[PATCH] x86, mm: Avoid step_size overflow and add comments

[PATCH] x86, mm: Avoid step_size overflow and add comments

[Yinghai Lu]

Current code
1. shift step_size directly, and this could have overflow problem
   after we shift step_size several times.
2. use MACRO to have shift to set to 5, but there is not
   explanation about selection.

We can fix them with:
1. Add get_new_step_size(), and check overflow by comparing new_size
   with old step_size.
2. add comment in it about why we are using 5.

That will make code more readable while not increasing size of
init_mem_mapping().

-v2: According to Ingo, update changelog and comments.

Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Reviewed-by: Tang Chen <tangchen@cn.fujitsu.com>
Tested-by: Tang Chen <tangchen@cn.fujitsu.com>

---
 arch/x86/mm/init.c |   21 ++++++++++++++++++---
 1 file changed, 18 insertions(+), 3 deletions(-)

Index: linux-2.6/arch/x86/mm/init.c
===================================================================
--- linux-2.6.orig/arch/x86/mm/init.c
+++ linux-2.6/arch/x86/mm/init.c
@@ -399,8 +399,23 @@ static unsigned long __init init_range_m
 	return mapped_ram_size;
 }
 
-/* (PUD_SHIFT-PMD_SHIFT)/2 */
-#define STEP_SIZE_SHIFT 5
+static unsigned long __init get_new_step_size(unsigned long step_size)
+{
+	/*
+	 * initial mapped size is PMD_SIZE (2M).
+	 * We can not set step_size to be PUD_SIZE (1G) yet.
+	 * In worse case, when we cross the 1G boundary, and
+	 * PG_LEVEL_2M is not set, we will need 1+1+512 pages (2M + 8k)
+	 * to map 1G range with PTE. Use 5 as shift for now.
+	 */
+	unsigned long new_step_size = step_size << 5;
+
+	if (new_step_size > step_size)
+		step_size = new_step_size;
+
+	return  step_size;
+}
+
 void __init init_mem_mapping(void)
 {
 	unsigned long end, real_end, start, last_start;
@@ -449,7 +464,7 @@ void __init init_mem_mapping(void)
 		min_pfn_mapped = last_start >> PAGE_SHIFT;
 		/* only increase step_size after big range get mapped */
 		if (new_mapped_ram_size > mapped_ram_size)
-			step_size <<= STEP_SIZE_SHIFT;
+			step_size = get_new_step_size(step_size);
 		mapped_ram_size += new_mapped_ram_size;
 	}
 

Re: [PATCH] x86, mm: Avoid step_size overflow and add comments

[Ingo Molnar]

* Yinghai Lu <yinghai@kernel.org> wrote:

> Current code

"Current code has two problems:"

> 1. shift step_size directly, and this could have overflow problem
>    after we shift step_size several times.

The most important piece of information is not explained: 
under what circumstances can this happen currently, what 
are the symptoms if this happens, roughly what kind of 
systems are affected, etc?

I mean the kind of information you'd put into a BIOS errata 
description or into any software changelog. You wouldn't 
primarily write about the mechanics of the change, but 
about the _actual impact_ of the change, to end users, 
right?

> 2. use MACRO to have shift to set to 5, but there is not
>    explanation about selection.
> 
> We can fix them with:
> 1. Add get_new_step_size(), and check overflow by comparing new_size
>    with old step_size.
> 2. add comment in it about why we are using 5.
> 
> That will make code more readable while not increasing size of
> init_mem_mapping().

Well, it gets not just more readable at a small size 
increase, but also the above unexplained overflow bug will 
be fixed as well, right?

Thanks,

	Ingo