Merge to M46: upgrade openjpeg to commit# cf352af

third_party/libopenjpeg20/dwt.c

 /*
  * The copyright in this software is being made available under the 2-clauses 
  * BSD License, included below. This software may be subject to other third 
  * party and contributor rights, including patent rights, and no such rights
  * are granted under this license.
  *
  * Copyright (c) 2002-2014, Universite catholique de Louvain (UCL), Belgium
  * Copyright (c) 2002-2014, Professor Benoit Macq
  * Copyright (c) 2001-2003, David Janssens
  * Copyright (c) 2002-2003, Yannick Verschueren
@@ skipping 107 matching lines @@
 /**
 Explicit calculation of the Quantization Stepsizes 
 */
 static void opj_dwt_encode_stepsize(OPJ_INT32 stepsize, OPJ_INT32 numbps, opj_stepsize_t *bandno_stepsize);
 /**
 Inverse wavelet transform in 2-D.
 */
 static OPJ_BOOL opj_dwt_decode_tile(const opj_tcd_tilecomp_t* tilec, OPJ_UINT32 i, DWT1DFN fn);
 
 static OPJ_BOOL opj_dwt_encode_procedure(const opj_tcd_tilecomp_t * tilec,
-                                                                                    void (*p_function)(OPJ_INT32 *, OPJ_SIZE_T,OPJ_INT32,OPJ_INT32,OPJ_INT32) );
+                                                                                    void(*p_function)(OPJ_INT32 *, OPJ_SIZE_T, OPJ_INT32, OPJ_INT32, OPJ_INT32));
 
 static OPJ_UINT32 opj_dwt_max_resolution(opj_tcd_resolution_t* restrict r, OPJ_UINT32 i);
 
 /* <summary>                             */
 /* Inverse 9-7 wavelet transform in 1-D. */
 /* </summary>                            */
 static void opj_v4dwt_decode(opj_v4dwt_t* restrict dwt);
 
 static void opj_v4dwt_interleave_h(opj_v4dwt_t* restrict w, OPJ_FLOAT32* restrict a, OPJ_INT32 x, OPJ_INT32 size);
 
 static void opj_v4dwt_interleave_v(opj_v4dwt_t* restrict v , OPJ_FLOAT32* restrict a , OPJ_INT32 x, OPJ_INT32 nb_elts_read);
 
 #ifdef __SSE__
 static void opj_v4dwt_decode_step1_sse(opj_v4_t* w, OPJ_INT32 count, const __m128 c);
 
 static void opj_v4dwt_decode_step2_sse(opj_v4_t* l, opj_v4_t* w, OPJ_INT32 k, OPJ_INT32 m, __m128 c);
 
 #else
 static void opj_v4dwt_decode_step1(opj_v4_t* w, OPJ_INT32 count, const OPJ_FLOAT32 c);
 
 static void opj_v4dwt_decode_step2(opj_v4_t* l, opj_v4_t* w, OPJ_INT32 k, OPJ_INT32 m, OPJ_FLOAT32 c);
 
 #endif
 
 /*@}*/
 
 /*@}*/
 
 #define IDX_S(i) (i)*2
-#define IDX_D(i) 1+(i)*2
-#define UNDERFLOW_SN(i) ((i)>=sn&&sn>0)
-#define UNDERFLOW_DN(i) ((i)>=dn&&dn>0)
-#define OVERFLOW_S(i) (IDX_S(i)>=a_count)
-#define OVERFLOW_D(i) (IDX_D(i)>=a_count)
+#define IDX_D(i) 1 + (i)* 2
+#define UNDERFLOW_SN(i) ((i) >= sn&&sn>0)
+#define UNDERFLOW_DN(i) ((i) >= dn&&dn>0)
+#define OVERFLOW_S(i) (IDX_S(i) >= a_count)
+#define OVERFLOW_D(i) (IDX_D(i) >= a_count)
 
