[patch 1/3] mm: remove unused get_vm_area_node

[patch 1/3] mm: remove unused get_vm_area_node

[David Rientjes]

get_vm_area_node() is unused in the kernel and can thus be removed.

Signed-off-by: David Rientjes <rientjes@google.com>
---
 include/linux/vmalloc.h |    3 ---
 mm/vmalloc.c            |    7 -------
 2 files changed, 0 insertions(+), 10 deletions(-)

diff --git a/include/linux/vmalloc.h b/include/linux/vmalloc.h
--- a/include/linux/vmalloc.h
+++ b/include/linux/vmalloc.h
@@ -92,9 +92,6 @@ extern struct vm_struct *__get_vm_area_caller(unsigned long size,
 					unsigned long flags,
 					unsigned long start, unsigned long end,
 					void *caller);
-extern struct vm_struct *get_vm_area_node(unsigned long size,
-					  unsigned long flags, int node,
-					  gfp_t gfp_mask);
 extern struct vm_struct *remove_vm_area(const void *addr);
 
 extern int map_vm_area(struct vm_struct *area, pgprot_t prot,
diff --git a/mm/vmalloc.c b/mm/vmalloc.c
--- a/mm/vmalloc.c
+++ b/mm/vmalloc.c
@@ -1309,13 +1309,6 @@ struct vm_struct *get_vm_area_caller(unsigned long size, unsigned long flags,
 						-1, GFP_KERNEL, caller);
 }
 
-struct vm_struct *get_vm_area_node(unsigned long size, unsigned long flags,
-				   int node, gfp_t gfp_mask)
-{
-	return __get_vm_area_node(size, 1, flags, VMALLOC_START, VMALLOC_END,
-				  node, gfp_mask, __builtin_return_address(0));
-}
-
 static struct vm_struct *find_vm_area(const void *addr)
 {
 	struct vmap_area *va;

--
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 policy in Canada: sign http://dissolvethecrtc.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

[patch 2/3] mm: remove gfp mask from pcpu_get_vm_areas

[David Rientjes]

pcpu_get_vm_areas() only uses GFP_KERNEL allocations, so remove the gfp_t
formal and use the mask internally.

Signed-off-by: David Rientjes <rientjes@google.com>
---
 include/linux/vmalloc.h |    2 +-
 mm/percpu-vm.c          |    2 +-
 mm/vmalloc.c            |   21 +++++++++------------
 3 files changed, 11 insertions(+), 14 deletions(-)

diff --git a/include/linux/vmalloc.h b/include/linux/vmalloc.h
--- a/include/linux/vmalloc.h
+++ b/include/linux/vmalloc.h
@@ -119,7 +119,7 @@ extern __init void vm_area_register_early(struct vm_struct *vm, size_t align);
 #ifdef CONFIG_SMP
 struct vm_struct **pcpu_get_vm_areas(const unsigned long *offsets,
 				     const size_t *sizes, int nr_vms,
-				     size_t align, gfp_t gfp_mask);
+				     size_t align);
 
 void pcpu_free_vm_areas(struct vm_struct **vms, int nr_vms);
 #endif
diff --git a/mm/percpu-vm.c b/mm/percpu-vm.c
--- a/mm/percpu-vm.c
+++ b/mm/percpu-vm.c
@@ -421,7 +421,7 @@ static struct pcpu_chunk *pcpu_create_chunk(void)
 		return NULL;
 
