[Ffmpeg-devel] [PATCH] simple internal lzo decoder
Reimar Döffinger
Reimar.Doeffinger
Sun Jan 15 18:27:55 CET 2006
Hi,
On Sat, Jan 14, 2006 at 04:21:44PM +0100, Reimar D?ffinger wrote:
> Also instead of the last patch I'd like to propose the attached one,
A last try, the speed is really nice (though only tested on my system).
But it needs a bit of padding for the buffers - since it is done in
other place I guess this is okay even if I personally don't like it
much.
Ok, here are the values: first comes liblzo, then minilzo and then this
implementation with the patch. And then the same but for gcc 4.0.0 (I
know, outdated, but that's what I had).
file all zeroes, 56325B comp:
8.90 8.90 0.80 8.87 8.46 0.75
nwnmod, 2675408B comp:
7.39 6.98 6.87 7.35 7.18 6.39
/dev/urandom, 10527740B comp:
4.28 4.13 1.72 4.28 4.24 1.73
Oh, and the image of my RE drawing is up as well :-) :
http://www.stud.uni-karlsruhe.de/~uvhe/LZODraw_bw.png
Greetings,
Reimar D?ffinger
-------------- next part --------------
Index: libavcodec/cscd.c
===================================================================
RCS file: /cvsroot/ffmpeg/ffmpeg/libavcodec/cscd.c,v
retrieving revision 1.3
diff -u -r1.3 cscd.c
--- libavcodec/cscd.c 12 Jan 2006 22:43:14 -0000 1.3
+++ libavcodec/cscd.c 15 Jan 2006 17:17:11 -0000
@@ -232,7 +232,7 @@
c->linelen = avctx->width * avctx->bits_per_sample / 8;
c->height = avctx->height;
c->decomp_size = c->height * c->linelen;
- c->decomp_buf = av_malloc(c->decomp_size);
+ c->decomp_buf = av_malloc(c->decomp_size + LZO_OUTPUT_PADDING);
if (!c->decomp_buf) {
av_log(avctx, AV_LOG_ERROR, "Can't allocate decompression buffer.\n");
return 1;
Index: libavcodec/lzo.c
===================================================================
RCS file: /cvsroot/ffmpeg/ffmpeg/libavcodec/lzo.c,v
retrieving revision 1.4
diff -u -r1.4 lzo.c
--- libavcodec/lzo.c 14 Jan 2006 14:59:11 -0000 1.4
+++ libavcodec/lzo.c 15 Jan 2006 17:17:15 -0000
@@ -17,12 +17,18 @@
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
*/
#include "common.h"
+//! avoid e.g. MPlayers fast_memcpy, it slows things down here
+#undef memcpy
+#include <string.h>
#include "lzo.h"
+//! define if we may write up to 12 bytes beyond the output buffer
+#define OUTBUF_PADDED 1
+//! define if we may read up to 4 bytes beyond the input buffer
+#define INBUF_PADDED 1
typedef struct LZOContext {
uint8_t *in, *in_end;
- uint8_t *out, *out_end;
- int out_size;
+ uint8_t *out_start, *out, *out_end;
int error;
} LZOContext;
@@ -57,17 +63,29 @@
* \param cnt number of bytes to copy, must be > 0
*/
static inline void copy(LZOContext *c, int cnt) {
- if (c->in + cnt > c->in_end) {
- cnt = c->in_end - c->in;
+ register uint8_t *src = c->in;
+ register uint8_t *dst = c->out;
+ if (src + cnt > c->in_end) {
+ cnt = c->in_end - src;
c->error |= LZO_INPUT_DEPLETED;
}
- if (c->out + cnt > c->out_end) {
- cnt = c->out_end - c->out;
+ if (dst + cnt > c->out_end) {
+ cnt = c->out_end - dst;
c->error |= LZO_OUTPUT_FULL;
}
- do {
- *c->out++ = *c->in++;
- } while (--cnt);
+#if defined(INBUF_PADDED) && defined(OUTBUF_PADDED)
+ dst[0] = src[0];
+ dst[1] = src[1];
+ dst[2] = src[2];
+ dst[3] = src[3];
+ src += 4;
+ dst += 4;
+ cnt -= 4;
+ if (cnt > 0)
+#endif
+ memcpy(dst, src, cnt);
+ c->in = src + cnt;
+ c->out = dst + cnt;
}
/**
@@ -78,17 +96,55 @@
* cnt > back is valid, this will copy the bytes we just copied.
