I am trying to debug an extern on linux using gdb and this tutorial https://puredata.info/docs/developer/DebuggingPdExternals (for MacOS and some years old) does not work on my machine. In fact, I've tried this a few times in the past with little success. The watchdog just hangs when I break and gdb and Pd can't do anything.
I remember Miller once suggesting that a way to run gdb on an extern is to build it into Pd with debugging symbols on. I am trying to do this, but having trouble there too. I added 'my_extern.c' to the src folder and to the list of .c files that comprise SRC in makefile.gnu. I also added -g to CFLAGS. Everything compiles and I see 'my_extern.o' in the obj folder, but the extern doesn't load when I create the object in Pd.
When I run pd from gdb, i can set a break point on my_extern_new, but when I create the object, I just get a ...cannot create message in the console and no debugging symbols. I feel like I'm missing a step somewhere.
Any help is kindly appreciated.
-David
Did you add the two lines to m_conf.c ?
cheers Miller
On Wed, Apr 11, 2018 at 07:03:20PM +0200, David Medine wrote:
I am trying to debug an extern on linux using gdb and this tutorial https://puredata.info/docs/developer/DebuggingPdExternals (for MacOS and some years old) does not work on my machine. In fact, I've tried this a few times in the past with little success. The watchdog just hangs when I break and gdb and Pd can't do anything.
I remember Miller once suggesting that a way to run gdb on an extern is to build it into Pd with debugging symbols on. I am trying to do this, but having trouble there too. I added 'my_extern.c' to the src folder and to the list of .c files that comprise SRC in makefile.gnu. I also added -g to CFLAGS. Everything compiles and I see 'my_extern.o' in the obj folder, but the extern doesn't load when I create the object in Pd.
When I run pd from gdb, i can set a break point on my_extern_new, but when I create the object, I just get a ...cannot create message in the console and no debugging symbols. I feel like I'm missing a step somewhere.
Any help is kindly appreciated.
-David
Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://lists.puredata.info/listinfo/pd-list
On 04/11/2018 07:03 PM, David Medine wrote:
I am trying to debug an extern on linux using gdb and this tutorial https://puredata.info/docs/developer/DebuggingPdExternals (for MacOS and some years old) does not work on my machine. In fact, I've tried this a few times in the past with little success. The watchdog just hangs when I break and gdb and Pd can't do anything.
I remember Miller once suggesting that a way to run gdb on an extern is to build it into Pd with debugging symbols on. I am trying to do this, but having trouble there too. I added 'my_extern.c' to the src folder and to the list of .c files that comprise SRC in makefile.gnu. I also added -g to CFLAGS. Everything compiles and I see 'my_extern.o' in the obj folder, but the extern doesn't load when I create the object in Pd.
When I run pd from gdb, i can set a break point on my_extern_new, but when I create the object, I just get a ...cannot create message in the console and no debugging symbols. I feel like I'm missing a step somewhere.
i never had problems with using gdb and "external" objects (that is: objects, not copmiled into Pd).
i'm still using gdb, as i'm on linux. on macOS i hear that you should use the llvm based debugger instead.
a few tricks:
- compile your external with "-g" (so it includes debugging symbols; i
still think that it is a shame that pd-lib-builder doesn't do that by default; practically all my projects compile with "-g" all the time)
- make sure that you are running Pd with "-nrt" (so you don't have
"watchdog...signalling Pd" messages all over the place once the execution is halted)
- run pd with "-stderr"
- before setting any breakpoints, do a "warm up" run: just run Pd
(inside the debugger) and load the external. after this, the debugger should be able to resolve all the symbols.
gfmdsar IOhannes
@Miller, no I hadn't. Thanks! I knew there was something like that I had neglected. Now I am able to debug with gdb.
@iohannes, I am still having trouble with this technique. If I run Pd from the gdb prompt (with -nrt and -stderr) and a break point within the extern (meaning object not compiled into pd) nothing happens when I run the function where I want to break. If I follow the instructions on Hans' tutorial (run pd in a separate process then attach gdb to it) Pd simply hangs when I hit the break point and gdb doesn't appear to do anything.
On 04/11/2018 08:21 PM, IOhannes m zmölnig wrote:
On 04/11/2018 07:03 PM, David Medine wrote:
I am trying to debug an extern on linux using gdb and this tutorial https://puredata.info/docs/developer/DebuggingPdExternals (for MacOS and some years old) does not work on my machine. In fact, I've tried this a few times in the past with little success. The watchdog just hangs when I break and gdb and Pd can't do anything.
