Intel PT support added by Bojan Novković

Regen

Regenerate

In D44084#1005680, @dev_submerge.ch wrote:

I'm still testing this, but I haven't found any problems in the last few days. It's hard to tell whether this approach is robust against missed interrupts, because I never encountered any.

Ruslan, does the calculation of position and length look reasonable to you? We have to store the playback writing position extra because we don't set it in the sndbuf structure.

In D43798#1002933, @br wrote:

All seems to work. I have only line output connected here at home, so only hear sound on first two playback ports in qjackctl.
I can check phone (port 3,4) output on my different machine during the week.

In D43798#1002444, @dev_submerge.ch wrote:

In D43798#1002381, @br wrote:

In D43798#999017, @dev_submerge.ch wrote:

This was tested with a HDSPe RayDAT. @br, could you at least test the analog outputs of the AIO? You probably need audio/jack or audio/virtual_oss to handle that many channels. Just ask me if you need help with the setup.

I have set hw.hdspe.unified_pcm=1 and tested playback mono (s32le:1.0) and stereo on Line output

help needed with jackit config to test this properly!

My general notes on Jack are here, Jack settings described here, and there's man jackd.

For a quick test I would do (roughly):

1. Set hw.hdspe.unified_pcm="1" in /boot/loader.conf.
2. Reboot.
3. Check cat /dev/sndstat | grep HDSPe, should be one pcm device named "all".
4. sysctl dev.hdspe.0.speedshould be 0.
5. sysctl dev.hdspe.0.clock_source should be internal.
6. sysctl dev.pcm.0.bitperfect should be 1 (use pcm number of HDSPe AIO).
7. Set dev.pcm.0.rec.vchans=0 and dev.pcm.0.play.vchans=0.
8. Start jackd -d oss --rate 48000 -p 1024 -w 32 --inchannels 14 --outchannels 16 --capture /dev/dsp0 --playback /dev/dsp0.
9. Open Hydrogen (audio/hydrogen) and load a demo from the examples.
10. Use the connection editor of QJackCtl (audio/qjackctl) to plug Hydrogen output into other playback ports (phones should be 3 and 4).

Sorry, I just see that the number of outchannels (playback) and inchannels (recording) in the man page are wrong.
I think the correct ones would be 16 | 14 up to 48kHz, 12 | 10 up to 96kHz, 10 | 8 up to 192kHz. Could you verify that? I'll fix the man page tomorrow.

Hope this helps, if you are stuck somewhere I need more details.

In D43798#999017, @dev_submerge.ch wrote:

This was tested with a HDSPe RayDAT. @br, could you at least test the analog outputs of the AIO? You probably need audio/jack or audio/virtual_oss to handle that many channels. Just ask me if you need help with the setup.

Seems to work. I hear gaps and stuttering when I force period to 4096. On default (32 to 256) it works very well, even up to 2048.

looks very good

Note that some lines don't fit 80 chars per style (9) and some have too much whitespace after wrap.

Interesting!

Looks good. but I noticed these things just before committing

great work, just minor things

Pass ctx to backend_deinit().

Pass ctx to backend_deinit().

rebase for freebsd

rebase for freebsd

rebase for freebsd

User provides cpusetsize together with *cpuset_t.

Fix typo.

Replace most of spin locks to standard mtx. Use refcounting in some places

change backend lock, ident set lock, ownerhash lock to standard mtx

remove hwt_libxo.h as not needed

Regenerate

Regenerate

obtain ETM IDR registers from sysctl. This is needed for Juno

Allow attaching to an existed proc;
Add support for libxo

Allow attaching to an existed proc

Add description for CPU mode of operation into hwt.c

Fix context ID allocation, do not allocate twice and check the status of allocation

regenerate

cleanup

Regenerate

Regenerate

regenerate

regenerate

decode multiple cpus at the same time

Regenerate

Fix tracing cpu_id > 0

Add -R (fs_root) flag.

hwt kernel module added

ability to build as a module

Code cleanup. Implement some TODOs

Regenerate

format flag added

Allow tracing multiple cpus in CPU(system) mode

allow tracing multiple cpus in system mode

allow tracing multiple cpus in system mode

Remove cheribsd-specific "size" member from hwt_record as not needed

Kernel-mode tracing support added.

Kernel-mode tracing support added.

Add kernel-mode tracing support.

Initial work on cpu/system mode of operation

hwt_backend interface changed

usage() added

Abstract tracing devices

regenerate

regenerate

suspend specific thread on mmap, not entire proc

suspend a single thread on mmap, not entire proc

ip range filtering (draft) added

coresight_setup() method added

ip range filtering added

Regenerate

In D40466#921159, @andrew wrote:

This looks like a kernel with HWT_HOOKS but not HWT wouldn't link as, for example, hwt_switch_in will be missing. HWPMC handles this by adding a function pointer and protects it with an epoch.

I added this functionality. I am not sure if epoch is needed. PMC have several macroses for hook invocation, only some of them require epoch.

Make coresight optional. Depends on hwt

Move changes related to coresight to another review

Move hwt files to conf/files.

Add '-t' argument allowing to specify thread index we want to read trace for.

Add short overview and context lifecycle.

Add hwt_hook variable and HWT_CALL_HOOK macro

Add hwt_hook variable and HWT_CALL_HOOK macro

Remove mtx_threads and mtx_records. Use mtx of entire context

Split out ownerhash/contexthash code to a separate files.