> Le 28 mars 2018 à 13:33, Oleg Nesterov <o...@redhat.com> a écrit : > > and I think there are more problems with overflows/conversions. Say, > > process = @(100000000000000.0); > > doesn't crash faust but the error message doesn't look right: > > ERROR : possible negative values of : -2147483648 > used in delay expression : IN[0]@-2147483648 > interval(-2.14748e+09, -2.14748e+09) > > I guess this is because of sigFixDelay() -> sigIntCast() which hits the > same double-to-int problem. > > > > > OTOH, this code > > process = @(100000000000000); > > compiles, but the cpp code is wrong: > > float fVec0[536870912]; > > probably this is simply because atoi() used by the parser doesn't report an > error. > > O just > process = 100000000000000; > > this compiles to > > output0[i] = FAUSTFLOAT(276447232); > > without any warning. Not that I think this is the serious problem, but still. > > Oleg.
Yann just pushed a quick fix for checkDelayInterval (version 2.5.30). But your raised a more fundamental issue. I’ve started an Issue to follow it here : https://github.com/grame-cncm/faust/issues/170 Fell free to complete it, thanks ! Stéphane ------------------------------------------------------------------------------ Check out the vibrant tech community on one of the world's most engaging tech sites, Slashdot.org! http://sdm.link/slashdot _______________________________________________ Faudiostream-users mailing list Faudiostream-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/faudiostream-users
