Open hholst80 opened 7 months ago
hholst80
commented
7 months ago I do not think the '--no-hw' path is especially attractive solution anyway, but the error message was more informative to why 'No usable encoding profile found' was returned in the default setup.
russelltg
commented
7 months ago Retrospectively --no-hw isn't a great name for the flag, as it still does the pixel format conversion in HW. Does wf-recorder with vaapi work for you? Hard to imagine it does considering vainfo doesn't show any h264 support.....
hholst80
commented
7 months ago I find it likely that the reason why it works on my other Fedora box is that it uses an Intel GPU and here on the workstation I have a fancy AMD GPU.
The error message "failed to create encoder: Function not implemented" is perhaps not the best, but besides that I do not think there is much you can help me with here.
"Fedora 37 drops VA-API accelerated hardware video decoding support"
https://www.reddit.com/r/linux/comments/xpyggi/comment/iq6nkhn/
russelltg
commented
7 months ago Hmm, this is encode and that is decode. Other people have reported this working fine on AMD GPUs, but I don't have one to test.
Unfortunately that error is from ffmpeg.
russelltg
commented
7 months ago I have happy users with AMD GPU's (#1, #39, #33). What GPU do you have? Does ffmpeg -vaapi_device /dev/dri/renderD128 -f lavfi -i rgbtestsrc=duration=5:size=1280x720:rate=30 -vf hwupload,crop=100:100:200:200,scale_vaapi=format=nv12:w=100:h=100 -c:v h264_vaapi testout.mp4 work?
hholst80
commented
7 months ago The situation seems to be improving on my side. I got a lot of things in order on the new system. So, I wanted of course to go back and revisit this. This issue was the reason why I changed from Fedora to Arch Linux (Artix flavor) to begin with. :-)
[I] root@neo ~
# ffmpeg -vaapi_device /dev/dri/renderD128 -f lavfi -i rgbtestsrc=duration=5:size=1280x720:rate=30 -vf hwupload,crop=100:100:200:200,scale_vaapi=format=nv12:w=100:h=100 -c:v h264_vaapi testout.mp4
ffmpeg version 6.1-static https://johnvansickle.com/ffmpeg/ Copyright (c) 2000-2023 the FFmpeg developers
built with gcc 8 (Debian 8.3.0-6)
configuration: --enable-gpl --enable-version3 --enable-static --disable-debug --disable-ffplay --disable-indev=sndio --disable-outdev=sndio --cc=gcc --enable-fontconfig --enable-frei0r --enable-gnutls --enable-gmp --enable-libgme --enable-gray --enable-libaom --enable-libfribidi --enable-libass --enable-libvmaf --enable-libfreetype --enable-libmp3lame --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libopenjpeg --enable-librubberband --enable-libsoxr --enable-libspeex --enable-libsrt --enable-libvorbis --enable-libopus --enable-libtheora --enable-libvidstab --enable-libvo-amrwbenc --enable-libvpx --enable-libwebp --enable-libx264 --enable-libx265 --enable-libxml2 --enable-libdav1d --enable-libxvid --enable-libzvbi --enable-libzimg
libavutil 58. 29.100 / 58. 29.100
libavcodec 60. 31.102 / 60. 31.102
libavformat 60. 16.100 / 60. 16.100
libavdevice 60. 3.100 / 60. 3.100
libavfilter 9. 12.100 / 9. 12.100
libswscale 7. 5.100 / 7. 5.100
libswresample 4. 12.100 / 4. 12.100
libpostproc 57. 3.100 / 57. 3.100
Unrecognized option 'vaapi_device'.
Error splitting the argument list: Option not found
[I] root@neo ~ [8]
# [I] root@neo ~
# wf-recorder -f test-vaapi.mp4 -c h264_vaapi -d /dev/dri/renderD128
Output file "test-vaapi.mp4" exists. Overwrite? Y/n: y
using VA-API, trying to enable DMA-BUF capture...
enabled DMA-BUF capture, device /dev/dri/renderD128
selected region 0,0 0x0
Using video filter: scale_vaapi=format=nv12:out_range=full
[h264_vaapi @ 0x7f1168002300] No quality level set; using default (20).
