interlace filter crash

[alexpigment]

Summary of the bug:

When using a real 1080p60 video source (i.e. not lavfi), attempting to both crop to 1440x1080 and then interlace causes FFMPEG to crash. The same video can be either cropped or interlaced individually with success. Also, doing the cropping and interlacing in two subsequent encodes works as expected. Interestingly, if I crop, scale to 720x480, then interlace in a single encode, this works as expected.

How to reproduce:

Use included sample (or perhaps another 1080p60 video) and then use the following command line:

% ffmpeg -i 1080p60sample.mp4 -vf crop=1440:1080,interlace=scan=tff -c:v libx264 -x264-params tff -flags +ilme+ildct -an output.mp4

ffmpeg version N-86537-gae6f6d4
built on 2017-06-20

[alexpigment]

1080p60 video file

[llogan]

I didn't test for regression.

(gdb) r -f lavfi -i testsrc=s=hd1080,format=yuv420p -vf crop=1440:1080,interlace -f null -
Starting program: ffmpeg_g -f lavfi -i testsrc=s=hd1080,format=yuv420p -vf crop=1440:1080,interlace -f null -
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/usr/lib/libthread_db.so.1".
ffmpeg version N-86620-g3594788b71 Copyright (c) 2000-2017 the FFmpeg developers
  built with gcc 7.1.1 (GCC) 20170621
  configuration: --enable-gpl
  libavutil      55. 66.100 / 55. 66.100
  libavcodec     57. 99.102 / 57. 99.102
  libavformat    57. 75.100 / 57. 75.100
  libavdevice    57.  7.100 / 57.  7.100
  libavfilter     6. 94.100 /  6. 94.100
  libswscale      4.  7.101 /  4.  7.101
  libswresample   2.  8.100 /  2.  8.100
  libpostproc    54.  6.100 / 54.  6.100
[New Thread 0x7ffff3ec0700 (LWP 15562)]
[New Thread 0x7ffff36bf700 (LWP 15563)]
[New Thread 0x7ffff2ebe700 (LWP 15564)]
[New Thread 0x7ffff26bd700 (LWP 15565)]
[New Thread 0x7ffff1ebc700 (LWP 15566)]
[New Thread 0x7ffff16bb700 (LWP 15567)]
[New Thread 0x7ffff0eba700 (LWP 15568)]
[New Thread 0x7ffff06b9700 (LWP 15569)]
[New Thread 0x7fffefeb8700 (LWP 15570)]
Input #0, lavfi, from 'testsrc=s=hd1080,format=yuv420p':
  Duration: N/A, start: 0.000000, bitrate: N/A
    Stream #0:0: Video: rawvideo (I420 / 0x30323449), yuv420p, 1920x1080 [SAR 1:1 DAR 16:9], 25 tbr, 25 tbn, 25 tbc
Stream mapping:
  Stream #0:0 -> #0:0 (rawvideo (native) -> wrapped_avframe (native))
Press [q] to stop, [?] for help
[New Thread 0x7fffeebc5700 (LWP 15571)]
[New Thread 0x7fffee3c4700 (LWP 15572)]
[New Thread 0x7fffedbc3700 (LWP 15573)]
[New Thread 0x7fffed3c2700 (LWP 15574)]
[New Thread 0x7fffecbc1700 (LWP 15575)]
[New Thread 0x7fffec3c0700 (LWP 15576)]
[New Thread 0x7fffebbbf700 (LWP 15577)]
[New Thread 0x7fffeb3be700 (LWP 15578)]
[New Thread 0x7fffeabbd700 (LWP 15579)]
Output #0, null, to 'pipe:':
  Metadata:
    encoder         : Lavf57.75.100
    Stream #0:0: Video: wrapped_avframe, yuv420p, 1440x1080 [SAR 1:1 DAR 4:3], q=2-31, 200 kb/s, 12.50 fps, 12.50 tbn, 12.50 tbc
    Metadata:
      encoder         : Lavc57.99.102 wrapped_avframe