 #define OPJ_S(i) a[IDX_S(i)]
 #define OPJ_D(i) a[IDX_D(i)]
-#define OPJ_S_(i) ((i)<0?OPJ_S(0):(UNDERFLOW_SN(i)?OPJ_S(sn-1):OVERFLOW_S(i)?OPJ_S(i-1):OPJ_S(i)))
-#define OPJ_D_(i) ((i)<0?OPJ_D(0):(UNDERFLOW_DN(i)?OPJ_D(dn-1):OVERFLOW_D(i)?OPJ_D(i-1):OPJ_D(i)))
+#define OPJ_S_(i) ((i)<0 ? OPJ_S(0) : (UNDERFLOW_SN(i) ? OPJ_S(sn - 1) : OVERFLOW_S(i) ? OPJ_S(i - 1) : OPJ_S(i)))
+#define OPJ_D_(i) ((i)<0 ? OPJ_D(0) : (UNDERFLOW_DN(i) ? OPJ_D(dn - 1) : OVERFLOW_D(i) ? OPJ_D(i - 1) : OPJ_D(i)))
 /* new */
-#define OPJ_SS_(i) ((i)<0?OPJ_S(0):(UNDERFLOW_DN(i)?OPJ_S(dn-1):OVERFLOW_S(i)?OPJ_S(i-1):OPJ_S(i)))
-#define OPJ_DD_(i) ((i)<0?OPJ_D(0):(UNDERFLOW_SN(i)?OPJ_D(sn-1):OVERFLOW_D(i)?OPJ_D(i-1):OPJ_D(i)))
+#define OPJ_SS_(i) ((i)<0 ? OPJ_S(0) : (UNDERFLOW_DN(i) ? OPJ_S(dn - 1) : OVERFLOW_S(i) ? OPJ_S(i - 1) : OPJ_S(i)))
+#define OPJ_DD_(i) ((i)<0 ? OPJ_D(0) : (UNDERFLOW_SN(i) ? OPJ_D(sn - 1) : OVERFLOW_D(i) ? OPJ_D(i - 1) : OPJ_D(i)))
 
 /* <summary>                                                              */
 /* This table contains the norms of the 5-3 wavelets for different bands. */
 /* </summary>                                                             */
 static const OPJ_FLOAT64 opj_dwt_norms[4][10] = {
         {1.000, 1.500, 2.750, 5.375, 10.68, 21.34, 42.67, 85.33, 170.7, 341.3},
         {1.038, 1.592, 2.919, 5.703, 11.33, 22.64, 45.25, 90.48, 180.9},
         {1.038, 1.592, 2.919, 5.703, 11.33, 22.64, 45.25, 90.48, 180.9},
         {.7186, .9218, 1.586, 3.043, 6.019, 12.01, 24.00, 47.97, 95.93}
 };
@@ skipping 10 matching lines @@
 
 /* 
 ==========================================================
    local functions
 ==========================================================
 */
 
 /* <summary>                                     */
 /* Forward lazy transform (horizontal).  */
 /* </summary>                            */ 
-void opj_dwt_deinterleave_h(OPJ_INT32 *a, OPJ_INT32 *b, OPJ_INT32 dn, OPJ_INT32 sn, OPJ_INT32 cas) {
+static void opj_dwt_deinterleave_h(OPJ_INT32 *a, OPJ_INT32 *b, OPJ_INT32 dn, OPJ_INT32 sn, OPJ_INT32 cas) {
         OPJ_INT32 i;
         OPJ_INT32 * l_dest = b;
         OPJ_INT32 * l_src = a+cas;
 
     for (i=0; i<sn; ++i) {
                 *l_dest++ = *l_src;
                 l_src += 2;
         }
         
     l_dest = b + sn;
         l_src = a + 1 - cas;
 
     for (i=0; i<dn; ++i)  {
                 *l_dest++=*l_src;
                 l_src += 2;
         }
 }
 
 /* <summary>                             */  
 /* Forward lazy transform (vertical).    */
 /* </summary>                            */ 
-void opj_dwt_deinterleave_v(OPJ_INT32 *a, OPJ_INT32 *b, OPJ_INT32 dn, OPJ_INT32 sn, OPJ_INT32 x, OPJ_INT32 cas) {
+static void opj_dwt_deinterleave_v(OPJ_INT32 *a, OPJ_INT32 *b, OPJ_INT32 dn, OPJ_INT32 sn, OPJ_INT32 x, OPJ_INT32 cas) {
     OPJ_INT32 i = sn;
         OPJ_INT32 * l_dest = b;
         OPJ_INT32 * l_src = a+cas;
 
     while (i--) {
                 *l_dest = *l_src;
                 l_dest += x;
                 l_src += 2;
                 } /* b[i*x]=a[2*i+cas]; */
 
         l_dest = b + sn * x;
         l_src = a + 1 - cas;
         
         i = dn;
     while (i--) {
                 *l_dest = *l_src;
                 l_dest += x;
                 l_src += 2;
         } /*b[(sn+i)*x]=a[(2*i+1-cas)];*/
 }
 
