* [PATCH] bitmap: speedup in bitmap_find_free_region when order is 0
@ 2013-04-08 2:23 Chanho Min
2013-04-08 21:30 ` Andrew Morton
0 siblings, 1 reply; 3+ messages in thread
From: Chanho Min @ 2013-04-08 2:23 UTC (permalink / raw)
To: Andrew Morton, Nadia Yvette Chambers, Jiri Kosina, Guennadi Liakhovetski
Cc: linux-kernel, Chanho Min
If bitmap_find_free_region() is called with order=0, We can reduce
for-loops to find 1 free bit. First, It scans bitmap array by the
increment of long type, then find 1 free bit within 1 long type value.
In 32 bits system and 1024 bits size, in the worst case, We need 1024
for-loops to find 1 free bit. But, If This is applied, it takes
64 for-loops. Instead, if free bit is in the first index of the bitmaps,
It will be needed additional 1 for-loop. But from second index, It
will be speed up significantly.
Signed-off-by: Chanho Min <chanho.min@lge.com>
---
lib/bitmap.c | 34 ++++++++++++++++++++++++++++++++++
1 file changed, 34 insertions(+)
diff --git a/lib/bitmap.c b/lib/bitmap.c
index 06f7e4f..0c1f03a 100644
--- a/lib/bitmap.c
+++ b/lib/bitmap.c
@@ -1099,6 +1099,37 @@ done:
}
/**
+ * bitmap_find_free_one - find a mem region
+ * @bitmap: array of unsigned longs corresponding to the bitmap
+ * @bits: number of bits in the bitmap
+ *
+ * Find one of free (zero) bits in a @bitmap of @bits bits and
+ * allocate them (set them to one).
+ *
+ * Return the bit offset in bitmap of the allocated region,
+ * or -errno on failure.
+ */
+static int __bitmap_find_free_one(unsigned long *bitmap, int bits)
+{
+ int pos, i;
+ unsigned long mask = (unsigned long)(~((unsigned long) 0));
+ int nlongs_reg = BITS_TO_LONGS(bits);
+
+ for (i = 0 ; i < nlongs_reg ; i++) {
+ if ((bitmap[i] & mask) != mask) {
+ for (pos = 0 ; pos < BITS_PER_LONG ; pos++) {
+ if (!__reg_op(&bitmap[i], pos, 0,
+ REG_OP_ISFREE))
+ continue;
+ __reg_op(&bitmap[i], pos, 0, REG_OP_ALLOC);
+ return pos;
+ }
+ }
+ }
+ return -ENOMEM;
+}
+
+/**
* bitmap_find_free_region - find a contiguous aligned mem region
* @bitmap: array of unsigned longs corresponding to the bitmap
* @bits: number of bits in the bitmap
@@ -1116,6 +1147,9 @@ int bitmap_find_free_region(unsigned long *bitmap, int bits, int order)
{
int pos, end; /* scans bitmap by regions of size order */
+ if (order == 0)
+ return __bitmap_find_free_one(bitmap, bits);
+
for (pos = 0 ; (end = pos + (1 << order)) <= bits; pos = end) {
if (!__reg_op(bitmap, pos, order, REG_OP_ISFREE))
continue;
--
1.7.9.5
^ permalink raw reply [flat|nested] 3+ messages in thread* Re: [PATCH] bitmap: speedup in bitmap_find_free_region when order is 0
2013-04-08 2:23 [PATCH] bitmap: speedup in bitmap_find_free_region when order is 0 Chanho Min
@ 2013-04-08 21:30 ` Andrew Morton
2013-04-08 21:35 ` Joe Perches
0 siblings, 1 reply; 3+ messages in thread
From: Andrew Morton @ 2013-04-08 21:30 UTC (permalink / raw)
To: Chanho Min
Cc: Nadia Yvette Chambers, Jiri Kosina, Guennadi Liakhovetski, linux-kernel
On Mon, 8 Apr 2013 11:23:48 +0900 Chanho Min <chanho.min@lge.com> wrote:
> If bitmap_find_free_region() is called with order=0, We can reduce
> for-loops to find 1 free bit. First, It scans bitmap array by the
> increment of long type, then find 1 free bit within 1 long type value.
