[Discuss-gnuradio] segfault in volk_32fc_x2_multiply_32fc_a_sse3 using current master branch

[Ben Reynwar]

I'm seeing a segfault in volk_32fc_x2_multiply_32fc_a_sse3 and am
using the current master branch.

It occurs when I connect a gr.multiply_cc to a running flowgraph
(using tb.lock() and tb.unlock()).  If the multiply_cc block is
connected before the flowgraph starts running then I don't see a
segfault.

I have attached:
  - a minimal python script that reproduces the problem for me.
  - the output from gdb (backtrace and disassemble)

OS is 32 bit Ubuntu 11.04 beta.
cpu is Intel(R) Core(TM) i3-2100 CPU @ 3.10GHz

minimal_segfault.py
Description: Binary data

gdb.log
Description: Binary data