[h264_vaapi @ 0x7f1168002300] Driver does not support some wanted packed headers (wanted 0xd, found 0).
[h264_vaapi @ 0x7f1168002300] Driver does not support packed sequence headers, but a global header is requested.
[h264_vaapi @ 0x7f1168002300] No global header will be written: this may result in a stream which is not usable for some purposes (e.g. not muxable to some containers).
Output #0, mp4, to 'test-vaapi.mp4':
Stream #0:0: Video: h264 (High), vaapi(pc), 3840x2160 [SAR 1:1 DAR 16:9], q=2-31
^C⏎ [I] root@neo ~
# wl-screenrec --dri-device /dev/dri/renderD128
Using output DP-1
Opening libva device from DRM device /dev/dri/renderD128
[h264_vaapi @ 0x564efc28fac0] No usable encoding entrypoint found for profile VAProfileH264High (7).
failed to open encoder in low_power mode (Function not implemented), trying non low_power mode. if you have an intel iGPU, set enable_guc=2 in the i915 module to use the fixed function encoder. pass --low-power=off to suppress this warning
[h264_vaapi @ 0x564efc28fac0] Driver does not support some wanted packed headers (wanted 0xd, found 0).
[h264_vaapi @ 0x564efc28fac0] Driver does not support packed sequence headers, but a global header is requested.
[h264_vaapi @ 0x564efc28fac0] No global header will be written: this may result in a stream which is not usable for some purposes (e.g. not muxable to some containers).
[h264_vaapi @ 0x564efc28fac0] Failed to upload encode parameters: 2 (resource allocation failed).
[h264_vaapi @ 0x564efc28fac0] Encode failed: -5.
thread 'main' panicked at src/main.rs:1426:51:
called `Result::unwrap()` on an `Err` value: ffmpeg::Error(5: Input/output error)
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace
[I] root@neo ~ [101]
# vainfo
Trying display: wayland
vainfo: VA-API version: 1.20 (libva 2.20.0)
vainfo: Driver version: Mesa Gallium driver 23.2.1-arch1.2 for AMD Radeon VII (vega20, LLVM 16.0.6, DRM 3.49, 6.1.64-1-lts)
vainfo: Supported profile and entrypoints
VAProfileMPEG2Simple : VAEntrypointVLD
VAProfileMPEG2Main : VAEntrypointVLD
VAProfileVC1Simple : VAEntrypointVLD
VAProfileVC1Main : VAEntrypointVLD
VAProfileVC1Advanced : VAEntrypointVLD
VAProfileH264ConstrainedBaseline: VAEntrypointVLD
VAProfileH264ConstrainedBaseline: VAEntrypointEncSlice
VAProfileH264Main : VAEntrypointVLD
VAProfileH264Main : VAEntrypointEncSlice
VAProfileH264High : VAEntrypointVLD
VAProfileH264High : VAEntrypointEncSlice
VAProfileHEVCMain : VAEntrypointVLD
VAProfileHEVCMain : VAEntrypointEncSlice
VAProfileHEVCMain10 : VAEntrypointVLD
VAProfileNone : VAEntrypointVideoProc
[I] root@neo ~
# [I] root@neo ~ [244]
# ffmpeg -init_hw_device vaapi=hw:/dev/dri/renderD128 -filter_hw_device hw -f lavfi -i rgbtestsrc=duration=5:size=1280x720:rate=30 -vf hwupload,crop=100:100:200:200,scale_vaapi=format=nv12:w=100:h=100 -c:v h264_vaapi
testout.mp4
ffmpeg version 6.1-static https://johnvansickle.com/ffmpeg/ Copyright (c) 2000-2023 the FFmpeg developers
built with gcc 8 (Debian 8.3.0-6)
configuration: --enable-gpl --enable-version3 --enable-static --disable-debug --disable-ffplay --disable-indev=sndio --disable-outdev=sndio --cc=gcc --enable-fontconfig --enable-frei0r --enable-gnutls --enable-gmp --enable-libgme --enable-gray --enable-libaom --enable-libfribidi --enable-libass --enable-libvmaf --enable-libfreetype --enable-libmp3lame --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libopenjpeg --enable-librubberband --enable-libsoxr --enable-libspeex --enable-libsrt --enable-libvorbis --enable-libopus --enable-libtheora --enable-libvidstab --enable-libvo-amrwbenc --enable-libvpx --enable-libwebp --enable-libx264 --enable-libx265 --enable-libxml2 --enable-libdav1d --enable-libxvid --enable-libzvbi --enable-libzimg
libavutil 58. 29.100 / 58. 29.100
libavcodec 60. 31.102 / 60. 31.102
libavformat 60. 16.100 / 60. 16.100
libavdevice 60. 3.100 / 60. 3.100
libavfilter 9. 12.100 / 9. 12.100
libswscale 7. 5.100 / 7. 5.100
libswresample 4. 12.100 / 4. 12.100
libpostproc 57. 3.100 / 57. 3.100
Device creation failed: -12.
