Merge remote-tracking branch 'qatar/master'

* qatar/master:
  mpegvideo: drop unnecessary arguments to hpel_motion()
  mpegvideo: drop 'inline' from some functions
  nellymoserdec: drop support for s16 output.
  bmpdec: only initialize palette for pal8.
  build: Properly remove object files while cleaning
  flacdsp: arm optimised lpc filter
  compat/vsnprintf: return number of bytes required on truncation.

Merged-by: Michael Niedermayer <michaelni@gmx.at>

10 files changed, 229 insertions, 65 deletions

diff --git a/common.mak b/common.mak
index 172664011c..dd5de7a508 100644
--- a/common.mak
+++ b/common.mak
@@ -124,4 +124,7 @@ CLEANSUFFIXES     = *.d *.o *~ *.h.c *.map *.ver *.ho *.gcno *.gcda
 DISTCLEANSUFFIXES = *.pc
 LIBSUFFIXES       = *.a *.lib *.so *.so.* *.dylib *.dll *.def *.dll.a
 
+clean::
+	$(RM) $(OBJS) $(OBJS:.o=.d)
+
 -include $(wildcard $(OBJS:.o=.d) $(HOSTOBJS:.o=.d) $(TESTOBJS:.o=.d) $(HOBJS:.o=.d))
diff --git a/libavcodec/arm/Makefile b/libavcodec/arm/Makefile
index a8e531cf18..745a5bdfe2 100644
--- a/libavcodec/arm/Makefile
+++ b/libavcodec/arm/Makefile
@@ -8,6 +8,9 @@ OBJS-$(CONFIG_DCA_DECODER)             += arm/dcadsp_init_arm.o         \
 
 ARMV6-OBJS-$(CONFIG_AC3DSP)            += arm/ac3dsp_armv6.o
 
+OBJS-$(CONFIG_FLAC_DECODER)            += arm/flacdsp_init_arm.o        \
+                                          arm/flacdsp_arm.o             \
+
 OBJS-$(CONFIG_MPEGAUDIODSP)            += arm/mpegaudiodsp_init_arm.o
 ARMV6-OBJS-$(CONFIG_MPEGAUDIODSP)      += arm/mpegaudiodsp_fixed_armv6.o
 
