From: Janne Grunau <janne-x264@jannau.net>
Date: Sun, 20 Jul 2014 11:32:10 +0000 (+0200)
Subject: arm: check if the assembler supports the '.func' directive
X-Git-Url: https://git.sesse.net/?a=commitdiff_plain;h=d72760401cb0602b8bf86037988e66cdc810681c;p=x264

arm: check if the assembler supports the '.func' directive

The integrated assembler in llvm trunk (to be released as 3.5) is
otherwise capable enough to assemble the arm asm correctly.
---

diff --git a/common/arm/asm.S b/common/arm/asm.S
index 66568589..3fb11b86 100644
--- a/common/arm/asm.S
+++ b/common/arm/asm.S
@@ -50,6 +50,12 @@
 #   define ELF @
 #endif
 
+#if HAVE_AS_FUNC
+#   define FUNC
+#else
+#   define FUNC @
+#endif
+
 .macro require8, val=1
 ELF     .eabi_attribute 24, \val
 .endm
@@ -59,17 +65,22 @@ ELF     .eabi_attribute 25, \val
 .endm
 
 .macro function name, export=1
+    .macro endfunc
+ELF     .size   \name, . - \name
+FUNC    .endfunc
+        .purgem endfunc
+    .endm
         .align  2
 .if \export == 1
         .global EXTERN_ASM\name
 ELF     .hidden EXTERN_ASM\name
 ELF     .type   EXTERN_ASM\name, %function
-        .func   EXTERN_ASM\name
+FUNC    .func   EXTERN_ASM\name
 EXTERN_ASM\name:
 .else
 ELF     .hidden \name
 ELF     .type   \name, %function
-        .func   \name
+FUNC    .func   \name
 \name:
 .endif
 .endm
diff --git a/common/arm/cpu-a.S b/common/arm/cpu-a.S
index 7fc273a1..9285219d 100644
--- a/common/arm/cpu-a.S
+++ b/common/arm/cpu-a.S
@@ -32,7 +32,7 @@
 function x264_cpu_neon_test
     vadd.i16    q0, q0, q0
     bx          lr
-.endfunc
+endfunc
 
 // return: 0 on success
 //         1 if counters were already enabled
@@ -48,14 +48,14 @@ function x264_cpu_enable_armv7_counter, export=0
     mov         r2, #1 << 31                // enable cycle counter
     mcr         p15, 0, r2, c9, c12, 1      // write CNTENS
     bx          lr
-.endfunc
+endfunc
 
 function x264_cpu_disable_armv7_counter, export=0
     mrc         p15, 0, r0, c9, c12, 0      // read PMNC
     bic         r0, r0, #1                  // disable counters
     mcr         p15, 0, r0, c9, c12, 0      // write PMNC
     bx          lr
-.endfunc
+endfunc
 
 
 .macro READ_TIME r
@@ -105,4 +105,4 @@ average_loop:
     cmp         r0, #10
     movgt       r0, #0
     pop         {r4-r6,pc}
-.endfunc
+endfunc
diff --git a/common/arm/dct-a.S b/common/arm/dct-a.S
index 9e7d75f1..f8d1ccfd 100644
--- a/common/arm/dct-a.S
+++ b/common/arm/dct-a.S
@@ -80,7 +80,7 @@ function x264_dct4x4dc_neon
     vrhadd.s16      d3,  d6,  d7
     vst1.64         {d0-d3}, [r0,:128]
     bx              lr
-.endfunc
+endfunc
 
 function x264_idct4x4dc_neon
     vld1.64         {d0-d3}, [r0,:128]
@@ -92,7 +92,7 @@ function x264_idct4x4dc_neon
     HADAMARD        2, sumsub, d3, d2, d6, d7
     vst1.64         {d0-d3}, [r0,:128]
     bx              lr
-.endfunc
+endfunc
 
 
 .macro DCT_1D d0 d1 d2 d3  d4 d5 d6 d7
@@ -127,7 +127,7 @@ function x264_sub4x4_dct_neon
     DCT_1D          d4, d5, d6, d7, d0, d1, d2, d3
     vst1.64         {d4-d7}, [r0,:128]
     bx              lr
-.endfunc
+endfunc
 
 function x264_sub8x4_dct_neon, export=0
     vld1.64         {d0}, [r1,:64], r3
@@ -163,7 +163,7 @@ function x264_sub8x4_dct_neon, export=0
     vst1.64         {d4-d5}, [r0,:128]!
     vst1.64         {d6-d7}, [r0,:128]!
     bx              lr
-.endfunc
+endfunc
 
 function x264_sub8x8_dct_neon
     push            {lr}
@@ -172,7 +172,7 @@ function x264_sub8x8_dct_neon
     bl              x264_sub8x4_dct_neon
     pop             {lr}
     b               x264_sub8x4_dct_neon
-.endfunc
+endfunc
 
 function x264_sub16x16_dct_neon
     push            {lr}
@@ -193,7 +193,7 @@ function x264_sub16x16_dct_neon
     bl              x264_sub8x4_dct_neon
     pop             {lr}
     b               x264_sub8x4_dct_neon
-.endfunc
+endfunc
 
 
 .macro DCT8_1D type
@@ -277,7 +277,7 @@ function x264_sub8x8_dct8_neon
     vst1.64         {d24-d27}, [r0,:128]!
     vst1.64         {d28-d31}, [r0,:128]!
     bx              lr
-.endfunc
+endfunc
 
 function x264_sub16x16_dct8_neon
     push            {lr}
@@ -292,7 +292,7 @@ function x264_sub16x16_dct8_neon
     sub             r1,  r1,  #FENC_STRIDE*8 - 8
     sub             r2,  r2,  #FDEC_STRIDE*8 - 8
     b               X(x264_sub8x8_dct8_neon)
-.endfunc
+endfunc
 