Failed to set value 'vaapi=hw:/dev/dri/renderD128' for option 'init_hw_device': Cannot allocate memory
Error parsing global options: Cannot allocate memory
[I] root@neo ~ [244]
# The error message looks strange because radeontop tells me there is plenty of memory.
Ristovski
commented
7 months ago Just to be safe, can you show the output of ffmpeg -hwaccels?
Should look something like this:
Hardware acceleration methods:
vdpau
vaapi
drm
vulkan(not sure if this has been merged into upstream ffmpeg, but the "Cannot allocate memory" message might be bogus as per https://patchwork.ffmpeg.org/project/ffmpeg/patch/484c6644-dfcd-1ad9-95fe-4bf2a06b30e7@jkqxz.net/)
hholst80
commented
7 months ago It only shows vdpau ... ?
Just to be clear, I have an [AMD/ATI] Vega 20 [Radeon VII] (rev c1) GPU. I have disabled the Intel integrated GPU such that the OS only sees the PEG device. vainfo seems to work fine and also wf-recorder seems to be using the GPU (both CPU and GPU utilization is low during recording.)
Ristovski
commented
7 months ago It only shows vdpau ... ?
@hholst80 That is weird, its as if your ffmpeg build does not have vaapi enabled for some reason. vdpau is only for hw accelerated decoding.
Can you check the output of ffmpeg -buildconf | grep -Ei "vdpau|vaapi"?
For me it shows both --enable-vaapi and --enable-vdpau (which correlates to my ffmpeg -hwaccels output)
hholst80
commented
7 months ago Classic "well, this is embarrassing" situation.
ffmpeg version 6.1-static https://johnvansickle.com/ffmpeg/ Copyright (c) 2000-2023 the FFmpeg developers
built with gcc 8 (Debian 8.3.0-6)[I] root@neo ~
# which ffmpeg
/home/root/.local/bin/ffmpeg
# mv /home/root/.local/bin/ffmpeg{,-static}
#Fixing that I get:
Hardware acceleration methods:
vdpau
cuda
vaapi
qsv
drm
opencl
vulkan# I to increase w and h to 300 because driver complined 100x100 was too small.
# ffmpeg -init_hw_device vaapi=hw:/dev/dri/renderD128 -filter_hw_device hw -f lavfi -i rgbtestsrc=duration=5:size=1280x720:rate=30 -vf hwupload,crop=100:100:200:200,scale_vaapi=format=nv12:w=300:h=300 -c:v h264_vaapi testout.mp4Could wl-screenrec be using the wrong ffmpeg? ldd $(which wl-screenrec)
hholst80
commented
7 months ago Looks like ffmpeg and wl-screenrec are using the same libs.
diff:
https://github.com/russelltg/wl-screenrec/assets/6200749/51677f50-8fda-48ad-b16b-2af32c63bbdb
hholst80
commented
7 months ago Perhaps this is just due to a more permissive error handling in wf-recorder vs. wl-screenrec?
