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 7.0 has been released! | Please read ffmpeg.org/developer.html#Code-of-conduct
<Lynne> elenril: you broke something in lavfi
<Lynne> before, using av_buffersink_get_frame_flags() with AV_BUFFERSINK_FLAG_NO_REQUEST would eventually get you an AVERROR_EOF
<Lynne> but now, AVERROR_EOF is never returned
<Lynne> elenril: correction, what happens, is that with AV_BUFFERSINK_FLAG_NO_REQUEST, EOF is never propagated
<Lynne> in fact it always returns EAGAIN despite having hundreds of megabytes of frames
<Lynne> perhaps I was misusing the API, calling avfilter_graph_request_oldest after every av_buffersrc_add_frame_flags does work
<Lynne> but why did it work before?
aaabbb has quit [Ping timeout: 246 seconds]
lexano has quit [Ping timeout: 260 seconds]
iive has quit [Quit: They came for me...]
thilo has quit [Ping timeout: 256 seconds]
thilo has joined #ffmpeg-devel
thilo has joined #ffmpeg-devel
jamrial has quit [Read error: Connection reset by peer]
jamrial_ has joined #ffmpeg-devel
lemourin has quit [Quit: The Lounge - https://thelounge.chat]
lemourin has joined #ffmpeg-devel
arch1t3cht1 has joined #ffmpeg-devel
arch1t3cht has quit [Ping timeout: 260 seconds]
arch1t3cht1 is now known as arch1t3cht
cone-731 has joined #ffmpeg-devel
<cone-731> ffmpeg Frank Plowman master:2d79ae3f8a33: lavc/vvc: Error if SPS ID is duplicated within CVS
MisterMinister has joined #ffmpeg-devel
MisterMinister has quit [Ping timeout: 268 seconds]
jamrial_ has quit []
Martchus has joined #ffmpeg-devel
Martchus_ has quit [Ping timeout: 268 seconds]
ringo has quit [Ping timeout: 240 seconds]
ringo has joined #ffmpeg-devel
ringo is now known as Guest3678
bossjones has quit [Ping timeout: 240 seconds]
bossjones has joined #ffmpeg-devel
System_Error has quit [Remote host closed the connection]
System_Error has joined #ffmpeg-devel
cone-731 has quit [Quit: transmission timeout]
<elenril> Lynne: what makes you think it was me?
<Lynne> I was thinking it may have been the filter graph changes
<elenril> which changes?
<elenril> I don't think I touched that part of lavfi recently
<Lynne> not sure in that case, could have been elon musk
<Lynne> it does make sense that you have to trigger the filter graph yourself though, no idea why it happened before
<elenril> nobody understands how that part of lavfi works
<elenril> it should be replaced entirely
<Lynne> yeah, sorry to jump to conclusions
<Lynne> we still have two scheduling systems for multi-input frames, by two authors, with one more planned
<Lynne> s/frames/filters
<elenril> the main problem is that it's entangled with filters themselves
darkapex has quit [Ping timeout: 246 seconds]
darkapex has joined #ffmpeg-devel
Poorvagaikar2003 has joined #ffmpeg-devel
<j-b> good morning
kasper93 has quit [Ping timeout: 264 seconds]
kasper93 has joined #ffmpeg-devel
mkver has joined #ffmpeg-devel
Raz- has joined #ffmpeg-devel
<Lynne> are the chinese planning to stick with codecid=12 or are they switching to ehnanced rtmp?
Livio has joined #ffmpeg-devel
Poorva_ has joined #ffmpeg-devel
Poorvagaikar2003 has quit [Ping timeout: 240 seconds]
Livio has quit [Ping timeout: 268 seconds]
Poorva_ has quit [Quit: Leaving]
kepstin has quit [Remote host closed the connection]
kepstin has joined #ffmpeg-devel
kurosu has joined #ffmpeg-devel
kasper93 has quit [Ping timeout: 268 seconds]
kasper93_ has joined #ffmpeg-devel
scat117 has joined #ffmpeg-devel
kasper93 has joined #ffmpeg-devel
kasper93_ has quit [Ping timeout: 256 seconds]
j45_ has joined #ffmpeg-devel
<kierank> ffmpeg now needs "Ubuntu Pro" 22.04?
<kierank> mkver: do you know about this?
j45 has quit [Ping timeout: 268 seconds]
j45_ is now known as j45
j45 has joined #ffmpeg-devel
j45 has quit [Changing host]
<JEEB> why would it? :D
<mkver> kierank: No.
<JEEB> ubuntu pro means that you get extra support, so if FFmpeg is within packages that are no longer supported by normal support period then yes - any updates for that would come from the extended support stuff
<kierank> there is a user complaining on X about this
<kierank> he is on 22.04
<kierank> saying it says it needs pro to update
<JEEB> you can list the support period for packages with some command
<BtbN> Yes, I saw that too recently. "Hey, there's packages with known security issues on your system: ... libav* ...", "Sign up to Ubuntu Pro to update them!"