Thread 1 "ffmpeg_g" received signal SIGSEGV, Segmentation fault.
0x00000000005f8b63 in ff_lowpass_line_sse2 ()
(gdb) bt
#0  0x00000000005f8b63 in ff_lowpass_line_sse2 ()
#1  0x0000000000533a50 in copy_picture_field (src_frame=0x2046960, dst_frame=dst_frame@entry=0x2045720, 
    inlink=inlink@entry=0x20443c0, field_type=field_type@entry=FIELD_UPPER, lowpass=1, s=<optimized out>)
    at libavfilter/vf_interlace.c:193
#2  0x0000000000533bce in filter_frame (inlink=inlink@entry=0x20443c0, buf=<optimized out>) at libavfilter/vf_interlace.c:257
#3  0x00000000004bd60b in ff_filter_frame_framed (frame=<optimized out>, link=0x20443c0) at libavfilter/avfilter.c:1116
#4  ff_filter_frame_to_filter (link=0x20443c0) at libavfilter/avfilter.c:1264
#5  ff_filter_activate_default (filter=<optimized out>) at libavfilter/avfilter.c:1315
#6  ff_filter_activate (filter=<optimized out>) at libavfilter/avfilter.c:1476
#7  0x00000000004c0bdc in ff_filter_graph_run_once (graph=<optimized out>) at libavfilter/avfiltergraph.c:1449
#8  0x00000000004c1206 in get_frame_internal (samples=<optimized out>, flags=1, frame=<optimized out>, ctx=0x2045300)
    at libavfilter/buffersink.c:110
#9  av_buffersink_get_frame_flags (ctx=0x2045300, frame=frame@entry=0x0, flags=flags@entry=1) at libavfilter/buffersink.c:121
#10 0x00000000004c0929 in avfilter_graph_request_oldest (graph=0x2043400) at libavfilter/avfiltergraph.c:1402
#11 0x000000000047e89b in transcode_from_filter (best_ist=<synthetic pointer>, graph=<optimized out>) at ffmpeg.c:4455
#12 transcode_step () at ffmpeg.c:4521
#13 transcode () at ffmpeg.c:4597
#14 main (argc=<optimized out>, argv=<optimized out>) at ffmpeg.c:4803

(gdb) disass $pc-32,$pc+32
Dump of assembler code from 0x5f8b43 to 0x5f8b83:
   0x00000000005f8b43:  nopl   %cs:0x0(%rax,%rax,1)
   0x00000000005f8b4c:  nopl   0x0(%rax)
   0x00000000005f8b50 <ff_lowpass_line_sse2+0>: add    %rsi,%rdi
   0x00000000005f8b53 <ff_lowpass_line_sse2+3>: add    %rsi,%rdx
   0x00000000005f8b56 <ff_lowpass_line_sse2+6>: add    %rdx,%rcx
   0x00000000005f8b59 <ff_lowpass_line_sse2+9>: add    %rdx,%r8
   0x00000000005f8b5c <ff_lowpass_line_sse2+12>:        neg    %rsi
   0x00000000005f8b5f <ff_lowpass_line_sse2+15>:        pcmpeqb %xmm6,%xmm6
=> 0x00000000005f8b63 <ff_lowpass_line_sse2+19>:        movdqa (%rcx,%rsi,1),%xmm0
   0x00000000005f8b68 <ff_lowpass_line_sse2+24>:        movdqa 0x10(%rcx,%rsi,1),%xmm1
   0x00000000005f8b6e <ff_lowpass_line_sse2+30>:        pavgb  (%r8,%rsi,1),%xmm0
   0x00000000005f8b74 <ff_lowpass_line_sse2+36>:        pavgb  0x10(%r8,%rsi,1),%xmm1
   0x00000000005f8b7b <ff_lowpass_line_sse2+43>:        pxor   %xmm6,%xmm0
   0x00000000005f8b7f <ff_lowpass_line_sse2+47>:        pxor   %xmm6,%xmm1
End of assembler dump.

