[PD] pdmax - broken pipe in macos?
Christof Ressi
info at christofressi.com
Wed Nov 17 15:10:49 CET 2021
D'oh. I don't know why I though Joao was on Windows... Thanks for the
correction!
Note that lldb is the clang debugger and it works slightly different
than GDB. But you can do the same things.
On 17.11.2021 15:07, Lucas Cordiviola wrote:
>
> @Christof,
>
> I think Joao is needing the *macOS* pdmax.
>
> @Joao,
>
> Everything that Christof said applies if you follow the macOS part:
>
> https://github.com/pure-data/pure-data/blob/master/INSTALL.txt
>
> you just need to install xcode with this command on a terminal:
>
> xcode-select --install
>
> macOS debugger is included and its called (and invoked with) "lldb".
>
> I used it one or two times only. I'm not a mac user.
>
>
> --
>
> Mensaje telepatico asistido por maquinas.
> On 11/17/2021 10:32 AM, Christof Ressi wrote:
>>
>> You would have to build the pdmax external from source with debug
>> symbols to get meaningful information in the debugger. This means you
>> would need to setup Msys2, as described in the "Windows" section in
>> https://github.com/pure-data/pure-data/blob/master/INSTALL.txt.
>>
>> Actually, if the aim is to only get a stacktrace, it is not necessary
>> to install a graphical frontend for GDB, you can also run GDB from
>> the command line.
>>
>> I you never compiled somthing from source it can be a bit
>> overwhelming at first, but it's a valuable skill to have. There are a
>> few Windows people on the list who can help you out (including myself).
>>
>> So I would say it's not trivial, but also not impossible :-)
>>
>> ---
>>
>> That's what I'm doing to debug subprocesses (I also needed to do this
>> with [vstplugin~] a few times):
>>
>> 1) run Pd with DSP off and start the subprocess (here: [pd~ start()
>>
>> 2) find the PID of the subprocess, e.g. with "tasklist | grep <name>"
>> or with ProcessExplorer
>>
>> 3) attach to the subprocess; terminal:
>> https://stackoverflow.com/questions/14370972/how-to-attach-a-process-in-gdb;
>> Qt Creator: "Start Debuggin" -> "Attach to Running Application"
>>
>> 4) turn on DSP and wait for the subprocess to crash
>>
>> If the subprocess crashes immediately even with DSP turned off,
>> things are a bit harder.
>>
>> ---
>>
>> If you can build pdmax with Visual Studio (I don't know if this is
>> supported?) then things are even easier because VS has a very good
>> built-in graphical debugger that can also attach to running applications.
>>
>> Christof
>>
>> On 17.11.2021 14:04, João Pais wrote:
>>> is that something a "normal user" like me (someone that only uses
>>> the surface of the software) can do, or it has to do with C or any
>>> other custom compiled/prepared versions?
>>>
>>> On Wed, 17 Nov 2021 at 01:06, Christof Ressi
>>> <info at christofressi.com> wrote:
>>>
>>> You can attach GDB to a running or even a not-yet-running
>>> application.
>>> When that application crashes, you get a backtrace. That's how I
>>> found
>>> the crash in the [pd~] subprocess.
>>>
>>> Personally, I use Qt Creator because it has a nice graphical
>>> front-end
>>> for GDB where I can easily step through functions while looking
>>> at the
>>> code, watch local variables, etc.
>>>
>>> Christof
>>>
>>> On 17.11.2021 00:49, Miller Puckette via Pd-list wrote:
>>> > If by any chance you're using pd 0.52 (test version) then I think
>>> > it can crash when used as a subprocess (this should be fixed
>>> for the next
>>> > test release). If you're using Pd 0.51-4 the problem is
>>> something else.
>>> >
>>> > The "broken pipe" message means the subprocess died suddenly
>>> for some reason.
>>> > But I can't think of an easy way to figure out what killed it.
>>> >
>>> > cheers
>>> > Miller
>>> >
>>> > On Sat, Nov 06, 2021 at 08:45:28PM +0100, João Pais wrote:
>>> >> Hello list,
>>> >>
>>> >> I have a max patch with pd inside through the pd~ object. But
>>> I'm getting
>>> >> the "broken pipe" error in some cases in a mac, such as:
>>> >>
>>> >> - in a compiled standalone it usually works
>>> >>
>>> >> - in the original patch it doesn't
>>> >>
>>> >> - both patches were made in the same system, with the latest
>>> pdmax and pd
>>> >> versions were used. I'm not sure anymore which max version or
>>> macos system
>>> >> was used to compile the standalone, if that's important (it
>>> was either max 7
>>> >> or 8).
>>> >>
>>> >> I don't have access to a mac myself, but the system where it
>>> was tested was
>>> >> a 11.6, with Max 8. It seems to me that in some systems it
>>> works, and in
>>> >> other it doesn't - but I don't have enough hard data to prove it.
>>> >>
>>> >> Best,
>>> >>
>>> >> Joao
>>> >>
>>> >>
>>> >>
>>> >>
>>> >> _______________________________________________
>>> >> Pd-list at lists.iem.at mailing list
>>> >> UNSUBSCRIBE and account-management ->
>>> https://urldefense.proofpoint.com/v2/url?u=https-3A__lists.puredata.info_listinfo_pd-2Dlist&d=DwICAg&c=-35OiAkTchMrZOngvJPOeA&r=XprZV3Fxus2L1LCw80hE4Q&m=sR96adpeL4KB76yEfQG-D8WqWrtlw2sW2qCwz_nEjLiuLt0KgKMIh-Lc-HCrSJcC&s=K9LPmjo4sdbRhF9wuUqbYoL5l1WxeUKav5yEeAse22o&e=
>>> <https://urldefense.proofpoint.com/v2/url?u=https-3A__lists.puredata.info_listinfo_pd-2Dlist&d=DwICAg&c=-35OiAkTchMrZOngvJPOeA&r=XprZV3Fxus2L1LCw80hE4Q&m=sR96adpeL4KB76yEfQG-D8WqWrtlw2sW2qCwz_nEjLiuLt0KgKMIh-Lc-HCrSJcC&s=K9LPmjo4sdbRhF9wuUqbYoL5l1WxeUKav5yEeAse22o&e=>
>>>
>>>
>>>
>>> _______________________________________________
>>> Pd-list at lists.iem.at mailing list
>>> UNSUBSCRIBE and account-management ->
>>> https://lists.puredata.info/listinfo/pd-list
>>>
>>
>> _______________________________________________
>> Pd-list at lists.iem.at mailing list
>> UNSUBSCRIBE and account-management ->https://lists.puredata.info/listinfo/pd-list
>
> _______________________________________________
> Pd-list at lists.iem.at mailing list
> UNSUBSCRIBE and account-management ->https://lists.puredata.info/listinfo/pd-list
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.puredata.info/pipermail/pd-list/attachments/20211117/10168754/attachment-0001.htm>
More information about the Pd-list
mailing list