 /* <summary>                             */
 /* Inverse lazy transform (horizontal).  */
 /* </summary>                            */
-void opj_dwt_interleave_h(opj_dwt_t* h, OPJ_INT32 *a) {
+static void opj_dwt_interleave_h(opj_dwt_t* h, OPJ_INT32 *a) {
     OPJ_INT32 *ai = a;
     OPJ_INT32 *bi = h->mem + h->cas;
     OPJ_INT32  i        = h->sn;
     while( i-- ) {
       *bi = *(ai++);
           bi += 2;
     }
     ai  = a + h->sn;
     bi  = h->mem + 1 - h->cas;
     i   = h->dn ;
     while( i-- ) {
       *bi = *(ai++);
           bi += 2;
     }
 }
 
 /* <summary>                             */  
 /* Inverse lazy transform (vertical).    */
 /* </summary>                            */ 
-void opj_dwt_interleave_v(opj_dwt_t* v, OPJ_INT32 *a, OPJ_INT32 x) {
+static void opj_dwt_interleave_v(opj_dwt_t* v, OPJ_INT32 *a, OPJ_INT32 x) {
     OPJ_INT32 *ai = a;
     OPJ_INT32 *bi = v->mem + v->cas;
     OPJ_INT32  i = v->sn;
     while( i-- ) {
       *bi = *ai;
           bi += 2;
           ai += x;
     }
     ai = a + (v->sn * x);
     bi = v->mem + 1 - v->cas;
     i = v->dn ;
     while( i-- ) {
       *bi = *ai;
           bi += 2;  
           ai += x;
     }
 }
 
 
 /* <summary>                            */
 /* Forward 5-3 wavelet transform in 1-D. */
 /* </summary>                           */
-void opj_dwt_encode_1(OPJ_INT32 *a, OPJ_SIZE_T a_count, OPJ_INT32 dn, OPJ_INT32 sn, OPJ_INT32 cas) {
+static void opj_dwt_encode_1(OPJ_INT32 *a, OPJ_SIZE_T a_count, OPJ_INT32 dn, OPJ_INT32 sn, OPJ_INT32 cas) {
         OPJ_INT32 i;
         
         if (!cas) {
                 if ((dn > 0) || (sn > 1)) {     /* NEW :  CASE ONE ELEMENT */
                         for (i = 0; i < dn; i++) OPJ_D(i) -= (OPJ_S_(i) + OPJ_S_(i + 1)) >> 1;
                         for (i = 0; i < sn; i++) OPJ_S(i) += (OPJ_D_(i - 1) + OPJ_D_(i) + 2) >> 2;
                 }
         } else {
                 if (!sn && dn == 1)                 /* NEW :  CASE ONE ELEMENT */
                         OPJ_S(0) *= 2;
                 else {
                         for (i = 0; i < dn; i++) OPJ_S(i) -= (OPJ_DD_(i) + OPJ_DD_(i - 1)) >> 1;
                         for (i = 0; i < sn; i++) OPJ_D(i) += (OPJ_SS_(i) + OPJ_SS_(i + 1) + 2) >> 2;
                 }
         }
 }
 
 /* <summary>                            */
 /* Inverse 5-3 wavelet transform in 1-D. */
 /* </summary>                           */ 
-void opj_dwt_decode_1_(OPJ_INT32 *a, OPJ_SIZE_T a_count, OPJ_INT32 dn, OPJ_INT32 sn, OPJ_INT32 cas) {
+static void opj_dwt_decode_1_(OPJ_INT32 *a, OPJ_SIZE_T a_count, OPJ_INT32 dn, OPJ_INT32 sn, OPJ_INT32 cas) {
         OPJ_INT32 i;
         
         if (!cas) {
                 if ((dn > 0) || (sn > 1)) { /* NEW :  CASE ONE ELEMENT */
                         for (i = 0; i < sn; i++) OPJ_S(i) -= (OPJ_D_(i - 1) + OPJ_D_(i) + 2) >> 2;
                         for (i = 0; i < dn; i++) OPJ_D(i) += (OPJ_S_(i) + OPJ_S_(i + 1)) >> 1;
                 }
         } else {
                 if (!sn  && dn == 1)          /* NEW :  CASE ONE ELEMENT */
                         OPJ_S(0) /= 2;
                 else {
                         for (i = 0; i < sn; i++) OPJ_D(i) -= (OPJ_SS_(i) + OPJ_SS_(i + 1) + 2) >> 2;
                         for (i = 0; i < dn; i++) OPJ_S(i) += (OPJ_DD_(i) + OPJ_DD_(i - 1)) >> 1;
                 }
         }
 }
 