(gdb) info all-registers
rax            0x2042260        33825376
rbx            0x10e    270
rcx            0x7fffeedc0758   140737200785240
rdx            0x7fffeedc0758   140737200785240
rsi            0xfffffffffffffd30       -720
rdi            0x7ffff7e65310   140737352454928
rbp            0x10e    0x10e
rsp            0x7fffffffdfd8   0x7fffffffdfd8
r8             0x7fffeedc0b18   140737200786200
r9             0x10e    270
r10            0x22     34
r11            0x246    582
r12            0x2d0    720
r13            0x7ffff7e65040   140737352454208
r14            0x2      2
r15            0x7fffeedc0488   140737200784520
rip            0x5f8b63 0x5f8b63 <ff_lowpass_line_sse2+19>
eflags         0x10287  [ CF PF SF IF RF ]
cs             0x33     51
ss             0x2b     43
ds             0x0      0
es             0x0      0
fs             0x0      0
gs             0x0      0
st0            0        (raw 0x00000000000000000000)
st1            0        (raw 0x00000000000000000000)
st2            0        (raw 0x00000000000000000000)
st3            0        (raw 0x00000000000000000000)
st4            0        (raw 0x00000000000000000000)
st5            0        (raw 0x00000000000000000000)
st6            0        (raw 0x00000000000000000000)
st7            0        (raw 0x00000000000000000000)
fctrl          0x37f    895
fstat          0x0      0
ftag           0xffff   65535
fiseg          0x0      0
fioff          0x0      0
foseg          0x0      0
fooff          0x0      0
fop            0x0      0
xmm0           {v4_float = {0x0, 0x0, 0x0, 0x0}, v2_double = {0x0, 0x0}, v16_int8 = {0xaa <repeats 16 times>}, v8_int16 = {
    0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa}, v4_int32 = {0xaaaaaaaa, 0xaaaaaaaa, 0xaaaaaaaa, 
    0xaaaaaaaa}, v2_int64 = {0xaaaaaaaaaaaaaaaa, 0xaaaaaaaaaaaaaaaa}, uint128 = 0xaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa}
xmm1           {v4_float = {0x0, 0x0, 0x0, 0x0}, v2_double = {0x0, 0x0}, v16_int8 = {0xaa <repeats 16 times>}, v8_int16 = {
    0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa, 0xaaaa}, v4_int32 = {0xaaaaaaaa, 0xaaaaaaaa, 0xaaaaaaaa, 
    0xaaaaaaaa}, v2_int64 = {0xaaaaaaaaaaaaaaaa, 0xaaaaaaaaaaaaaaaa}, uint128 = 0xaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa}
xmm2           {v4_float = {0x55500000, 0x55500000, 0x55500000, 0x55500000}, v2_double = {0x8000000000000000, 
    0x8000000000000000}, v16_int8 = {0x55 <repeats 16 times>}, v8_int16 = {0x5555, 0x5555, 0x5555, 0x5555, 0x5555, 0x5555, 
    0x5555, 0x5555}, v4_int32 = {0x55555555, 0x55555555, 0x55555555, 0x55555555}, v2_int64 = {0x5555555555555555, 
    0x5555555555555555}, uint128 = 0x55555555555555555555555555555555}
xmm3           {v4_float = {0x55500000, 0x55500000, 0x55500000, 0x55500000}, v2_double = {0x8000000000000000, 
    0x8000000000000000}, v16_int8 = {0x55 <repeats 16 times>}, v8_int16 = {0x5555, 0x5555, 0x5555, 0x5555, 0x5555, 0x5555, 
---Type <return> to continue, or q <return> to quit---
    0x5555, 0x5555}, v4_int32 = {0x55555555, 0x55555555, 0x55555555, 0x55555555}, v2_int64 = {0x5555555555555555, 
    0x5555555555555555}, uint128 = 0x55555555555555555555555555555555}
xmm4           {v4_float = {0x0, 0x0, 0x0, 0x0}, v2_double = {0x0, 0x0}, v16_int8 = {0x80 <repeats 16 times>}, v8_int16 = {
    0x8080, 0x8080, 0x8080, 0x8080, 0x8080, 0x8080, 0x8080, 0x8080}, v4_int32 = {0x80808080, 0x80808080, 0x80808080, 
    0x80808080}, v2_int64 = {0x8080808080808080, 0x8080808080808080}, uint128 = 0x80808080808080808080808080808080}
xmm5           {v4_float = {0x0, 0x0, 0x0, 0x0}, v2_double = {0x0, 0x0}, v16_int8 = {0x80 <repeats 16 times>}, v8_int16 = {
    0x8080, 0x8080, 0x8080, 0x8080, 0x8080, 0x8080, 0x8080, 0x8080}, v4_int32 = {0x80808080, 0x80808080, 0x80808080, 
    0x80808080}, v2_int64 = {0x8080808080808080, 0x8080808080808080}, uint128 = 0x80808080808080808080808080808080}
xmm6           {v4_float = {0x0, 0x0, 0x0, 0x0}, v2_double = {0x8000000000000000, 0x8000000000000000}, v16_int8 = {