# wf-recorder -f test-vaapi.mp4 -c h264_vaapi -d /dev/dri/renderD128
Output file "test-vaapi.mp4" exists. Overwrite? Y/n: y
using VA-API, trying to enable DMA-BUF capture...
enabled DMA-BUF capture, device /dev/dri/renderD128
selected region 0,0 0x0
Using video filter: scale_vaapi=format=nv12:out_range=full
[h264_vaapi @ 0x7f38c0002300] No quality level set; using default (20).
[h264_vaapi @ 0x7f38c0002300] Driver does not support some wanted packed headers (wanted 0xd, found 0).
[h264_vaapi @ 0x7f38c0002300] Driver does not support packed sequence headers, but a global header is requested.
[h264_vaapi @ 0x7f38c0002300] No global header will be written: this may result in a stream which is not usable for some purposes (e.g. not muxable to some containers).
Output #0, mp4, to 'test-vaapi.mp4':
Stream #0:0: Video: h264 (High), vaapi(pc), 3840x2160 [SAR 1:1 DAR 16:9], q=2-31
^C⏎ [I] root@neo ~
# + wl-screenrec --low-power=off --encode-resolution 1920x1080 --audio
Using output DP-1
Opening libva device from DRM device /dev/dri/renderD128
[h264_vaapi @ 0x55f9bf3ef4c0] Driver does not support some wanted packed headers (wanted 0xd, found 0).
[h264_vaapi @ 0x55f9bf3ef4c0] Driver does not support packed sequence headers, but a global header is requested.
[h264_vaapi @ 0x55f9bf3ef4c0] No global header will be written: this may result in a stream which is not usable for some purposes (e.g. not muxable to some containers).
0 fps
[h264_vaapi @ 0x55f9bf3ef4c0] Failed to upload encode parameters: 2 (resource allocation failed).
[h264_vaapi @ 0x55f9bf3ef4c0] Encode failed: -5.
thread 'main' panicked at src/main.rs:1426:51:
called `Result::unwrap()` on an `Err` value: ffmpeg::Error(5: Input/output error)
stack backtrace:
0: 0x55f9bdc94ee1 - std::backtrace_rs::backtrace::libunwind::trace::h938efb7ac3b4b9bb
at /builddir/build/BUILD/rustc-1.74.0-src/library/std/src/../../backtrace/src/backtrace/libunwind.rs:93:5
1: 0x55f9bdc94ee1 - std::backtrace_rs::backtrace::trace_unsynchronized::hdb3130afb00196b2
at /builddir/build/BUILD/rustc-1.74.0-src/library/std/src/../../backtrace/src/backtrace/mod.rs:66:5
2: 0x55f9bdc94ee1 - std::sys_common::backtrace::_print_fmt::h99aab20d4d0542ff
at /builddir/build/BUILD/rustc-1.74.0-src/library/std/src/sys_common/backtrace.rs:67:5
3: 0x55f9bdc94ee1 - <std::sys_common::backtrace::_print::DisplayBacktrace as core::fmt::Display>::fmt::h91f0c6dac97584ac
at /builddir/build/BUILD/rustc-1.74.0-src/library/std/src/sys_common/backtrace.rs:44:22
4: 0x55f9bdc5be4c - core::fmt::rt::Argument::fmt::hccd1db59d85bec3c
at /builddir/build/BUILD/rustc-1.74.0-src/library/core/src/fmt/rt.rs:138:9
5: 0x55f9bdc5be4c - core::fmt::write::hadb576d8cc22952d
at /builddir/build/BUILD/rustc-1.74.0-src/library/core/src/fmt/mod.rs:1114:21
6: 0x55f9bdc92b6d - std::io::Write::write_fmt::hd3784cb8817c312c
at /builddir/build/BUILD/rustc-1.74.0-src/library/std/src/io/mod.rs:1763:15
7: 0x55f9bdc94cc5 - std::sys_common::backtrace::_print::h330d475718a82b03
at /builddir/build/BUILD/rustc-1.74.0-src/library/std/src/sys_common/backtrace.rs:47:5
8: 0x55f9bdc94cc5 - std::sys_common::backtrace::print::hfd68e0b6fc4a7be7