<BtbN> super scummy behaviour
<kierank> should I tweet ubuntu about this?
<nevcairiel> dont they offer 5 years of updates for mainline, and only require pro after
<BtbN> Only for the main repo
<JEEB> teh support has always been split into repos
<BtbN> universe/multiverse, nope. And ffmpeg is in there.
<JEEB> yea
<kierank> is that to trollish?
<kierank> j-b: ^
<BtbN> I mean, it's factually incorrect
<BtbN> It's not marked as needing Ubuntu Pro. Just security updates are.
<kierank> that's crazy
<kierank> people package security updates and ubuntu charge for them
<JEEB> well, by itself it's nothing new; non-main packages have always had shorter support period
<BtbN> Also, those updates don't even bump the version. It's just a bunch of Ubuntu-Maintainer hand-picket backports.
<BtbN> *picked
<kierank> the guy dmming me is right
<JEEB> it's just that they now actively try to get you to pay for longer support
<kierank> it's confusing as hell
<JEEB> so they are showing the stuff in various places
<BtbN> The super shitty thing about this is: If you want to stay on LTS on a Desktop... you can't.
<JEEB> ubuntu-support-status used to be a command that shows that stuff, I think?
<BtbN> Or even on a server that needs media components
<BtbN> Cause 22.04, the last LTS, is pulling this shit before the next LTS is even released
<BtbN> This completely disqualifies Ubuntu for anything but toy-use imo
<kierank> what do we do about this?
<kierank> before the twitter nuclear option
<BtbN> I don't think we can do anything about it. It's just shiity company policy, and I'm sure they checked with a lawyer that it's fine.
<JEEB> BtbN: yea it seems like even in 2013 they had this split
<BtbN> Just use Debian instead really
<BtbN> Ubuntu is migrating everything to Snaps anyway, which makes it useless
<JEEB> but yea, the only change is that they are just advertising the paid option everywhere now
<JEEB> so people notice it
<JEEB> but in actuality they had almost always stopped supporting non-main packages much more quickly than 5 years
<kierank> seems so sketch
<kierank> just no security updates
<BtbN> Yeah, "You can use this distro as a desktop safely for 5 years"
<BtbN> But we stop updating critical media components that parse tons of random stuff from the web after less than two years.
<BtbN> So that means, you just can't use Ubuntu safely on a Desktop if you don't do every single minor edition upgrade
Krowl has joined #ffmpeg-devel
<Sebastinas> Is the ppc asm still maintained? There's #9604 unanswered for 2 years and with 7.0 there is now also #10955.
<mkver> Sebastinas: I recently nuked ppc asm that failed checkasm; the tests your ticket refer to fail depending upon the seed.
<mkver> I don't remember anyone properly working on ppc asm. So yeah, it is unmaintained.
arch1t3cht has quit [Ping timeout: 264 seconds]
arch1t3cht has joined #ffmpeg-devel
<Sebastinas> Okay, then I'll build with --disable-asm on that arch for now.
jamrial has joined #ffmpeg-devel
Krowl has quit [Read error: Connection reset by peer]
<BBB> you could ask lu_zero, he worked on ppc asm at some point
<BBB> but I think he moved on from there
<sdc> does anyone know why the ffmpeg/ffmpeg_g binaries might not be made when I try to disable pthreads / os2 threads?
<sdc> this is on linux
<mkver> sdc: The ffmpeg cli tools relies on threading support.
<mkver> sdc: Since 760ce4bc0bd11f74f0851c0a662dd5cae888df83.
<sdc> ahh okay thanks
iive has joined #ffmpeg-devel
<sdc> So I've been trying to write an x86 SAD implementation. I've been trying to write just a dummy function that does random things and then returns the value but keep getting an error
<sdc> I'm guess there's something obvious I'm missing here?
<sdc> the function call is
<sdc> min_sad = fc->vvcdsp.inter.sad(tmp[L0], tmp[L1], dx, dy, block_w, block_h);
<sdc> min_sad = fc->vvcdsp.inter.mysad(tmp[L0], tmp[L1], 4, 6, block_w, block_h);
<Lynne> sdc: in the function definition, you're excluding block_w/h from being counted as arguments
<Lynne> 6 vector regs, 8 gprs, 4 of the gpurs are arguments, is what you've defined
<sdc> OH
<jamrial> Lynne: no, you're reading it wrong. he declared 6 arguments, 8 gpr, and 4 xmm
<Lynne> oh, right, the order is reversed
<Lynne> in that case, maybe sign extension is the issue here
qeed_ has quit [Quit: qeed_]
lexano has joined #ffmpeg-devel
<Lynne> sdc: try loading everything with movsxd before adding?
<jamrial> you don't even need that random tmp gpr for what you're doing there
<jamrial> also, i'm very sure we have sad functions already
<Lynne> yeah, libavutil/pixelutils.h
<sdc> yeah more for practice / 16bit version eventually maybe
kurosu has quit [Quit: Connection closed for inactivity]
lexano has quit [Ping timeout: 260 seconds]
<jamrial> you don't need random, and you're requesting 8 gprs when you only need 7 (6 if you remove random)
<jamrial> and the disassembly on the right looks weird
<jamrial> why al/cl? and roll?
