Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > comp.os.linux.advocacy > #452872 > unrolled thread
| Started by | "F. Russell" <fr@random.info> |
|---|---|
| First post | 2018-05-31 13:10 +0000 |
| Last post | 2018-06-01 05:25 -0700 |
| Articles | 20 on this page of 37 — 8 participants |
Back to article view | Back to comp.os.linux.advocacy
Need Help -- Who's Got Nvidia? "F. Russell" <fr@random.info> - 2018-05-31 13:10 +0000
Re: Need Help -- Who's Got Nvidia? Melzzzzz <Melzzzzz@zzzzz.com> - 2018-05-31 16:37 +0000
Re: Need Help -- Who's Got Nvidia? "F. Russell" <fr@random.info> - 2018-05-31 20:00 +0000
Re: Need Help -- Who's Got Nvidia? Melzzzzz <Melzzzzz@zzzzz.com> - 2018-05-31 21:38 +0000
Re: Need Help -- Who's Got Nvidia? owl <owl@rooftop.invalid> - 2018-05-31 20:20 +0000
Re: Need Help -- Who's Got Nvidia? "F. Russell" <fr@random.info> - 2018-05-31 20:57 +0000
Re: Need Help -- Who's Got Nvidia? owl <owl@rooftop.invalid> - 2018-05-31 22:32 +0000
Re: Need Help -- Who's Got Nvidia? "F. Russell" <fr@random.info> - 2018-05-31 22:36 +0000
Re: Need Help -- Who's Got Nvidia? Melzzzzz <Melzzzzz@zzzzz.com> - 2018-05-31 22:40 +0000
Re: Need Help -- Who's Got Nvidia? Snit took things out of COLA and called the cops <frelwizzen@gmail.com> - 2018-05-31 16:10 -0700
Re: Need Help -- Who's Got Nvidia? Melzzzzz <Melzzzzz@zzzzz.com> - 2018-05-31 22:41 +0000
Re: Need Help -- Who's Got Nvidia? "F. Russell" <fr@random.info> - 2018-05-31 22:59 +0000
Re: Need Help -- Who's Got Nvidia? Melzzzzz <Melzzzzz@zzzzz.com> - 2018-05-31 23:39 +0000
Re: Need Help -- Who's Got Nvidia? owl <owl@rooftop.invalid> - 2018-06-01 00:00 +0000
Re: Need Help -- Who's Got Nvidia? [Solved -- Partially] "F. Russell" <fr@random.info> - 2018-06-01 00:21 +0000
Re: Need Help -- Who's Got Nvidia? Snit took things out of COLA and called the cops <frelwizzen@gmail.com> - 2018-05-31 16:25 -0700
Re: Need Help -- Who's Got Nvidia? vallor <vallor@cultnix.org> - 2018-05-31 21:09 +0000
Re: Need Help -- Who's Got Nvidia? "F. Russell" <fr@random.info> - 2018-05-31 21:39 +0000
Re: Need Help -- Who's Got Nvidia? "F. Russell" <fr@random.info> - 2018-05-31 21:46 +0000
Re: Need Help -- Who's Got Nvidia? Steve Carroll <fretwizzer@gmail.com> - 2018-05-31 15:02 -0700
Re: Need Help -- Who's Got Nvidia? "F. Russell" <fr@random.info> - 2018-05-31 22:39 +0000
Re: Need Help -- Who's Got Nvidia? Steve Carroll <fretwizzer@gmail.com> - 2018-05-31 16:10 -0700
Re: Need Help -- Who's Got Nvidia? Snit <usenet@gallopinginsanity.com> - 2018-05-31 16:59 -0700
Re: Need Help -- Who's Got Nvidia? vallor <vallor@cultnix.org> - 2018-05-31 23:11 +0000
Re: Need Help -- Who's Got Nvidia? "F. Russell" <fr@random.info> - 2018-05-31 23:27 +0000
Re: Need Help -- Who's Got Nvidia? Snit <usenet@gallopinginsanity.com> - 2018-05-31 17:02 -0700
Re: Need Help -- Who's Got Nvidia? Snit <usenet@gallopinginsanity.com> - 2018-05-31 17:16 -0700
Re: Need Help -- Who's Got Nvidia? Steve Carroll <fretwizzer@gmail.com> - 2018-05-31 17:09 -0700
Re: Need Help -- Who's Got Nvidia? Snit <usenet@gallopinginsanity.com> - 2018-05-31 16:54 -0700