 
 // First part of IDCT (minus final SUMSUB_BA)
@@ -334,7 +334,7 @@ function x264_add4x4_idct_neon
     vst1.32         {d2[1]}, [r0,:32], r2
     vst1.32         {d2[0]}, [r0,:32], r2
     bx              lr
-.endfunc
+endfunc
 
 function x264_add8x4_idct_neon, export=0
     vld1.64         {d0-d3}, [r1,:128]!
@@ -374,7 +374,7 @@ function x264_add8x4_idct_neon, export=0
     vst1.32         {d2}, [r0,:64], r2
     vst1.32         {d3}, [r0,:64], r2
     bx              lr
-.endfunc
+endfunc
 
 function x264_add8x8_idct_neon
     mov             r2, #FDEC_STRIDE
@@ -382,7 +382,7 @@ function x264_add8x8_idct_neon
     bl              x264_add8x4_idct_neon
     mov             lr, ip
     b               x264_add8x4_idct_neon
-.endfunc
+endfunc
 
 function x264_add16x16_idct_neon
     mov             r2, #FDEC_STRIDE
@@ -399,7 +399,7 @@ function x264_add16x16_idct_neon
     bl              x264_add8x4_idct_neon
     mov             lr, ip
     b               x264_add8x4_idct_neon
-.endfunc
+endfunc
 
 
 .macro IDCT8_1D type
@@ -496,7 +496,7 @@ function x264_add8x8_idct8_neon
     vst1.64         {d6}, [r0,:64], r2
     vst1.64         {d7}, [r0,:64], r2
     bx              lr
-.endfunc
+endfunc
 
 function x264_add16x16_idct8_neon
     mov             ip,  lr
@@ -508,7 +508,7 @@ function x264_add16x16_idct8_neon
     sub             r0,  r0,  #8*FDEC_STRIDE-8
     mov             lr,  ip
     b               X(x264_add8x8_idct8_neon)
-.endfunc
+endfunc
 
 
 function x264_add8x8_idct_dc_neon
@@ -560,7 +560,7 @@ function x264_add8x8_idct_dc_neon
     vst1.64         {d6}, [r0,:64], r2
     vst1.64         {d7}, [r0,:64], r2
     bx              lr
-.endfunc
+endfunc
 
 .macro ADD16x4_IDCT_DC dc
     vld1.64         {d16-d17}, [r0,:128], r3
@@ -608,7 +608,7 @@ function x264_add16x16_idct_dc_neon
     ADD16x4_IDCT_DC d2
     ADD16x4_IDCT_DC d3
     bx              lr
-.endfunc
+endfunc
 
 function x264_sub8x8_dct_dc_neon
     mov             r3,  #FENC_STRIDE
@@ -656,7 +656,7 @@ function x264_sub8x8_dct_dc_neon
     vpadd.s16       d0,  d0,  d1
     vst1.64         {d0}, [r0,:64]
     bx              lr
-.endfunc
+endfunc
 
 
 function x264_zigzag_scan_4x4_frame_neon
@@ -669,4 +669,4 @@ function x264_zigzag_scan_4x4_frame_neon
     vtbl.8      d7, {d2-d3}, d19
     vst1.64     {d4-d7},   [r0,:128]
     bx          lr
-.endfunc
+endfunc
diff --git a/common/arm/deblock-a.S b/common/arm/deblock-a.S
index 2d0ef9ff..59977b4b 100644
--- a/common/arm/deblock-a.S
+++ b/common/arm/deblock-a.S
@@ -140,7 +140,7 @@ function x264_deblock_v_luma_neon
 
     align_pop_regs
     bx              lr
-.endfunc
+endfunc
 
 function x264_deblock_h_luma_neon
     h264_loop_filter_start
@@ -192,7 +192,7 @@ function x264_deblock_h_luma_neon
 
     align_pop_regs
     bx              lr
-.endfunc
+endfunc
 
 .macro h264_loop_filter_chroma
     vdup.8          q11, r2         // alpha
@@ -253,7 +253,7 @@ function x264_deblock_v_chroma_neon
     vst2.8          {d0, d1},  [r0,:128], r1
 
     bx              lr
-.endfunc
+endfunc
 
 function x264_deblock_h_chroma_neon
     h264_loop_filter_start
@@ -301,7 +301,7 @@ function x264_deblock_h_chroma_neon
     vst1.8          {d3},  [r0], r1
 
     bx              lr
-.endfunc
+endfunc
 
 function x264_deblock_strength_neon
     ldr             ip,  [sp]
@@ -407,4 +407,4 @@ lists:
 
     vst1.8          {q8}, [r3,:128]     @ bs[0]
     bx              lr
-.endfunc
+endfunc
diff --git a/common/arm/mc-a.S b/common/arm/mc-a.S
index 3a16d0db..cd57920c 100644
--- a/common/arm/mc-a.S
+++ b/common/arm/mc-a.S
@@ -49,7 +49,7 @@ function x264_prefetch_ref_arm
     pld         [r3, r1, lsl #1]
     pld         [r3, r2]
     bx          lr
-.endfunc
+endfunc
 
 // void prefetch_fenc( uint8_t *pix_y,  intptr_t stride_y,
 //                     uint8_t *pix_uv, intptr_t stride_uv, int mb_x )
@@ -75,7 +75,7 @@ function x264_prefetch_fenc_arm
     pld         [ip]
     pld         [ip, r3]
     pop         {pc}
-.endfunc
+endfunc
 
 
 // void *x264_memcpy_aligned( void *dst, const void *src, size_t n )
@@ -84,7 +84,7 @@ function x264_memcpy_aligned_neon
     movrel      ip,  memcpy_table
     and         r3,  r3,  #0xc
     ldr         pc,  [ip, r3]
-.endfunc
+endfunc
 
 .macro MEMCPY_ALIGNED srcalign dstalign
 function memcpy_aligned_\dstalign\()_\srcalign\()_neon, export=0
@@ -126,7 +126,7 @@ function memcpy_aligned_\dstalign\()_\srcalign\()_neon, export=0
     vst1.64     {d0}, [r3,:64]!
 .endif
     bx          lr
-.endfunc
+endfunc
 .endm
 
 MEMCPY_ALIGNED 16, 16
@@ -155,7 +155,7 @@ memzero_loop:
 .endr
     bgt         memzero_loop
     bx          lr
-.endfunc
+endfunc
 
 
 // void pixel_avg( uint8_t *dst,  intptr_t dst_stride,
@@ -174,7 +174,7 @@ function x264_pixel_avg_\w\()x\h\()_neon
     cmp         ip,  #0
     bge         x264_pixel_avg_weight_w\w\()_add_add_neon
     b           x264_pixel_avg_weight_w\w\()_sub_add_neon     // weight < 0
-.endfunc
+endfunc
 .endm
 
