diff --git a/libswresample/arm/resample.S b/libswresample/arm/resample.S
index 791f4cc016..e48df3c8fb 100644
--- a/libswresample/arm/resample.S
+++ b/libswresample/arm/resample.S
@@ -21,57 +21,57 @@
 #include "libavutil/arm/asm.S"
 
 function ff_resample_common_apply_filter_x4_float_neon, export=1
-    vmov.f32            q0, #0.0                                       @ accumulator
-1:  vld1.32             {q1}, [r1]!                                    @ src[0..3]
-    vld1.32             {q2}, [r2]!                                    @ filter[0..3]
-    vmla.f32            q0, q1, q2                                     @ accumulator += src[0..3] * filter[0..3]
-    subs                r3, #4                                         @ filter_length -= 4
-    bgt                 1b                                             @ loop until filter_length
-    vpadd.f32           d0, d0, d1                                     @ pair adding of the 4x32-bit accumulated values
-    vpadd.f32           d0, d0, d0                                     @ pair adding of the 4x32-bit accumulator values
-    vst1.32             {d0[0]}, [r0]                                  @ write accumulator
-    bx                  lr
+        vmov.f32        q0, #0.0                                       @ accumulator
+1:      vld1.32         {q1}, [r1]!                                    @ src[0..3]
+        vld1.32         {q2}, [r2]!                                    @ filter[0..3]
+        vmla.f32        q0, q1, q2                                     @ accumulator += src[0..3] * filter[0..3]
+        subs            r3, #4                                         @ filter_length -= 4
+        bgt             1b                                             @ loop until filter_length
+        vpadd.f32       d0, d0, d1                                     @ pair adding of the 4x32-bit accumulated values
+        vpadd.f32       d0, d0, d0                                     @ pair adding of the 4x32-bit accumulator values
+        vst1.32         {d0[0]}, [r0]                                  @ write accumulator
+        bx              lr
 endfunc
 
 function ff_resample_common_apply_filter_x8_float_neon, export=1
-    vmov.f32            q0, #0.0                                       @ accumulator
-1:  vld1.32             {q1}, [r1]!                                    @ src[0..3]
-    vld1.32             {q2}, [r2]!                                    @ filter[0..3]
-    vld1.32             {q8}, [r1]!                                    @ src[4..7]
-    vld1.32             {q9}, [r2]!                                    @ filter[4..7]
-    vmla.f32            q0, q1, q2                                     @ accumulator += src[0..3] * filter[0..3]
-    vmla.f32            q0, q8, q9                                     @ accumulator += src[4..7] * filter[4..7]
-    subs                r3, #8                                         @ filter_length -= 8
-    bgt                 1b                                             @ loop until filter_length
-    vpadd.f32           d0, d0, d1                                     @ pair adding of the 4x32-bit accumulated values
-    vpadd.f32           d0, d0, d0                                     @ pair adding of the 4x32-bit accumulator values
-    vst1.32             {d0[0]}, [r0]                                  @ write accumulator
-    bx                  lr
+        vmov.f32        q0, #0.0                                       @ accumulator
+1:      vld1.32         {q1}, [r1]!                                    @ src[0..3]
+        vld1.32         {q2}, [r2]!                                    @ filter[0..3]
+        vld1.32         {q8}, [r1]!                                    @ src[4..7]
+        vld1.32         {q9}, [r2]!                                    @ filter[4..7]
+        vmla.f32        q0, q1, q2                                     @ accumulator += src[0..3] * filter[0..3]
+        vmla.f32        q0, q8, q9                                     @ accumulator += src[4..7] * filter[4..7]
+        subs            r3, #8                                         @ filter_length -= 8
+        bgt             1b                                             @ loop until filter_length
+        vpadd.f32       d0, d0, d1                                     @ pair adding of the 4x32-bit accumulated values
+        vpadd.f32       d0, d0, d0                                     @ pair adding of the 4x32-bit accumulator values
+        vst1.32         {d0[0]}, [r0]                                  @ write accumulator
+        bx              lr
 endfunc
 
 function ff_resample_common_apply_filter_x4_s16_neon, export=1
-    vmov.s32            q0, #0                                         @ accumulator
-1:  vld1.16             {d2}, [r1]!                                    @ src[0..3]
-    vld1.16             {d4}, [r2]!                                    @ filter[0..3]
-    vmlal.s16           q0, d2, d4                                     @ accumulator += src[0..3] * filter[0..3]
-    subs                r3, #4                                         @ filter_length -= 4
-    bgt                 1b                                             @ loop until filter_length
-    vpadd.s32           d0, d0, d1                                     @ pair adding of the 4x32-bit accumulated values
-    vpadd.s32           d0, d0, d0                                     @ pair adding of the 4x32-bit accumulator values
-    vst1.32             {d0[0]}, [r0]                                  @ write accumulator
-    bx                  lr
+        vmov.s32        q0, #0                                         @ accumulator
+1:      vld1.16         {d2}, [r1]!                                    @ src[0..3]
+        vld1.16         {d4}, [r2]!                                    @ filter[0..3]
+        vmlal.s16       q0, d2, d4                                     @ accumulator += src[0..3] * filter[0..3]
+        subs            r3, #4                                         @ filter_length -= 4
+        bgt             1b                                             @ loop until filter_length
+        vpadd.s32       d0, d0, d1                                     @ pair adding of the 4x32-bit accumulated values
+        vpadd.s32       d0, d0, d0                                     @ pair adding of the 4x32-bit accumulator values
+        vst1.32         {d0[0]}, [r0]                                  @ write accumulator
+        bx              lr
 endfunc
 
 function ff_resample_common_apply_filter_x8_s16_neon, export=1
-    vmov.s32            q0, #0                                         @ accumulator
-1:  vld1.16             {q1}, [r1]!                                    @ src[0..7]
-    vld1.16             {q2}, [r2]!                                    @ filter[0..7]
-    vmlal.s16           q0, d2, d4                                     @ accumulator += src[0..3] * filter[0..3]
-    vmlal.s16           q0, d3, d5                                     @ accumulator += src[4..7] * filter[4..7]
-    subs                r3, #8                                         @ filter_length -= 8
-    bgt                 1b                                             @ loop until filter_length
-    vpadd.s32           d0, d0, d1                                     @ pair adding of the 4x32-bit accumulated values
-    vpadd.s32           d0, d0, d0                                     @ pair adding of the 4x32-bit accumulator values
-    vst1.32             {d0[0]}, [r0]                                  @ write accumulator
-    bx                  lr
+        vmov.s32        q0, #0                                         @ accumulator
+1:      vld1.16         {q1}, [r1]!                                    @ src[0..7]
+        vld1.16         {q2}, [r2]!                                    @ filter[0..7]
+        vmlal.s16       q0, d2, d4                                     @ accumulator += src[0..3] * filter[0..3]
+        vmlal.s16       q0, d3, d5                                     @ accumulator += src[4..7] * filter[4..7]
+        subs            r3, #8                                         @ filter_length -= 8
+        bgt             1b                                             @ loop until filter_length
+        vpadd.s32       d0, d0, d1                                     @ pair adding of the 4x32-bit accumulated values
+        vpadd.s32       d0, d0, d0                                     @ pair adding of the 4x32-bit accumulator values
+        vst1.32         {d0[0]}, [r0]                                  @ write accumulator
+        bx              lr
 endfunc