Re: Need Help -- Who's Got Nvidia? "F. Russell" <fr@random.info> - 2018-06-01 00:40 +0000
Re: Need Help -- Who's Got Nvidia? Steve Carroll <fretwizzer@gmail.com> - 2018-05-31 17:42 -0700
Re: Need Help -- Who's Got Nvidia? "F. Russell" <fr@random.info> - 2018-06-01 00:48 +0000
Re: Need Help -- Who's Got Nvidia? Snit calls the cops when he steps on his dick <frelwizzen@gmail.com> - 2018-06-01 02:16 -0700
Re: Need Help -- Who's Got Nvidia? owl <owl@rooftop.invalid> - 2018-06-01 00:22 +0000
Re: Need Help -- Who's Got Nvidia? "F. Russell" <fr@random.info> - 2018-06-01 08:14 +0000
Re: Need Help -- Who's Got Nvidia? Snit calls the cops when he steps on his dick <frelwizzen@gmail.com> - 2018-06-01 01:45 -0700
Re: Need Help -- Who's Got Nvidia? Snit calls the cops when he steps on his dick <frelwizzen@gmail.com> - 2018-06-01 05:25 -0700
Page 1 of 2 [1] 2 Next page →
| From | "F. Russell" <fr@random.info> |
|---|---|
| Date | 2018-05-31 13:10 +0000 |
| Subject | Need Help -- Who's Got Nvidia? |
| Message-ID | <peos8t01td@news2.newsguy.com> |
On my system, ffmpeg fails with Nvidia nvenc: [h264_nvenc @ 0xf0cf00] Cannot init CUDA Is this ffmpeg? Is this nvidia driver? If you have an nvenc capable card try to transcode or record with ffmpeg nvenc and report the results. One of many possible commands would be: ffmpeg -i video.mp4 -vcodec h264_nvenc video.avi To check if nvenc is supported use: ffmpeg -encoders | grep nvenc
[toc] | [next] | [standalone]
| From | Melzzzzz <Melzzzzz@zzzzz.com> |
|---|---|
| Date | 2018-05-31 16:37 +0000 |
| Message-ID | <hxVPC.369900$RH3.260396@fx27.am4> |
| In reply to | #452872 |
On 2018-05-31, F. Russell <fr@random.info> wrote: > On my system, ffmpeg fails with Nvidia nvenc: > > [h264_nvenc @ 0xf0cf00] Cannot init CUDA > > Is this ffmpeg? Is this nvidia driver? > > If you have an nvenc capable card try to transcode or record > with ffmpeg nvenc and report the results. One of many possible > commands would be: > > ffmpeg -i video.mp4 -vcodec h264_nvenc video.avi It works. -- press any key to continue or any other to quit...
[toc] | [prev] | [next] | [standalone]
| From | "F. Russell" <fr@random.info> |
|---|---|
| Date | 2018-05-31 20:00 +0000 |
| Message-ID | <pepk8r0bg9@news2.newsguy.com> |
| In reply to | #452873 |
On Thu, 31 May 2018 16:37:33 +0000, Melzzzzz wrote: > > It works. > What is the output of the following commands: lspci | grep -i nvidia and ffmpeg -encoders | grep nvenc
[toc] | [prev] | [next] | [standalone]
| From | Melzzzzz <Melzzzzz@zzzzz.com> |
|---|---|
| Date | 2018-05-31 21:38 +0000 |
| Message-ID | <eXZPC.225694$%L2.198435@fx40.am4> |
| In reply to | #452899 |
On 2018-05-31, F. Russell <fr@random.info> wrote: > On Thu, 31 May 2018 16:37:33 +0000, Melzzzzz wrote: > >> >> It works. >> > > What is the output of the following commands: > > lspci | grep -i nvidia 01:00.0 VGA compatible controller: NVIDIA Corporation GM107 [GeForce GTX 750 Ti] (rev a2) 01:00.1 Audio device: NVIDIA Corporation Device 0fbc (rev a1) > > and > > ffmpeg -encoders | grep nvenc > ffmpeg version 4.0 Copyright (c) 2000-2018 the FFmpeg developers built with gcc 8.1.0 (GCC) configuration: --prefix=/usr --disable-debug --disable-static --disable-stripping --enable-avresample --enable-fontconfig --enable-gmp --enable-gnutls --enable-gpl --enable-ladspa --enable-libass --enable-libbluray --enable-libdrm --enable-libfreetype --enable-libfribidi --enable-libgsm --enable-libiec61883 --enable-libmodplug --enable-libmp3lame --enable-libopencore_amrnb --enable-libopencore_amrwb --enable-libopenjpeg --enable-libopus --enable-libpulse --enable-libsoxr --enable-libspeex --enable-libssh --enable-libtheora --enable-libv4l2 --enable-libvidstab --enable-libvorbis --enable-libvpx --enable-libwebp --enable-libx264 --enable-libx265 --enable-libxcb --enable-libxml2 --enable-libxvid --enable-nvenc --enable-omx --enable-shared --enable-version3 libavutil 56. 14.100 / 56. 14.100 libavcodec 58. 18.100 / 58. 18.100 libavformat 58. 12.100 / 58. 12.100 libavdevice 58. 3.100 / 58. 3.100 libavfilter 7. 16.100 / 7. 16.100 libavresample 4. 0. 0 / 4. 0. 0 libswscale 5. 1.100 / 5. 1.100 libswresample 3. 1.100 / 3. 1.100 libpostproc 55. 1.100 / 55. 1.100 V..... h264_nvenc NVIDIA NVENC H.264 encoder (codec h264) V..... nvenc NVIDIA NVENC H.264 encoder (codec h264) V..... nvenc_h264 NVIDIA NVENC H.264 encoder (codec h264) V..... nvenc_hevc NVIDIA NVENC hevc encoder (codec hevc) V..... hevc_nvenc NVIDIA NVENC hevc encoder (codec hevc) -- press any key to continue or any other to quit...
[toc] | [prev] | [next] | [standalone]
| From | owl <owl@rooftop.invalid> |
|---|---|
| Date | 2018-05-31 20:20 +0000 |
| Message-ID | <9a98b03a.go3o@rooftop.invalid> |
| In reply to | #452872 |
F. Russell <fr@random.info> wrote: > On my system, ffmpeg fails with Nvidia nvenc: > > [h264_nvenc @ 0xf0cf00] Cannot init CUDA > > Is this ffmpeg? The message is from ffmpeg's nvenc.c
[toc] | [prev] | [next] | [standalone]
| From | "F. Russell" <fr@random.info> |
|---|---|
| Date | 2018-05-31 20:57 +0000 |
| Message-ID | <pepnk00qvg@news1.newsguy.com> |
| In reply to | #452904 |
On Thu, 31 May 2018 20:20:31 +0000, owl wrote:
>
> The message is from ffmpeg's nvenc.c
>
Yes, the message is from ffmpeg, It is prefaced by
the ffmpeg encoder h264_nvenc.
But I suppose that I should have asked if it is ffmpeg
or the nvidia driver that is ultimately responsible.
The use of the strace utility shows the following:
stat("/dev/nvidia-uvm", {st_mode=S_IFCHR|0666, st_rdev=makedev(249, 0), ...}) = 0
stat("/dev/nvidia-uvm-tools", {st_mode=S_IFCHR|0666, st_rdev=makedev(249, 1), ...}) = 0
openat(AT_FDCWD, "/dev/nvidia-uvm", O_RDWR|O_CLOEXEC) = -1 EIO (Input/output error)
openat(AT_FDCWD, "/dev/nvidia-uvm", O_RDWR) = -1 EIO (Input/output error)
ioctl(-5, _IOC(0, 0x00, 0x01, 0x3000), 0x7ffcaa636160) = -1 EBADF (Bad file descriptor)
ioctl(-5, _IOC(0, 0x00, 0x02, 0x3000), 0) = -1 EBADF (Bad file descriptor)
close(-5) = -1 EBADF (Bad file descriptor)
write(2, "\33[48;5;0m\33[38;5;51m[h264_nvenc @"..., 48[h264_nvenc @ 0x2277d00] ) = 48
write(2, "\33[48;5;0m\33[38;5;196mCannot init "..., 41Cannot init CUDA
Something (ffmpeg or nvidia?) tris to open /dev/nvidia-uvm and gets
an error and then the program terminates.