 AVGH  4, 2
@@ -252,7 +252,7 @@ function x264_pixel_avg_weight_w4_\ext\()_neon, export=0
     vst1.32         {d1[0]}, [r0,:32], r1
     bgt             1b
     pop             {r4-r6,pc}
-.endfunc
+endfunc
 
 function x264_pixel_avg_weight_w8_\ext\()_neon, export=0
     load_weights_\ext
@@ -276,7 +276,7 @@ function x264_pixel_avg_weight_w8_\ext\()_neon, export=0
     vst1.64         {d3}, [r0,:64], r1
     bgt             1b
     pop             {r4-r6,pc}
-.endfunc
+endfunc
 
 function x264_pixel_avg_weight_w16_\ext\()_neon, export=0
     load_weights_\ext
@@ -296,7 +296,7 @@ function x264_pixel_avg_weight_w16_\ext\()_neon, export=0
     vst1.64         {d2-d3}, [r0,:128], r1
     bgt             1b
     pop             {r4-r6,pc}
-.endfunc
+endfunc
 .endm
 
 AVG_WEIGHT add_add
@@ -315,7 +315,7 @@ function x264_pixel_avg_w4_neon, export=0
     vst1.32     {d1[0]}, [r0,:32], r1
     bgt         x264_pixel_avg_w4_neon
     pop         {r4-r6,pc}
-.endfunc
+endfunc
 
 function x264_pixel_avg_w8_neon, export=0
     subs        lr,  lr,  #4
@@ -337,7 +337,7 @@ function x264_pixel_avg_w8_neon, export=0
     vst1.64     {d3}, [r0,:64], r1
     bgt         x264_pixel_avg_w8_neon
     pop         {r4-r6,pc}
-.endfunc
+endfunc
 
 function x264_pixel_avg_w16_neon, export=0
     subs        lr,  lr,  #4
@@ -359,7 +359,7 @@ function x264_pixel_avg_w16_neon, export=0
     vst1.64     {d6-d7}, [r0,:128], r1
     bgt         x264_pixel_avg_w16_neon
     pop         {r4-r6,pc}
-.endfunc
+endfunc
 
 
 function x264_pixel_avg2_w4_neon
@@ -378,7 +378,7 @@ avg2_w4_loop:
     vst1.32     {d1[0]}, [r0,:32], r1
     bgt         avg2_w4_loop
     pop         {pc}
-.endfunc
+endfunc
 
 function x264_pixel_avg2_w8_neon
     ldr         ip,  [sp, #4]
@@ -396,7 +396,7 @@ avg2_w8_loop:
     vst1.64     {d1}, [r0,:64], r1
     bgt         avg2_w8_loop
     pop         {pc}
-.endfunc
+endfunc
 
 function x264_pixel_avg2_w16_neon
     ldr         ip,  [sp, #4]
@@ -414,7 +414,7 @@ avg2_w16_loop:
     vst1.64     {d4-d5}, [r0,:128], r1
     bgt         avg2_w16_loop
     pop         {pc}
-.endfunc
+endfunc
 
 function x264_pixel_avg2_w20_neon
     ldr         ip,  [sp, #4]
@@ -437,7 +437,7 @@ avg2_w20_loop:
     vst1.32     {d6[0]},  [r0,:32], r1
     bgt         avg2_w20_loop
     pop         {pc}
-.endfunc
+endfunc
 
 
 .macro weight_prologue type
@@ -498,7 +498,7 @@ weight20_loop:
     vst1.32     {d20[1]},  [r0,:32], r1
     bgt         weight20_loop
     pop         {r4-r5,pc}
-.endfunc
+endfunc
 
 function x264_mc_weight_w16_neon
     weight_prologue full
@@ -530,7 +530,7 @@ weight16_loop:
     vst1.8      {d18-d19}, [r0,:128], r1
     bgt         weight16_loop
     pop         {r4-r5,pc}
-.endfunc
+endfunc
 
 function x264_mc_weight_w8_neon
     weight_prologue full
@@ -552,7 +552,7 @@ weight8_loop:
     vst1.8      {d18}, [r0,:64], r1
     bgt         weight8_loop
     pop         {r4-r5,pc}
-.endfunc
+endfunc
 
 function x264_mc_weight_w4_neon
     weight_prologue full
@@ -571,7 +571,7 @@ weight4_loop:
     vst1.32     {d16[1]}, [r0,:32], r1
     bgt         weight4_loop
     pop         {r4-r5,pc}
-.endfunc
+endfunc
 
 function x264_mc_weight_w20_nodenom_neon
     weight_prologue nodenom
@@ -608,7 +608,7 @@ weight20_nodenom_loop:
     vst1.32     {d20[1]},  [r0,:32], r1
     bgt         weight20_nodenom_loop
     pop         {r4-r5,pc}
-.endfunc
+endfunc
 
 function x264_mc_weight_w16_nodenom_neon
     weight_prologue nodenom
@@ -636,7 +636,7 @@ weight16_nodenom_loop:
     vst1.8      {d18-d19}, [r0,:128], r1
     bgt         weight16_nodenom_loop
     pop         {r4-r5,pc}
-.endfunc
+endfunc
 
 function x264_mc_weight_w8_nodenom_neon
     weight_prologue nodenom
@@ -656,7 +656,7 @@ weight8_nodenom_loop:
     vst1.8      {d17}, [r0,:64], r1
     bgt         weight8_nodenom_loop
     pop         {r4-r5,pc}
-.endfunc
+endfunc
 
 function x264_mc_weight_w4_nodenom_neon
     weight_prologue nodenom
@@ -674,7 +674,7 @@ weight4_nodenom_loop:
     vst1.32     {d16[1]}, [r0,:32], r1
     bgt         weight4_nodenom_loop
     pop         {r4-r5,pc}
-.endfunc
+endfunc
 