Livio has joined #ffmpeg-devel
<Lynne> sdc: I meant all variables, including the ones you add
<jamrial> he doesn't need sign extension if he's never using 64bit values or regs
<jamrial> if there's a problem, is either in the assembler, or the function prototype
<another|> sdc: dx, dy swapped
lexano has joined #ffmpeg-devel
kurosu has joined #ffmpeg-devel
lexano has quit [Ping timeout: 272 seconds]
* Sean_McG peeks in
lexano has joined #ffmpeg-devel
* another| peeks out
System_Error has quit [Ping timeout: 260 seconds]
Krowl has joined #ffmpeg-devel
System_Error has joined #ffmpeg-devel
Krowl has quit [Quit: Krowl]
Krowl has joined #ffmpeg-devel
<sdc> holy crap I'm dumb... I forgot SECTION .text .....
<sdc> thanks Lynne and James for helping to troubleshoot
<sdc> and another as well
<Lynne> ah...
<Lynne> I've been there, the other way around, disassembling rodata and wondering where the nonsense instructions I was looking at were coming from
rvalue- has joined #ffmpeg-devel
rvalue has quit [Ping timeout: 256 seconds]
rvalue- is now known as rvalue
<Sean_McG> I'm having renovations done at my house so my FATE nodes will probably be down for a few days.
<kierank> sdc: do not worry about it
<kierank> we have all been there
lexano has quit [Ping timeout: 256 seconds]
Sean_McG has quit [Quit: leaving]
zsoltiv_ has joined #ffmpeg-devel
Krowl has quit [Read error: Connection reset by peer]
lexano has joined #ffmpeg-devel
dykai has joined #ffmpeg-devel
dykai has quit [Ping timeout: 260 seconds]
Livio has quit [Ping timeout: 260 seconds]
Livio has joined #ffmpeg-devel
kasper93 has quit [Ping timeout: 256 seconds]
cone-297 has joined #ffmpeg-devel
<cone-297> ffmpeg Andreas Rheinhardt master:eef5d60ac620: avcodec/huffyuv: Inline common alloc/free functions in their callers
<cone-297> ffmpeg Andreas Rheinhardt master:cebf1d59a569: avcodec/huffyuvdec: Don't zero unnecessarily
<cone-297> ffmpeg Andreas Rheinhardt master:cf96c0295ea9: avcodec/huffyuv(dec|enc): Use union for temp/temp16
<cone-297> ffmpeg Andreas Rheinhardt master:d517c9e51be5: avcodec/huffyuvenc: Avoid duplicate variables
<cone-297> ffmpeg Andreas Rheinhardt master:0e5af493fc13: avcodec/huffyuv: Return proper error code
<cone-297> ffmpeg Andreas Rheinhardt master:5110d16f5ad5: avcodec/huffyuvenc: Avoid code duplication
<cone-297> ffmpeg Andreas Rheinhardt master:a6979e3bd2b3: avcodec/huffyuvenc: Deduplicate options
<cone-297> ffmpeg Andreas Rheinhardt master:a6d6c8442ce9: avcodec/huffyuvdec: Use bytestream API for byte-aligned reads
<cone-297> ffmpeg Andreas Rheinhardt master:f3c8d0399acd: avcodec/huffyuvdec: Use assert to check for things that can't fail
<cone-297> ffmpeg Andreas Rheinhardt master:2d33d6bfcc2a: avcodec/dv: Don't pretend initializing work chunks can fail
<andrewrk> sdc: if you think you're dumb look at this bug that took me 3 days to track down: https://codeberg.org/andrewrk/player/commit/36190af657dfae09933bd99d30190709f9e799a5
kasper93 has joined #ffmpeg-devel
AbleBacon has joined #ffmpeg-devel
kasper93 has quit [Ping timeout: 256 seconds]
Raz- has quit [Ping timeout: 240 seconds]
System_Error has quit [Remote host closed the connection]
System_Error has joined #ffmpeg-devel
<sdc> Haha yeah definitely more relieved to finally get it resolved!
kurosu has quit [Quit: Connection closed for inactivity]
Livio has quit [Ping timeout: 260 seconds]
sudden has quit [Ping timeout: 260 seconds]
kasper93 has joined #ffmpeg-devel
sudden has joined #ffmpeg-devel
cone-297 has quit [Quit: transmission timeout]
Marth64 has joined #ffmpeg-devel
<Marth64> <3 ffmpeg. Car crashed into mine today (luckily everybody okay). Needed to cleanup and salvage footage from proprietary dashcam. Least stressful part of the day was using ffmpeg to do that and hstack it.
<Marth64> you all have a good day. be back soon
Marth64 has quit [Remote host closed the connection]
lexano has quit [Ping timeout: 272 seconds]