Strace only shows system calls so I can't tell if ffmpeg or
nvidia makes the system calls.
Ffmpeg may just get an error code from nvidia.
[toc] | [prev] | [next] | [standalone]
| From | owl <owl@rooftop.invalid> |
|---|---|
| Date | 2018-05-31 22:32 +0000 |
| Message-ID | <azb9083tga.hou@rooftop.invalid> |
| In reply to | #452906 |
F. Russell <fr@random.info> wrote:
> On Thu, 31 May 2018 20:20:31 +0000, owl wrote:
>
>>
>> The message is from ffmpeg's nvenc.c
>>
>
> Yes, the message is from ffmpeg, It is prefaced by
> the ffmpeg encoder h264_nvenc.
>
> But I suppose that I should have asked if it is ffmpeg
> or the nvidia driver that is ultimately responsible.
>
> The use of the strace utility shows the following:
>
> stat("/dev/nvidia-uvm", {st_mode=S_IFCHR|0666, st_rdev=makedev(249, 0), ...}) = 0
> stat("/dev/nvidia-uvm-tools", {st_mode=S_IFCHR|0666, st_rdev=makedev(249, 1), ...}) = 0
> openat(AT_FDCWD, "/dev/nvidia-uvm", O_RDWR|O_CLOEXEC) = -1 EIO (Input/output error)
> openat(AT_FDCWD, "/dev/nvidia-uvm", O_RDWR) = -1 EIO (Input/output error)
> ioctl(-5, _IOC(0, 0x00, 0x01, 0x3000), 0x7ffcaa636160) = -1 EBADF (Bad file descriptor)
> ioctl(-5, _IOC(0, 0x00, 0x02, 0x3000), 0) = -1 EBADF (Bad file descriptor)
> close(-5) = -1 EBADF (Bad file descriptor)
> write(2, "\33[48;5;0m\33[38;5;51m[h264_nvenc @"..., 48[h264_nvenc @ 0x2277d00] ) = 48
> write(2, "\33[48;5;0m\33[38;5;196mCannot init "..., 41Cannot init CUDA
>
> Something (ffmpeg or nvidia?) tris to open /dev/nvidia-uvm and gets
> an error and then the program terminates.
>
> Strace only shows system calls so I can't tell if ffmpeg or
> nvidia makes the system calls.
>
No mention of 'nvidia-uvm' in ffmpeg source.
> Ffmpeg may just get an error code from nvidia.
>
>
ffmpeg generates the error message because:
(dl_fn->cu_init(0)) != CUDA_SUCCESS
[toc] | [prev] | [next] | [standalone]
| From | "F. Russell" <fr@random.info> |
|---|---|
| Date | 2018-05-31 22:36 +0000 |
| Message-ID | <pepte40g0k@news2.newsguy.com> |
| In reply to | #452925 |
On Thu, 31 May 2018 22:32:38 +0000, owl wrote: > > ffmpeg generates the error message because: > (dl_fn->cu_init(0)) != CUDA_SUCCESS > It's that proprietary nvidia driver written by highly overpaid incompetent programmers who can't even adequately support their own hardware.
[toc] | [prev] | [next] | [standalone]
| From | Melzzzzz <Melzzzzz@zzzzz.com> |
|---|---|
| Date | 2018-05-31 22:40 +0000 |
| Message-ID | <qR_PC.225696$%L2.125471@fx40.am4> |
| In reply to | #452926 |
On 2018-05-31, F. Russell <fr@random.info> wrote: > On Thu, 31 May 2018 22:32:38 +0000, owl wrote: > >> >> ffmpeg generates the error message because: >> (dl_fn->cu_init(0)) != CUDA_SUCCESS >> > > It's that proprietary nvidia driver written by highly overpaid > incompetent programmers who can't even adequately support their > own hardware. > What version? I have latest beta. 396.24 I think. > -- press any key to continue or any other to quit...