diff --git a/libavcodec/arm/flacdsp_arm.S b/libavcodec/arm/flacdsp_arm.S
new file mode 100644
index 0000000000..f8861c5967
--- /dev/null
+++ b/libavcodec/arm/flacdsp_arm.S
@@ -0,0 +1,146 @@
+/*
+ * Copyright (c) 2012 Mans Rullgard <mans@mansr.com>
+ *
+ * This file is part of FFmpeg
+ *
+ * FFmpeg is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2.1 of the License, or (at your option) any later version.
+ *
+ * FFmpeg is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with FFmpeg; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
+ */
+
+#include "libavutil/arm/asm.S"
+
+function flac_lpc_16_1_arm
+        ldr             r12, [sp]
+        push            {r4, lr}
+        ldr             r1,  [r1]
+        subs            r12, r12, #2
+        ldr             lr,  [r0], #4
+        beq             2f
+        it              lt
+        poplt           {r4, pc}
+1:
+        mul             r4,  lr,  r1
+        ldm             r0,  {r2, lr}
+        add_sh          r2,  r2,  r4,  asr r3
+        mul             r4,  r2,  r1
+        subs            r12, r12, #2
+        add_sh          lr,  lr,  r4,  asr r3
+        stm             r0!, {r2, lr}
+        bgt             1b
+        it              lt
+        poplt           {r4, pc}
+2:
+        mul             r4,  lr,  r1
+        ldr             r2,  [r0]
+        add_sh          r2,  r2,  r4,  asr r3
+        str             r2,  [r0]
+        pop             {r4, pc}
+endfunc
+
+function flac_lpc_16_2_arm
+        ldr             r12, [sp]
+        subs            r12, r12, r2
+        it              le
+        bxle            lr
+
+        push            {r4-r9, lr}
+        ldm             r0!, {r6, r7}
+        ldm             r1,  {r8, r9}
+        subs            r12, r12, #1
+        beq             2f
+1:
+        mul             r4,  r6,  r8
+        mul             r5,  r7,  r8
+        mla             r4,  r7,  r9,  r4
+        ldm             r0,  {r6, r7}
+        add_sh          r6,  r6,  r4,  asr r3
+        mla             r5,  r6,  r9,  r5
+        add_sh          r7,  r7,  r5,  asr r3
+        stm             r0!, {r6, r7}
+        subs            r12, r12, #2
+        bgt             1b
+        it              lt
+        poplt           {r4-r9, pc}
+2:
+        mul             r4,  r6,  r8
+        mla             r4,  r7,  r9,  r4
+        ldr             r5,  [r0]
+        add_sh          r5,  r5,  r4,  asr r3
+        str             r5,  [r0]
+        pop             {r4-r9, pc}
+endfunc
+
+function ff_flac_lpc_16_arm, export=1
+        cmp             r2,  #2
+        blt             flac_lpc_16_1_arm
+        beq             flac_lpc_16_2_arm
+
+        ldr             r12, [sp]
+        subs            r12, r12, r2
+        it              le
+        bxle            lr
+
+        push            {r4-r9, lr}
+
+        subs            r12, r12, #1
+        beq             3f
+1:
+        sub             lr,  r2,  #2
+        mov             r4,  #0
+        mov             r5,  #0
+
+        ldr             r7,  [r0], #4
+        ldr             r9,  [r1], #4
+2:
+        mla             r4,  r7,  r9,  r4
+        ldm             r0!, {r6, r7}
+        mla             r5,  r6,  r9,  r5
+        ldm             r1!, {r8, r9}
+        mla             r4,  r6,  r8,  r4
+        subs            lr,  lr,  #2
+        mla             r5,  r7,  r8,  r5
+        bgt             2b
+        blt             6f
+
+        mla             r4,  r7,  r9,  r4
+        ldr             r7,  [r0], #4
+        mla             r5,  r7,  r9,  r5
+        ldr             r9,  [r1], #4
+6:
+        mla             r4,  r7,  r9,  r4
+        ldm             r0,  {r6, r7}
+        add_sh          r6,  r6,  r4,  asr r3
+        mla             r5,  r6,  r9,  r5
+        add_sh          r7,  r7,  r5,  asr r3
+        stm             r0!, {r6, r7}
+        sub             r0,  r0,  r2,  lsl #2
+        sub             r1,  r1,  r2,  lsl #2
+
+        subs            r12, r12, #2
+        bgt             1b
+        it              lt
+        poplt           {r4-r9, pc}
+3:
+        mov             r4,  #0
+4:
+        ldr             r5,  [r1], #4
+        ldr             r6,  [r0], #4
+        mla             r4,  r5,  r6,  r4
+        subs            r2,  r2,  #1
+        bgt             4b
+        ldr             r5,  [r0]
+        add_sh          r5,  r5,  r4,  asr r3
+        str             r5,  [r0]
+        pop             {r4-r9, pc}
+endfunc
diff --git a/libavcodec/arm/flacdsp_init_arm.c b/libavcodec/arm/flacdsp_init_arm.c
new file mode 100644
index 0000000000..9b9394280f
--- /dev/null
+++ b/libavcodec/arm/flacdsp_init_arm.c
@@ -0,0 +1,32 @@
+/*
+ * Copyright (c) 2012 Mans Rullgard <mans@mansr.com>
+ *
+ * This file is part of FFmpeg.
+ *
+ * FFmpeg is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU Lesser General Public
+ * License as published by the Free Software Foundation; either
+ * version 2.1 of the License, or (at your option) any later version.
+ *
+ * FFmpeg is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+ * Lesser General Public License for more details.
+ *
+ * You should have received a copy of the GNU Lesser General Public
+ * License along with FFmpeg; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
+ */
+
+#include "libavcodec/flacdsp.h"
+#include "config.h"
+
+void ff_flac_lpc_16_arm(int32_t *samples, const int coeffs[32], int order,
+                        int qlevel, int len);
+
+av_cold void ff_flacdsp_init_arm(FLACDSPContext *c, enum AVSampleFormat fmt,
+                                 int bps)
+{
+    if (bps <= 16)
+        c->lpc = ff_flac_lpc_16_arm;
+}
diff --git a/libavcodec/bmp.c b/libavcodec/bmp.c
index e97038c5fe..7e160ae556 100644
--- a/libavcodec/bmp.c
+++ b/libavcodec/bmp.c
@@ -232,9 +232,6 @@ static int bmp_decode_frame(AVCodecContext *avctx,
     if(comp == BMP_RLE4 || comp == BMP_RLE8)
         memset(p->data[0], 0, avctx->height * p->linesize[0]);
 