 	vms = pcpu_get_vm_areas(pcpu_group_offsets, pcpu_group_sizes,
-				pcpu_nr_groups, pcpu_atom_size, GFP_KERNEL);
+				pcpu_nr_groups, pcpu_atom_size);
 	if (!vms) {
 		pcpu_free_chunk(chunk);
 		return NULL;
diff --git a/mm/vmalloc.c b/mm/vmalloc.c
--- a/mm/vmalloc.c
+++ b/mm/vmalloc.c
@@ -2190,17 +2190,16 @@ static unsigned long pvm_determine_end(struct vmap_area **pnext,
  * @sizes: array containing size of each area
  * @nr_vms: the number of areas to allocate
  * @align: alignment, all entries in @offsets and @sizes must be aligned to this
- * @gfp_mask: allocation mask
  *
  * Returns: kmalloc'd vm_struct pointer array pointing to allocated
  *	    vm_structs on success, %NULL on failure
  *
  * Percpu allocator wants to use congruent vm areas so that it can
  * maintain the offsets among percpu areas.  This function allocates
- * congruent vmalloc areas for it.  These areas tend to be scattered
- * pretty far, distance between two areas easily going up to
- * gigabytes.  To avoid interacting with regular vmallocs, these areas
- * are allocated from top.
+ * congruent vmalloc areas for it with GFP_KERNEL.  These areas tend to
+ * be scattered pretty far, distance between two areas easily going up
+ * to gigabytes.  To avoid interacting with regular vmallocs, these
+ * areas are allocated from top.
  *
  * Despite its complicated look, this allocator is rather simple.  It
  * does everything top-down and scans areas from the end looking for
@@ -2211,7 +2210,7 @@ static unsigned long pvm_determine_end(struct vmap_area **pnext,
  */
 struct vm_struct **pcpu_get_vm_areas(const unsigned long *offsets,
 				     const size_t *sizes, int nr_vms,
-				     size_t align, gfp_t gfp_mask)
+				     size_t align)
 {
 	const unsigned long vmalloc_start = ALIGN(VMALLOC_START, align);
 	const unsigned long vmalloc_end = VMALLOC_END & ~(align - 1);
@@ -2221,8 +2220,6 @@ struct vm_struct **pcpu_get_vm_areas(const unsigned long *offsets,
 	unsigned long base, start, end, last_end;
 	bool purged = false;
 
-	gfp_mask &= GFP_RECLAIM_MASK;
-
 	/* verify parameters and allocate data structures */
 	BUG_ON(align & ~PAGE_MASK || !is_power_of_2(align));
 	for (last_area = 0, area = 0; area < nr_vms; area++) {
@@ -2255,14 +2252,14 @@ struct vm_struct **pcpu_get_vm_areas(const unsigned long *offsets,
 		return NULL;
 	}
 
-	vms = kzalloc(sizeof(vms[0]) * nr_vms, gfp_mask);
-	vas = kzalloc(sizeof(vas[0]) * nr_vms, gfp_mask);
+	vms = kzalloc(sizeof(vms[0]) * nr_vms, GFP_KERNEL);
+	vas = kzalloc(sizeof(vas[0]) * nr_vms, GFP_KERNEL);
 	if (!vas || !vms)
 		goto err_free;
 
 	for (area = 0; area < nr_vms; area++) {
-		vas[area] = kzalloc(sizeof(struct vmap_area), gfp_mask);
-		vms[area] = kzalloc(sizeof(struct vm_struct), gfp_mask);
+		vas[area] = kzalloc(sizeof(struct vmap_area), GFP_KERNEL);
+		vms[area] = kzalloc(sizeof(struct vm_struct), GFP_KERNEL);
 		if (!vas[area] || !vms[area])
 			goto err_free;
 	}

--
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 policy in Canada: sign http://dissolvethecrtc.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

[patch 3/3] mm: unify module_alloc code for vmalloc

[David Rientjes]

Four architectures (arm, mips, sparc, x86) use __vmalloc_area() for
module_init().  Much of the code is duplicated and can be generalized in
a globally accessible function, __vmalloc_node_range().

__vmalloc_node() now calls into __vmalloc_node_range() with a range of
[VMALLOC_START, VMALLOC_END) for functionally equivalent behavior.

Each architecture may then use __vmalloc_node_range() directly to remove
the duplication of code.

