michaelni changed the topic of #ffmpeg-devel to: Welcome to the FFmpeg development channel | Questions about using FFmpeg or developing with libav* libs should be asked in #ffmpeg | This channel is publicly logged | FFmpeg 7.1 has been released! | Please read ffmpeg.org/developer.html#Code-of-conduct
zenmov has quit [Ping timeout: 252 seconds]
names_are_hard has quit [Quit: Leaving]
^Neo has joined #ffmpeg-devel
^Neo has quit [Changing host]
^Neo has joined #ffmpeg-devel
iive has quit [Quit: They came for me...]
System_Error has quit [Remote host closed the connection]
IndecisiveTurtle has quit [Quit: IndecisiveTurtle]
System_Error has joined #ffmpeg-devel
arch1t3cht5 has joined #ffmpeg-devel
arch1t3cht has quit [Ping timeout: 248 seconds]
arch1t3cht5 is now known as arch1t3cht
zenmov has joined #ffmpeg-devel
mkver has quit [Ping timeout: 276 seconds]
thilo has quit [Ping timeout: 272 seconds]
thilo has joined #ffmpeg-devel
zenmov_ has joined #ffmpeg-devel
zenmov has quit [Ping timeout: 276 seconds]
^Neo has quit [Ping timeout: 252 seconds]
Marth64 has joined #ffmpeg-devel
feiw has quit [Ping timeout: 252 seconds]
feiw has joined #ffmpeg-devel
kasper93 has quit [Remote host closed the connection]
zenmov_ has quit [Read error: Connection reset by peer]
jamrial has quit []
Everything has quit [Quit: leaving]
zenmov has joined #ffmpeg-devel
feiw has quit [Ping timeout: 252 seconds]
feiw has joined #ffmpeg-devel
KillerWasp has joined #ffmpeg-devel
\\Mr_C\\ has quit [Remote host closed the connection]
<KillerWasp>
hello, where i can find ffformatcontext() function? what's it? seem give a FFFormatContext structure, but i don't know what is it also.
<KillerWasp>
oooooh.... i found it....
<KillerWasp>
Definitely a ffmpeg bug, it is related to the input data, specifically a temporary packet. But since I am not using input data, only output, then it gives segfault.
<KillerWasp>
The error appear in L1253:./libavformat/mux.c
<KillerWasp>
AVPacket *const pkt = si->parse_pkt;
Martchus_ has joined #ffmpeg-devel
Martchus has quit [Ping timeout: 276 seconds]
Traneptora has joined #ffmpeg-devel
<Marth64>
KillerWasp:What are you trying to do?
<Marth64>
(surrounding code or CLI command, etc)
cone-401 has joined #ffmpeg-devel
<cone-401>
ffmpeg Frank Plowman master:1e3dc705dfb9: lavc/x86/videodsp: Drop MMX usage
Sean_McG has quit [Ping timeout: 246 seconds]
Sean_McG has joined #ffmpeg-devel
rvalue- has joined #ffmpeg-devel
rvalue has quit [Ping timeout: 255 seconds]
rvalue- is now known as rvalue
Marth64 has quit [Quit: Leaving]
cone-401 has quit [Quit: transmission timeout]
lexano has quit [Ping timeout: 265 seconds]
lexano has joined #ffmpeg-devel
Krowl has joined #ffmpeg-devel
Krowl has quit [Read error: Connection reset by peer]
Krowl has joined #ffmpeg-devel
kepstin has quit [Remote host closed the connection]
kepstin has joined #ffmpeg-devel
IndecisiveTurtle has joined #ffmpeg-devel
kasper93 has joined #ffmpeg-devel
blb has quit [Ping timeout: 252 seconds]
blb has joined #ffmpeg-devel
Krowl has quit [Read error: Connection reset by peer]
^Neo has joined #ffmpeg-devel
^Neo has quit [Changing host]
^Neo has joined #ffmpeg-devel
jamrial has joined #ffmpeg-devel
<ramiro>
michaelni: I still don't understand how that chunk of code you reviewed can trigger ub. dstU[i] is int16_t, multiplying it by 4663 doesn't overflow, subtracting 9289992 doesn't either, and the right shift propagates the msb
<ramiro>
I feel like I'm missing some fundamental understanding of these undefined behaviour fixes.
<michaelni>
ramiro, you are correct, please disregard my comment
<michaelni>
i must have mixed something up when i looked at it
<ramiro>
michaelni: ok, thanks. maybe you meant chrRangeToJpeg16_c instead of chrRangeToJpeg_c. the input in this case is int32_t, so there might be an ub there.