 .macro weight_simple_prologue
     push        {lr}
@@ -698,7 +698,7 @@ weight20_\name\()_loop:
     vst1.8      {d19-d21}, [r0,:64], r1
     bgt         weight20_\name\()_loop
     pop         {pc}
-.endfunc
+endfunc
 
 function x264_mc_weight_w16_\name\()_neon
     weight_simple_prologue
@@ -712,7 +712,7 @@ weight16_\name\()_loop:
     vst1.8      {d18-d19}, [r0,:128], r1
     bgt         weight16_\name\()_loop
     pop         {pc}
-.endfunc
+endfunc
 
 function x264_mc_weight_w8_\name\()_neon
     weight_simple_prologue
@@ -725,7 +725,7 @@ weight8_\name\()_loop:
     vst1.8      {d17}, [r0,:64], r1
     bgt         weight8_\name\()_loop
     pop         {pc}
-.endfunc
+endfunc
 
 function x264_mc_weight_w4_\name\()_neon
     weight_simple_prologue
@@ -738,7 +738,7 @@ weight4_\name\()_loop:
     vst1.32     {d17[0]}, [r0,:32], r1
     bgt         weight4_\name\()_loop
     pop         {pc}
-.endfunc
+endfunc
 .endm
 
 weight_simple offsetadd, vqadd.u8
@@ -760,7 +760,7 @@ copy_w4_loop:
     vst1.32     {d3[0]}, [r0,:32], r1
     bgt         copy_w4_loop
     bx          lr
-.endfunc
+endfunc
 
 function x264_mc_copy_w8_neon
     ldr         ip,  [sp]
@@ -776,7 +776,7 @@ copy_w8_loop:
     vst1.32     {d3}, [r0,:64], r1
     bgt         copy_w8_loop
     bx          lr
-.endfunc
+endfunc
 
 function x264_mc_copy_w16_neon
     ldr         ip,  [sp]
@@ -792,7 +792,7 @@ copy_w16_loop:
     vst1.32     {d6-d7}, [r0,:128], r1
     bgt         copy_w16_loop
     bx          lr
-.endfunc
+endfunc
 
 function x264_mc_copy_w16_aligned_neon
     ldr         ip,  [sp]
@@ -808,7 +808,7 @@ copy_w16_aligned_loop:
     vst1.32     {d6-d7}, [r0,:128], r1
     bgt         copy_w16_aligned_loop
     bx          lr
-.endfunc
+endfunc
 
 
 // void x264_mc_chroma_neon( uint8_t *dst, intptr_t i_dst_stride,
@@ -1158,7 +1158,7 @@ mc_chroma_w8:
     vpop            {d8-d11}
     pop             {r4-r8, pc}
 
-.endfunc
+endfunc
 
 
 // hpel_filter_v( uint8_t *dst, uint8_t *src, int16_t *buf, intptr_t stride, int width )
@@ -1199,7 +1199,7 @@ filter_v_loop:
     vst1.64         {d0-d1},   [r0,:128]!
     bgt             filter_v_loop
     pop             {pc}
-.endfunc
+endfunc
 
 // hpel_filter_c( uint8_t *dst, int16_t *buf, int width );
 function x264_hpel_filter_c_neon
@@ -1284,7 +1284,7 @@ filter_c_loop:
     vst1.64         {d30-d31}, [r0,:128]!
     bgt             filter_c_loop
     bx              lr
-.endfunc
+endfunc
 
 // hpel_filter_h( uint8_t *dst, uint8_t *src, int width );
 function x264_hpel_filter_h_neon
@@ -1371,7 +1371,7 @@ filter_h_loop:
     vst1.64         {d6-d7}, [r0,:128]!
     bgt             filter_h_loop
     bx              lr
-.endfunc
+endfunc
 
 
 // frame_init_lowres_core( uint8_t *src0, uint8_t *dst0, uint8_t *dsth, uint8_t *dstv,
@@ -1463,7 +1463,7 @@ lowres_xloop_end:
 
     vpop            {d8-d15}
     pop             {r4-r10,pc}
-.endfunc
+endfunc
 
 function x264_load_deinterleave_chroma_fdec_neon
     mov             ip,  #FDEC_STRIDE/2
@@ -1476,7 +1476,7 @@ function x264_load_deinterleave_chroma_fdec_neon
     bgt             1b
 
     bx              lr
-.endfunc
+endfunc
 
 function x264_load_deinterleave_chroma_fenc_neon
     mov             ip,  #FENC_STRIDE/2
@@ -1489,7 +1489,7 @@ function x264_load_deinterleave_chroma_fenc_neon
     bgt             1b
 
     bx              lr
-.endfunc
+endfunc
 
 function x264_plane_copy_deinterleave_neon
     push            {r4-r7, lr}
@@ -1515,7 +1515,7 @@ block:
     bgt             block
 
     pop             {r4-r7, pc}
-.endfunc
+endfunc
 
 function x264_plane_copy_deinterleave_rgb_neon
     push            {r4-r8, r10, r11, lr}
@@ -1567,7 +1567,7 @@ block4:
     bgt             block4
 
     pop             {r4-r8, r10, r11, pc}
-.endfunc
+endfunc
 
 function x264_plane_copy_interleave_neon
     push            {r4-r7, lr}
@@ -1594,7 +1594,7 @@ blocki:
     bgt             blocki
 
     pop             {r4-r7, pc}
-.endfunc
+endfunc
 
 function x264_store_interleave_chroma_neon
     push            {lr}
@@ -1608,4 +1608,4 @@ function x264_store_interleave_chroma_neon
     bgt             1b
 
     pop             {pc}
-.endfunc
+endfunc
diff --git a/common/arm/pixel-a.S b/common/arm/pixel-a.S
index e288bcf6..80b8b700 100644
--- a/common/arm/pixel-a.S
+++ b/common/arm/pixel-a.S
@@ -61,7 +61,7 @@ function x264_pixel_sad_4x\h\()_armv6
 .endr
     usada8      r0, r6, lr, ip
     pop         {r4-r6,pc}
-.endfunc
+endfunc
 .endm
 
 SAD4_ARMV6 4
@@ -138,7 +138,7 @@ function x264_pixel_sad\name\()_\w\()x\h\()_neon
     vpaddl.u16  d0,  d0
     vmov.u32    r0,  d0[0]
     bx          lr
-.endfunc
+endfunc
 .endm
 
 SAD_FUNC  4,  4
@@ -223,7 +223,7 @@ function x264_pixel_sad_aligned_\w\()x\h\()_neon_dual
     vpaddl.u16  d0,  d0
     vmov.u32    r0,  d0[0]
     bx          lr
-.endfunc
+endfunc
 .endm
 