I remember Miller once suggesting that a way to run gdb on an extern is to build it into Pd with debugging symbols on. I am trying to do this, but having trouble there too. I added 'my_extern.c' to the src folder and to the list of .c files that comprise SRC in makefile.gnu. I also added -g to CFLAGS. Everything compiles and I see 'my_extern.o' in the obj folder, but the extern doesn't load when I create the object in Pd.
When I run pd from gdb, i can set a break point on my_extern_new, but when I create the object, I just get a ...cannot create message in the console and no debugging symbols. I feel like I'm missing a step somewhere.
i never had problems with using gdb and "external" objects (that is: objects, not copmiled into Pd).
i'm still using gdb, as i'm on linux. on macOS i hear that you should use the llvm based debugger instead.
a few tricks:
- compile your external with "-g" (so it includes debugging symbols; i
still think that it is a shame that pd-lib-builder doesn't do that by default; practically all my projects compile with "-g" all the time)
- make sure that you are running Pd with "-nrt" (so you don't have
"watchdog...signalling Pd" messages all over the place once the execution is halted)
- run pd with "-stderr"
- before setting any breakpoints, do a "warm up" run: just run Pd
(inside the debugger) and load the external. after this, the debugger should be able to resolve all the symbols.
gfmdsar IOhannes
Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://lists.puredata.info/listinfo/pd-list
On 04/12/2018 11:18 AM, David Medine wrote:
@Miller, no I hadn't. Thanks! I knew there was something like that I had neglected. Now I am able to debug with gdb.
@iohannes, I am still having trouble with this technique. If I run Pd from the gdb prompt (with -nrt and -stderr) and a break point within the extern (meaning object not compiled into pd) nothing happens when I run the function where I want to break.
but the breakpoint gets resolved correctly? e.g. after a warm-up run, setting the breakpoint yields an address:
$ gdb --args pd -nrt -stderr -lib zexy ./reference/tabdump-help.pd [...] (gdb) run [...] (gdb) break tabdump_bang Breakpoint 1 at 0x7ffff5cb6730: file tabdump.c, line 33. (gdb) run [ ...send "bang" to tabdump... ] Thread 1 "pd" hit Breakpoint 1, tabdump_bang (x=0x555555971ef0) at tabdump.c:33 33 { (gdb) print(x) $1 = (t_tabdump *) 0x555555971ef0 (gdb)
gfdsmrt IOhannes
After starting gdb --args etc. as in your example, and after calling run I get this:
[Thread debugging using libthread_db enabled] Using host libthread_db library "/lib64/libthread_db.so.1". Detaching after fork from child process 3889. [New Thread 0x7ffff61ff700 (LWP 3891)]
(Pd is on PID 3888 in this case)
and there is no gdb prompt. If I put a break point in before I hit 'run' I get the same, and the process doesn't break, even though the function I wanted to break is surely called.
On 04/12/2018 12:19 PM, IOhannes m zmölnig wrote:
On 04/12/2018 11:18 AM, David Medine wrote:
@Miller, no I hadn't. Thanks! I knew there was something like that I had neglected. Now I am able to debug with gdb.
@iohannes, I am still having trouble with this technique. If I run Pd from the gdb prompt (with -nrt and -stderr) and a break point within the extern (meaning object not compiled into pd) nothing happens when I run the function where I want to break.
but the breakpoint gets resolved correctly? e.g. after a warm-up run, setting the breakpoint yields an address:
$ gdb --args pd -nrt -stderr -lib zexy ./reference/tabdump-help.pd [...] (gdb) run [...] (gdb) break tabdump_bang Breakpoint 1 at 0x7ffff5cb6730: file tabdump.c, line 33. (gdb) run [ ...send "bang" to tabdump... ] Thread 1 "pd" hit Breakpoint 1, tabdump_bang (x=0x555555971ef0) at tabdump.c:33 33 { (gdb) print(x) $1 = (t_tabdump *) 0x555555971ef0 (gdb)
gfdsmrt IOhannes
Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://lists.puredata.info/listinfo/pd-list
On 04/12/2018 01:11 PM, David Medine wrote:
After starting gdb --args etc. as in your example, and after calling run I get this:
[Thread debugging using libthread_db enabled] Using host libthread_db library "/lib64/libthread_db.so.1". Detaching after fork from child process 3889. [New Thread 0x7ffff61ff700 (LWP 3891)]
(Pd is on PID 3888 in this case)
and there is no gdb prompt.
i might have forgotten to mention that: after the initial "run", i manually terminate Pd (so it "hangs" for me as well and there is no gdb prompt, until i quit Pd). this first run is only there to make all the symbols known to gdb.
gfdsamr Iohannes
Hello,
Somme interrogation here with ALSA loopback and Pd.
