search

PhilippeBekaert / snd-hdspe

New linux driver and tools for RME HDSPe sound cards and extension modules
GNU General Public License v3.0
51 stars 18 forks source link

Compilation fails #5

Open alowell-ssl opened 2 years ago

alowell-ssl commented 2 years ago

Hi, I just cloned commit cc8621559834a7b18de7285c518bddcf26b522b8. I'm on Ubuntu 20.04 LTS, kernel is 5.4.0-91-lowlatency

Going through your instructions, compilation fails with: image

PhilippeBekaert commented 2 years ago

Strange. I’m also on Ubuntu 20.04 lowlatency. I’m currently out of office but will check asap what has happened with snd_card.sync_irq. Thanks for reporting. Happy new year. Ph

Op 30 dec. 2021 om 17:17 heeft alowell-ssl @.***> het volgende geschreven:

 Hi, I just cloned commit cc86215. I'm on Ubuntu 20.04 LTS, kernel is 5.4.0-91-lowlatency

Going through your instructions, compilation fails with:

— Reply to this email directly, view it on GitHub, or unsubscribe. Triage notifications on the go with GitHub Mobile for iOS or Android. You are receiving this because you are subscribed to this thread.

PhilippeBekaert commented 2 years ago

I see I'm on a later version of the kernel: 5.11.0-43. Is it an option for you to update/upgrade your system?

Or you could try commenting out that offending line 464. It's an addition to audio drivers dating back to the end of 2019. (Just that single line added to the hdspm driver at that time. hdspe is based on hdspm.)

However, you may end up finding other compilation issues due to changes in the kernel alsa subsystem since your kernel version.

Let me know how it works out, and thanks again for reporting.

Best,

Philippe.

alowell-ssl commented 2 years ago

Hi Philippe,

Thanks for your attention! I commented out the line and the compilation was successful. I'll give it a shot soon. I think I got everything working stably with the stock hdsp kernel module (this took 1-2 days of messing with my system), but I'm still curious to try your module. Do you think your module offers anything in the way of reducing xruns compared to the stock module?

Happy new year, Alex

On Mon, Jan 3, 2022 at 9:01 AM Philippe Bekaert @.***> wrote:

I see I'm on a later version of the kernel: 5.11.0-43. Is it an option for you to update/upgrade your system?

Or you could try commenting out that offending line 464. It's an addition to audio drivers dating back to the end of 2019. (Just that single line added to the hdspm driver at that time. hdspe is based on hdspm.)

However, you may end up finding other compilation issues due to changes in the kernel alsa subsystem since your kernel version.

Let me know how it works out, and thanks again for reporting.

Best,

Philippe.

— Reply to this email directly, view it on GitHub https://github.com/PhilippeBekaert/snd-hdspe/issues/5#issuecomment-1004226058, or unsubscribe https://github.com/notifications/unsubscribe-auth/AGBCCSFFJ7VRINZ4H2LHYE3UUHI6HANCNFSM5K762KLA . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

You are receiving this because you authored the thread.Message ID: @.***>

PhilippeBekaert commented 2 years ago

What card are you using the driver with? Hdsp and hdspm were already pretty optimal wrt xruns. My driver will be similar I expect. In any case it was not worse in my tests. I fixed several issues, especially wrt double and quad speed, implemented support for the new aio pro, completed the feature set and wrote a user space configuration tool similar to those of rme on other platforms, but in terms of xrun performance I do not think there was much to improve. (I have no xruns btw even on small buffer sizes, but that may be different on older or less powerful hardware). Best, ph

Op 3 jan. 2022 om 18:37 heeft alowell-ssl @.***> het volgende geschreven:

 Hi Philippe,

Thanks for your attention! I commented out the line and the compilation was successful. I'll give it a shot soon. I think I got everything working stably with the stock hdsp kernel module (this took 1-2 days of messing with my system), but I'm still curious to try your module. Do you think your module offers anything in the way of reducing xruns compared to the stock module?

Happy new year, Alex

On Mon, Jan 3, 2022 at 9:01 AM Philippe Bekaert @.***> wrote:

I see I'm on a later version of the kernel: 5.11.0-43. Is it an option for you to update/upgrade your system?

Or you could try commenting out that offending line 464. It's an addition to audio drivers dating back to the end of 2019. (Just that single line added to the hdspm driver at that time. hdspe is based on hdspm.)

However, you may end up finding other compilation issues due to changes in the kernel alsa subsystem since your kernel version.

Let me know how it works out, and thanks again for reporting.

Best,

Philippe.