at /builddir/build/BUILD/rustc-1.74.0-src/library/std/src/sys_common/backtrace.rs:34:9
9: 0x55f9bdc961b3 - std::panicking::default_hook::{{closure}}::h50dd6a5c8f27eca4
at /builddir/build/BUILD/rustc-1.74.0-src/library/std/src/panicking.rs:272:22
10: 0x55f9bdc95f02 - std::panicking::default_hook::hfcea74576cf59f61
at /builddir/build/BUILD/rustc-1.74.0-src/library/std/src/panicking.rs:292:9
11: 0x55f9bdc9669b - std::panicking::rust_panic_with_hook::hd001ca82f62340ad
at /builddir/build/BUILD/rustc-1.74.0-src/library/std/src/panicking.rs:731:13
12: 0x55f9bdc9659e - std::panicking::begin_panic_handler::{{closure}}::h815b59c3e229799b
at /builddir/build/BUILD/rustc-1.74.0-src/library/std/src/panicking.rs:609:13
13: 0x55f9bdc953c6 - std::sys_common::backtrace::__rust_end_short_backtrace::h4a443d64544d51c7
at /builddir/build/BUILD/rustc-1.74.0-src/library/std/src/sys_common/backtrace.rs:170:18
14: 0x55f9bdc96342 - rust_begin_unwind
at /builddir/build/BUILD/rustc-1.74.0-src/library/std/src/panicking.rs:597:5
15: 0x55f9bdbff855 - core::panicking::panic_fmt::he1d13495b4d13340
at /builddir/build/BUILD/rustc-1.74.0-src/library/core/src/panicking.rs:72:14
16: 0x55f9bdbffcf3 - core::result::unwrap_failed::h416d273107a0926d
at /builddir/build/BUILD/rustc-1.74.0-src/library/core/src/result.rs:1652:5
17: 0x55f9bdcd6aa0 - wl_screenrec::EncState::process_ready::hb8c0be31b72cbe8f
18: 0x55f9bdccda9a - <wl_screenrec::State as wayland_client::event_queue::Dispatch<wayland_protocols_wlr::screencopy::v1::generated::client::zwlr_screencopy_frame_v1::ZwlrScreencopyFrameV1,()>>::event::h8d49ba548dfd364f
19: 0x55f9bdce6b91 - wayland_client::event_queue::queue_callback::h12f54bcc24edc7dc
20: 0x55f9bdce8a95 - wayland_client::event_queue::EventQueue<State>::dispatching_impl::h0a6f117edb30a3be
21: 0x55f9bdce8dcf - wayland_client::event_queue::EventQueue<State>::blocking_dispatch::h1514ecc02112bec1
22: 0x55f9bdcd94be - wl_screenrec::main::h5606d1205275c550
23: 0x55f9bdd07753 - std::sys_common::backtrace::__rust_begin_short_backtrace::hd914cf76d5a3b902
24: 0x55f9bdd00639 - std::rt::lang_start::{{closure}}::hb732c73c09ab0374
25: 0x55f9bdc8e55f - core::ops::function::impls::<impl core::ops::function::FnOnce<A> for &F>::call_once::hcab3a09e04d63fe7
at /builddir/build/BUILD/rustc-1.74.0-src/library/core/src/ops/function.rs:284:13
26: 0x55f9bdc8e55f - std::panicking::try::do_call::h8427747ad0e09c1b
at /builddir/build/BUILD/rustc-1.74.0-src/library/std/src/panicking.rs:504:40
27: 0x55f9bdc8e55f - std::panicking::try::h82c1b474e0405a64
at /builddir/build/BUILD/rustc-1.74.0-src/library/std/src/panicking.rs:468:19
28: 0x55f9bdc8e55f - std::panic::catch_unwind::h91cc7393872dbf91
at /builddir/build/BUILD/rustc-1.74.0-src/library/std/src/panic.rs:142:14
29: 0x55f9bdc8e55f - std::rt::lang_start_internal::{{closure}}::h39807600b3c8b3ee
at /builddir/build/BUILD/rustc-1.74.0-src/library/std/src/rt.rs:148:48
30: 0x55f9bdc8e55f - std::panicking::try::do_call::hf61bc514e36736df
at /builddir/build/BUILD/rustc-1.74.0-src/library/std/src/panicking.rs:504:40