-    if(depth == 4 || depth == 8)
-        memset(p->data[1], 0, 1024);
-
     if(height > 0){
         ptr = p->data[0] + (avctx->height - 1) * p->linesize[0];
         linesize = -p->linesize[0];
@@ -245,6 +242,9 @@ static int bmp_decode_frame(AVCodecContext *avctx,
 
     if(avctx->pix_fmt == PIX_FMT_PAL8){
         int colors = 1 << depth;
+
+        memset(p->data[1], 0, 1024);
+
         if(ihsize >= 36){
             int t;
             buf = buf0 + 46;
diff --git a/libavcodec/flacdsp.c b/libavcodec/flacdsp.c
index 6c2458fc10..e51a91a07c 100644
--- a/libavcodec/flacdsp.c
+++ b/libavcodec/flacdsp.c
@@ -21,6 +21,7 @@
 #include "libavutil/attributes.h"
 #include "libavutil/samplefmt.h"
 #include "flacdsp.h"
+#include "config.h"
 
 #define SAMPLE_SIZE 16
 #define PLANAR 0
@@ -119,4 +120,7 @@ av_cold void ff_flacdsp_init(FLACDSPContext *c, enum AVSampleFormat fmt,
         c->decorrelate[3] = flac_decorrelate_ms_c_16p;
         break;
     }
+
+    if (ARCH_ARM)
+        ff_flacdsp_init_arm(c, fmt, bps);
 }
diff --git a/libavcodec/flacdsp.h b/libavcodec/flacdsp.h
index efb5e02ef4..00be2659ce 100644
--- a/libavcodec/flacdsp.h
+++ b/libavcodec/flacdsp.h
@@ -30,5 +30,6 @@ typedef struct FLACDSPContext {
 } FLACDSPContext;
 
 void ff_flacdsp_init(FLACDSPContext *c, enum AVSampleFormat fmt, int bps);
+void ff_flacdsp_init_arm(FLACDSPContext *c, enum AVSampleFormat fmt, int bps);
 
 #endif /* AVCODEC_FLACDSP_H */
diff --git a/libavcodec/mpegvideo_motion.c b/libavcodec/mpegvideo_motion.c
index 354b60b386..2e5f7e611e 100644
--- a/libavcodec/mpegvideo_motion.c
+++ b/libavcodec/mpegvideo_motion.c
@@ -30,9 +30,9 @@
 #include "msmpeg4.h"
 #include <limits.h>
 
-static inline void gmc1_motion(MpegEncContext *s,
-                               uint8_t *dest_y, uint8_t *dest_cb, uint8_t *dest_cr,
-                               uint8_t **ref_picture)
+static void gmc1_motion(MpegEncContext *s,
+                        uint8_t *dest_y, uint8_t *dest_cb, uint8_t *dest_cr,
+                        uint8_t **ref_picture)
 {
     uint8_t *ptr;
     int offset, src_x, src_y, linesize, uvlinesize;
@@ -116,9 +116,9 @@ static inline void gmc1_motion(MpegEncContext *s,
     return;
 }
 
-static inline void gmc_motion(MpegEncContext *s,
-                               uint8_t *dest_y, uint8_t *dest_cb, uint8_t *dest_cr,
-                               uint8_t **ref_picture)
+static void gmc_motion(MpegEncContext *s,
+                       uint8_t *dest_y, uint8_t *dest_cb, uint8_t *dest_cr,
+                       uint8_t **ref_picture)
 {
     uint8_t *ptr;
     int linesize, uvlinesize;
@@ -174,11 +174,8 @@ static inline void gmc_motion(MpegEncContext *s,
 
 static inline int hpel_motion(MpegEncContext *s,
                                   uint8_t *dest, uint8_t *src,
-                                  int field_based, int field_select,
                                   int src_x, int src_y,
-                                  int width, int height, int stride,
-                                  int h_edge_pos, int v_edge_pos,
-                                  int w, int h, op_pixels_func *pix_op,
+                                  op_pixels_func *pix_op,
                                   int motion_x, int motion_y)
 {
     int dxy;
@@ -189,26 +186,24 @@ static inline int hpel_motion(MpegEncContext *s,
     src_y += motion_y >> 1;
 