 SAD_FUNC_DUAL  8,  4
@@ -369,7 +369,7 @@ function x264_pixel_sad_x\x\()_\w\()x\h\()_neon
     vst1.32     {d0-d1}, [r7]
 .endif
     pop         {r6-r7,pc}
-.endfunc
+endfunc
 .endm
 
 SAD_X_FUNC  3, 4,  4
@@ -478,7 +478,7 @@ function x264_pixel_ssd_\w\()x\h\()_neon
     vpadd.s32   d0, d0, d0
     vmov.32     r0, d0[0]
     bx          lr
-.endfunc
+endfunc
 .endm
 
 SSD_FUNC   4, 4
@@ -518,7 +518,7 @@ function x264_pixel_var_8x8_neon
     vld1.64         {d26}, [r0,:64], r1
     VAR_SQR_SUM     q2,  q10,  q15, d26
     b               x264_var_end
-.endfunc
+endfunc
 
 function x264_pixel_var_8x16_neon
     vld1.64         {d16}, [r0,:64], r1
@@ -550,7 +550,7 @@ function x264_pixel_var_8x16_neon
 2:
     VAR_SQR_SUM     q2,  q13,  q15, d22
     b               x264_var_end
-.endfunc
+endfunc
 
 function x264_pixel_var_16x16_neon
     vld1.64         {d16-d17}, [r0,:128], r1
@@ -574,7 +574,7 @@ var16_loop:
     VAR_SQR_SUM     q1,  q12,  q14, d18
     VAR_SQR_SUM     q2,  q13,  q15, d19
     bgt             var16_loop
-.endfunc
+endfunc
 
 function x264_var_end, export=0
     vpaddl.u16      q8,  q14
@@ -589,7 +589,7 @@ function x264_var_end, export=0
 
     vmov            r0,  r1,  d0
     bx              lr
-.endfunc
+endfunc
 
 .macro DIFF_SUM diff da db lastdiff
     vld1.64         {\da}, [r0,:64], r1
@@ -634,7 +634,7 @@ function x264_pixel_var2_8x8_neon
     mul             r0,  r0,  r0
     sub             r0,  r1,  r0,  lsr #6
     bx              lr
-.endfunc
+endfunc
 
 function x264_pixel_var2_8x16_neon
     vld1.64         {d16}, [r0,:64], r1
@@ -678,7 +678,7 @@ function x264_pixel_var2_8x16_neon
     mul             r0,  r0,  r0
     sub             r0,  r1,  r0,  lsr #7
     bx              lr
-.endfunc
+endfunc
 
 .macro LOAD_DIFF_8x4 q0 q1 q2 q3
     vld1.32     {d1}, [r2], r3
@@ -715,7 +715,7 @@ function x264_pixel_satd_4x4_neon
     HORIZ_ADD   d0,  d0,  d1
     vmov.32     r0,  d0[0]
     bx          lr
-.endfunc
+endfunc
 
 function x264_pixel_satd_4x8_neon
     vld1.32     {d1[]},  [r2], r3
@@ -742,7 +742,7 @@ function x264_pixel_satd_4x8_neon
     vsubl.u8    q3,  d6,  d7
     SUMSUB_AB   q10, q11, q2,  q3
     b           x264_satd_4x8_8x4_end_neon
-.endfunc
+endfunc
 
 function x264_pixel_satd_8x4_neon
     vld1.64     {d1}, [r2], r3
@@ -759,7 +759,7 @@ function x264_pixel_satd_8x4_neon
     vld1.64     {d6}, [r0,:64], r1
     vsubl.u8    q3,  d6,  d7
     SUMSUB_AB   q10, q11, q2,  q3
-.endfunc
+endfunc
 
 function x264_satd_4x8_8x4_end_neon, export=0
     vadd.s16    q0,  q8,  q10
@@ -786,7 +786,7 @@ function x264_satd_4x8_8x4_end_neon, export=0
     HORIZ_ADD   d0,  d0,  d1
     vmov.32     r0,  d0[0]
     bx          lr
-.endfunc
+endfunc
 
 function x264_pixel_satd_8x8_neon
     mov         ip,  lr
@@ -800,7 +800,7 @@ function x264_pixel_satd_8x8_neon
     mov         lr,  ip
     vmov.32     r0,  d0[0]
     bx          lr
-.endfunc
+endfunc
 
 function x264_pixel_satd_8x16_neon
     vpush       {d8-d11}
@@ -822,7 +822,7 @@ function x264_pixel_satd_8x16_neon
     mov         lr,  ip
     vmov.32     r0,  d0[0]
     bx          lr
-.endfunc
+endfunc
 
 function x264_satd_8x8_neon, export=0
     LOAD_DIFF_8x4 q8,  q9,  q10, q11
@@ -842,7 +842,7 @@ function x264_satd_8x8_neon, export=0
     SUMSUB_AB   q9,  q11, q1,  q3
     vld1.64     {d0}, [r0,:64], r1
     vsubl.u8    q15, d0,  d1
-.endfunc
+endfunc
 
 // one vertical hadamard pass and two horizontal
 function x264_satd_8x4v_8x8h_neon, export=0
@@ -871,7 +871,7 @@ function x264_satd_8x4v_8x8h_neon, export=0
     vmax.s16    q14, q8,  q10
     vmax.s16    q15, q9,  q11
     bx          lr
-.endfunc
+endfunc
 
 function x264_pixel_satd_16x8_neon
     vpush       {d8-d11}
@@ -893,7 +893,7 @@ function x264_pixel_satd_16x8_neon
     mov         lr,  ip
     vmov.32     r0,  d0[0]
     bx          lr
-.endfunc
+endfunc
 
 function x264_pixel_satd_16x16_neon
     vpush       {d8-d11}
@@ -927,7 +927,7 @@ function x264_pixel_satd_16x16_neon
     mov         lr,  ip
     vmov.32     r0,  d0[0]
     bx          lr
-.endfunc
+endfunc
 
 function x264_satd_16x4_neon, export=0
     vld1.64     {d2-d3}, [r2], r3
@@ -951,7 +951,7 @@ function x264_satd_16x4_neon, export=0
     SUMSUB_AB   q2,  q3,  q10, q11
     SUMSUB_ABCD q8,  q10, q9,  q11, q0,  q2,  q1,  q3
     b           x264_satd_8x4v_8x8h_neon
-.endfunc
+endfunc
 
 
 function x264_pixel_sa8d_8x8_neon
@@ -964,7 +964,7 @@ function x264_pixel_sa8d_8x8_neon
     add             r0,  r0,  #1
     lsr             r0,  r0,  #1
     bx              lr
-.endfunc
+endfunc
 
 function x264_pixel_sa8d_16x16_neon
     vpush           {d8-d11}
@@ -996,7 +996,7 @@ function x264_pixel_sa8d_16x16_neon
     add             r0,  r0,  #1
     lsr             r0,  r0,  #1
     bx              lr
-.endfunc
+endfunc
 