[toc] | [prev] | [next] | [standalone]
| From | Snit took things out of COLA and called the cops <frelwizzen@gmail.com> |
|---|---|
| Date | 2018-05-31 16:10 -0700 |
| Message-ID | <d80be82f-8bc1-4f3e-b21b-5d02cd202d59@googlegroups.com> |
| In reply to | #452927 |
On Thursday, May 31, 2018 at 3:40:30 PM UTC-7, Melzzzzz wrote: > On 2018-05-31, F. Russell <fr@random.info> wrote: > > On Thu, 31 May 2018 22:32:38 +0000, owl wrote: > > > >> > >> ffmpeg generates the error message because: > >> (dl_fn->cu_init(0)) != CUDA_SUCCESS > >> > > > > It's that proprietary nvidia driver written by highly overpaid > > incompetent programmers who can't even adequately support their > > own hardware. > > > > What version? I have latest beta. 396.24 I think. > > > > > -- > press any key to continue or any other to quit... Times change. Maybe the idea of a usenet group is getting a little dated. Obviously, the sole concern that is important to Peter Kohlmann is thinking he is "the winner", and if he can't have that he will post more evidence to beat "others" down, which only serves to destroy discussion. Any horribly moronic insomniac could easily do the same. Just look at what Peter Kohlmann has posted and look at RonB's, there is nothing for me to learn from a jerk like Peter Kohlmann. But knock yourself out, let him keep making a blockhead of himself. It is not like anyone believes him. Peter Kohlmann must recognize that others can go get whatever software he has found, right, medhead? On top of that, anyone can can set a KF, which renders his usenet group attack harmless, just like Peter Kohlmann. So what is Peter Kohlmann's method for the flooding insanity? mySQL? That is the only language he knows, or admits to. He must be abusing it to produce these goofy "attack" idiotic threads. Here is my idea, Peter Kohlmann took RonB's code that he hacked and blamed RonB for... he is feeding it Google Groups data, grabbing random paragraphs, then mangling those using Markova chains and then he actively posts them because his insanity enables him to do that nonstop. Peter Kohlmann is being controlled by Apple. I have known bricks who argue better than Peter Kohlmann does. -- I Left My Husband & Daughter At Home And THIS happened! http://www.5z8.info/--INITIATE-CREDIT-CARD-XFER--_k7s1jt_inject_worm https://youtu.be/D_so1dvjeyI Jonas Eklundh Communication AB
[toc] | [prev] | [next] | [standalone]
| From | Melzzzzz <Melzzzzz@zzzzz.com> |
|---|---|
| Date | 2018-05-31 22:41 +0000 |
| Message-ID | <dS_PC.225697$%L2.61668@fx40.am4> |
| In reply to | #452925 |
On 2018-05-31, owl <owl@rooftop.invalid> wrote:
> F. Russell <fr@random.info> wrote:
>> On Thu, 31 May 2018 20:20:31 +0000, owl wrote:
>>
>>>
>>> The message is from ffmpeg's nvenc.c
>>>
>>
>> Yes, the message is from ffmpeg, It is prefaced by
>> the ffmpeg encoder h264_nvenc.
>>
>> But I suppose that I should have asked if it is ffmpeg
>> or the nvidia driver that is ultimately responsible.
>>
>> The use of the strace utility shows the following:
>>
>> stat("/dev/nvidia-uvm", {st_mode=S_IFCHR|0666, st_rdev=makedev(249, 0), ...}) = 0
>> stat("/dev/nvidia-uvm-tools", {st_mode=S_IFCHR|0666, st_rdev=makedev(249, 1), ...}) = 0
>> openat(AT_FDCWD, "/dev/nvidia-uvm", O_RDWR|O_CLOEXEC) = -1 EIO (Input/output error)
>> openat(AT_FDCWD, "/dev/nvidia-uvm", O_RDWR) = -1 EIO (Input/output error)
>> ioctl(-5, _IOC(0, 0x00, 0x01, 0x3000), 0x7ffcaa636160) = -1 EBADF (Bad file descriptor)
>> ioctl(-5, _IOC(0, 0x00, 0x02, 0x3000), 0) = -1 EBADF (Bad file descriptor)
>> close(-5) = -1 EBADF (Bad file descriptor)
>> write(2, "\33[48;5;0m\33[38;5;51m[h264_nvenc @"..., 48[h264_nvenc @ 0x2277d00] ) = 48
>> write(2, "\33[48;5;0m\33[38;5;196mCannot init "..., 41Cannot init CUDA
>>
>> Something (ffmpeg or nvidia?) tris to open /dev/nvidia-uvm and gets
>> an error and then the program terminates.