 /* <summary>                            */
 /* Inverse 5-3 wavelet transform in 1-D. */
 /* </summary>                           */ 
-void opj_dwt_decode_1(opj_dwt_t *v) {
+static void opj_dwt_decode_1(opj_dwt_t *v) {
         opj_dwt_decode_1_(v->mem, v->mem_count, v->dn, v->sn, v->cas);
 }
 
 /* <summary>                             */
 /* Forward 9-7 wavelet transform in 1-D. */
 /* </summary>                            */
-void opj_dwt_encode_1_real(OPJ_INT32 *a, OPJ_SIZE_T a_count, OPJ_INT32 dn, OPJ_INT32 sn, OPJ_INT32 cas) {
+static void opj_dwt_encode_1_real(OPJ_INT32 *a, OPJ_SIZE_T a_count, OPJ_INT32 dn, OPJ_INT32 sn, OPJ_INT32 cas) {
         OPJ_INT32 i;
         if (!cas) {
                 if ((dn > 0) || (sn > 1)) {     /* NEW :  CASE ONE ELEMENT */
                         for (i = 0; i < dn; i++)
                                 OPJ_D(i) -= opj_int_fix_mul(OPJ_S_(i) + OPJ_S_(i + 1), 12993);
                         for (i = 0; i < sn; i++)
                                 OPJ_S(i) -= opj_int_fix_mul(OPJ_D_(i - 1) + OPJ_D_(i), 434);
                         for (i = 0; i < dn; i++)
                                 OPJ_D(i) += opj_int_fix_mul(OPJ_S_(i) + OPJ_S_(i + 1), 7233);
                         for (i = 0; i < sn; i++)
@@ skipping 14 matching lines @@
                         for (i = 0; i < sn; i++)
                                 OPJ_D(i) += opj_int_fix_mul(OPJ_SS_(i) + OPJ_SS_(i + 1), 3633);
                         for (i = 0; i < dn; i++)
                                 OPJ_S(i) = opj_int_fix_mul(OPJ_S(i), 5038);     /*5038 */
                         for (i = 0; i < sn; i++)
                                 OPJ_D(i) = opj_int_fix_mul(OPJ_D(i), 6659);     /*6660 */
                 }
         }
 }
 
-void opj_dwt_encode_stepsize(OPJ_INT32 stepsize, OPJ_INT32 numbps, opj_stepsize_t *bandno_stepsize) {
+static void opj_dwt_encode_stepsize(OPJ_INT32 stepsize, OPJ_INT32 numbps, opj_stepsize_t *bandno_stepsize) {
         OPJ_INT32 p, n;
         p = opj_int_floorlog2(stepsize) - 13;
         n = 11 - opj_int_floorlog2(stepsize);
         bandno_stepsize->mant = (n < 0 ? stepsize >> -n : stepsize << n) & 0x7ff;
         bandno_stepsize->expn = numbps - p;
 }
 
 /* 
 ==========================================================
    DWT interface
 ==========================================================
 */
 
 
 /* <summary>                            */
 /* Forward 5-3 wavelet transform in 2-D. */
 /* </summary>                           */
-INLINE OPJ_BOOL opj_dwt_encode_procedure(const opj_tcd_tilecomp_t * tilec,void (*p_function)(OPJ_INT32 *, OPJ_SIZE_T,OPJ_INT32,OPJ_INT32,OPJ_INT32) )
+static INLINE OPJ_BOOL opj_dwt_encode_procedure(const opj_tcd_tilecomp_t * tilec, void(*p_function)(OPJ_INT32 *, OPJ_SIZE_T, OPJ_INT32, OPJ_INT32, OPJ_INT32))
 {
         OPJ_INT32 i, j, k;
         OPJ_INT32 *a = 00;
         OPJ_INT32 *aj = 00;
         OPJ_INT32 *bj = 00;
         OPJ_INT32 w, l;
 
         OPJ_INT32 rw;                   /* width of the resolution level computed   */
         OPJ_INT32 rh;                   /* height of the resolution level computed  */
         OPJ_SIZE_T l_data_count;
         OPJ_SIZE_T l_data_size;
 
         opj_tcd_resolution_t * l_cur_res = 0;
         opj_tcd_resolution_t * l_last_res = 0;
 
         w = tilec->x1-tilec->x0;
         l = (OPJ_INT32)tilec->numresolutions-1;
         a = tilec->data;
 
         l_cur_res = tilec->resolutions + l;
         l_last_res = l_cur_res - 1;
 