 .macro HADAMARD4_V r1, r2, r3, r4, t1, t2, t3, t4
     SUMSUB_ABCD \t1, \t2, \t3, \t4, \r1, \r2, \r3, \r4
@@ -1059,7 +1059,7 @@ function x264_sa8d_8x8_neon, export=0
     vadd.i16        q8,  q8,  q9
     vadd.i16        q9,  q10, q11
     bx              lr
-.endfunc
+endfunc
 
 
 .macro HADAMARD_AC w h
@@ -1095,7 +1095,7 @@ function x264_pixel_hadamard_ac_\w\()x\h\()_neon
     lsr             r0,  r0,  #1
     lsr             r1,  r1,  #2
     bx              lr
-.endfunc
+endfunc
 .endm
 
 HADAMARD_AC  8, 8
@@ -1190,7 +1190,7 @@ function x264_hadamard_ac_8x8_neon, export=0
     vadd.s16        q2,  q2,  q14
     vpadal.u16      q5,  q2
     bx              lr
-.endfunc
+endfunc
 
 
 .macro SSIM_ITER n ssa s12 ssb lastssa lasts12 lastssb da db dnext
@@ -1244,7 +1244,7 @@ function x264_pixel_ssim_4x4x2_core_neon
 
     vst4.32     {d0-d3}, [ip]
     bx          lr
-.endfunc
+endfunc
 
 // FIXME: see about doing 16x16 -> 32 bit multiplies for s1/s2
 function x264_pixel_ssim_end4_neon
@@ -1315,4 +1315,4 @@ ssim_skip:
     vpadd.f32   d0,  d0,  d0
     vmov.32     r0,  d0[0]
     bx          lr
-.endfunc
+endfunc
diff --git a/common/arm/predict-a.S b/common/arm/predict-a.S
index 8af861bd..593d4030 100644
--- a/common/arm/predict-a.S
+++ b/common/arm/predict-a.S
@@ -75,7 +75,7 @@ function x264_predict_4x4_h_armv6
     add     ip, ip, ip, lsl #16
     str     ip, [r0, #3*FDEC_STRIDE]
     bx      lr
-.endfunc
+endfunc
 
 function x264_predict_4x4_v_armv6
     ldr     r1,  [r0, #0 - 1 * FDEC_STRIDE]
@@ -84,7 +84,7 @@ function x264_predict_4x4_v_armv6
     str     r1,  [r0, #0 + 2 * FDEC_STRIDE]
     str     r1,  [r0, #0 + 3 * FDEC_STRIDE]
     bx      lr
-.endfunc
+endfunc
 
 function x264_predict_4x4_dc_armv6
     mov     ip, #0
@@ -107,7 +107,7 @@ function x264_predict_4x4_dc_armv6
     str     r1, [r0, #2*FDEC_STRIDE]
     str     r1, [r0, #3*FDEC_STRIDE]
     bx      lr
-.endfunc
+endfunc
 
 function x264_predict_4x4_dc_top_neon
     mov         r12, #FDEC_STRIDE
@@ -122,7 +122,7 @@ function x264_predict_4x4_dc_top_neon
     vst1.32     d1[0], [r0,:32], r12
     vst1.32     d1[0], [r0,:32], r12
     bx          lr
-.endfunc
+endfunc
 
 // return a1 = (a1+2*b1+c1+2)>>2  a2 = (a2+2*b2+c2+2)>>2
 .macro PRED4x4_LOWPASS a1 b1 c1 a2 b2 c2 pb_1
@@ -165,7 +165,7 @@ function x264_predict_4x4_ddr_armv6
     add     r5, r5, r4, lsr #8
     str     r5, [r0, #3*FDEC_STRIDE]
     pop     {r4-r6,pc}
-.endfunc
+endfunc
 
 function x264_predict_4x4_ddl_neon
     sub         r0, #FDEC_STRIDE
@@ -184,7 +184,7 @@ function x264_predict_4x4_ddl_neon
     vst1.32     {d2[0]}, [r0,:32], ip
     vst1.32     {d3[0]}, [r0,:32], ip
     bx          lr
-.endfunc
+endfunc
 
 function x264_predict_8x8_dc_neon
     mov     ip, #0
@@ -208,7 +208,7 @@ function x264_predict_8x8_dc_neon
     vst1.64 {d0}, [r0,:64], ip
 .endr
     pop    {r4-r5,pc}
-.endfunc
+endfunc
 
 function x264_predict_8x8_h_neon
     add         r1, r1, #7
@@ -231,7 +231,7 @@ function x264_predict_8x8_h_neon
     vst1.64     {d6}, [r0,:64], ip
     vst1.64     {d7}, [r0,:64], ip
     bx          lr
-.endfunc
+endfunc
 
 function x264_predict_8x8_v_neon
     add         r1, r1, #16
@@ -241,7 +241,7 @@ function x264_predict_8x8_v_neon
     vst1.8      {d0}, [r0,:64], r12
 .endr
     bx          lr
-.endfunc
+endfunc
 
 function x264_predict_8x8_ddl_neon
     add         r1, #16
@@ -269,7 +269,7 @@ function x264_predict_8x8_ddl_neon
     vst1.8      d2, [r0,:64], r12
     vst1.8      d1, [r0,:64], r12
     bx          lr
-.endfunc
+endfunc
 
 function x264_predict_8x8_ddr_neon
     vld1.8      {d0-d3}, [r1,:128]
@@ -299,7 +299,7 @@ function x264_predict_8x8_ddr_neon
     vst1.8      {d4}, [r0,:64], r12
     vst1.8      {d5}, [r0,:64], r12
     bx          lr
-.endfunc
+endfunc
 
