diff --git a/ChangeLog b/ChangeLog
index 870dfe64..bde04546 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,5 +1,20 @@
 # Note that this ChangeLog covers only main changes to the release branch
 
+2019-11-12 14:48 Isibaar
+	* xvidcore/src/bitstream/bitstream.c
+	* xvidcore/src/bitstream/bitstream.h
+	* xvidcore/src/bitstream/mbcoding.c
+	* xvidcore/src/decoder.c
+	* xvidcore/src/global.h
+	* xvidcore/src/image/colorspace.c
+	* xvidcore/src/motion/gmc.c
+	* xvidcore/src/quant/quant_matrix.c
+	* xvidcore/src/xvid.h: Backporting patches from HEAD
+
+2019-01-17 10:33 Isibaar
+	* xvidcore/src/nasm.inc: always use .text for machine (Patch 
+	  by Peter Ross)
+
 2017-07-12 08:23 Isibaar
 	* xvidcore/vfw/src/codec.c: Re-add support to decode raw YV12 
 	  input FourCC video
diff --git a/build/generic/configure.in b/build/generic/configure.in
index 902fc418..a3c0c34d 100644
--- a/build/generic/configure.in
+++ b/build/generic/configure.in
@@ -8,7 +8,7 @@ dnl ==========================================================================
 
 AC_PREREQ([2.50])
 
-AC_INIT([Xvid], [1.3.5], [xvid-devel@xvid.org])
+AC_INIT([Xvid], [1.3.6], [xvid-devel@xvid.org])
 AC_CONFIG_SRCDIR(configure.in)
 
 dnl Do not forget to increase that when needed.
diff --git a/src/bitstream/bitstream.c b/src/bitstream/bitstream.c
index 5f318d36..886a0927 100644
--- a/src/bitstream/bitstream.c
+++ b/src/bitstream/bitstream.c
@@ -153,6 +153,8 @@ read_video_packet_header(Bitstream *bs,
 		READ_MARKER();
 		if (dec->time_inc_bits)
 			time_increment = (BitstreamGetBits(bs, dec->time_inc_bits));	/* vop_time_increment */
+		else
+			time_increment = 0;
 		READ_MARKER();
 		DPRINTF(XVID_DEBUG_HEADER,"time %i:%i\n", time_base, time_increment);
 
@@ -398,7 +400,6 @@ BitstreamReadHeaders(Bitstream * bs,
 	uint32_t start_code;
 	uint32_t time_incr = 0;
 	int32_t time_increment = 0;
-	int resize = 0;
 
 	while ((BitstreamPos(bs) >> 3) + 4 <= bs->length) {
 
@@ -466,6 +467,7 @@ BitstreamReadHeaders(Bitstream * bs,
 			BitstreamSkip(bs, 32);	/* video_object_start_code */
 
 		} else if ((start_code & ~VIDOBJLAY_START_CODE_MASK) == VIDOBJLAY_START_CODE) {
+			uint32_t width = 0, height = 0;
 
 			DPRINTF(XVID_DEBUG_STARTCODE, "<video_object_layer>\n");
 			DPRINTF(XVID_DEBUG_HEADER, "vol id %i\n", start_code & VIDOBJLAY_START_CODE_MASK);
@@ -566,8 +568,6 @@ BitstreamReadHeaders(Bitstream * bs,
 			if (dec->shape != VIDOBJLAY_SHAPE_BINARY_ONLY) {
 
 				if (dec->shape == VIDOBJLAY_SHAPE_RECTANGULAR) {
-					uint32_t width, height;
-
 					READ_MARKER();
 					width = BitstreamGetBits(bs, 13);	/* video_object_layer_width */
 					READ_MARKER();
@@ -576,18 +576,6 @@ BitstreamReadHeaders(Bitstream * bs,
 
