[PATCH 1/1] mmap.c: find_vma: remove if(mm) check

[PATCH 1/1] mmap.c: find_vma: remove if(mm) check

[Kautuk Consul]

find_vma is called from kernel code where it is absolutely
sure that the mm_struct arg being passed to it is non-NULL.

Remove the if(mm) check.
This will also serve the purpose of mandating that the execution
context(user-mode/kernel-mode) be known before find_vma is called.

Also fixed 2 checkpatch.pl errors in the declaration
of the rb_node and vma_tmp local variables.

I have tested this patch on my x86 PC and there are no crashes
due to this in the course of normal desktop execution.

Signed-off-by: Kautuk Consul <consul.kautuk@gmail.com>
---
 mm/mmap.c |   54 ++++++++++++++++++++++++++----------------------------
 1 files changed, 26 insertions(+), 28 deletions(-)

diff --git a/mm/mmap.c b/mm/mmap.c
index a7bf6a3..2b2fe67 100644
--- a/mm/mmap.c
+++ b/mm/mmap.c
@@ -1587,35 +1587,33 @@ EXPORT_SYMBOL(get_unmapped_area);
 /* Look up the first VMA which satisfies  addr < vm_end,  NULL if none. */
 struct vm_area_struct *find_vma(struct mm_struct *mm, unsigned long addr)
 {
-	struct vm_area_struct *vma = NULL;
-
-	if (mm) {
-		/* Check the cache first. */
-		/* (Cache hit rate is typically around 35%.) */
-		vma = mm->mmap_cache;
-		if (!(vma && vma->vm_end > addr && vma->vm_start <= addr)) {
-			struct rb_node * rb_node;
-
-			rb_node = mm->mm_rb.rb_node;
-			vma = NULL;
-
-			while (rb_node) {
-				struct vm_area_struct * vma_tmp;
-
-				vma_tmp = rb_entry(rb_node,
-						struct vm_area_struct, vm_rb);
-
-				if (vma_tmp->vm_end > addr) {
-					vma = vma_tmp;
-					if (vma_tmp->vm_start <= addr)
-						break;
-					rb_node = rb_node->rb_left;
-				} else
-					rb_node = rb_node->rb_right;
-			}
-			if (vma)
-				mm->mmap_cache = vma;
+	struct vm_area_struct *vma;
+
+	/* Check the cache first. */
+	/* (Cache hit rate is typically around 35%.) */
+	vma = mm->mmap_cache;
+	if (!(vma && vma->vm_end > addr && vma->vm_start <= addr)) {
+		struct rb_node *rb_node;
+
+		rb_node = mm->mm_rb.rb_node;
+		vma = NULL;
+
+		while (rb_node) {
+			struct vm_area_struct *vma_tmp;
+
+			vma_tmp = rb_entry(rb_node,
+					   struct vm_area_struct, vm_rb);
+
+			if (vma_tmp->vm_end > addr) {
+				vma = vma_tmp;
+				if (vma_tmp->vm_start <= addr)
+					break;
+				rb_node = rb_node->rb_left;
+			} else
+				rb_node = rb_node->rb_right;
 		}
+		if (vma)
+			mm->mmap_cache = vma;
 	}
 	return vma;
 }
-- 
1.7.5.4

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

Re: [PATCH 1/1] mmap.c: find_vma: remove if(mm) check

[Andrew Morton]

On Mon, 26 Mar 2012 19:49:27 -0400
Kautuk Consul <consul.kautuk@gmail.com> wrote:

> find_vma is called from kernel code where it is absolutely
> sure that the mm_struct arg being passed to it is non-NULL.
> 
> Remove the if(mm) check.

It's odd that the if(mm) test exists - I wonder why it was originally
added.  My repo only goes back ten years, and it's there in 2.4.18.

Any code which calls find_vma() without an mm is surely pretty busted?


Still, I think I'd prefer to do

	if (WARN_ON_ONCE(!mm))
		return NULL;

then let that bake for a kernel release, just to find out if we have a
weird caller out there, such as a function which is called by both user
threads and by kernel threads.

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

Re: [PATCH 1/1] mmap.c: find_vma: remove if(mm) check

[Kautuk Consul]

On Wed, Mar 28, 2012 at 3:42 AM, Andrew Morton
<akpm@linux-foundation.org> wrote:
> On Mon, 26 Mar 2012 19:49:27 -0400
> Kautuk Consul <consul.kautuk@gmail.com> wrote:
>
>> find_vma is called from kernel code where it is absolutely
>> sure that the mm_struct arg being passed to it is non-NULL.
>>
>> Remove the if(mm) check.
>
> It's odd that the if(mm) test exists - I wonder why it was originally
> added.  My repo only goes back ten years, and it's there in 2.4.18.
>
> Any code which calls find_vma() without an mm is surely pretty busted?
>
>
> Still, I think I'd prefer to do
>
>        if (WARN_ON_ONCE(!mm))
>                return NULL;
>

yes, I agree. that is safe for now as there are a huge number of calls
to this API.

> then let that bake for a kernel release, just to find out if we have a
> weird caller out there, such as a function which is called by both user
> threads and by kernel threads.

ok. I'll spin another one and send it to you with your suggestions in a day or
two when I go back home after my day job.

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Fight unfair telecom internet charges in Canada: sign http://stopthemeter.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>