Signed-off-by: David Rientjes <rientjes@google.com>
---
 arch/arm/kernel/module.c   |   14 ++---------
 arch/mips/kernel/module.c  |   14 ++---------
 arch/sparc/kernel/module.c |   14 +++--------
 arch/x86/kernel/module.c   |   17 +++-----------
 include/linux/vmalloc.h    |    5 ++-
 mm/vmalloc.c               |   50 +++++++++++++++++++++++++------------------
 6 files changed, 46 insertions(+), 68 deletions(-)

diff --git a/arch/arm/kernel/module.c b/arch/arm/kernel/module.c
--- a/arch/arm/kernel/module.c
+++ b/arch/arm/kernel/module.c
@@ -38,17 +38,9 @@
 #ifdef CONFIG_MMU
 void *module_alloc(unsigned long size)
 {
-	struct vm_struct *area;
-
-	size = PAGE_ALIGN(size);
-	if (!size)
-		return NULL;
-
-	area = __get_vm_area(size, VM_ALLOC, MODULES_VADDR, MODULES_END);
-	if (!area)
-		return NULL;
-
-	return __vmalloc_area(area, GFP_KERNEL, PAGE_KERNEL_EXEC);
+	return __vmalloc_node_range(size, 1, MODULES_VADDR, MODULES_END,
+				GFP_KERNEL, PAGE_KERNEL_EXEC, -1,
+				__builtin_return_address(0));
 }
 #else /* CONFIG_MMU */
 void *module_alloc(unsigned long size)
diff --git a/arch/mips/kernel/module.c b/arch/mips/kernel/module.c
--- a/arch/mips/kernel/module.c
+++ b/arch/mips/kernel/module.c
@@ -46,17 +46,9 @@ static DEFINE_SPINLOCK(dbe_lock);
 void *module_alloc(unsigned long size)
 {
 #ifdef MODULE_START
-	struct vm_struct *area;
-
-	size = PAGE_ALIGN(size);
-	if (!size)
-		return NULL;
-
-	area = __get_vm_area(size, VM_ALLOC, MODULE_START, MODULE_END);
-	if (!area)
-		return NULL;
-
-	return __vmalloc_area(area, GFP_KERNEL, PAGE_KERNEL);
+	return __vmalloc_node_range(size, 1, MODULE_START, MODULE_END,
+				GFP_KERNEL, PAGE_KERNEL, -1,
+				__builtin_return_address(0));
 #else
 	if (size == 0)
 		return NULL;
diff --git a/arch/sparc/kernel/module.c b/arch/sparc/kernel/module.c
--- a/arch/sparc/kernel/module.c
+++ b/arch/sparc/kernel/module.c
@@ -23,17 +23,11 @@
 
 static void *module_map(unsigned long size)
 {
-	struct vm_struct *area;
-
-	size = PAGE_ALIGN(size);
-	if (!size || size > MODULES_LEN)
-		return NULL;
-
-	area = __get_vm_area(size, VM_ALLOC, MODULES_VADDR, MODULES_END);
-	if (!area)
+	if (PAGE_ALIGN(size) > MODULES_LEN)
 		return NULL;
-
-	return __vmalloc_area(area, GFP_KERNEL, PAGE_KERNEL);
+	return __vmalloc_node_range(size, 1, MODULES_VADDR, MODULES_END,
+				GFP_KERNEL, PAGE_KERNEL, -1,
+				__builtin_return_address(0));
 }
 
 static char *dot2underscore(char *name)
diff --git a/arch/x86/kernel/module.c b/arch/x86/kernel/module.c
--- a/arch/x86/kernel/module.c
+++ b/arch/x86/kernel/module.c
@@ -37,20 +37,11 @@
 
 void *module_alloc(unsigned long size)
 {
-	struct vm_struct *area;
-
-	if (!size)
-		return NULL;
-	size = PAGE_ALIGN(size);
-	if (size > MODULES_LEN)
+	if (PAGE_ALIGN(size) > MODULES_LEN)
 		return NULL;
-
-	area = __get_vm_area(size, VM_ALLOC, MODULES_VADDR, MODULES_END);
-	if (!area)
-		return NULL;
-
-	return __vmalloc_area(area, GFP_KERNEL | __GFP_HIGHMEM,
-					PAGE_KERNEL_EXEC);
+	return __vmalloc_node_range(size, 1, MODULES_VADDR, MODULES_END,
+				GFP_KERNEL | __GFP_HIGHMEM, PAGE_KERNEL_EXEC,
+				-1, __builtin_return_address(0));
 }
 