— Reply to this email directly, view it on GitHub https://github.com/PhilippeBekaert/snd-hdspe/issues/5#issuecomment-1004226058, or unsubscribe https://github.com/notifications/unsubscribe-auth/AGBCCSFFJ7VRINZ4H2LHYE3UUHI6HANCNFSM5K762KLA . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

You are receiving this because you authored the thread.Message ID: @.***>

— Reply to this email directly, view it on GitHub, or unsubscribe. Triage notifications on the go with GitHub Mobile for iOS or Android. You are receiving this because you commented.

alowell-ssl commented 2 years ago

Hi Philippe,

Thanks, I'm using the AIO (non-pro). I have a pretty modern desktop, and finally got rid of xruns (mostly) after fiddling with many settings. I still get some if I'm using certain browser apps (for example, just loading a big google doc will cause a huge burst of xruns), but otherwise it runs smooth even with high DSP load. The final source of xruns that I tracked down was probably not related to the driver. Instead, I think I just wasn't running jackd with high enough priority, so background system tasks were preempting and causing xruns (figured this out by watching the xrun counter and journalctl -f at the same time).

Just curious, how does one go about writing driver code for commercial hardware? Did RME give you information on which registers do what? Or was this all reverse engineered?

Thanks, Alex

On Mon, Jan 3, 2022 at 11:19 AM Philippe Bekaert @.***> wrote:

What card are you using the driver with? Hdsp and hdspm were already pretty optimal wrt xruns. My driver will be similar I expect. In any case it was not worse in my tests. I fixed several issues, especially wrt double and quad speed, implemented support for the new aio pro, completed the feature set and wrote a user space configuration tool similar to those of rme on other platforms, but in terms of xrun performance I do not think there was much to improve. (I have no xruns btw even on small buffer sizes, but that may be different on older or less powerful hardware). Best, ph

Op 3 jan. 2022 om 18:37 heeft alowell-ssl @.***> het volgende geschreven:

 Hi Philippe,

Thanks for your attention! I commented out the line and the compilation was successful. I'll give it a shot soon. I think I got everything working stably with the stock hdsp kernel module (this took 1-2 days of messing with my system), but I'm still curious to try your module. Do you think your module offers anything in the way of reducing xruns compared to the stock module?

Happy new year, Alex

On Mon, Jan 3, 2022 at 9:01 AM Philippe Bekaert @.***> wrote:

I see I'm on a later version of the kernel: 5.11.0-43. Is it an option for you to update/upgrade your system?

Or you could try commenting out that offending line 464. It's an addition to audio drivers dating back to the end of 2019. (Just that single line added to the hdspm driver at that time. hdspe is based on hdspm.)

However, you may end up finding other compilation issues due to changes in the kernel alsa subsystem since your kernel version.

Let me know how it works out, and thanks again for reporting.

Best,

Philippe.

— Reply to this email directly, view it on GitHub < https://github.com/PhilippeBekaert/snd-hdspe/issues/5#issuecomment-1004226058 , or unsubscribe < https://github.com/notifications/unsubscribe-auth/AGBCCSFFJ7VRINZ4H2LHYE3UUHI6HANCNFSM5K762KLA

. Triage notifications on the go with GitHub Mobile for iOS < https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675

or Android < https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub .

You are receiving this because you authored the thread.Message ID: @.***>

— Reply to this email directly, view it on GitHub, or unsubscribe. Triage notifications on the go with GitHub Mobile for iOS or Android. You are receiving this because you commented.

— Reply to this email directly, view it on GitHub https://github.com/PhilippeBekaert/snd-hdspe/issues/5#issuecomment-1004302283, or unsubscribe https://github.com/notifications/unsubscribe-auth/AGBCCSBDUFQOOGFPNS4KHJLUUHZDHANCNFSM5K762KLA . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

You are receiving this because you authored the thread.Message ID: @.***>

PhilippeBekaert commented 2 years ago

Hi Alex,

The AIO is supported perfectly well. I've been using these cards for many years myself too, including television broadcasting and live streaming. Finding out RME discontinued the production of the AIO and now offers the AIO Pro, which was not supported by the hdspm driver, was the trigger that made me develop the new driver. jack thread priority helps, and a low latency kernel too (ubuntu studio e.g.). There is no reverse engineering in this driver. RME was so kind to share their windows and mac driver source code with me. Best, Philippe. On Jan 3 2022, at 8:40 pm, alowell-ssl @.***> wrote:

Hi Philippe, Thanks, I'm using the AIO (non-pro). I have a pretty modern desktop, and finally got rid of xruns (mostly) after fiddling with many settings. I still get some if I'm using certain browser apps (for example, just loading a big google doc will cause a huge burst of xruns), but otherwise it runs smooth even with high DSP load. The final source of xruns that I tracked down was probably not related to the driver. Instead, I think I just wasn't running jackd with high enough priority, so background system tasks were preempting and causing xruns (figured this out by watching the xrun counter and journalctl -f at the same time).