 					DPRINTF(XVID_DEBUG_HEADER, "width %i\n", width);
 					DPRINTF(XVID_DEBUG_HEADER, "height %i\n", height);
-
-					if (dec->width != width || dec->height != height)
-					{
-						if (dec->fixed_dimensions)
-						{
-							DPRINTF(XVID_DEBUG_ERROR, "decoder width/height does not match bitstream\n");
-							return -1;
-						}
-						resize = 1;
-						dec->width = width;
-						dec->height = height;
-					}
 				}
 
 				dec->interlacing = BitstreamGetBit(bs);
@@ -764,7 +752,19 @@ BitstreamReadHeaders(Bitstream * bs,
 
 			}
 
-			return (resize ? -3 : -2 );	/* VOL */
+			if (((width > 0) && (height > 0)) && (dec->width != width || dec->height != height))
+			{
+				if (dec->fixed_dimensions)
+				{
+					DPRINTF(XVID_DEBUG_ERROR, "decoder width/height does not match bitstream\n");
+					return -1;
+				}
+				dec->width = width;
+				dec->height = height;
+				return -3;
+			}
+
+			return -2;	/* VOL */
 
 		} else if (start_code == GRPOFVOP_START_CODE) {
 
@@ -860,8 +860,9 @@ BitstreamReadHeaders(Bitstream * bs,
 				dec->shape == VIDOBJLAY_SHAPE_RECTANGULAR &&
 				(coding_type == P_VOP || coding_type == I_VOP)) {
 
-				if (BitstreamGetBit(bs));
+				if (BitstreamGetBit(bs)) {
 					DPRINTF(XVID_DEBUG_ERROR, "RRV not supported (anymore)\n");
+                                }
 			}
 
 			if (dec->shape != VIDOBJLAY_SHAPE_RECTANGULAR) {
@@ -919,7 +920,7 @@ BitstreamReadHeaders(Bitstream * bs,
 
 				int i;
 
-				for (i = 0 ; i < dec->sprite_warping_points; i++)
+				for (i = 0 ; i < MIN(4, dec->sprite_warping_points); i++)
 				{
 					int length;
 					int x = 0, y = 0;
@@ -981,7 +982,7 @@ BitstreamReadHeaders(Bitstream * bs,
 
 		} else if (start_code == USERDATA_START_CODE) {
 			char tmp[256];
-		    int i, version, build;
+                        int i, version = 0, build = 0;
 			char packed;
 
 			BitstreamSkip(bs, 32);	/* user_data_start_code */
@@ -989,7 +990,7 @@ BitstreamReadHeaders(Bitstream * bs,
 			memset(tmp, 0, 256);
 			tmp[0] = BitstreamShowBits(bs, 8);
 
-			for(i = 1; i < 256; i++){
+			for(i = 1; i < 255; i++){
 				tmp[i] = (BitstreamShowBits(bs, 16) & 0xFF);
 
 				if(tmp[i] == 0)
diff --git a/src/bitstream/bitstream.h b/src/bitstream/bitstream.h
index 2bf6b3f0..73b11f9d 100644
--- a/src/bitstream/bitstream.h
+++ b/src/bitstream/bitstream.h
@@ -171,7 +171,9 @@ void write_video_packet_header(Bitstream * const bs,
 the end of the buffer. Padding might be appropriate. If only chunks
 of 4bytes are applicable, define XVID_SAFE_BS_TAIL. Note that this will
 slow decoding, so consider this as a last-resort solution */
-/* #define XVID_SAFE_BS_TAIL */
+#ifndef XVID_HAVE_PADDED_BS_BUFFER
+#define XVID_SAFE_BS_TAIL
+#endif
 
 /* initialise bitstream structure */
 
diff --git a/src/bitstream/mbcoding.c b/src/bitstream/mbcoding.c
index 06c0d0d4..679809e6 100644
--- a/src/bitstream/mbcoding.c
+++ b/src/bitstream/mbcoding.c
@@ -901,8 +901,8 @@ get_mv_data(Bitstream * bs)
 
 	index -= 4;
 