 function x264_predict_8x8_vl_neon
     add         r1, #16
@@ -330,7 +330,7 @@ function x264_predict_8x8_vl_neon
     vst1.8      {d3}, [r0,:64], r12
     vst1.8      {d2}, [r0,:64], r12
     bx          lr
-.endfunc
+endfunc
 
 function x264_predict_8x8_vr_neon
     add         r1, #8
@@ -362,7 +362,7 @@ function x264_predict_8x8_vr_neon
     vst1.8      {d6}, [r0,:64], r12
     vst1.8      {d3}, [r0,:64], r12
     bx          lr
-.endfunc
+endfunc
 
 function x264_predict_8x8_hd_neon
     mov         r12, #FDEC_STRIDE
@@ -395,7 +395,7 @@ function x264_predict_8x8_hd_neon
     vst1.8      {d16}, [r0,:64], r12
 
     bx          lr
-.endfunc
+endfunc
 
 function x264_predict_8x8_hu_neon
     mov         r12, #FDEC_STRIDE
@@ -428,7 +428,7 @@ function x264_predict_8x8_hu_neon
     vst1.8      {d7}, [r0,:64], r12
     vst1.8      {d17}, [r0,:64]
     bx          lr
-.endfunc
+endfunc
 
 function x264_predict_8x8c_dc_top_neon
     sub         r2,  r0,  #FDEC_STRIDE
@@ -441,7 +441,7 @@ function x264_predict_8x8c_dc_top_neon
     vdup.8      d0,  d0[0]
     vtrn.32     d0,  d1
     b           pred8x8_dc_end
-.endfunc
+endfunc
 
 function x264_predict_8x8c_dc_left_neon
     mov         r1,  #FDEC_STRIDE
@@ -453,7 +453,7 @@ function x264_predict_8x8c_dc_left_neon
     vdup.8      d1,  d0[1]
     vdup.8      d0,  d0[0]
     b           pred8x8_dc_end
-.endfunc
+endfunc
 
 function x264_predict_8x8c_dc_neon
     sub         r2,  r0,  #FDEC_STRIDE
@@ -479,7 +479,7 @@ pred8x8_dc_end:
     vst1.8      {d1}, [r2,:64], r1
 .endr
     bx          lr
-.endfunc
+endfunc
 
 function x264_predict_8x8c_h_neon
     sub         r1, r0, #1
@@ -491,7 +491,7 @@ function x264_predict_8x8c_h_neon
     vst1.64     {d2}, [r0,:64], ip
 .endr
     bx          lr
-.endfunc
+endfunc
 
 function x264_predict_8x8c_v_neon
     sub         r0, r0, #FDEC_STRIDE
@@ -501,7 +501,7 @@ function x264_predict_8x8c_v_neon
     vst1.64     {d0}, [r0,:64], ip
 .endr
     bx          lr
-.endfunc
+endfunc
 
 function x264_predict_8x8c_p_neon
     sub         r3,  r0,  #FDEC_STRIDE
@@ -554,7 +554,7 @@ function x264_predict_8x8c_p_neon
     subs        r3,  r3,  #1
     bne         1b
     bx          lr
-.endfunc
+endfunc
 
 
 function x264_predict_16x16_dc_top_neon
@@ -565,7 +565,7 @@ function x264_predict_16x16_dc_top_neon
     vrshrn.u16  d0,  q0,  #4
     vdup.8      q0,  d0[0]
     b           pred16x16_dc_end
-.endfunc
+endfunc
 
 function x264_predict_16x16_dc_left_neon
     mov         r1,  #FDEC_STRIDE
@@ -576,7 +576,7 @@ function x264_predict_16x16_dc_left_neon
     vrshrn.u16  d0,  q0,  #4
     vdup.8      q0,  d0[0]
     b           pred16x16_dc_end
-.endfunc
+endfunc
 
 function x264_predict_16x16_dc_neon
     sub         r3, r0, #FDEC_STRIDE
@@ -614,7 +614,7 @@ pred16x16_dc_end:
     vst1.64     {d0-d1}, [r0,:128], r1
 .endr
     bx          lr
-.endfunc
+endfunc
 
 function x264_predict_16x16_h_neon
     sub         r1, r0, #1
@@ -628,7 +628,7 @@ function x264_predict_16x16_h_neon
     vst1.64     {d2-d3}, [r0,:128], ip
 .endr
     bx          lr
-.endfunc
+endfunc
 
 function x264_predict_16x16_v_neon
     sub         r0, r0, #FDEC_STRIDE
@@ -638,7 +638,7 @@ function x264_predict_16x16_v_neon
     vst1.64     {d0-d1}, [r0,:128], ip
 .endr
     bx          lr
-.endfunc
+endfunc
 
 function x264_predict_16x16_p_neon
     sub         r3,  r0,  #FDEC_STRIDE
@@ -695,4 +695,4 @@ function x264_predict_16x16_p_neon
     subs        r3,  r3,  #1
     bne         1b
     bx          lr
-.endfunc
+endfunc
diff --git a/common/arm/quant-a.S b/common/arm/quant-a.S
index 374796c7..d22a10ec 100644
--- a/common/arm/quant-a.S
+++ b/common/arm/quant-a.S
@@ -78,7 +78,7 @@ function x264_quant_2x2_dc_neon
     vsub.s16    d3,  d3,  d0
     vst1.64     {d3}, [r0,:64]
     QUANT_END   d3
-.endfunc
+endfunc
 
 // quant_4x4_dc( int16_t dct[16], int mf, int bias )
 function x264_quant_4x4_dc_neon
@@ -90,7 +90,7 @@ function x264_quant_4x4_dc_neon
     QUANT_TWO   q0,  q0,  d4,  d5,  d4,  d5,  q0
     vorr        d0,  d0,  d1
     QUANT_END   d0
-.endfunc
+endfunc
 
 // quant_4x4( int16_t dct[16], uint16_t mf[16], uint16_t bias[16] )
 function x264_quant_4x4_neon
@@ -102,7 +102,7 @@ function x264_quant_4x4_neon
     QUANT_TWO   q0,  q1,  d4,  d5,  d6,  d7, q0
     vorr        d0,  d0,  d1
     QUANT_END   d0
-.endfunc
+endfunc
 