I used to use ALSA loopback to route audio on my Linux and it works like a charm with ffmpeg : I send the audio to hw:1,0,0 and get it on hw:1,1,0. And the sound sounds good (no glitch).
With Pd, it is not the same. First I need to send the sound on hw:1,1,0 (for instance with ffmpeg) and get it on the card Loopback (in Pd). The sound is noisy.
So, my three questions :
Why the sound is noisy using Pd and how to solve this ? Why I need to send the sound to hw:1,1,0 (and not hw:1,0,0) to get it with Pd ? (not a problem actually) How can I get access to the other device of Loopback with Pd ?
It seems I miss something somewhere...
++
Jack
Le 12/04/2018 à 15:06, Jack a écrit :
Hello,
Somme interrogation here with ALSA loopback and Pd.
I used to use ALSA loopback to route audio on my Linux and it works like a charm with ffmpeg : I send the audio to hw:1,0,0 and get it on hw:1,1,0. And the sound sounds good (no glitch).
With Pd, it is not the same. First I need to send the sound on hw:1,1,0 (for instance with ffmpeg) and get it on the card Loopback (in Pd). The sound is noisy.
So, my three questions :
Why the sound is noisy using Pd and how to solve this ?
Adding latency solves this problem. ++
Jack
Why I need to send the sound to hw:1,1,0 (and not hw:1,0,0) to get it with Pd ? (not a problem actually) How can I get access to the other device of Loopback with Pd ?
It seems I miss something somewhere...
++
Jack
Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://lists.puredata.info/listinfo/pd-list
Hello,
I have a problem to get all usable channels of the ALSA loopback in Pd.
Here the output of : $ aplay -l **** List of PLAYBACK Hardware Devices **** card 0: PCH [HDA Intel PCH], device 0: ALC3266 Analog [ALC3266 Analog] Subdevices: 0/1 Subdevice #0: subdevice #0 card 0: PCH [HDA Intel PCH], device 3: HDMI 0 [HDMI 0] Subdevices: 1/1 Subdevice #0: subdevice #0 card 0: PCH [HDA Intel PCH], device 7: HDMI 1 [HDMI 1] Subdevices: 1/1 Subdevice #0: subdevice #0 card 0: PCH [HDA Intel PCH], device 8: HDMI 2 [HDMI 2] Subdevices: 1/1 Subdevice #0: subdevice #0 card 0: PCH [HDA Intel PCH], device 9: HDMI 3 [HDMI 3] Subdevices: 1/1 Subdevice #0: subdevice #0 card 0: PCH [HDA Intel PCH], device 10: HDMI 4 [HDMI 4] Subdevices: 1/1 Subdevice #0: subdevice #0 card 1: Loopback [Loopback], device 0: Loopback PCM [Loopback PCM] Subdevices: 7/8 Subdevice #0: subdevice #0 Subdevice #1: subdevice #1 Subdevice #2: subdevice #2 Subdevice #3: subdevice #3 Subdevice #4: subdevice #4 Subdevice #5: subdevice #5 Subdevice #6: subdevice #6 Subdevice #7: subdevice #7 card 1: Loopback [Loopback], device 1: Loopback PCM [Loopback PCM] Subdevices: 7/8 Subdevice #0: subdevice #0 Subdevice #1: subdevice #1 Subdevice #2: subdevice #2 Subdevice #3: subdevice #3 Subdevice #4: subdevice #4 Subdevice #5: subdevice #5 Subdevice #6: subdevice #6 Subdevice #7: subdevice #7
and of : $ arecord -l **** List of CAPTURE Hardware Devices **** card 0: PCH [HDA Intel PCH], device 0: ALC3266 Analog [ALC3266 Analog] Subdevices: 1/1 Subdevice #0: subdevice #0 card 1: Loopback [Loopback], device 0: Loopback PCM [Loopback PCM] Subdevices: 7/8 Subdevice #0: subdevice #0 Subdevice #1: subdevice #1 Subdevice #2: subdevice #2 Subdevice #3: subdevice #3 Subdevice #4: subdevice #4 Subdevice #5: subdevice #5 Subdevice #6: subdevice #6 Subdevice #7: subdevice #7 card 1: Loopback [Loopback], device 1: Loopback PCM [Loopback PCM] Subdevices: 7/8 Subdevice #0: subdevice #0 Subdevice #1: subdevice #1 Subdevice #2: subdevice #2 Subdevice #3: subdevice #3 Subdevice #4: subdevice #4 Subdevice #5: subdevice #5 Subdevice #6: subdevice #6 Subdevice #7: subdevice #7
With Pd, if I select Loopback (hardware) with 8 in/out, i get in console : ALSA: set input channels to 2 ALSA: set output channels to 2
How can I recover the 8 channels of ALSA loopback to get all in/out ? Any help would be appreciated. ++
Jack
I don't understand how Pd or ALSA manage in/out. But I found a (weird or not ?) solution : I have to use several devices in Pd with 2 outputs each. So to get 6 inputs from ALSA loopback, I need 3 Loopback devices with 2 input channels in Pd (see small screen capture attached). Someone can confirm if this is the way to do things with Pd and ALSA loopback ?