-	BitstreamSkip(bs, TMNMVtab2[index].len);
-	return TMNMVtab2[index].code;
+	BitstreamSkip(bs, TMNMVtab2[index&0x7f].len);
+	return TMNMVtab2[index&0x7f].code;
 
 }
 
@@ -1965,7 +1965,8 @@ VLC const TMNMVtab2[] = {
 	{14, 10}, {14, 10}, {14, 10}, {14, 10},
 	{-14, 10}, {-14, 10}, {-14, 10}, {-14, 10},
 	{13, 10}, {13, 10}, {13, 10}, {13, 10},
-	{-13, 10}, {-13, 10}, {-13, 10}, {-13, 10}
+	{-13, 10}, {-13, 10}, {-13, 10}, {-13, 10},
+        {0, 0}, {0, 0}, {0, 0}, {0, 0}
 };
 
 short const dc_threshold[] = {
diff --git a/src/decoder.c b/src/decoder.c
index e91fbe1e..f8611fcd 100644
--- a/src/decoder.c
+++ b/src/decoder.c
@@ -149,6 +149,7 @@ decoder_resize(DECODER * dec)
 int
 decoder_create(xvid_dec_create_t * create)
 {
+  int ret = 0;
   DECODER *dec;
 
   if (XVID_VERSION_MAJOR(create->version) != 1) /* v1.x.x */
@@ -169,8 +170,8 @@ decoder_create(xvid_dec_create_t * create)
 
   create->handle = dec;
 
-  dec->width = create->width;
-  dec->height = create->height;
+  dec->width = MAX(0, create->width);
+  dec->height = MAX(0, create->height);
 
   dec->num_threads = MAX(0, create->num_threads);
 
@@ -209,13 +210,10 @@ decoder_create(xvid_dec_create_t * create)
 
   dec->fixed_dimensions = (dec->width > 0 && dec->height > 0);
 
-  if (dec->fixed_dimensions) {
-    int ret = decoder_resize(dec);
-    if (ret == XVID_ERR_MEMORY) create->handle = NULL;
-    return ret;
-  }
-  else
-    return 0;
+  ret = decoder_resize(dec);
+  if (ret == XVID_ERR_MEMORY) create->handle = NULL;
+
+  return ret;
 }
 
 
@@ -266,7 +264,7 @@ decoder_mbintra(DECODER * dec,
   uint32_t stride2 = stride / 2;
   uint32_t next_block = stride * 8;
   uint32_t i;
-  uint32_t iQuant = pMB->quant;
+  uint32_t iQuant = MAX(1, pMB->quant);
   uint8_t *pY_Cur, *pU_Cur, *pV_Cur;
 
   pY_Cur = dec->cur.y + (y_pos << 4) * stride + (x_pos << 4);
@@ -363,7 +361,7 @@ decoder_mb_decode(DECODER * dec,
 
   int stride = dec->edged_width;
   int i;
-  const uint32_t iQuant = pMB->quant;
+  const uint32_t iQuant = MAX(1, pMB->quant);
   const int direction = dec->alternate_vertical_scan ? 2 : 0;
   typedef void (*get_inter_block_function_t)(
       Bitstream * bs,
@@ -1540,9 +1538,11 @@ static void decoder_output(DECODER * dec, IMAGE * img, MACROBLOCK * mbs,
     img = &dec->tmp;
   }
 
-  image_output(img, dec->width, dec->height,
-         dec->edged_width, (uint8_t**)frame->output.plane, frame->output.stride,
-         frame->output.csp, dec->interlacing);
+  if ((frame->output.plane[0] != NULL) && (frame->output.stride[0] >= dec->width)) {
+    image_output(img, dec->width, dec->height,
+           dec->edged_width, (uint8_t**)frame->output.plane, frame->output.stride,
+           frame->output.csp, dec->interlacing);
+  }
 
   if (stats) {
     stats->type = coding2type(coding_type);
@@ -1565,19 +1565,20 @@ decoder_decode(DECODER * dec,
 {
 
