[PATCH] x86: fix pte allocation in "x86: introduce init_memory_mapping for 32bit"

[PATCH] x86: fix pte allocation in "x86: introduce init_memory_mapping for 32bit"

[Jeremy Fitzhardinge]

The patch "x86: introduce init_memory_mapping for 32bit" does not allocate 
enough space for PTEs if the CPU does not implement PSE.

Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
Cc: Yinghai Lu <yhlu.kernel@gmail.com>
---
 arch/x86/mm/init_32.c |    5 +++++
 1 file changed, 5 insertions(+)

===================================================================
--- a/arch/x86/mm/init_32.c
+++ b/arch/x86/mm/init_32.c
@@ -733,6 +733,11 @@
 	pmds = (end + PMD_SIZE - 1) >> PMD_SHIFT;
 	tables += PAGE_ALIGN(pmds * sizeof(pmd_t));
 
+	if (!cpu_has_pse) {
+		int ptes = (end + PAGE_SIZE - 1) >> PAGE_SHIFT;
+		tables += PAGE_ALIGN(ptes * sizeof(pte_t));
+	}
+
 	/*
 	 * RED-PEN putting page tables only on node 0 could
 	 * cause a hotspot and fill up ZONE_DMA. The page tables

Re: [PATCH] x86: fix pte allocation in "x86: introduce init_memory_mapping for 32bit"

[Yinghai Lu]

On Tue, Jun 24, 2008 at 2:32 PM, Jeremy Fitzhardinge <jeremy@goop.org> wrote:
> The patch "x86: introduce init_memory_mapping for 32bit" does not allocate
> enough space for PTEs if the CPU does not implement PSE.
>
> Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
> Cc: Yinghai Lu <yhlu.kernel@gmail.com>
> ---
> arch/x86/mm/init_32.c |    5 +++++
> 1 file changed, 5 insertions(+)
>
> ===================================================================
> --- a/arch/x86/mm/init_32.c
> +++ b/arch/x86/mm/init_32.c
> @@ -733,6 +733,11 @@
>        pmds = (end + PMD_SIZE - 1) >> PMD_SHIFT;
>        tables += PAGE_ALIGN(pmds * sizeof(pmd_t));
>
> +       if (!cpu_has_pse) {
> +               int ptes = (end + PAGE_SIZE - 1) >> PAGE_SHIFT;
> +               tables += PAGE_ALIGN(ptes * sizeof(pte_t));
> +       }
> +
>        /*
>         * RED-PEN putting page tables only on node 0 could
>         * cause a hotspot and fill up ZONE_DMA. The page tables
>
thanks.

my test system never use the area from table_start to table_end. and I
wonder why init_pgtable is that big..

YH

Re: [PATCH] x86: fix pte allocation in "x86: introduce init_memory_mapping for 32bit"

[Jeremy Fitzhardinge]

Yinghai Lu wrote:
> thanks.
>
> my test system never use the area from table_start to table_end. and I
> wonder why init_pgtable is that big..

This code looks just like the 64-bit code (I guess it's becoming prime 
for unification), and I'd just spent some time debugging a bug in Xen64 
caused by an overrun of the allocated area when creating non-PSE 
mappings of physical memory.  I didn't observe a specific problem with 
this code, but consistency and pattern matching made it seem like a 
necessary addition.  It only ends up reserving the amount of space 
actually used, so there's no wasted memory.

If you use PSE to create the mapping and have enough mapped pages anyway 
(small kernel), then you may not end up using any of this space.

    J

Re: [PATCH] x86: fix pte allocation in "x86: introduce init_memory_mapping for 32bit"

[Ingo Molnar]

* Jeremy Fitzhardinge <jeremy@goop.org> wrote:

> The patch "x86: introduce init_memory_mapping for 32bit" does not 
> allocate enough space for PTEs if the CPU does not implement PSE.

applied to tip/x86/unify-setup - thanks Jeremy!

	Ingo