>>
>> Strace only shows system calls so I can't tell if ffmpeg or
>> nvidia makes the system calls.
>>
>
> No mention of 'nvidia-uvm' in ffmpeg source.
>
>
>> Ffmpeg may just get an error code from nvidia.
>>
>>
>
> ffmpeg generates the error message because:
> (dl_fn->cu_init(0)) != CUDA_SUCCESS
cu_init is from libcuda?
>
>
--
press any key to continue or any other to quit...
[toc] | [prev] | [next] | [standalone]
| From | "F. Russell" <fr@random.info> |
|---|---|
| Date | 2018-05-31 22:59 +0000 |
| Message-ID | <pepunq0uis@news1.newsguy.com> |
| In reply to | #452929 |
On Thu, 31 May 2018 22:41:13 +0000, Melzzzzz wrote: > > cu_init is from libcuda? > It's a struct within ffmpeg that contains pointers (and other stuff) to functions of the Nvidia API. dl_fn means "Dynamic Link Functions"
[toc] | [prev] | [next] | [standalone]
| From | Melzzzzz <Melzzzzz@zzzzz.com> |
|---|---|
| Date | 2018-05-31 23:39 +0000 |
| Message-ID | <3J%PC.235884$7B3.120175@fx13.am4> |
| In reply to | #452931 |
On 2018-05-31, F. Russell <fr@random.info> wrote: > On Thu, 31 May 2018 22:41:13 +0000, Melzzzzz wrote: > >> >> cu_init is from libcuda? >> > > It's a struct within ffmpeg that contains pointers > (and other stuff) to functions of the Nvidia API. > > dl_fn means "Dynamic Link Functions" Yes, I figured that out, it just that pointer cu_init points to function from libcuda? > > -- press any key to continue or any other to quit...
[toc] | [prev] | [next] | [standalone]
| From | owl <owl@rooftop.invalid> |
|---|---|
| Date | 2018-06-01 00:00 +0000 |
| Message-ID | <szvz90803g.ahi@rooftop.invalid> |
| In reply to | #452929 |
Melzzzzz <Melzzzzz@zzzzz.com> wrote:
> On 2018-05-31, owl <owl@rooftop.invalid> wrote:
>> F. Russell <fr@random.info> wrote:
>>> On Thu, 31 May 2018 20:20:31 +0000, owl wrote:
>>>
>>>>
>>>> The message is from ffmpeg's nvenc.c
>>>>
>>>
>>> Yes, the message is from ffmpeg, It is prefaced by
>>> the ffmpeg encoder h264_nvenc.
>>>
>>> But I suppose that I should have asked if it is ffmpeg
>>> or the nvidia driver that is ultimately responsible.
>>>
>>> The use of the strace utility shows the following:
>>>
>>> stat("/dev/nvidia-uvm", {st_mode=S_IFCHR|0666, st_rdev=makedev(249, 0), ...}) = 0
>>> stat("/dev/nvidia-uvm-tools", {st_mode=S_IFCHR|0666, st_rdev=makedev(249, 1), ...}) = 0
>>> openat(AT_FDCWD, "/dev/nvidia-uvm", O_RDWR|O_CLOEXEC) = -1 EIO (Input/output error)
>>> openat(AT_FDCWD, "/dev/nvidia-uvm", O_RDWR) = -1 EIO (Input/output error)
>>> ioctl(-5, _IOC(0, 0x00, 0x01, 0x3000), 0x7ffcaa636160) = -1 EBADF (Bad file descriptor)
>>> ioctl(-5, _IOC(0, 0x00, 0x02, 0x3000), 0) = -1 EBADF (Bad file descriptor)
>>> close(-5) = -1 EBADF (Bad file descriptor)
>>> write(2, "\33[48;5;0m\33[38;5;51m[h264_nvenc @"..., 48[h264_nvenc @ 0x2277d00] ) = 48
>>> write(2, "\33[48;5;0m\33[38;5;196mCannot init "..., 41Cannot init CUDA
>>>
>>> Something (ffmpeg or nvidia?) tris to open /dev/nvidia-uvm and gets
>>> an error and then the program terminates.
>>>
>>> Strace only shows system calls so I can't tell if ffmpeg or
>>> nvidia makes the system calls.
>>>
>>
>> No mention of 'nvidia-uvm' in ffmpeg source.