<michaelni>
hmm, in chrRangeToJpeg16_c() it uses unsigned so its not undefined, but that should make the >>12 unsigned too which is ok if only the low 16 bit is used, bit the FFMIN there could have unexpected results maybe if the value is negative
<michaelni>
but even before it used 32bit
<michaelni>
so hmm
<michaelni>
no it did cast to int before the shift before the change
<michaelni>
i think you should leave the cast before the shift
<michaelni>
ramiro, the cast is there so the shift is signed and not unsigned, it becomes unsigned because of the * 4663U
<michaelni>
but its interresting that this is not piccked up by any tests, maybe its not so easy to be negative there
mkver has joined #ffmpeg-devel
zenmov has quit [Quit: leaving]
zenmov has joined #ffmpeg-devel
zenmov has quit [Quit: leaving]
zenmov has joined #ffmpeg-devel
sudden has quit [Ping timeout: 252 seconds]
sudden has joined #ffmpeg-devel
cone-218 has joined #ffmpeg-devel
<cone-218>
ffmpeg James Almer release/7.1:3513ccccb650: avformat/movenc: don't write a calculated avgBitrate when the provided one is unset
<cone-218>
ffmpeg James Almer release/7.1:33588a37497b: avformat/movenc: write stss boxes for xHE-AAC
<cone-218>
ffmpeg James Almer release/7.1:cb997ccd0973: avcodec/libfdk-aacenc: export CPB properties
<cone-218>
ffmpeg James Almer release/7.1:8fa5fa146302: avformat/movenc: don't disable edit lists when writing CMAF output
<cone-218>
ffmpeg James Almer release/7.1:e07d7c75dbfd: avformat/spdifdec: parse headers for audio codecs
<cone-218>
ffmpeg James Almer release/7.1:13282e61b808: avformat/test/movenc: set audio packets as key frames
<cone-218>
ffmpeg James Almer release/7.1:f1592a7ff1d3: avformat/matroskadec: set all frames in a keyframe simple box as keyframes
<cone-218>
ffmpeg James Almer release/7.1:dc9d9b8636f1: avformat/mov: don't unconditionally set all audio packets in fragments as key frames
<cone-218>
ffmpeg James Almer release/7.1:075c6d325ac1: avcodec/aac_parser: set key_frame and profile
<cone-218>
ffmpeg James Almer release/7.1:1b5bdce86a76: avcodec/aac/aacdec: set keyframe flag in output frames
<cone-218>
ffmpeg James Almer release/7.1:302f1b388253: avcodec/aacenc: set keyframe flag in output packets
<cone-218>
ffmpeg James Almer release/7.1:07a92c1a640d: avcodec/audiotoolboxdec: set set keyframe flag in output frames
<cone-218>
ffmpeg James Almer release/7.1:f7d59ab17a75: avcodec/audiotoolboxnec: set set keyframe flag in output packets
<cone-218>
ffmpeg James Almer release/7.1:c13a0a2f59bd: avcodec/libfdk-aacdec: set keyframe flag and profile in output frames
<cone-218>
ffmpeg James Almer release/7.1:26bc7827e4be: avcodec/libfdk-aacenc: set keyframe in output packets
<cone-218>
ffmpeg James Almer release/7.1:4b801baa4673: avcodec/mediacodecdec: set set keyframe flag in output frames
<cone-218>
ffmpeg James Almer release/7.1:f265f9c9d048: avcodec/codec_desc: remove Intra Only prop for AAC
<cone-218>
ffmpeg James Almer release/7.1:364f6a5f118d: avcodec/ac3dec: fix downmix logic for eac3
Marth64 has joined #ffmpeg-devel
DauntlessOne4 has quit [Ping timeout: 244 seconds]
DauntlessOne4 has joined #ffmpeg-devel
<kasper93>
Marth64: seeking with angle set to not 1, doesn't seem to work
<kasper93>
seems to always jump to begining of chapter
<kasper93>
(there is one chapter with alternative angles)
<kasper93>
works fine with default angle
iive has joined #ffmpeg-devel
<Marth64>
kasper93: I will check thanks
zsoltiv_ has joined #ffmpeg-devel
<fflogger>
[newticket] RmBeer: Ticket #11329 ([avformat] Give segfault where you don't use input datas) created https://trac.ffmpeg.org/ticket/11329
<haasn>
Am I being daft?
IndecisiveTurtle has quit [Ping timeout: 276 seconds]
<haasn>
Why does everyone seem to think that if (!ptr) return NULL; else return ptr; is somehow meaningfully different from just return ptr?
<sfan5>
do people think that?
<beastd>
haasn: not sure, maybe somehow mixed up that the if check would avoid deref, but that is the other if check before that you left untouched
<beastd>
or it's some other thing that is not checked yet and should be checked in future, at least from the code I cannot deduce that your version is functionally different
HarshK23 has quit [Quit: Connection closed for inactivity]
cone-218 has quit [Quit: transmission timeout]
cone-029 has joined #ffmpeg-devel
<cone-029>
ffmpeg Marth64 master:a5a899bfcbe4: avformat/dvdvideodec: fix missing last chapter marker due to off-by-one
<cone-029>
ffmpeg Marth64 master:e256761defe5: avformat/dvdvideodec: check return code of ff_dvdclut_yuv_to_rgb()