   Bitstream bs;
-  uint32_t rounding;
+  uint32_t rounding = 0;
   uint32_t quant = 2;
-  uint32_t fcode_forward;
-  uint32_t fcode_backward;
-  uint32_t intra_dc_threshold;
+  uint32_t fcode_forward = 0;
+  uint32_t fcode_backward = 0;
+  uint32_t intra_dc_threshold = 0;
   WARPPOINTS gmc_warp;
-  int coding_type;
+  int coding_type = -1;
   int success, output, seen_something;
 
   if (XVID_VERSION_MAJOR(frame->version) != 1 || (stats && XVID_VERSION_MAJOR(stats->version) != 1))  /* v1.x.x */
     return XVID_ERR_VERSION;
 
   start_global_timer();
+  memset((void *)&gmc_warp, 0, sizeof(WARPPOINTS));
 
   dec->low_delay_default = (frame->general & XVID_LOWDELAY);
   if ((frame->general & XVID_DISCONTINUITY))
@@ -1664,7 +1665,7 @@ decoder_decode(DECODER * dec,
     goto repeat;
   }
 
-  if(dec->frames == 0 && coding_type != I_VOP) {
+  if((dec->frames == 0 && coding_type != I_VOP) || (!dec->width || !dec->height)) {
     /* 1st frame is not an i-vop */
     goto repeat;
   }
diff --git a/src/global.h b/src/global.h
index 67e1d37d..83fe5ba4 100644
--- a/src/global.h
+++ b/src/global.h
@@ -81,7 +81,7 @@ VECTOR;
 
 typedef struct
 {
-	VECTOR duv[3];
+	VECTOR duv[4];
 }
 WARPPOINTS;
 
diff --git a/src/image/colorspace.c b/src/image/colorspace.c
index d72f6c80..01438e69 100644
--- a/src/image/colorspace.c
+++ b/src/image/colorspace.c
@@ -102,6 +102,7 @@ NAME(uint8_t * x_ptr, int x_stride,	\
 	int y_dif = y_stride - fixed_width;				\
 	int uv_dif = uv_stride - (fixed_width / 2);		\
 	int x, y;										\
+	if ((x_ptr == NULL) || (x_dif < 0)) return;		\
 	if (vflip) {								\
 		x_ptr += (height - 1) * x_stride;			\
 		x_dif = -(SIZE)*fixed_width - x_stride;		\
diff --git a/src/image/image.c b/src/image/image.c
index b61f957f..a3d1323d 100644
--- a/src/image/image.c
+++ b/src/image/image.c
@@ -415,10 +415,10 @@ safe_packed_conv(uint8_t * x_ptr, int x_stride,
 
     /* packed conversions require height to be divisable by 2
        (or even by 4 for interlaced conversion) */
-    if (interlacing)
-        height_opt = height & (~3);
-    else
-        height_opt = height & (~1);
+       if (interlacing)
+               height_opt = height & (~3);
+       else
+               height_opt = height & (~1);
 
 	func_opt(x_ptr, x_stride,
 			y_ptr, u_ptr, v_ptr, y_stride, uv_stride,
diff --git a/src/motion/gmc.c b/src/motion/gmc.c
index b1780b8a..24192f44 100644
--- a/src/motion/gmc.c
+++ b/src/motion/gmc.c
@@ -269,7 +269,7 @@ void Predict_1pt_16x16_C(const NEW_GMC_DATA * const This,
 {
 	const int W	 = This->sW;
 	const int H	 = This->sH;
-	const int rho = 3-This->accuracy;
+	const int rho = 3-MIN(This->accuracy, 3);
 	const int32_t Rounder = ( 128 - (rounding<<(2*rho)) ) << 16;
 
