-11T22:46:15+00:00

Signed-off-by: Ralf Baechle

[MIPS] Optimize __alloc_zeroed_user_highpage implementation.

2007-10-11T22:46:12+00:00

Signed-off-by: Ralf Baechle

[MIPS] Rename CONFIG_BUILD_ELF64 into KBUILD_64BIT_SYM32

2007-10-11T22:45:59+00:00

This patch renames it for 3 reasons:

    - "CONFIG" pattern is used by Kconfig. Now this macro is
      no more defined by Kconfig but by Kbuild itself make this
      clear by translating "CONFIG" into "KBUILD".

    - "ELF32" word is improper because it is irrelevant to ELF
      format and it makes confusion with CONFIG_BOOT_ELF32. So
      translate it with SYM32.

    - Add "64BIT" part to make clear that this macro implies a
      64 bits kernel.

Signed-off-by: Franck Bui-Huu 
Signed-off-by: Ralf Baechle

[MIPS] Automatically set CONFIG_BUILD_ELF64

2007-10-11T22:45:59+00:00

We do not rely on user anymore to setup this config correctly.
Instead we make our choice depending on the load address.

If we want to force Kbuild to use ELF64 format whatever
the load address we can still do:

        $ make BUILD_ELF32=no

Signed-off-by: Franck Bui-Huu 
Signed-off-by: Ralf Baechle

[MIPS] Fix CONFIG_BUILD_ELF64 kernels with symbols in CKSEG0.

2007-09-27T22:19:16+00:00

The __pa() for those did assume that all symbols have XKPHYS values and
the math fails for any other address range.

Signed-off-by: Ralf Baechle

[MIPS] Fix PHYS_OFFSET for 64-bits kernels with 32-bits symbols

2007-07-10T16:32:57+00:00

The current implementation of __pa() for 64-bits kernels with 32-bits
symbols is broken. In this configuration, we need 2 values for
PAGE_OFFSET, one in XKPHYS and the other in CKSEG0 space.

When the value in CKSEG0 space is used, it doesn't take into account
of PHYS_OFFSET. Even worse we can't redefine this value.

The patch restores CPHYSADDR() but in __pa()'s implementation because
it removes the need of 2 PAGE_OFFSET.

OTOH, CPHYSADDR() is quite bad when dealing with mapped kernels. So
this patch assumes there's no need to deal with such kernel in 64-bits
world.

Signed-off-by: Franck Bui-Huu 
Signed-off-by: Ralf Baechle

[MIPS] Move PHY_OFFSET definition in spaces.h