31: 0x55f9bdc8e55f - std::panicking::try::hb982ff816c4ba4b3
at /builddir/build/BUILD/rustc-1.74.0-src/library/std/src/panicking.rs:468:19
32: 0x55f9bdc8e55f - std::panic::catch_unwind::h6298355f030b7719
at /builddir/build/BUILD/rustc-1.74.0-src/library/std/src/panic.rs:142:14
33: 0x55f9bdc8e55f - std::rt::lang_start_internal::h3650e9192e2d4912
at /builddir/build/BUILD/rustc-1.74.0-src/library/std/src/rt.rs:148:20
34: 0x55f9bdcdfd45 - main
35: 0x7fcd21445cd0 - <unknown>
36: 0x7fcd21445d8a - __libc_start_main
37: 0x55f9bdc04625 - _start
38: 0x0 - <unknown>
[I] root@neo ~
# I don't think so. You're welcome to test this patch against latest main:
diff --git a/src/main.rs b/src/main.rs
index fd62896..a2790ad 100644
--- a/src/main.rs
+++ b/src/main.rs
@@ -1423,7 +1423,9 @@ impl EncState {
.is_ok()
{
// encoder has same time base as the filter, so don't do any time scaling
- self.enc_video.send_frame(&yuv_frame).unwrap();
+ if let Err(e) = self.enc_video.send_frame(&yuv_frame) {
+ eprintln!("Failed to encode: {e}")
+ }
}
let mut encoded = Packet::empty();I'll probably push this patch anyways as it's less nasty than a panic, and I suppose maybe sometimes that error isn't fatal. It's hard to say.
I would be interested in full strace outputs from those two commands to see exactly what is failing.
hholst80
commented
4 months ago on git sha 6110ba66495c65b913cf50afb595eab7b992cb34 I get this.
Seems to be the same error message as I got on the "new machine" (called "goblin") in #64
This is on the same hardware as before on my "neo" machine.
[I] root@neo ~
# wl-screenrec --low-power=off --encode-resolution 1920x1080 --audio
Using output DP-1
Opening libva device from DRM device /dev/dri/renderD128
[Parsed_crop_0 @ 0x57ec399cf580] Invalid too big or non positive size for width '3840' or height '2160'
[Parsed_crop_0 @ 0x57ec399cf580] Failed to configure input pad on Parsed_crop_0
thread 'main' panicked at src/main.rs:1706:18:
called `Result::unwrap()` on an `Err` value: ffmpeg::Error(22: Invalid argument)
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace
[I] root@neo ~ [101]
#[I] root@neo ~
# vainfo
Trying display: wayland
vainfo: VA-API version: 1.20 (libva 2.20.1)
vainfo: Driver version: Mesa Gallium driver 24.0.1-arch1.1 for AMD Radeon VII (radeonsi, vega20, LLVM 16.0.6, DRM 3.57, 6.7.5-zen1-1-zen)
vainfo: Supported profile and entrypoints
VAProfileMPEG2Simple : VAEntrypointVLD
VAProfileMPEG2Main : VAEntrypointVLD
VAProfileVC1Simple : VAEntrypointVLD
VAProfileVC1Main : VAEntrypointVLD
VAProfileVC1Advanced : VAEntrypointVLD
VAProfileH264ConstrainedBaseline: VAEntrypointVLD
VAProfileH264ConstrainedBaseline: VAEntrypointEncSlice
VAProfileH264Main : VAEntrypointVLD
VAProfileH264Main : VAEntrypointEncSlice
VAProfileH264High : VAEntrypointVLD
VAProfileH264High : VAEntrypointEncSlice
VAProfileHEVCMain : VAEntrypointVLD
VAProfileHEVCMain : VAEntrypointEncSlice
VAProfileHEVCMain10 : VAEntrypointVLD
VAProfileNone : VAEntrypointVideoProc
[I] root@neo ~
#Aha! This might be related to non-integer scaling (I am exploring the best way to do HiDPI in a portable way between my many machines.).