 
diff --git a/src/nasm.inc b/src/nasm.inc
index e5181c93..547011c9 100644
--- a/src/nasm.inc
+++ b/src/nasm.inc
@@ -1,221 +1,213 @@
-;/****************************************************************************
-; *
-; *  XVID MPEG-4 VIDEO CODEC
-; *  - NASM common header -
-; *
-; *  Copyright (C) 2008 Michael Militzer
-; *
-; *  This program is free software ; you can redistribute it and/or modify
-; *  it under the terms of the GNU General Public License as published by
-; *  the Free Software Foundation ; either version 2 of the License, or
-; *  (at your option) any later version.
-; *
-; *  This program 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 General Public License for more details.
-; *
-; *  You should have received a copy of the GNU General Public License
-; *  along with this program ; if not, write to the Free Software
-; *  Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307 USA
-; *
-; * $Id: nasm.inc,v 1.7.2.2 2011-02-25 12:40:25 Isibaar Exp $
-; *
-; ***************************************************************************/
-
-%ifdef ARCH_IS_X86_64
-
-BITS 64
-DEFAULT REL
-
-%define SECTION_ALIGN 32
-
-%ifdef WINDOWS
-
-%define prm1 rcx
-%define prm2 rdx
-%define prm3 r8
-%define prm4 r9
-%define prm5 [rsp+40]
-%define prm6 [rsp+48]
-%define prm7 [rsp+56]
-%define prm8 [rsp+64]
-
-%define prm1d ecx
-%define prm2d edx
-%define prm3d r8d
-%define prm4d r9d
-%define prm5d dword prm5
-%define prm6d dword prm6
-%define prm7d dword prm7
-%define prm8d dword prm8
-
-%macro PUSH_XMM6_XMM7 0
-      movdqa [_ESP+PTR_SIZE], xmm6
-      movdqa [_ESP+PTR_SIZE+16], xmm7
-%endmacro
-
-%macro POP_XMM6_XMM7 0
-      movdqa xmm6, [_ESP+PTR_SIZE]
-      movdqa xmm7, [_ESP+PTR_SIZE+16]
-%endmacro
-
-%else ; Linux
-
-%define prm1 rdi
-%define prm2 rsi
-%define prm3 rdx
-%define prm4 rcx
-%define prm5 r8
-%define prm6 r9
-%define prm7 [rsp+8]
-%define prm8 [rsp+16]
-
-%define prm1d edi
-%define prm2d esi
-%define prm3d edx
-%define prm4d ecx
-%define prm5d r8d
-%define prm6d r9d
-%define prm7d dword prm7
-%define prm8d dword prm8
-
-%define PUSH_XMM6_XMM7
-%define POP_XMM6_XMM7
-
-%endif
-
-%define _EAX rax
-%define _EBX rbx
-%define _ECX rcx
-%define _EDX rdx
-%define _ESI rsi
-%define _EDI rdi
-%define _EBP rbp
-%define _ESP rsp
-
-%define TMP0 r10
-%define TMP1 r11
-
-%define TMP0d r10d
-%define TMP1d r11d
-
-%define PTR_SIZE 8
-%define PTR_TYPE qword
-
-%ifdef __YASM_VERSION_ID__
-%define XVID_MOVSXD movsxd
-%else
-%define XVID_MOVSXD movsx
-%endif
-
-%else
-
-%define SECTION_ALIGN 16
-
-BITS 32
-
-%define prm1 [esp +  4]
-%define prm2 [esp +  8]
-%define prm3 [esp + 12]
-%define prm4 [esp + 16]
-%define prm5 [esp + 20]
-%define prm6 [esp + 24]
-%define prm7 [esp + 28]
-%define prm8 [esp + 32]
-
-%define prm1d dword prm1
-%define prm2d dword prm2
-%define prm3d dword prm3
-%define prm4d dword prm4
-%define prm5d dword prm5
-%define prm6d dword prm6
-%define prm7d dword prm7
-%define prm8d dword prm8
-
-%define _EAX eax
-%define _EBX ebx
-%define _ECX ecx