-        l_data_count = opj_dwt_max_resolution( tilec->resolutions,tilec->numresolutions);
+        l_data_count = opj_dwt_max_resolution(tilec->resolutions, tilec->numresolutions) * (OPJ_UINT32)sizeof(OPJ_INT32);
         l_data_size = l_data_count * (OPJ_UINT32)sizeof(OPJ_INT32);
         bj = (OPJ_INT32*)opj_malloc(l_data_size);
         if (! bj) {
                 return OPJ_FALSE;
         }
         i = l;
 
         while (i--) {
                 OPJ_INT32 rw1;          /* width of the resolution level once lower than computed one                                       */
                 OPJ_INT32 rh1;          /* height of the resolution level once lower than computed one                                      */
@@ skipping 114 matching lines @@
                         OPJ_FLOAT64 norm = opj_dwt_norms_real[orient][level];
                         stepsize = (1 << (gain)) / norm;
                 }
                 opj_dwt_encode_stepsize((OPJ_INT32) floor(stepsize * 8192.0), (OPJ_INT32)(prec + gain), &tccp->stepsizes[bandno]);
         }
 }
 
 /* <summary>                             */
 /* Determine maximum computed resolution level for inverse wavelet transform */
 /* </summary>                            */
-OPJ_UINT32 opj_dwt_max_resolution(opj_tcd_resolution_t* restrict r, OPJ_UINT32 i) {
+static OPJ_UINT32 opj_dwt_max_resolution(opj_tcd_resolution_t* restrict r, OPJ_UINT32 i) {
         OPJ_UINT32 mr   = 0;
         OPJ_UINT32 w;
         while( --i ) {
                 ++r;
                 if( mr < ( w = (OPJ_UINT32)(r->x1 - r->x0) ) )
                         mr = w ;
                 if( mr < ( w = (OPJ_UINT32)(r->y1 - r->y0) ) )
                         mr = w ;
         }
         return mr ;
 }
 
 /* <summary>                            */
 /* Inverse wavelet transform in 2-D.     */
 /* </summary>                           */
-OPJ_BOOL opj_dwt_decode_tile(const opj_tcd_tilecomp_t* tilec, OPJ_UINT32 numres, DWT1DFN dwt_1D) {
+static OPJ_BOOL opj_dwt_decode_tile(const opj_tcd_tilecomp_t* tilec, OPJ_UINT32 numres, DWT1DFN dwt_1D) {
         opj_dwt_t h;
         opj_dwt_t v;
 
         opj_tcd_resolution_t* tr = tilec->resolutions;
 
         OPJ_UINT32 rw = (OPJ_UINT32)(tr->x1 - tr->x0);  /* width of the resolution level computed */
         OPJ_UINT32 rh = (OPJ_UINT32)(tr->y1 - tr->y0);  /* height of the resolution level computed */
 
         OPJ_UINT32 w = (OPJ_UINT32)(tilec->x1 - tilec->x0);
 
@@ skipping 36 matching lines @@
                         (dwt_1D)(&v);
                         for(k = 0; k < rh; ++k) {
                                 tiledp[k * w + j] = v.mem[k];
                         }
                 }
         }
         opj_aligned_free(h.mem);
         return OPJ_TRUE;
 }
 
-void opj_v4dwt_interleave_h(opj_v4dwt_t* restrict w, OPJ_FLOAT32* restrict a, OPJ_INT32 x, OPJ_INT32 size){
+static void opj_v4dwt_interleave_h(opj_v4dwt_t* restrict w, OPJ_FLOAT32* restrict a, OPJ_INT32 x, OPJ_INT32 size){
         OPJ_FLOAT32* restrict bi = (OPJ_FLOAT32*) (w->wavelet + w->cas);
         OPJ_INT32 count = w->sn;
         OPJ_INT32 i, k;
 
         for(k = 0; k < 2; ++k){
                 if ( count + 3 * x < size && ((size_t) a & 0x0f) == 0 && ((size_t) bi & 0x0f) == 0 && (x & 0x0f) == 0 ) {
                         /* Fast code path */
                         for(i = 0; i < count; ++i){
                                 OPJ_INT32 j = i;
                                 bi[i*8    ] = a[j];
@@ skipping 22 matching lines @@
                         }
                 }
 
                 bi = (OPJ_FLOAT32*) (w->wavelet + 1 - w->cas);
                 a += w->sn;
                 size -= w->sn;
                 count = w->dn;
         }
 }
 