If I use an integer scaling, like 2 instead of 2.4 like I am using now on both neo and goblin, I get this more familiar error message:
[I] root@neo ~/Apps/wl-screenrec (main)
# wl-screenrec --low-power=off --encode-resolution 1920x1080 --audio
Opening libva device from DRM device /dev/dri/renderD128
[h264_vaapi @ 0x588bbf9c5cc0] Driver does not support some wanted packed headers (wanted 0xd, found 0x1).
[h264_vaapi @ 0x588bbf9c5cc0] Failed to upload encode parameters: 2 (resource allocation failed).
[h264_vaapi @ 0x588bbf9c5cc0] Encode failed: -5.
thread 'main' panicked at src/main.rs:1505:51:
called `Result::unwrap()` on an `Err` value: ffmpeg::Error(5: Input/output error)
note: run with `RUST_BACKTRACE=1` environment variable to display a backtrace
[I] root@neo ~/Apps/wl-screenrec (main) [101]
#Another codec will work, it records a video, but with some yellow warnings:
mp
russelltg
commented
4 months ago Does
ffmpeg -vaapi_device /dev/dri/renderD128 -hwaccel vaapi -f lavfi -i testsrc=duration=10:size=1280x720 -filter_complex 'hwupload,scale_vaapi=format=nv12' -c:v h264_vaapi -loglevel 80 test.mp4work and create a valid video?
SArpnt
commented
1 month ago i've recorded videos with this program before, but i started having this exact issue some time ago and i'm not sure what caused it.
sarpnt@mime-cube
----------------
OS: Arch Linux x86_64
Kernel: 6.9.2-zen1-1-zen
Uptime: 1 hour, 10 mins
Packages: 1298 (pacman)
Shell: fish 3.7.1
Resolution: 1920x1080
WM: sway
Theme: oomox-darker-than-fusion [GTK2/3]
Icons: breeze-dark [GTK2/3]
Terminal: foot
CPU: Intel i5-4690K (4) @ 3.900GHz
GPU: Intel HD Graphics
GPU: AMD ATI Radeon RX 470/480/570/570X/580/580X/590
Memory: 3708MiB / 11828MiB sarpnt@mime-cube ~> wl-screenrec
Opening libva device from DRM device /dev/dri/renderD129
[h264_vaapi @ 0x5604f2686d40] No usable encoding entrypoint found for profile VAProfileH264High (7).
failed to open encoder in low_power mode (Function not implemented), trying non low_power mode. if you have an intel iGPU, set enable_guc=2 in the i915 module to use the fixed function encoder. pass --low-power=off to suppress this warning
[h264_vaapi @ 0x5604f2686d40] Driver does not support some wanted packed headers (wanted 0xd, found 0x1).
[h264_vaapi @ 0x5604f2686d40] Failed to upload encode parameters: 2 (resource allocation failed).
[h264_vaapi @ 0x5604f2686d40] Encode failed: -5.
thread 'main' panicked at src/main.rs:1564:51:
called `Result::unwrap()` on an `Err` value: ffmpeg::Error(5: Input/output error)
note: run with `RUST_BACKTRACE=1` environment variable to display a backtracesarpnt@mime-cube ~ [101]> wl-screenrec --no-hw
Opening libva device from DRM device /dev/dri/renderD129
[Parsed_scale_vaapi_1 @ 0x634217b0d580] Hardware does not support output format yuv420p.
[Parsed_scale_vaapi_1 @ 0x634217b0d580] Failed to configure output pad on Parsed_scale_vaapi_1
thread 'main' panicked at src/main.rs:1782:18:
called `Result::unwrap()` on an `Err` value: ffmpeg::Error(22: Invalid argument)
note: run with `RUST_BACKTRACE=1` environment variable to display a backtracevainfo shows the amd card should work and ffmpeg with vaapi encoding works fine
Not sure what is going on here. wl-record says vaapi does not support yuv420 but vainfo says it does?
wl-record v0.1.1 compiled via cargo + git.