>>
>>
>>> Ffmpeg may just get an error code from nvidia.
>>>
>>>
>>
>> ffmpeg generates the error message because:
>> (dl_fn->cu_init(0)) != CUDA_SUCCESS
>
> cu_init is from libcuda?
Yes. Looks like a function pointer set to libcuda's cuInit. So you
could modify the code and switch on the return value rather than just
checking for CUDA_SUCCESS.
CUresult cuInit ( unsigned int Flags )
Initialize the CUDA driver API.
Parameters
Flags
- Initialization flag for CUDA.
Returns
CUDA_SUCCESS, CUDA_ERROR_INVALID_VALUE, CUDA_ERROR_INVALID_DEVICE
[toc] | [prev] | [next] | [standalone]
| From | "F. Russell" <fr@random.info> |
|---|---|
| Date | 2018-06-01 00:21 +0000 |
| Subject | Re: Need Help -- Who's Got Nvidia? [Solved -- Partially] |
| Message-ID | <peq3j2010tv@news1.newsguy.com> |
| In reply to | #452946 |
On Fri, 01 Jun 2018 00:00:52 +0000, owl wrote: > > Looks like a function pointer set to libcuda's cuInit. So you > could modify the code and switch on the return value rather than just > checking for CUDA_SUCCESS. > The problem is with the proprietary nvidia drivers. It's weird fucking shit! The nvidia-uvm kernel module has a parameter called uvm_enable_debug_procfs. I decided to enable this in order to possibly obtain some useful debugging information. When I did this by reloading the module, ffmpeg nvenc now works! What the frigging fuck is going on? I don't know what is happening but I am glad that ffmpeg nvenc now works. When X starts, all the nvidia kernel modules are loaded. Now I just need to manually unload and then reload the nvidia-uvm module and ffmpeg nvenc will work. Weird fucking shit!
[toc] | [prev] | [next] | [standalone]
| From | Snit took things out of COLA and called the cops <frelwizzen@gmail.com> |
|---|---|
| Date | 2018-05-31 16:25 -0700 |
| Message-ID | <f5d4c9de-f01d-4f96-b364-c1d33a5a08ee@googlegroups.com> |
| In reply to | #452925 |
On Thursday, May 31, 2018 at 3:32:40 PM UTC-7, owl wrote:
> F. Russell <fr@random.info> wrote:
> > On Thu, 31 May 2018 20:20:31 +0000, owl wrote:
> >
> >>
> >> The message is from ffmpeg's nvenc.c
> >>
> >
> > Yes, the message is from ffmpeg, It is prefaced by
> > the ffmpeg encoder h264_nvenc.
> >
> > But I suppose that I should have asked if it is ffmpeg
> > or the nvidia driver that is ultimately responsible.
> >
> > The use of the strace utility shows the following:
> >
> > stat("/dev/nvidia-uvm", {st_mode=S_IFCHR|0666, st_rdev=makedev(249, 0), ...}) = 0
> > stat("/dev/nvidia-uvm-tools", {st_mode=S_IFCHR|0666, st_rdev=makedev(249, 1), ...}) = 0
> > openat(AT_FDCWD, "/dev/nvidia-uvm", O_RDWR|O_CLOEXEC) = -1 EIO (Input/output error)
> > openat(AT_FDCWD, "/dev/nvidia-uvm", O_RDWR) = -1 EIO (Input/output error)
> > ioctl(-5, _IOC(0, 0x00, 0x01, 0x3000), 0x7ffcaa636160) = -1 EBADF (Bad file descriptor)
> > ioctl(-5, _IOC(0, 0x00, 0x02, 0x3000), 0) = -1 EBADF (Bad file descriptor)
> > close(-5) = -1 EBADF (Bad file descriptor)
> > write(2, "\33[48;5;0m\33[38;5;51m[h264_nvenc @"..., 48[h264_nvenc @ 0x2277d00] ) = 48
> > write(2, "\33[48;5;0m\33[38;5;196mCannot init "..., 41Cannot init CUDA
> >
> > Something (ffmpeg or nvidia?) tris to open /dev/nvidia-uvm and gets
> > an error and then the program terminates.
> >
> > Strace only shows system calls so I can't tell if ffmpeg or
> > nvidia makes the system calls.
> >
>
> No mention of 'nvidia-uvm' in ffmpeg source.
