[PATCH 05/11] powerpc/8xx: Optimise access to swapper_pg_dir

[David Laight]

From: leroy christophe
> Le 06/01/2015 13:08, David Laight a écrit :
> >> On Tue, 2014-12-16 at 16:03 +0100, Christophe Leroy wrote:
> >>> All accessed to PGD entries are done via 0(r11).
> >>> By using lower part of swapper_pg_dir as load index to r11, we can remove the
> >>> ori instruction.
> >>>
> >>> Signed-off-by: Christophe Leroy <christophe.leroy at c-s.fr>
> >> Nice :)
> >> Acked-by: Joakim Tjernlund <joakim.tjernlund at transmode.se>
> >>
> >>> ---
> >>>   arch/powerpc/kernel/head_8xx.S | 22 ++++++++++------------
> >>>   1 file changed, 10 insertions(+), 12 deletions(-)
> >>>
> >>> diff --git a/arch/powerpc/kernel/head_8xx.S b/arch/powerpc/kernel/head_8xx.S
> >>> index ae05f28..aa45225 100644
> >>> --- a/arch/powerpc/kernel/head_8xx.S
> >>> +++ b/arch/powerpc/kernel/head_8xx.S
> >>> @@ -322,13 +322,12 @@ InstructionTLBMiss:
> >>>          mfspr   r11, SPRN_M_TW/* Get level 1 table base address */
> >>>   #ifdef CONFIG_MODULES
> >>>          beq     3f
> >>> -       lis     r11, (swapper_pg_dir-PAGE_OFFSET)@h
> >>> -       ori     r11, r11, (swapper_pg_dir-PAGE_OFFSET)@l
> >>> +       lis     r11, (swapper_pg_dir-PAGE_OFFSET)@ha
> >>>   3:
> >>>   #endif
> >>>          /* Insert level 1 index */
> >>>          rlwimi  r11, r10, 32 - ((PAGE_SHIFT - 2) << 1), (PAGE_SHIFT - 2) << 1, 29
> >>> -       lwz     r11, 0(r11)/* Get the level 1 entry */
> >>> +       lwz     r11, (swapper_pg_dir-PAGE_OFFSET)@l(r11)/* Get the level 1 entry */
> >>>
> > On the face of it that fragment doesn't look right when CONFIG_MODULES is undefined.
> >
> > 	David
> >
> I'm not sure I understand what you mean.
> 
> The other part of the patch adds the following:
> +    li    r5, (swapper_pg_dir-PAGE_OFFSET)@l
> +    sub    r4, r4, r5
> 
> r4 is the value put into SPRN_M_TW, so I don't see what may be wrong.

Ok, but it is completely non-obvious.
You probably need to change some comments somewhere about what is held
in SPRN_M_TW - since it is no longer the L1 base address, but that value
offset by some 'random' amount.

	David