 /* Free memory returned from module_alloc */
diff --git a/include/linux/vmalloc.h b/include/linux/vmalloc.h
--- a/include/linux/vmalloc.h
+++ b/include/linux/vmalloc.h
@@ -61,8 +61,9 @@ extern void *vmalloc_exec(unsigned long size);
 extern void *vmalloc_32(unsigned long size);
 extern void *vmalloc_32_user(unsigned long size);
 extern void *__vmalloc(unsigned long size, gfp_t gfp_mask, pgprot_t prot);
-extern void *__vmalloc_area(struct vm_struct *area, gfp_t gfp_mask,
-				pgprot_t prot);
+extern void *__vmalloc_node_range(unsigned long size, unsigned long align,
+			unsigned long start, unsigned long end, gfp_t gfp_mask,
+			pgprot_t prot, int node, void *caller);
 extern void vfree(const void *addr);
 
 extern void *vmap(struct page **pages, unsigned int count,
diff --git a/mm/vmalloc.c b/mm/vmalloc.c
--- a/mm/vmalloc.c
+++ b/mm/vmalloc.c
@@ -1524,25 +1524,12 @@ fail:
 	return NULL;
 }
 
-void *__vmalloc_area(struct vm_struct *area, gfp_t gfp_mask, pgprot_t prot)
-{
-	void *addr = __vmalloc_area_node(area, gfp_mask, prot, -1,
-					 __builtin_return_address(0));
-
-	/*
-	 * A ref_count = 3 is needed because the vm_struct and vmap_area
-	 * structures allocated in the __get_vm_area_node() function contain
-	 * references to the virtual address of the vmalloc'ed block.
-	 */
-	kmemleak_alloc(addr, area->size - PAGE_SIZE, 3, gfp_mask);
-
-	return addr;
-}
-
 /**
- *	__vmalloc_node  -  allocate virtually contiguous memory
+ *	__vmalloc_node_range  -  allocate virtually contiguous memory
  *	@size:		allocation size
  *	@align:		desired alignment
+ *	@start:		vm area range start
+ *	@end:		vm area range end
  *	@gfp_mask:	flags for the page level allocator
  *	@prot:		protection mask for the allocated pages
  *	@node:		node to use for allocation or -1
@@ -1552,9 +1539,9 @@ void *__vmalloc_area(struct vm_struct *area, gfp_t gfp_mask, pgprot_t prot)
  *	allocator with @gfp_mask flags.  Map them into contiguous
  *	kernel virtual space, using a pagetable protection of @prot.
  */
-static void *__vmalloc_node(unsigned long size, unsigned long align,
-			    gfp_t gfp_mask, pgprot_t prot,
-			    int node, void *caller)
+void *__vmalloc_node_range(unsigned long size, unsigned long align,
+			unsigned long start, unsigned long end, gfp_t gfp_mask,
+			pgprot_t prot, int node, void *caller)
 {
 	struct vm_struct *area;
 	void *addr;
@@ -1564,8 +1551,8 @@ static void *__vmalloc_node(unsigned long size, unsigned long align,
 	if (!size || (size >> PAGE_SHIFT) > totalram_pages)
 		return NULL;
 
-	area = __get_vm_area_node(size, align, VM_ALLOC, VMALLOC_START,
-				  VMALLOC_END, node, gfp_mask, caller);
+	area = __get_vm_area_node(size, align, VM_ALLOC, start, end, node,
+				  gfp_mask, caller);
 
 	if (!area)
 		return NULL;
@@ -1582,6 +1569,27 @@ static void *__vmalloc_node(unsigned long size, unsigned long align,
 	return addr;
 }
 
