Hi!
Attached is my attempt to fix ticket #4421 based on the
analysis by trac user lvqcl. If the patch is ok, I will
add the encoder check to the decoder and the version
bump.
Please review, Carl Eugen
TODO: Fix indentation in libavcodec/x86/flacdsp_init.c()
diff --git a/libavcodec/arm/flacdsp_init_arm.c
b/libavcodec/arm/flacdsp_init_arm.c
index 82a807f..564e3dc 100644
--- a/libavcodec/arm/flacdsp_init_arm.c
+++ b/libavcodec/arm/flacdsp_init_arm.c
@@ -27,6 +27,6 @@ void ff_flac_lpc_16_arm(int32_t *samples, const int
coeffs[32], int order,
av_cold void ff_flacdsp_init_arm(FLACDSPContext *c, enum AVSampleFormat fmt,
int channels,
int bps)
{
- if (CONFIG_FLAC_DECODER && bps <= 16)
- c->lpc = ff_flac_lpc_16_arm;
+ if (CONFIG_FLAC_DECODER)
+ c->lpc16 = ff_flac_lpc_16_arm;
}
diff --git a/libavcodec/flacdec.c b/libavcodec/flacdec.c
index 34a0a70..bfedb6d 100644
--- a/libavcodec/flacdec.c
+++ b/libavcodec/flacdec.c
@@ -338,7 +338,11 @@ static int decode_subframe_lpc(FLACContext *s, int32_t
*decoded, int pred_order,
if ((ret = decode_residuals(s, decoded, pred_order)) < 0)
return ret;
- s->dsp.lpc(decoded, coeffs, pred_order, qlevel, s->blocksize);
+ if (bps <= 16 && bps + coeff_prec + av_log2(pred_order) <= 32) {
+ s->dsp.lpc16(decoded, coeffs, pred_order, qlevel, s->blocksize);
+ } else {
+ s->dsp.lpc32(decoded, coeffs, pred_order, qlevel, s->blocksize);
+ }
return 0;
}
diff --git a/libavcodec/flacdsp.c b/libavcodec/flacdsp.c
index a83eb83..30b6648 100644
--- a/libavcodec/flacdsp.c
+++ b/libavcodec/flacdsp.c
@@ -88,13 +88,10 @@ static void flac_lpc_32_c(int32_t *decoded, const int
coeffs[32],
av_cold void ff_flacdsp_init(FLACDSPContext *c, enum AVSampleFormat fmt, int
channels,
int bps)
{
- if (bps > 16) {
- c->lpc = flac_lpc_32_c;
- c->lpc_encode = flac_lpc_encode_c_32;
- } else {
- c->lpc = flac_lpc_16_c;
- c->lpc_encode = flac_lpc_encode_c_16;
- }
+ c->lpc16 = flac_lpc_16_c;
+ c->lpc32 = flac_lpc_32_c;
+ c->lpc16_encode = flac_lpc_encode_c_16;
+ c->lpc32_encode = flac_lpc_encode_c_32;
switch (fmt) {
case AV_SAMPLE_FMT_S32:
diff --git a/libavcodec/flacdsp.h b/libavcodec/flacdsp.h
index 417381c..f5cbd94 100644
--- a/libavcodec/flacdsp.h
+++ b/libavcodec/flacdsp.h
@@ -25,10 +25,14 @@
typedef struct FLACDSPContext {
void (*decorrelate[4])(uint8_t **out, int32_t **in, int channels,
int len, int shift);
- void (*lpc)(int32_t *samples, const int coeffs[32], int order,
- int qlevel, int len);
- void (*lpc_encode)(int32_t *res, const int32_t *smp, int len, int order,
- const int32_t coefs[32], int shift);
+ void (*lpc16)(int32_t *samples, const int coeffs[32], int order,
+ int qlevel, int len);
+ void (*lpc32)(int32_t *samples, const int coeffs[32], int order,
+ int qlevel, int len);
+ void (*lpc16_encode)(int32_t *res, const int32_t *smp, int len, int order,
+ const int32_t coefs[32], int shift);
+ void (*lpc32_encode)(int32_t *res, const int32_t *smp, int len, int order,
+ const int32_t coefs[32], int shift);
} FLACDSPContext;
void ff_flacdsp_init(FLACDSPContext *c, enum AVSampleFormat fmt, int channels,
int bps);
diff --git a/libavcodec/flacenc.c b/libavcodec/flacenc.c
index bc6d00a..b6dc4d5 100644
--- a/libavcodec/flacenc.c
+++ b/libavcodec/flacenc.c