-%define _EDX edx
-%define _ESI esi
-%define _EDI edi
-%define _EBP ebp
-%define _ESP esp
-
-%define TMP0 ecx
-%define TMP1 edx
-
-%define TMP0d ecx
-%define TMP1d edx
-
-%define PTR_SIZE 4
-%define PTR_TYPE dword
-
-%define PUSH_XMM6_XMM7
-%define POP_XMM6_XMM7
-
-%define XVID_MOVSXD movsx
-%endif
-
-
-%ifdef WINDOWS
-      %define PREFIX
-%endif
-
-%ifdef NO_PREFIX
-      %undef PREFIX
-%endif
-
-%macro DATA 0
-      %ifdef FORMAT_COFF
-            SECTION .rodata
-      %else
-            SECTION .rodata align=SECTION_ALIGN
-      %endif
-%endmacro
-
-%macro TEXT 0
-%ifidn __OUTPUT_FORMAT__,macho32
-      SECTION .text align=SECTION_ALIGN
-%else
-%ifidn __OUTPUT_FORMAT__,macho64
-      SECTION .text align=SECTION_ALIGN
-%else
-      SECTION .rotext align=SECTION_ALIGN
-%endif
-%endif
-%endmacro
-
-%macro cglobal 1
-	%ifdef PREFIX
-		%ifdef MARK_FUNCS
-			global _%1:function %1.endfunc-%1
-			%define %1 _%1:function %1.endfunc-%1
-			%define ENDFUNC .endfunc:
-		%else
-			global _%1
-			%define %1 _%1
-			%define ENDFUNC
-		%endif
-	%else
-		%ifdef MARK_FUNCS
-			global %1:function %1.endfunc-%1
-			%define ENDFUNC .endfunc:
-		%else
-			global %1
-			%define ENDFUNC
-		%endif
-	%endif
-%endmacro
-
-%macro NON_EXEC_STACK 0
-%ifidn __OUTPUT_FORMAT__,elf
-section .note.GNU-stack noalloc noexec nowrite progbits
-%endif
-%ifidn __OUTPUT_FORMAT__,elf32
-section .note.GNU-stack noalloc noexec nowrite progbits
-%endif
-%ifidn __OUTPUT_FORMAT__,elf64
-section .note.GNU-stack noalloc noexec nowrite progbits
-%endif
-%endmacro
+;/****************************************************************************
+; *
+; *  XVID MPEG-4 VIDEO CODEC
+; *  - NASM common header -
+; *
+; *  Copyright (C) 2008 Michael Militzer
+; *
+; *  This program is free software ; you can redistribute it and/or modify
+; *  it under the terms of the GNU General Public License as published by
+; *  the Free Software Foundation ; either version 2 of the License, or
+; *  (at your option) any later version.
+; *
+; *  This program 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 General Public License for more details.
+; *
+; *  You should have received a copy of the GNU General Public License
+; *  along with this program ; if not, write to the Free Software
+; *  Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307 USA
+; *
+; * $Id: nasm.inc,v 1.7.2.2 2011-02-25 12:40:25 Isibaar Exp $
+; *
+; ***************************************************************************/
+
+%ifdef ARCH_IS_X86_64
+
+BITS 64
+DEFAULT REL
+
+%define SECTION_ALIGN 32
+
+%ifdef WINDOWS
+
+%define prm1 rcx
+%define prm2 rdx
+%define prm3 r8
+%define prm4 r9
+%define prm5 [rsp+40]
+%define prm6 [rsp+48]
+%define prm7 [rsp+56]
+%define prm8 [rsp+64]
+
+%define prm1d ecx
+%define prm2d edx
+%define prm3d r8d
+%define prm4d r9d
+%define prm5d dword prm5
+%define prm6d dword prm6
+%define prm7d dword prm7
+%define prm8d dword prm8
+
+%macro PUSH_XMM6_XMM7 0
+      movdqa [_ESP+PTR_SIZE], xmm6
+      movdqa [_ESP+PTR_SIZE+16], xmm7