However, if the quality of the sound thru the first stereo input is good, this is not the case with the second one (lot of glitch/noise). Sure I made something wrong here. So, if you have some recommandations, I will be happy to test them. ++
Jack
Le 12/04/2018 à 19:15, Jack a écrit :
Hello,
I have a problem to get all usable channels of the ALSA loopback in Pd.
Here the output of : $ aplay -l **** List of PLAYBACK Hardware Devices **** card 0: PCH [HDA Intel PCH], device 0: ALC3266 Analog [ALC3266 Analog] Subdevices: 0/1 Subdevice #0: subdevice #0 card 0: PCH [HDA Intel PCH], device 3: HDMI 0 [HDMI 0] Subdevices: 1/1 Subdevice #0: subdevice #0 card 0: PCH [HDA Intel PCH], device 7: HDMI 1 [HDMI 1] Subdevices: 1/1 Subdevice #0: subdevice #0 card 0: PCH [HDA Intel PCH], device 8: HDMI 2 [HDMI 2] Subdevices: 1/1 Subdevice #0: subdevice #0 card 0: PCH [HDA Intel PCH], device 9: HDMI 3 [HDMI 3] Subdevices: 1/1 Subdevice #0: subdevice #0 card 0: PCH [HDA Intel PCH], device 10: HDMI 4 [HDMI 4] Subdevices: 1/1 Subdevice #0: subdevice #0 card 1: Loopback [Loopback], device 0: Loopback PCM [Loopback PCM] Subdevices: 7/8 Subdevice #0: subdevice #0 Subdevice #1: subdevice #1 Subdevice #2: subdevice #2 Subdevice #3: subdevice #3 Subdevice #4: subdevice #4 Subdevice #5: subdevice #5 Subdevice #6: subdevice #6 Subdevice #7: subdevice #7 card 1: Loopback [Loopback], device 1: Loopback PCM [Loopback PCM] Subdevices: 7/8 Subdevice #0: subdevice #0 Subdevice #1: subdevice #1 Subdevice #2: subdevice #2 Subdevice #3: subdevice #3 Subdevice #4: subdevice #4 Subdevice #5: subdevice #5 Subdevice #6: subdevice #6 Subdevice #7: subdevice #7
and of : $ arecord -l **** List of CAPTURE Hardware Devices **** card 0: PCH [HDA Intel PCH], device 0: ALC3266 Analog [ALC3266 Analog] Subdevices: 1/1 Subdevice #0: subdevice #0 card 1: Loopback [Loopback], device 0: Loopback PCM [Loopback PCM] Subdevices: 7/8 Subdevice #0: subdevice #0 Subdevice #1: subdevice #1 Subdevice #2: subdevice #2 Subdevice #3: subdevice #3 Subdevice #4: subdevice #4 Subdevice #5: subdevice #5 Subdevice #6: subdevice #6 Subdevice #7: subdevice #7 card 1: Loopback [Loopback], device 1: Loopback PCM [Loopback PCM] Subdevices: 7/8 Subdevice #0: subdevice #0 Subdevice #1: subdevice #1 Subdevice #2: subdevice #2 Subdevice #3: subdevice #3 Subdevice #4: subdevice #4 Subdevice #5: subdevice #5 Subdevice #6: subdevice #6 Subdevice #7: subdevice #7
With Pd, if I select Loopback (hardware) with 8 in/out, i get in console : ALSA: set input channels to 2 ALSA: set output channels to 2
How can I recover the 8 channels of ALSA loopback to get all in/out ? Any help would be appreciated. ++
Jack
Pd-list@lists.iem.at mailing list UNSUBSCRIBE and account-management -> https://lists.puredata.info/listinfo/pd-list
-
David Medine -
IOhannes m zmölnig -
Jack -
Miller Puckette