     /* WARNING: do no forget half pels */
-    src_x = av_clip(src_x, -16, width); //FIXME unneeded for emu?
-    if (src_x == width)
+    src_x = av_clip(src_x, -16, s->width); //FIXME unneeded for emu?
+    if (src_x == s->width)
         dxy &= ~1;
-    src_y = av_clip(src_y, -16, height);
-    if (src_y == height)
+    src_y = av_clip(src_y, -16, s->height);
+    if (src_y == s->height)
         dxy &= ~2;
-    src += src_y * stride + src_x;
+    src += src_y * s->linesize + src_x;
 
     if(s->unrestricted_mv && (s->flags&CODEC_FLAG_EMU_EDGE)){
-        if(   (unsigned)src_x > FFMAX(h_edge_pos - (motion_x&1) - w, 0)
-           || (unsigned)src_y > FFMAX(v_edge_pos - (motion_y&1) - h, 0)){
-            s->dsp.emulated_edge_mc(s->edge_emu_buffer, src, s->linesize, w+1, (h+1)<<field_based,
-                             src_x, src_y<<field_based, h_edge_pos, s->v_edge_pos);
+        if(   (unsigned)src_x > FFMAX(s->h_edge_pos - (motion_x&1) - 8, 0)
+           || (unsigned)src_y > FFMAX(s->v_edge_pos - (motion_y&1) - 8, 0)){
+            s->dsp.emulated_edge_mc(s->edge_emu_buffer, src, s->linesize, 9, 9,
+                             src_x, src_y, s->h_edge_pos, s->v_edge_pos);
             src= s->edge_emu_buffer;
             emu=1;
         }
     }
-    if(field_select)
-        src += s->linesize;
-    pix_op[dxy](dest, src, stride, h);
+    pix_op[dxy](dest, src, s->linesize, 8);
     return emu;
 }
 
@@ -447,11 +442,9 @@ static inline void obmc_motion(MpegEncContext *s,
             ptr[i]= ptr[MID];
         }else{
             ptr[i]= s->obmc_scratchpad + 8*(i&1) + s->linesize*8*(i>>1);
-            hpel_motion(s, ptr[i], src, 0, 0,
+            hpel_motion(s, ptr[i], src,
                         src_x, src_y,
-                        s->width, s->height, s->linesize,
-                        s->h_edge_pos, s->v_edge_pos,
-                        8, 8, pix_op,
+                        pix_op,
                         mv[i][0], mv[i][1]);
         }
     }
@@ -554,11 +547,12 @@ static inline void qpel_motion(MpegEncContext *s,
 /**
  * h263 chroma 4mv motion compensation.
  */
-static inline void chroma_4mv_motion(MpegEncContext *s,
-                                     uint8_t *dest_cb, uint8_t *dest_cr,
-                                     uint8_t **ref_picture,
-                                     op_pixels_func *pix_op,
-                                     int mx, int my){
+static void chroma_4mv_motion(MpegEncContext *s,
+                              uint8_t *dest_cb, uint8_t *dest_cr,
+                              uint8_t **ref_picture,
+                              op_pixels_func *pix_op,
+                              int mx, int my)
+{
     int dxy, emu=0, src_x, src_y, offset;
     uint8_t *ptr;
 
@@ -773,11 +767,9 @@ static av_always_inline void MPV_motion_internal(MpegEncContext *s,
         }else{
             for(i=0;i<4;i++) {
                 hpel_motion(s, dest_y + ((i & 1) * 8) + (i >> 1) * 8 * s->linesize,
-                            ref_picture[0], 0, 0,
+                            ref_picture[0],
                             mb_x * 16 + (i & 1) * 8, mb_y * 16 + (i >>1) * 8,
-                            s->width, s->height, s->linesize,
-                            s->h_edge_pos, s->v_edge_pos,
-                            8, 8, pix_op[1],
+                            pix_op[1],
                             s->mv[dir][i][0], s->mv[dir][i][1]);
 
                 mx += s->mv[dir][i][0];
diff --git a/libavcodec/nellymoserdec.c b/libavcodec/nellymoserdec.c
index 20ab8f84cb..ae98bf5c31 100644
--- a/libavcodec/nellymoserdec.c
+++ b/libavcodec/nellymoserdec.c
@@ -48,13 +48,11 @@
 typedef struct NellyMoserDecodeContext {
     AVCodecContext* avctx;
     AVFrame         frame;
-    float          *float_buf;
     AVLFG           random_state;
     GetBitContext   gb;
     float           scale_bias;
     DSPContext      dsp;
     FFTContext      imdct_ctx;
-    FmtConvertContext fmt_conv;
     DECLARE_ALIGNED(32, float, imdct_buf)[2][NELLY_BUF_LEN];
     float          *imdct_out;
     float          *imdct_prev;
@@ -124,19 +122,8 @@ static av_cold int decode_init(AVCodecContext * avctx) {
 
     ff_dsputil_init(&s->dsp, avctx);
 