+%endmacro
+
+%macro POP_XMM6_XMM7 0
+      movdqa xmm6, [_ESP+PTR_SIZE]
+      movdqa xmm7, [_ESP+PTR_SIZE+16]
+%endmacro
+
+%else ; Linux
+
+%define prm1 rdi
+%define prm2 rsi
+%define prm3 rdx
+%define prm4 rcx
+%define prm5 r8
+%define prm6 r9
+%define prm7 [rsp+8]
+%define prm8 [rsp+16]
+
+%define prm1d edi
+%define prm2d esi
+%define prm3d edx
+%define prm4d ecx
+%define prm5d r8d
+%define prm6d r9d
+%define prm7d dword prm7
+%define prm8d dword prm8
+
+%define PUSH_XMM6_XMM7
+%define POP_XMM6_XMM7
+
+%endif
+
+%define _EAX rax
+%define _EBX rbx
+%define _ECX rcx
+%define _EDX rdx
+%define _ESI rsi
+%define _EDI rdi
+%define _EBP rbp
+%define _ESP rsp
+
+%define TMP0 r10
+%define TMP1 r11
+
+%define TMP0d r10d
+%define TMP1d r11d
+
+%define PTR_SIZE 8
+%define PTR_TYPE qword
+
+%ifdef __YASM_VERSION_ID__
+%define XVID_MOVSXD movsxd
+%else
+%define XVID_MOVSXD movsx
+%endif
+
+%else
+
+%define SECTION_ALIGN 16
+
+BITS 32
+
+%define prm1 [esp +  4]
+%define prm2 [esp +  8]
+%define prm3 [esp + 12]
+%define prm4 [esp + 16]
+%define prm5 [esp + 20]
+%define prm6 [esp + 24]
+%define prm7 [esp + 28]
+%define prm8 [esp + 32]
+
+%define prm1d dword prm1
+%define prm2d dword prm2
+%define prm3d dword prm3
+%define prm4d dword prm4
+%define prm5d dword prm5
+%define prm6d dword prm6
+%define prm7d dword prm7
+%define prm8d dword prm8
+
+%define _EAX eax
+%define _EBX ebx
+%define _ECX ecx
+%define _EDX edx
+%define _ESI esi
+%define _EDI edi
+%define _EBP ebp
+%define _ESP esp
+
+%define TMP0 ecx
+%define TMP1 edx
+
+%define TMP0d ecx
+%define TMP1d edx
+
+%define PTR_SIZE 4
+%define PTR_TYPE dword
+
+%define PUSH_XMM6_XMM7
+%define POP_XMM6_XMM7
+
+%define XVID_MOVSXD movsx
+%endif
+
+
+%ifdef WINDOWS
+      %define PREFIX
+%endif
+
+%ifdef NO_PREFIX
+      %undef PREFIX
+%endif
+
+%macro DATA 0
+      %ifdef FORMAT_COFF
+            SECTION .rodata
+      %else
+            SECTION .rodata align=SECTION_ALIGN
+      %endif
+%endmacro
+
+%macro TEXT 0
+      SECTION .text align=SECTION_ALIGN
+%endmacro
+
+%macro cglobal 1
+	%ifdef PREFIX
+		%ifdef MARK_FUNCS
+			global _%1:function %1.endfunc-%1
+			%define %1 _%1:function %1.endfunc-%1
+			%define ENDFUNC .endfunc:
+		%else
+			global _%1
+			%define %1 _%1
+			%define ENDFUNC
+		%endif
+	%else
+		%ifdef MARK_FUNCS
+			global %1:function %1.endfunc-%1
+			%define ENDFUNC .endfunc:
+		%else
+			global %1
+			%define ENDFUNC
+		%endif
+	%endif
+%endmacro
+
+%macro NON_EXEC_STACK 0
+%ifidn __OUTPUT_FORMAT__,elf
+section .note.GNU-stack noalloc noexec nowrite progbits
+%endif
+%ifidn __OUTPUT_FORMAT__,elf32
+section .note.GNU-stack noalloc noexec nowrite progbits
+%endif
+%ifidn __OUTPUT_FORMAT__,elf64
+section .note.GNU-stack noalloc noexec nowrite progbits
+%endif
+%endmacro
diff --git a/src/quant/quant_matrix.c b/src/quant/quant_matrix.c
index 385858c5..74ef2208 100644
--- a/src/quant/quant_matrix.c
+++ b/src/quant/quant_matrix.c
@@ -24,6 +24,7 @@
  *
  ****************************************************************************/
 
