[Jack-Devel] Problem with Jack-Brutefir-Atom board

Hello

I'm trying to set-up a music server based on MPD, Jack1 and Brutefir. The 
project is based on a fanless ATOM board (D525 proc). Everything seems to work 
ok, except when the music server stops ("mpc stop" for ex.). After a little 
time, errors ocurr, like:

 1- $ pp: cannot clean up byte from graph wait fd (Resource temporarily 
unavailable) 
 2- $ bad status (-2) for client mpd_jack handling event (type = 5) 
 3- $ subgraph starting at brutefir timed out (subgraph_wait_fd=9, status = 0, 
state = Finished, pollret = 0 revents = 0x0)

I've tried with different distros (gentoo and debian) and doesn't solve the 
problem. But doing the same set-up in different computers, this problem only 
happens when the processor is an Atom.

Any idea what's happening?

PD: Jack in verbose mode:

load = 0.9733 max usecs: 230.000, spare = 22989.000
load = 0.9259 max usecs: 204.000, spare = 23015.000
pp: cannot clean up byte from graph wait fd (Resource temporarily unavailable)
waking server thread
server thread back from poll
trying to lock graph to remove 1 problems
we have problem clients (problems = 1
++ Removing failed clients ...
client alsa_pcm error status 0
client mpd_jack error status 0
client brutefir error status 1
client failure: client brutefir state = Finished errors = 1
removing client "brutefir" from the processing chain
DIS-connect brutefir:tweeter_R and system:playback_2
++ jack_sort_graph
++ jack_rechain_graph():
+++ client is now brutefir active ? 1
client brutefir: start_fd=7, execution_order=0.
+++ client is now mpd_jack active ? 0
client brutefir: wait_fd=9, execution_order=1 (last client).
-- jack_rechain_graph()
-- jack_sort_graph
DIS-connect brutefir:tweeter_L and system:playback_1
++ jack_sort_graph
++ jack_rechain_graph():
+++ client is now mpd_jack active ? 0
+++ client is now brutefir active ? 1
client brutefir: start_fd=7, execution_order=0.
client brutefir: wait_fd=9, execution_order=1 (last client).
-- jack_rechain_graph()
-- jack_sort_graph
+++ deactivate brutefir
++ jack_sort_graph
++ jack_rechain_graph():
+++ client is now brutefir active ? 0
+++ client is now mpd_jack active ? 0
-- jack_rechain_graph()
-- jack_sort_graph
-- Removing failed clients ...
after removing clients, problems = 0
start poll on 5 fd's
load = 0.4802 max usecs: 8.000, spare = 23211.000
jack main caught signal 15
starting server engine shutdown
stopping driver
server thread back from poll
unloading driver
freeing shared port segments
stopping server thread
stopping watchdog thread
last xrun delay: 0.000 usecs
max delay reported by backend: 0.000 usecs
freeing engine shared memory
max usecs: 312.000, engine deleted

1324293627.29281_0.ltw:2,a <loom.20111219T121546-83 at post dot gmane dot org>