    0xff <repeats 16 times>}, v8_int16 = {0xffff, 0xffff, 0xffff, 0xffff, 0xffff, 0xffff, 0xffff, 0xffff}, v4_int32 = {
    0xffffffff, 0xffffffff, 0xffffffff, 0xffffffff}, v2_int64 = {0xffffffffffffffff, 0xffffffffffffffff}, 
  uint128 = 0xffffffffffffffffffffffffffffffff}
xmm7           {v4_float = {0x0, 0x0, 0x0, 0x0}, v2_double = {0x0, 0x0}, v16_int8 = {0x0 <repeats 16 times>}, v8_int16 = {0x0, 
    0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0}, v4_int32 = {0x0, 0x0, 0x0, 0x0}, v2_int64 = {0x0, 0x0}, 
  uint128 = 0x00000000000000000000000000000000}
xmm8           {v4_float = {0x0, 0x0, 0x0, 0x0}, v2_double = {0x0, 0x0}, v16_int8 = {0x80 <repeats 16 times>}, v8_int16 = {
    0x8080, 0x8080, 0x8080, 0x8080, 0x8080, 0x8080, 0x8080, 0x8080}, v4_int32 = {0x80808080, 0x80808080, 0x80808080, 
    0x80808080}, v2_int64 = {0x8080808080808080, 0x8080808080808080}, uint128 = 0x80808080808080808080808080808080}
xmm9           {v4_float = {0x0, 0x0, 0x0, 0x0}, v2_double = {0x0, 0x0}, v16_int8 = {0x88, 0xc, 0x0, 0x0, 0x87, 0x40, 0x88, 0xc, 
    0x88, 0xc, 0x0, 0x0, 0x87, 0x40, 0x88, 0xc}, v8_int16 = {0xc88, 0x0, 0x4087, 0xc88, 0xc88, 0x0, 0x4087, 0xc88}, v4_int32 = {
    0xc88, 0xc884087, 0xc88, 0xc884087}, v2_int64 = {0xc88408700000c88, 0xc88408700000c88}, 
  uint128 = 0x0c88408700000c880c88408700000c88}
xmm10          {v4_float = {0x0, 0x0, 0x0, 0x0}, v2_double = {0x0, 0x0}, v16_int8 = {0x2, 0x80, 0x3, 0x80, 0x4, 0x80, 0x5, 0x80, 
    0x8, 0x80, 0x9, 0x80, 0xa, 0x80, 0xb, 0x80}, v8_int16 = {0x8002, 0x8003, 0x8004, 0x8005, 0x8008, 0x8009, 0x800a, 0x800b}, 
  v4_int32 = {0x80038002, 0x80058004, 0x80098008, 0x800b800a}, v2_int64 = {0x8005800480038002, 0x800b800a80098008}, 
  uint128 = 0x800b800a800980088005800480038002}
xmm11          {v4_float = {0x0, 0x0, 0x0, 0x0}, v2_double = {0x0, 0x0}, v16_int8 = {0x0, 0x0, 0x0, 0xff, 0x0, 0x0, 0x0, 0x0, 
    0x0, 0xff, 0x0, 0x0, 0x0, 0xff, 0x0, 0x0}, v8_int16 = {0x0, 0xff00, 0x0, 0x0, 0xff00, 0x0, 0xff00, 0x0}, v4_int32 = {
    0xff000000, 0x0, 0xff00, 0xff00}, v2_int64 = {0xff000000, 0xff000000ff00}, uint128 = 0x0000ff000000ff0000000000ff000000}
xmm12          {v4_float = {0x0, 0x0, 0x0, 0x0}, v2_double = {0x0, 0x0}, v16_int8 = {0x66, 0x94, 0x48, 0xf, 0x11, 0x14, 0xe3, 
    0xb9, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0}, v8_int16 = {0x9466, 0xf48, 0x1411, 0xb9e3, 0x0, 0x0, 0x0, 0x0}, v4_int32 = {
    0xf489466, 0xb9e31411, 0x0, 0x0}, v2_int64 = {0xb9e314110f489466, 0x0}, uint128 = 0x0000000000000000b9e314110f489466}
xmm13          {v4_float = {0x0, 0x0, 0x0, 0x0}, v2_double = {0x0, 0x0}, v16_int8 = {0x40, 0xfe, 0xe8, 0x3a, 0x73, 0x76, 0x56, 
    0xbc, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0}, v8_int16 = {0xfe40, 0x3ae8, 0x7673, 0xbc56, 0x0, 0x0, 0x0, 0x0}, v4_int32 = {
    0x3ae8fe40, 0xbc567673, 0x0, 0x0}, v2_int64 = {0xbc5676733ae8fe40, 0x0}, uint128 = 0x0000000000000000bc5676733ae8fe40}
xmm14          {v4_float = {0x0, 0x0, 0x0, 0x0}, v2_double = {0x0, 0x0}, v16_int8 = {0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x80, 
    0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0}, v8_int16 = {0x0, 0x0, 0x0, 0x8000, 0x0, 0x0, 0x0, 0x0}, v4_int32 = {0x0, 
    0x80000000, 0x0, 0x0}, v2_int64 = {0x8000000000000000, 0x0}, uint128 = 0x00000000000000008000000000000000}
xmm15          {v4_float = {0x0, 0x0, 0x0, 0x0}, v2_double = {0x0, 0x0}, v16_int8 = {0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x80, 
    0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0}, v8_int16 = {0x0, 0x0, 0x0, 0x8000, 0x0, 0x0, 0x0, 0x0}, v4_int32 = {0x0, 
    0x80000000, 0x0, 0x0}, v2_int64 = {0x8000000000000000, 0x0}, uint128 = 0x00000000000000008000000000000000}

[Carl Eugen Hoyos]

Can be considered a regression since 6f373d75 / 2e170405

[mvignali]

Fixe in
​49dced9fd0c8a8d2f3676533914fda7417435df2

and similar case (with low_pass_complex) in
​3c6dc270355f27645cf931fae1ed2dc1405507f8