+#include "../global.h"
 #include "quant_matrix.h"
 
 #define FIX(X)   (((X)==1) ? 0xFFFF : ((1UL << 16) / (X) + 1))
@@ -114,7 +115,7 @@ set_intra_matrix(uint16_t * mpeg_quant_matrices, const uint8_t * matrix)
 	uint16_t *intra_matrix = mpeg_quant_matrices + 0*64;
 
 	for (i = 0; i < 64; i++) {
-		intra_matrix[i] = (!i) ? (uint16_t)8: (uint16_t)matrix[i];
+		intra_matrix[i] = (!i) ? (uint16_t)8: (uint16_t)MAX(1, matrix[i]);
 	}
 }
 
@@ -141,7 +142,7 @@ set_inter_matrix(uint16_t * mpeg_quant_matrices, const uint8_t * matrix)
 	uint16_t *inter_matrix_fixl = mpeg_quant_matrices + 7*64;
 
 	for (i = 0; i < 64; i++) {
-		inter_matrix1[i] = ((inter_matrix[i] = (int16_t) matrix[i])>>1);
+		inter_matrix1[i] = ((inter_matrix[i] = (int16_t)MAX(1, (matrix[i])>>1)));
 		inter_matrix1[i] += ((inter_matrix[i] == 1) ? 1: 0);
 		inter_matrix_fix[i] = (uint16_t) FIX(inter_matrix[i]);
 		inter_matrix_fixl[i] = (uint16_t) FIXL(inter_matrix[i]);
diff --git a/src/xvid.c b/src/xvid.c
index ba31c465..9dee35ab 100644
--- a/src/xvid.c
+++ b/src/xvid.c
@@ -699,7 +699,7 @@ xvid_gbl_info(xvid_gbl_info_t * info)
 		return XVID_ERR_VERSION;
 
 	info->actual_version = XVID_VERSION;
-	info->build = "xvid-1.3.5";
+	info->build = "xvid-1.3.6";
 	info->cpu_flags = detect_cpu_flags();
 	info->num_threads = 0; /* single-thread */
 
diff --git a/src/xvid.h b/src/xvid.h
index 2165cf88..648de1b2 100644
--- a/src/xvid.h
+++ b/src/xvid.h
@@ -57,7 +57,7 @@ extern "C" {
 #define XVID_API_MAJOR(a)        (((a)>>16) & 0xff)
 #define XVID_API_MINOR(a)        (((a)>> 0) & 0xff)
 
-#define XVID_VERSION             XVID_MAKE_VERSION(1,3,5)
+#define XVID_VERSION             XVID_MAKE_VERSION(1,3,6)
 #define XVID_API                 XVID_MAKE_API(4, 3)
 
 /* Bitstream Version
@@ -71,7 +71,7 @@ extern "C" {
  * doesnt hurt but not increasing it could cause difficulty for decoders in the
  * future
  */
-#define XVID_BS_VERSION 67 
+#define XVID_BS_VERSION 68 
 
 /*****************************************************************************
  * error codes
@@ -447,8 +447,8 @@ typedef struct
 	int sse_v;              /* [out] V plane's sse */
 /* End of duplicated data, kept only for binary compatibility */
 
-	int bquant_ratio;       /* [in] */
-	int bquant_offset;      /* [in] */
+	int bquant_ratio;       /* [out] */
+	int bquant_offset;      /* [out] */
 
 	xvid_enc_stats_t stats; /* [out] frame statistics */
 } xvid_plg_data_t;