That seems sensible. I assume-without-checking that single-bit is a
common case?
> In 32 bits system and 1024 bits size, in the worst case, We need 1024
> for-loops to find 1 free bit. But, If This is applied, it takes
> 64 for-loops. Instead, if free bit is in the first index of the bitmaps,
> It will be needed additional 1 for-loop. But from second index, It
> will be speed up significantly.
>
> --- a/lib/bitmap.c
> +++ b/lib/bitmap.c
> @@ -1099,6 +1099,37 @@ done:
> }
>
> /**
> + * bitmap_find_free_one - find a mem region
> + * @bitmap: array of unsigned longs corresponding to the bitmap
> + * @bits: number of bits in the bitmap
> + *
> + * Find one of free (zero) bits in a @bitmap of @bits bits and
> + * allocate them (set them to one).
> + *
> + * Return the bit offset in bitmap of the allocated region,
> + * or -errno on failure.
> + */
> +static int __bitmap_find_free_one(unsigned long *bitmap, int bits)
> +{
> + int pos, i;
> + unsigned long mask = (unsigned long)(~((unsigned long) 0));
That seems unnecessarily complicated. "unsigned long mask = -1;" works :)
> + int nlongs_reg = BITS_TO_LONGS(bits);
> +
> + for (i = 0 ; i < nlongs_reg ; i++) {
> + if ((bitmap[i] & mask) != mask) {
But here we could just do "if (bitmap[i] != -1)". Or ~0UL.
> + for (pos = 0 ; pos < BITS_PER_LONG ; pos++) {
> + if (!__reg_op(&bitmap[i], pos, 0,
> + REG_OP_ISFREE))
> + continue;
> + __reg_op(&bitmap[i], pos, 0, REG_OP_ALLOC);
> + return pos;
> + }
> + }
> + }
> + return -ENOMEM;
> +}
afacit the code is buggy - if `bits' is not an exact multiple of
BITS_PER_LONG, this search will wander off the end of the specified
region?
> +/**
> * bitmap_find_free_region - find a contiguous aligned mem region
> * @bitmap: array of unsigned longs corresponding to the bitmap
> * @bits: number of bits in the bitmap
> @@ -1116,6 +1147,9 @@ int bitmap_find_free_region(unsigned long *bitmap, int bits, int order)
> {
> int pos, end; /* scans bitmap by regions of size order */
>
> + if (order == 0)
> + return __bitmap_find_free_one(bitmap, bits);
> +
> for (pos = 0 ; (end = pos + (1 << order)) <= bits; pos = end) {
> if (!__reg_op(bitmap, pos, order, REG_OP_ISFREE))
> continue;
^ permalink raw reply [flat|nested] 3+ messages in thread* Re: [PATCH] bitmap: speedup in bitmap_find_free_region when order is 0
2013-04-08 21:30 ` Andrew Morton
@ 2013-04-08 21:35 ` Joe Perches
0 siblings, 0 replies; 3+ messages in thread
From: Joe Perches @ 2013-04-08 21:35 UTC (permalink / raw)
To: Andrew Morton
Cc: Chanho Min, Nadia Yvette Chambers, Jiri Kosina,
Guennadi Liakhovetski, linux-kernel
On Mon, 2013-04-08 at 14:30 -0700, Andrew Morton wrote:
> On Mon, 8 Apr 2013 11:23:48 +0900 Chanho Min <chanho.min@lge.com> wrote:
> > --- a/lib/bitmap.c
[]
> > +static int __bitmap_find_free_one(unsigned long *bitmap, int bits)
[]
> > + unsigned long mask = (unsigned long)(~((unsigned long) 0));
>
> That seems unnecessarily complicated. "unsigned long mask = -1;" works :)
So would ULONG_MAX or just ~0UL
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2013-04-08 21:35 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2013-04-08 2:23 [PATCH] bitmap: speedup in bitmap_find_free_region when order is 0 Chanho Min
2013-04-08 21:30 ` Andrew Morton
2013-04-08 21:35 ` Joe Perches
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®