@@ -838,8 +838,13 @@ static int encode_residual_ch(FlacEncodeContext *s, int ch)
order = av_clip(order, min_order - 1, max_order - 1);
if (order == last_order)
continue;
- s->flac_dsp.lpc_encode(res, smp, n, order+1, coefs[order],
- shift[order]);
+ if (s->bps_code * 4 + s->options.lpc_coeff_precision +
av_log2(order) <= 32) {
+ s->flac_dsp.lpc16_encode(res, smp, n, order+1, coefs[order],
+ shift[order]);
+ } else {
+ s->flac_dsp.lpc32_encode(res, smp, n, order+1, coefs[order],
+ shift[order]);
+ }
bits[i] = find_subframe_rice_params(s, sub, order+1);
if (bits[i] < bits[opt_index]) {
opt_index = i;
@@ -853,7 +858,11 @@ static int encode_residual_ch(FlacEncodeContext *s, int ch)
opt_order = 0;
bits[0] = UINT32_MAX;
for (i = min_order-1; i < max_order; i++) {
- s->flac_dsp.lpc_encode(res, smp, n, i+1, coefs[i], shift[i]);
+ if (s->bps_code * 4 + s->options.lpc_coeff_precision + av_log2(i)
<= 32) {
+ s->flac_dsp.lpc16_encode(res, smp, n, i+1, coefs[i], shift[i]);
+ } else {
+ s->flac_dsp.lpc32_encode(res, smp, n, i+1, coefs[i], shift[i]);
+ }
bits[i] = find_subframe_rice_params(s, sub, i+1);
if (bits[i] < bits[opt_order])
opt_order = i;
@@ -871,7 +880,11 @@ static int encode_residual_ch(FlacEncodeContext *s, int ch)
for (i = last-step; i <= last+step; i += step) {
if (i < min_order-1 || i >= max_order || bits[i] < UINT32_MAX)
continue;
- s->flac_dsp.lpc_encode(res, smp, n, i+1, coefs[i], shift[i]);
+ if (s->bps_code * 4 + s->options.lpc_coeff_precision +
av_log2(i) <= 32) {
+ s->flac_dsp.lpc32_encode(res, smp, n, i+1, coefs[i],
shift[i]);
+ } else {
+ s->flac_dsp.lpc16_encode(res, smp, n, i+1, coefs[i],
shift[i]);
+ }
bits[i] = find_subframe_rice_params(s, sub, i+1);
if (bits[i] < bits[opt_order])
opt_order = i;
@@ -886,7 +899,11 @@ static int encode_residual_ch(FlacEncodeContext *s, int ch)
for (i = 0; i < sub->order; i++)
sub->coefs[i] = coefs[sub->order-1][i];
- s->flac_dsp.lpc_encode(res, smp, n, sub->order, sub->coefs, sub->shift);
+ if (s->bps_code * 4 + s->options.lpc_coeff_precision + av_log2(opt_order)
<= 32) {
+ s->flac_dsp.lpc16_encode(res, smp, n, sub->order, sub->coefs,
sub->shift);
+ } else {
+ s->flac_dsp.lpc32_encode(res, smp, n, sub->order, sub->coefs,
sub->shift);
+ }
find_subframe_rice_params(s, sub, sub->order);
diff --git a/libavcodec/x86/flacdsp_init.c b/libavcodec/x86/flacdsp_init.c
index d04af45..f988b06 100644
--- a/libavcodec/x86/flacdsp_init.c
+++ b/libavcodec/x86/flacdsp_init.c
@@ -85,8 +85,7 @@ av_cold void ff_flacdsp_init_x86(FLACDSPContext *c, enum
AVSampleFormat fmt, int
}
}
if (EXTERNAL_SSE4(cpu_flags)) {
- if (bps > 16)
- c->lpc = ff_flac_lpc_32_sse4;
+ c->lpc32 = ff_flac_lpc_32_sse4;
}
if (EXTERNAL_AVX(cpu_flags)) {
if (fmt == AV_SAMPLE_FMT_S16) {
@@ -102,15 +101,14 @@ av_cold void ff_flacdsp_init_x86(FLACDSPContext *c, enum
AVSampleFormat fmt, int
}
}
if (EXTERNAL_XOP(cpu_flags)) {
- if (bps > 16)
- c->lpc = ff_flac_lpc_32_xop;
+ c->lpc32 = ff_flac_lpc_32_xop;
}
#endif
#if CONFIG_FLAC_ENCODER
if (EXTERNAL_SSE4(cpu_flags)) {
- if (CONFIG_GPL && bps == 16)
- c->lpc_encode = ff_flac_enc_lpc_16_sse4;
+ if (CONFIG_GPL)
+ c->lpc16_encode = ff_flac_enc_lpc_16_sse4;
}
#endif
#endif /* HAVE_YASM */
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
