ARM: LPAE: Invalidate the TLB for module addresses during translation…

… fault

During the free_pgtables() call all user and modules/pkmap entries are
removed. If a translation fault for the modules/pkmap area occurs before
we switched away from the current pgd, do_translation_fault() would copy
the init_mm pud into the user pud.

There is a small window between pud_clear() and pmd_free_tlb() in
free_pmd_range() where the pud entry was cleared but the TLB has not
been invalidated yet and the CPU may have cached the original (valid)
pud entry in the TLB. A scenario like below would get stuck in
continuous prefetch abort:

1. Current process exiting. The modules pmd entries not populated
2. exit_mmap() -> ... -> pmd_free_tlb()
3. pud_clear() for the 1GB pud containing user stack and modules (no TLB
   invalidation yet)
4. Interrupt -> module interrupt routine
5. Level 2 (pmd) translation fault occurs when executing the module
   interrupt routine. The CPU previously cached (TLB) the old valid pud
   value for the modules area, so we don't get a level 1 translation
   fault
6. do_translation fault() copies the pud_k into the pud
7. Linux returns to the faulty instruction. Goes back to 5

At point 7, since the CPU still has the old pud value, it goes back to
point 5 and never gets out of this loop. With this patch, the stale pud
TLB entry is invalidated after point 6 and the subsequent prefetch abort
doesn't occur.

Reported-by: Tony Thompson <Anthony.Thompson@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>

UBUNTU: Ubuntu-3.0.0-1280.6

Signed-off-by: Sebastien Jan <s-jan@ti.com>

UBUNTU: Ubuntu-3.0.0-1280.5

Signed-off-by: Sebastien Jan <s-jan@ti.com>

UBUNTU: Ubuntu-3.0.0-1280.4

Signed-off-by: Sebastien Jan <s-jan@ti.com>

UBUNTU: Ubuntu-3.0.0-1280.3

Signed-off-by: Sebastien Jan <s-jan@ti.com>

UBUNTU: Ubuntu-3.0.0-1280.2

Signed-off-by: Sebastien Jan <s-jan@ti.com>

UBUNTU: Ubuntu-3.0.0-1280.1

Signed-off-by: Sebastien Jan <s-jan@ti.com>

UBUNTU: Ubuntu-3.0.0-1280.1

Signed-off-by: Sebastien Jan <s-jan@ti.com>

OMAP: DSS2: DISPC: Fix minimum PCD value

The current driver had a hardcoded minimum value of 2 for pixel clock
divisor (PCD). This doesn't seem to be right.

OMAP4 TRM says that PCD can be 1 when not downscaling, and inverted
pixel clock (IPC) is off.

OMAP3 TRM says the same, but also in the register descriptions that PCD
value 1 is invalid.

OMAP2 TRM says PCD 2 is the minimum.

OMAP2 is still untested, but for both OMAP3 and OMAP4 PCD of 1 seems to
work fine.

This patch adds a new DSS feature, FEAT_PARAM_DSS_PCD, which is used to
find the minimum and maximum PCD. The minimum is set to 2 for OMAP2, and
1 for OMAP3/4.

Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>

Linux 3.1-rc4