 // quant_4x4x4( int16_t dct[4][16], uint16_t mf[16], uint16_t bias[16] )
 function x264_quant_4x4x4_neon
@@ -143,7 +143,7 @@ function x264_quant_4x4x4_neon
     orrne       r0,  #8
     vpop        {d8-d15}
     bx          lr
-.endfunc
+endfunc
 
 // quant_8x8( int16_t dct[64], uint16_t mf[64], uint16_t bias[64] )
 function x264_quant_8x8_neon
@@ -163,7 +163,7 @@ function x264_quant_8x8_neon
 .endr
     vorr        d0,  d0,  d1
     QUANT_END   d0
-.endfunc
+endfunc
 
 .macro DEQUANT_START mf_size offset dc=no
     mov         r3,  #0x2b
@@ -255,7 +255,7 @@ dequant_\size\()_rshift_loop:
     bgt         dequant_\size\()_rshift_loop
 .endif
     bx          lr
-.endfunc
+endfunc
 .endm
 
 DEQUANT 4x4, 4
@@ -305,7 +305,7 @@ dequant_4x4_dc_rshift:
     vmovn.s32   d3,  q13
     vst1.16     {d0-d3},   [r0,:128]
     bx          lr
-.endfunc
+endfunc
 
 
 // int coeff_last( int16_t *l )
@@ -317,7 +317,7 @@ function x264_coeff_last4_arm
     lsrs        r2,  r2,  #16
     addne       r0,  r0,  #1
     bx          lr
-.endfunc
+endfunc
 
 function x264_coeff_last8_arm
     ldrd        r2,  r3,  [r0, #8]
@@ -331,7 +331,7 @@ function x264_coeff_last8_arm
     lsrs        r2,  r2,  #16
     addne       r0,  r0,  #1
     bx          lr
-.endfunc
+endfunc
 
 .macro COEFF_LAST_1x size
 function x264_coeff_last\size\()_neon
@@ -356,7 +356,7 @@ function x264_coeff_last\size\()_neon
     subslt      r0,  r3,  r0,  lsr #2
     movlt       r0,  #0
     bx          lr
-.endfunc
+endfunc
 .endm
 
 COEFF_LAST_1x 15
@@ -405,4 +405,4 @@ function x264_coeff_last64_neon
     subslt      r0,  ip,  r0
     movlt       r0,  #0
     bx          lr
-.endfunc
+endfunc
diff --git a/configure b/configure
index 7d56fd8a..ebf737a7 100755
--- a/configure
+++ b/configure
@@ -197,8 +197,9 @@ cpp_check() {
 
 as_check() {
     log_check "whether $AS supports $1"
-    echo "$1" > conftest.asm
-    if $AS conftest.asm $ASFLAGS $2 -o conftest.o >conftest.log 2>&1; then
+    echo "$1" > conftest$AS_EXT
+    as_cmd="$AS conftest$AS_EXT $ASFLAGS $2 -o conftest.o"
+    if $as_cmd >conftest.log 2>&1; then
         res=$?
         log_ok
     else
@@ -206,12 +207,12 @@ as_check() {
         log_fail
         log_msg "Failed commandline was:"
         log_msg "--------------------------------------------------"
-        log_msg "$AS conftest.asm $ASFLAGS $2 -o conftest.o"
+        log_msg "$as_cmd"
         cat conftest.log >> config.log
         log_msg "--------------------------------------------------"
         log_msg "Failed program was:"
         log_msg "--------------------------------------------------"
-        cat conftest.asm >> config.log
+        cat conftest$AS_EXT >> config.log
         log_msg "--------------------------------------------------"
     fi
     return $res
@@ -302,10 +303,13 @@ HAVE_GETOPT_LONG=1
 cross_prefix=""
 
 EXE=""
+AS_EXT=".S"
+NL="
+"
 
 # list of all preprocessor HAVE values we can define
 CONFIG_HAVE="MALLOC_H ALTIVEC ALTIVEC_H MMX ARMV6 ARMV6T2 NEON BEOSTHREAD POSIXTHREAD WIN32THREAD THREAD LOG2F SWSCALE \
-             LAVF FFMS GPAC AVS GPL VECTOREXT INTERLACED CPU_COUNT OPENCL THP LSMASH X86_INLINE_ASM"
+             LAVF FFMS GPAC AVS GPL VECTOREXT INTERLACED CPU_COUNT OPENCL THP LSMASH X86_INLINE_ASM AS_FUNC"
 
 # parse options
 
@@ -586,6 +590,7 @@ case $host_cpu in
     i*86)
         ARCH="X86"
         AS="yasm"
+        AS_EXT=".asm"
         ASFLAGS="$ASFLAGS -O2 -DARCH_X86_64=0 -I\$(SRCPATH)/common/x86/"
         if [ $compiler = GNU ]; then
             if [[ "$asm" == auto && "$CFLAGS" != *-march* ]]; then
@@ -626,6 +631,7 @@ case $host_cpu in
     x86_64)
         ARCH="X86_64"
         AS="yasm"
+        AS_EXT=".asm"
         ASFLAGS="$ASFLAGS -DARCH_X86_64=1 -I\$(SRCPATH)/common/x86/"
         [ $compiler = GNU ] && CFLAGS="-m64 $CFLAGS" && LDFLAGS="-m64 $LDFLAGS"
         if [ "$SYS" = MACOSX ]; then
@@ -649,6 +655,7 @@ case $host_cpu in
         if [ $asm = auto ] ; then
             define HAVE_ALTIVEC
             AS="${AS-${CC}}"
+            AS_EXT=".c"
             if [ $SYS = MACOSX ] ; then
                 CFLAGS="$CFLAGS -faltivec -fastf -mcpu=G4"
             else
@@ -760,6 +767,11 @@ if [ $asm = auto -a $ARCH = ARM ] ; then
     fi
 fi
 
+if [ $asm = auto -a $ARCH = ARM ] ; then
+    # check if the assembler supports '.func' (clang 3.5 does not)
+    as_check ".func test${NL}.endfunc" && define HAVE_AS_FUNC 1
+fi
+
 [ $asm = no ] && AS=""
 [ "x$AS" = x ] && asm="no" || asm="yes"