-void opj_v4dwt_interleave_v(opj_v4dwt_t* restrict v , OPJ_FLOAT32* restrict a , OPJ_INT32 x, OPJ_INT32 nb_elts_read){
+static void opj_v4dwt_interleave_v(opj_v4dwt_t* restrict v , OPJ_FLOAT32* restrict a , OPJ_INT32 x, OPJ_INT32 nb_elts_read){
         opj_v4_t* restrict bi = v->wavelet + v->cas;
         OPJ_INT32 i;
 
         for(i = 0; i < v->sn; ++i){
                 memcpy(&bi[i*2], &a[i*x], (size_t)nb_elts_read * sizeof(OPJ_FLOAT32));
         }
 
         a += v->sn * x;
         bi = v->wavelet + 1 - v->cas;
 
         for(i = 0; i < v->dn; ++i){
                 memcpy(&bi[i*2], &a[i*x], (size_t)nb_elts_read * sizeof(OPJ_FLOAT32));
         }
 }
 
 #ifdef __SSE__
 
-void opj_v4dwt_decode_step1_sse(opj_v4_t* w, OPJ_INT32 count, const __m128 c){
+static void opj_v4dwt_decode_step1_sse(opj_v4_t* w, OPJ_INT32 count, const __m128 c){
         __m128* restrict vw = (__m128*) w;
         OPJ_INT32 i;
         /* 4x unrolled loop */
         for(i = 0; i < count >> 2; ++i){
                 *vw = _mm_mul_ps(*vw, c);
                 vw += 2;
                 *vw = _mm_mul_ps(*vw, c);
                 vw += 2;
                 *vw = _mm_mul_ps(*vw, c);
                 vw += 2;
@@ skipping 28 matching lines @@
         c = _mm_mul_ps(c, vl[0]);
         for(; m < k; ++m){
                 __m128 tmp = vw[-1];
                 vw[-1] = _mm_add_ps(tmp, c);
                 vw += 2;
         }
 }
 
 #else
 
-void opj_v4dwt_decode_step1(opj_v4_t* w, OPJ_INT32 count, const OPJ_FLOAT32 c)
+static void opj_v4dwt_decode_step1(opj_v4_t* w, OPJ_INT32 count, const OPJ_FLOAT32 c)
 {
         OPJ_FLOAT32* restrict fw = (OPJ_FLOAT32*) w;
         OPJ_INT32 i;
         for(i = 0; i < count; ++i){
                 OPJ_FLOAT32 tmp1 = fw[i*8    ];
                 OPJ_FLOAT32 tmp2 = fw[i*8 + 1];
                 OPJ_FLOAT32 tmp3 = fw[i*8 + 2];
                 OPJ_FLOAT32 tmp4 = fw[i*8 + 3];
                 fw[i*8    ] = tmp1 * c;
                 fw[i*8 + 1] = tmp2 * c;
                 fw[i*8 + 2] = tmp3 * c;
                 fw[i*8 + 3] = tmp4 * c;
         }
 }
 
-void opj_v4dwt_decode_step2(opj_v4_t* l, opj_v4_t* w, OPJ_INT32 k, OPJ_INT32 m, OPJ_FLOAT32 c)
+static void opj_v4dwt_decode_step2(opj_v4_t* l, opj_v4_t* w, OPJ_INT32 k, OPJ_INT32 m, OPJ_FLOAT32 c)
 {
         OPJ_FLOAT32* restrict fl = (OPJ_FLOAT32*) l;
         OPJ_FLOAT32* restrict fw = (OPJ_FLOAT32*) w;
         OPJ_INT32 i;
         for(i = 0; i < m; ++i){
                 OPJ_FLOAT32 tmp1_1 = fl[0];
                 OPJ_FLOAT32 tmp1_2 = fl[1];
                 OPJ_FLOAT32 tmp1_3 = fl[2];
                 OPJ_FLOAT32 tmp1_4 = fl[3];
                 OPJ_FLOAT32 tmp2_1 = fw[-4];
@@ skipping 168 matching lines @@
 
                         for(k = 0; k < rh; ++k){
                                 memcpy(&aj[k*w], &v.wavelet[k], (size_t)j * sizeof(OPJ_FLOAT32));
                         }
                 }
         }
 
         opj_aligned_free(h.wavelet);
         return OPJ_TRUE;
 }