seems to be that in src/output-jack.c:output_new_jack() we call
jack_client_open on our output_jack *a and then bail if it fails in goto out;
this then calls jack_ringbuffer_free on a->ringbuf if its not null, but ringbuf
isnt null, if we force a->ringbuf=NULL (as it indeed is) on the new 'a' obj
actually, its the call to jack_ringbuffer_free src/output-jackc:196
out:
if (a->ringbuf != NULL)
jack_ringbuffer_free(a->ringbuf);
if jack_ringbuffer_free is commented out it'll run, and leak no doubt :) but
somethings going wrong there
--
To UNSUBSCRIBE, email to d
hi hedeki. i've attached the debug output of naist-jdic-utf8.postinst and the
chasenrc-utf8
The problem seems to be that the script uses /bin/sh, which links to dash on a
default install.dash does not recognise the -a option on debian sid so the
script fails. it would work if the shebang was cha
hi daniel, thanks for looking at this. the previous patch didnt look quite
right to me either to be honest. i've given your patch some testing on muse and
it seems good, no problems so far with a reasonable amount of use. i'll keep
using it and add anything problematic here if it crops up.
than
4 matches
Mail list logo