+/**
+ *	__vmalloc_node  -  allocate virtually contiguous memory
+ *	@size:		allocation size
+ *	@align:		desired alignment
+ *	@gfp_mask:	flags for the page level allocator
+ *	@prot:		protection mask for the allocated pages
+ *	@node:		node to use for allocation or -1
+ *	@caller:	caller's return address
+ *
+ *	Allocate enough pages to cover @size from the page level
+ *	allocator with @gfp_mask flags.  Map them into contiguous
+ *	kernel virtual space, using a pagetable protection of @prot.
+ */
+static void *__vmalloc_node(unsigned long size, unsigned long align,
+			    gfp_t gfp_mask, pgprot_t prot,
+			    int node, void *caller)
+{
+	return __vmalloc_node_range(size, align, VMALLOC_START, VMALLOC_END,
+				gfp_mask, prot, node, caller);
+}
+
 void *__vmalloc(unsigned long size, gfp_t gfp_mask, pgprot_t prot)
 {
 	return __vmalloc_node(size, 1, gfp_mask, prot, -1,

--
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 policy in Canada: sign http://dissolvethecrtc.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

Re: [patch 2/3] mm: remove gfp mask from pcpu_get_vm_areas

[Tejun Heo]

On 11/17/2010 04:41 AM, David Rientjes wrote:
> pcpu_get_vm_areas() only uses GFP_KERNEL allocations, so remove the gfp_t
> formal and use the mask internally.
> 
> Signed-off-by: David Rientjes <rientjes@google.com>

Patch itself looks okay to me but why do you want to drop the
argument?

Thanks.

-- 
tejun

--
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 policy in Canada: sign http://dissolvethecrtc.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

Re: [patch 2/3] mm: remove gfp mask from pcpu_get_vm_areas

[David Rientjes]

On Wed, 17 Nov 2010, Tejun Heo wrote:

> > pcpu_get_vm_areas() only uses GFP_KERNEL allocations, so remove the gfp_t
> > formal and use the mask internally.
> > 
> > Signed-off-by: David Rientjes <rientjes@google.com>
> 
> Patch itself looks okay to me but why do you want to drop the
> argument?
> 

A recent thread[*] shows a problem whereas gfp masks may be passed into 
the vmalloc interface that restrict reclaim behavior, yet the underlying 
pte allocator unconditionally uses GFP_KERNEL.  This is a first-pass at an 
effort to remove all gfp_t formals from the vmalloc interface (and can be 
completed once gfs2, ntfs, and ceph have converted) and require them to 
use GFP_KERNEL.

Luckily for the per-cpu allocator, this was trivial since that happens to 
be the only use case already.

 [*] http://marc.info/?t=128942209500002

--
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 policy in Canada: sign http://dissolvethecrtc.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>

Re: [patch 2/3] mm: remove gfp mask from pcpu_get_vm_areas

[Tejun Heo]

Hello,

On 11/17/2010 09:32 PM, David Rientjes wrote:
> A recent thread[*] shows a problem whereas gfp masks may be passed into 
> the vmalloc interface that restrict reclaim behavior, yet the underlying 
> pte allocator unconditionally uses GFP_KERNEL.  This is a first-pass at an 
> effort to remove all gfp_t formals from the vmalloc interface (and can be 
> completed once gfs2, ntfs, and ceph have converted) and require them to 
> use GFP_KERNEL.

I see.

> Luckily for the per-cpu allocator, this was trivial since that happens to 
> be the only use case already.

per-cpu allocator intentionally only allowed GFP_KERNEL till now.
There were some requests about allowing GFP_ATOMIC allocations and
that's the reason why the @gfp is there for the vm function.  Anyways,
this looks like the nail in that coffin.

 Acked-by: Tejun Heo <tj@kernel.org>

Thanks.

-- 
tejun

--
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 policy in Canada: sign http://dissolvethecrtc.ca/
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>