-    if (avctx->request_sample_fmt == AV_SAMPLE_FMT_FLT) {
-        s->scale_bias = 1.0/(32768*8);
-        avctx->sample_fmt = AV_SAMPLE_FMT_FLT;
-    } else {
-        s->scale_bias = 1.0/(1*8);
-        avctx->sample_fmt = AV_SAMPLE_FMT_S16;
-        ff_fmt_convert_init(&s->fmt_conv, avctx);
-        s->float_buf = av_mallocz(NELLY_SAMPLES * sizeof(*s->float_buf));
-        if (!s->float_buf) {
-            av_log(avctx, AV_LOG_ERROR, "error allocating float buffer\n");
-            return AVERROR(ENOMEM);
-        }
-    }
+    s->scale_bias = 1.0/(32768*8);
+    avctx->sample_fmt = AV_SAMPLE_FMT_FLT;
 
     /* Generate overlap window */
     if (!ff_sine_128[127])
@@ -158,7 +145,6 @@ static int decode_tag(AVCodecContext *avctx, void *data,
     int buf_size = avpkt->size;
     NellyMoserDecodeContext *s = avctx->priv_data;
     int blocks, i, ret;
-    int16_t *samples_s16;
     float   *samples_flt;
 
     blocks     = buf_size / NELLY_BLOCK_LEN;
@@ -188,18 +174,11 @@ static int decode_tag(AVCodecContext *avctx, void *data,
         av_log(avctx, AV_LOG_ERROR, "get_buffer() failed\n");
         return ret;
     }
-    samples_s16 = (int16_t *)s->frame.data[0];
     samples_flt = (float   *)s->frame.data[0];
 
     for (i=0 ; i<blocks ; i++) {
-        if (avctx->sample_fmt == AV_SAMPLE_FMT_FLT) {
-            nelly_decode_block(s, buf, samples_flt);
-            samples_flt += NELLY_SAMPLES;
-        } else {
-            nelly_decode_block(s, buf, s->float_buf);
-            s->fmt_conv.float_to_int16(samples_s16, s->float_buf, NELLY_SAMPLES);
-            samples_s16 += NELLY_SAMPLES;
-        }
+        nelly_decode_block(s, buf, samples_flt);
+        samples_flt += NELLY_SAMPLES;
         buf += NELLY_BLOCK_LEN;
     }
 
@@ -212,7 +191,6 @@ static int decode_tag(AVCodecContext *avctx, void *data,
 static av_cold int decode_end(AVCodecContext * avctx) {
     NellyMoserDecodeContext *s = avctx->priv_data;
 
-    av_freep(&s->float_buf);
     ff_mdct_end(&s->imdct_ctx);
 
     return 0;
@@ -229,6 +207,5 @@ AVCodec ff_nellymoser_decoder = {
     .capabilities   = CODEC_CAP_DR1 | CODEC_CAP_PARAM_CHANGE,
     .long_name      = NULL_IF_CONFIG_SMALL("Nellymoser Asao"),
     .sample_fmts    = (const enum AVSampleFormat[]) { AV_SAMPLE_FMT_FLT,
-                                                      AV_SAMPLE_FMT_S16,
                                                       AV_SAMPLE_FMT_NONE },
 };
diff --git a/libavutil/arm/asm.S b/libavutil/arm/asm.S
index ccb52316c7..340ee18e8a 100644
--- a/libavutil/arm/asm.S
+++ b/libavutil/arm/asm.S
@@ -186,6 +186,12 @@ ELF     .size   \name, . - \name
 #endif
 .endm
 
+.macro  add_sh          rd,  rn,  rm,  sh:vararg
+A       add             \rd, \rn, \rm, \sh
+T       mov             \rm, \rm, \sh
+T       add             \rd, \rn, \rm
+.endm
+
 .macro  ldr_pre         rt,  rn,  rm:vararg
 A       ldr             \rt, [\rn, \rm]!
 T       add             \rn, \rn, \rm