*/
static inline void copy_backptr(LZOContext *c, int back, int cnt) {
- if (c->out - back < c->out_end - c->out_size) {
+ register uint8_t *src = &c->out[-back];
+ register uint8_t *dst = c->out;
+ if (src < c->out_start) {
c->error |= LZO_INVALID_BACKPTR;
return;
}
- if (c->out + cnt > c->out_end) {
- cnt = c->out_end - c->out;
+ if (dst + cnt > c->out_end) {
+ cnt = c->out_end - dst;
c->error |= LZO_OUTPUT_FULL;
}
- do {
- *c->out++ = c->out[-back];
- } while (--cnt);
+ if (back == 1) {
+ memset(dst, *src, cnt);
+ dst += cnt;
+ } else {
+#ifdef OUTBUF_PADDED
+ dst[0] = src[0];
+ dst[1] = src[1];
+ dst[2] = src[2];
+ dst[3] = src[3];
+ src += 4;
+ dst += 4;
+ cnt -= 4;
+ if (cnt > 0) {
+ dst[0] = src[0];
+ dst[1] = src[1];
+ dst[2] = src[2];
+ dst[3] = src[3];
+ dst[4] = src[4];
+ dst[5] = src[5];
+ dst[6] = src[6];
+ dst[7] = src[7];
+ src += 8;
+ dst += 8;
+ cnt -= 8;
+ }
+#endif
+ if (cnt > 0) {
+ int blocklen = back;
+ while (cnt > blocklen) {
+ memcpy(dst, src, blocklen);
+ dst += blocklen;
+ cnt -= blocklen;
+ blocklen <<= 1;
+ }
+ memcpy(dst, src, cnt);
+ }
+ dst += cnt;
+ }
+ c->out = dst;
}
/**
@@ -98,6 +154,9 @@
* \param in input buffer
* \param inlen size of input buffer, number of bytes left are returned here
* \return 0 on success, otherwise error flags, see lzo.h
+ *
+ * make sure all buffers are appropriately padded, in must provide
+ * LZO_INPUT_PADDING, out must provide LZO_OUTPUT_PADDING additional bytes
*/
int lzo1x_decode(void *out, int *outlen, void *in, int *inlen) {
enum {COPY, BACKPTR} state = COPY;
@@ -105,9 +164,8 @@
LZOContext c;
c.in = in;
c.in_end = in + *inlen;
- c.out = out;
+ c.out = c.out_start = out;
c.out_end = out + * outlen;
- c.out_size = *outlen;
c.error = 0;
x = get_byte(&c);
if (x > 17) {
Index: libavcodec/lzo.h
===================================================================
RCS file: /cvsroot/ffmpeg/ffmpeg/libavcodec/lzo.h,v
retrieving revision 1.1
diff -u -r1.1 lzo.h
--- libavcodec/lzo.h 12 Jan 2006 17:52:41 -0000 1.1
+++ libavcodec/lzo.h 15 Jan 2006 17:17:15 -0000
@@ -6,6 +6,9 @@
#define LZO_INVALID_BACKPTR 4
#define LZO_ERROR 8
+#define LZO_INPUT_PADDING 4
+#define LZO_OUTPUT_PADDING 12
+
int lzo1x_decode(void *out, int *outlen, void *in, int *inlen);
#endif
More information about the ffmpeg-devel
mailing list