| Message ID | 20170825091845.4120-6-nefelim4ag@gmail.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Fri, Aug 25, 2017 at 12:18:44PM +0300, Timofey Titovets wrote: > Calculate byte set size for data sample: > Calculate how many unique bytes has been in sample > By count all bytes in bucket with count > 0 > If byte set low (~25%), data are easily compressible > > Signed-off-by: Timofey Titovets <nefelim4ag@gmail.com> > --- > fs/btrfs/heuristic.c | 26 ++++++++++++++++++++++++++ > 1 file changed, 26 insertions(+) > > diff --git a/fs/btrfs/heuristic.c b/fs/btrfs/heuristic.c > index f1fa6e4f1c11..ef723e991576 100644 > --- a/fs/btrfs/heuristic.c > +++ b/fs/btrfs/heuristic.c > @@ -24,6 +24,7 @@ > #define ITER_SHIFT 256 > #define BUCKET_SIZE 256 > #define MAX_SAMPLE_SIZE (BTRFS_MAX_UNCOMPRESSED*READ_SIZE/ITER_SHIFT) > +#define BYTE_SET_THRESHOLD 64 Explanation needed, partially covered by the changelog but we need to see it in the code as well. > > struct bucket_item { > u32 count; > @@ -66,6 +67,27 @@ static struct list_head *heuristic_alloc_workspace(void) > return ERR_PTR(-ENOMEM); > } > > +static u32 byte_set_size(const struct workspace *ws) > +{ > + u32 a = 0; Please use 'i'. > + u32 byte_set_size = 0; > + > + for (; a < BYTE_SET_THRESHOLD; a++) { for (i = 0; ...) > + if (ws->bucket[a].count > 0) > + byte_set_size++; > + } > + > + for (; a < BUCKET_SIZE; a++) { So here the initialization is intentionally skipped, please add a comment that this is expected or explain like "continue past the first sample until ...". > + if (ws->bucket[a].count > 0) { > + byte_set_size++; > + if (byte_set_size > BYTE_SET_THRESHOLD) > + return byte_set_size; > + } > + } > + > + return byte_set_size; > +} > + > static bool sample_repeated_patterns(struct workspace *ws) > { > u32 i = 0; > @@ -138,6 +160,10 @@ static int heuristic(struct list_head *ws, struct inode *inode, > workspace->bucket[byte].count++; > } > > + a = byte_set_size(workspace); > + if (a > BYTE_SET_THRESHOLD) > + return 2; > + > return 1; > } > > -- > 2.14.1 > -- > To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
diff --git a/fs/btrfs/heuristic.c b/fs/btrfs/heuristic.c index f1fa6e4f1c11..ef723e991576 100644 --- a/fs/btrfs/heuristic.c +++ b/fs/btrfs/heuristic.c @@ -24,6 +24,7 @@ #define ITER_SHIFT 256 #define BUCKET_SIZE 256 #define MAX_SAMPLE_SIZE (BTRFS_MAX_UNCOMPRESSED*READ_SIZE/ITER_SHIFT) +#define BYTE_SET_THRESHOLD 64 struct bucket_item { u32 count; @@ -66,6 +67,27 @@ static struct list_head *heuristic_alloc_workspace(void) return ERR_PTR(-ENOMEM); } +static u32 byte_set_size(const struct workspace *ws) +{ + u32 a = 0; + u32 byte_set_size = 0; + + for (; a < BYTE_SET_THRESHOLD; a++) { + if (ws->bucket[a].count > 0) + byte_set_size++; + } + + for (; a < BUCKET_SIZE; a++) { + if (ws->bucket[a].count > 0) { + byte_set_size++; + if (byte_set_size > BYTE_SET_THRESHOLD) + return byte_set_size; + } + } + + return byte_set_size; +} + static bool sample_repeated_patterns(struct workspace *ws) { u32 i = 0; @@ -138,6 +160,10 @@ static int heuristic(struct list_head *ws, struct inode *inode, workspace->bucket[byte].count++; } + a = byte_set_size(workspace); + if (a > BYTE_SET_THRESHOLD) + return 2; + return 1; }
Calculate byte set size for data sample: Calculate how many unique bytes has been in sample By count all bytes in bucket with count > 0 If byte set low (~25%), data are easily compressible Signed-off-by: Timofey Titovets <nefelim4ag@gmail.com> --- fs/btrfs/heuristic.c | 26 ++++++++++++++++++++++++++ 1 file changed, 26 insertions(+) -- 2.14.1 -- To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html