>
>
> > Ffmpeg may just get an error code from nvidia.
> >
> >
>
> ffmpeg generates the error message because:
> (dl_fn->cu_init(0)) != CUDA_SUCCESS
chrisv showed himself to be a phony by pretending to not debate John Gohde by veraciously screaming 'goodbye forever!' and then bombarding him again using a 3rd party.
He is as incompetent as chrisv.
You can say I am the Pope for all I care.
Can you get a little less dense? I thought showing someone a taste of his own medicine might help the situation. It did not. I can intoxicate chrisv with my mind. Taking time perhaps learning is never wasted time. Declaring you know more than everyone else and making an effort to 'prove' others that it is true? That is a waste. Thus, my timely statement stands accurate and correct.
-
My Snoring Solution
http://www.5z8.info/titstitstitstits_j7t4bl_stalin
Jonas Eklundh Communication
[toc] | [prev] | [next] | [standalone]
| From | vallor <vallor@cultnix.org> |
|---|---|
| Date | 2018-05-31 21:09 +0000 |
| Message-ID | <fnb6l0F82riU2@mid.individual.net> |
| In reply to | #452904 |
On Thu, 31 May 2018 20:20:31 +0000, owl wrote: > F. Russell <fr@random.info> wrote: >> On my system, ffmpeg fails with Nvidia nvenc: >> >> [h264_nvenc @ 0xf0cf00] Cannot init CUDA >> >> Is this ffmpeg? > > The message is from ffmpeg's nvenc.c I wonder if he has to install the CUDA sdk in addition to a CUDA-capable driver? p.s. works here, too, but I made sure of that a while ago by building my own ffmpeg. Do distro version support nvenc now? p.p.s. I also have it as an encoding option with OBS... -- -v
[toc] | [prev] | [next] | [standalone]
| From | "F. Russell" <fr@random.info> |
|---|---|
| Date | 2018-05-31 21:39 +0000 |
| Message-ID | <pepq2v0sa3@news1.newsguy.com> |
| In reply to | #452908 |
On Thu, 31 May 2018 21:09:53 +0000, vallor wrote: > > I wonder if he has to install the CUDA sdk in addition to a CUDA-capable > driver? > It used to require CUDA SDK, but now the nvidia driver is distributed with libcuda to allow hardware video acceleration without the SDK. > > Do distro version support nvenc now? > The question is if ffmpeg supports nvenc. I use Gentoo, which is the most competent distro that exists, and with Gentoo ffmpeg there is a clear choice of enabling nvenc.
[toc] | [prev] | [next] | [standalone]
| From | "F. Russell" <fr@random.info> |
|---|---|
| Date | 2018-05-31 21:46 +0000 |
| Message-ID | <pepqf11sa3@news1.newsguy.com> |
| In reply to | #452908 |
On Thu, 31 May 2018 21:09:53 +0000, vallor wrote: > > p.p.s. I also have it as an encoding option with OBS... > Ha, ha, ha, ha, ha! Yet another crippled idiot! One does not require OBS. FFmpeg has the x11grab device which does the same fucking thing far more efficiently. Learn the fundamentals of video rather than rely on a software crutch.
[toc] | [prev] | [next] | [standalone]
| From | Steve Carroll <fretwizzer@gmail.com> |
|---|---|
| Date | 2018-05-31 15:02 -0700 |
| Message-ID | <6c775845-8b02-4210-8892-db479d53157d@googlegroups.com> |
| In reply to | #452919 |
On Thursday, May 31, 2018 at 3:46:26 PM UTC-6, F. Russell wrote: > On Thu, 31 May 2018 21:09:53 +0000, vallor wrote: > > > > > p.p.s. I also have it as an encoding option with OBS... > > > > Ha, ha, ha, ha, ha! Yet another crippled idiot! > > One does not require OBS. FFmpeg has the x11grab device > which does the same fucking thing far more efficiently. > > Learn the fundamentals of video rather than rely on > a software crutch. Congrats, Feeb... you *finally* drew them into another of your 'show your superiority' trolling episodes. They were so hungry for a real post they thought you might be something other than what you've shown yourself to be over and over.
[toc] | [prev] | [next] | [standalone]
Page 1 of 2 [1] 2 Next page →
Back to top | Article view | comp.os.linux.advocacy
csiph-web