[PlanetCCRMA] zita-convolver

[Brendan Jones]

On 01/25/2012 08:01 AM, craig stephenson wrote:
> I found the patch on the Debian list, installed a few dependencies I
> was missing, and got the following when I tried to build.  Any ideas?
>
> g++ ir.cc -Wall -I. -I/usr/include `pkg-config --cflags gtk+-2.0`
> `pkg-config --cflags gthread-2.0` -D__STDC_FORMAT_MACROS -O2
> -ffast-math -D_HAVE_GTK_ATLEAST_2_20 -c -fPIC -o ir.o
> ir.cc: In function ‘void free_conv_safely(Convproc*)’:
> ir.cc:168:9: error: ‘class Convproc’ has no member named ‘check’
> ir.cc: In function ‘void init_conv(IR*)’:
> ir.cc:561:23: error: no matching function for call to
> ‘Convproc::start_process(int)’
> ir.cc:561:23: note: candidate is:
> /usr/include/zita-convolver.h:414:9: note: int Convproc::start_process(int, int)
> /usr/include/zita-convolver.h:414:9: note:   candidate expects 2
> arguments, 1 provided
> make: *** [ir.o] Error 1
>
>
> On Tue, Jan 24, 2012 at 7:05 PM, craig stephenson
> <craigstep888 at gmail.com>  wrote:
>> Yes, it was compiled from source.  I will go find the patch.  Thank you.
>>
>> On Tue, Jan 24, 2012 at 6:34 PM, michael noble<looplog at gmail.com>  wrote:
>>> On Wed, Jan 25, 2012 at 10:26 AM, Fernando Lopez-Lezcano
>>> <nando at ccrma.stanford.edu>  wrote:
>>>> On 01/24/2012 01:30 PM, craig stephenson wrote:
>>>>> Hello all.  I've stumbled over another rock on the upgrade road from
>>>>> Fedora 14 to 16.  It seems the ir.lv2-1.2 plugin from
>>>>> http://factorial.hu/plugins/lv2/ir is broken using package
>>>>> zita-convolver-0:3.1.0-1.fc16.x86_64.
>>>>
>>>> Is the plugin something you compiled from source?
>>>> If so, you need to rebuild it with the new zita-convolver libraries.
>>>>
>>>> If it is part of a package (which one?) then that package needs to be
>>>> rebuilt.
>>>>
>>>
>>> The LV2.IR plugin source needs to be patched in order to build
>>> successfully with libzita-convolver-3. It won't build otherwise. I
>>> don't have the patch on this machine but if you search you may find
>>> it. I found it in a Debian mailing list I think. Ideally, the patch
>>> should be forwarded to the LV2.IR developer.
>>>
>>> -m
>
> _______________________________________________
> PlanetCCRMA mailing list
> PlanetCCRMA at ccrma.stanford.edu
> http://ccrma-mail.stanford.edu/mailman/listinfo/planetccrma

It looks as if your patch has not been applied.

diff -Nurp ir.lv2-1.3.1.orig/ir.cc ir.lv2-1.3.1/ir.cc
--- ir.lv2-1.3.1.orig/ir.cc	2011-02-04 18:36:58.000000000 +0100
+++ ir.lv2-1.3.1/ir.cc	2012-01-25 08:50:32.235930150 +0100
@@ -165,8 +165,8 @@ static void free_conv_safely(Convproc *
  		treq.tv_nsec = 10000000;
  		nanosleep(&treq, &trem);

-		conv->check();
-		state = conv->state();
+		if (conv->check_stop())
+    		state = conv->state();
  	}
  	delete conv;	
  }
@@ -558,7 +558,7 @@ static void init_conv(IR * ir) {
  			ir->nchan);
  	}

-	conv->start_process(0);
+	conv->start_process(0, SCHED_FIFO);
  	ir->conv_req_to_use = req_to_use;
  }
`