Just curious, how does one go about writing driver code for commercial hardware? Did RME give you information on which registers do what? Or was this all reverse engineered?

Thanks, Alex

On Mon, Jan 3, 2022 at 11:19 AM Philippe Bekaert @.***> wrote:

What card are you using the driver with? Hdsp and hdspm were already pretty optimal wrt xruns. My driver will be similar I expect. In any case it was not worse in my tests. I fixed several issues, especially wrt double and quad speed, implemented support for the new aio pro, completed the feature set and wrote a user space configuration tool similar to those of rme on other platforms, but in terms of xrun performance I do not think there was much to improve. (I have no xruns btw even on small buffer sizes, but that may be different on older or less powerful hardware). Best, ph

Op 3 jan. 2022 om 18:37 heeft alowell-ssl @.***> het volgende geschreven:

 Hi Philippe,

Thanks for your attention! I commented out the line and the compilation was successful. I'll give it a shot soon. I think I got everything working stably with the stock hdsp kernel module (this took 1-2 days of messing with my system), but I'm still curious to try your module. Do you think your module offers anything in the way of reducing xruns compared to the stock module?

Happy new year, Alex

On Mon, Jan 3, 2022 at 9:01 AM Philippe Bekaert @.***> wrote:

I see I'm on a later version of the kernel: 5.11.0-43. Is it an option for you to update/upgrade your system?

Or you could try commenting out that offending line 464. It's an addition to audio drivers dating back to the end of 2019. (Just that single line added to the hdspm driver at that time. hdspe is based on hdspm.)

However, you may end up finding other compilation issues due to changes in the kernel alsa subsystem since your kernel version.

Let me know how it works out, and thanks again for reporting.

Best,

Philippe.

— Reply to this email directly, view it on GitHub < https://github.com/PhilippeBekaert/snd-hdspe/issues/5#issuecomment-1004226058 , or unsubscribe < https://github.com/notifications/unsubscribe-auth/AGBCCSFFJ7VRINZ4H2LHYE3UUHI6HANCNFSM5K762KLA

. Triage notifications on the go with GitHub Mobile for iOS < https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675

or Android < https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub .

You are receiving this because you authored the thread.Message ID: @.***>

— Reply to this email directly, view it on GitHub, or unsubscribe. Triage notifications on the go with GitHub Mobile for iOS or Android. You are receiving this because you commented.

— Reply to this email directly, view it on GitHub https://github.com/PhilippeBekaert/snd-hdspe/issues/5#issuecomment-1004302283, or unsubscribe https://github.com/notifications/unsubscribe-auth/AGBCCSBDUFQOOGFPNS4KHJLUUHZDHANCNFSM5K762KLA . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

You are receiving this because you authored the thread.Message ID: @.***>

— Reply to this email directly, view it on GitHub @./0?redirect=https%3A%2F%2Fgithub.com%2FPhilippeBekaert%2Fsnd-hdspe%2Fissues%2F5%23issuecomment-1004313238&recipient=cmVwbHkrQVBDRVlMQUJQUVEySVJSWURXQ1czR043MzRaREJFVkJOSEhFQ0NTS1dJQHJlcGx5LmdpdGh1Yi5jb20%3D), or unsubscribe @./1?redirect=https%3A%2F%2Fgithub.com%2Fnotifications%2Funsubscribe-auth%2FAPCEYLA6WNA4SUYKPCCDSB3UUH3TBANCNFSM5K762KLA&recipient=cmVwbHkrQVBDRVlMQUJQUVEySVJSWURXQ1czR043MzRaREJFVkJOSEhFQ0NTS1dJQHJlcGx5LmdpdGh1Yi5jb20%3D). Triage notifications on the go with GitHub Mobile for iOS @./2?redirect=https%3A%2F%2Fapps.apple.com%2Fapp%2Fapple-store%2Fid1477376905%3Fct%3Dnotification-email%26mt%3D8%26pt%3D524675&recipient=cmVwbHkrQVBDRVlMQUJQUVEySVJSWURXQ1czR043MzRaREJFVkJOSEhFQ0NTS1dJQHJlcGx5LmdpdGh1Yi5jb20%3D) or Android @./3?redirect=https%3A%2F%2Fplay.google.com%2Fstore%2Fapps%2Fdetails%3Fid%3Dcom.github.android%26referrer%3Dutm_campaign%253Dnotification-email%2526utm_medium%253Demail%2526utm_source%253Dgithub&recipient=cmVwbHkrQVBDRVlMQUJQUVEySVJSWURXQ1czR043MzRaREJFVkJOSEhFQ0NTS1dJQHJlcGx5LmdpdGh1Yi5jb20%3D). You are receiving this because you commented.