cudaFreeHost cudaFreeHost may not work properly?

the following code:

domplex matrix;
cudaHostAlloc((void **)&matrix,9261
9261*sizeof(domplex),cudaHostAllocPortable|c
udaHostAllocWriteCombined);
cudaFreeHost((void *)matrix);

needs 325 seconds to run… and the 325 seconds are spent in the cudaFreeHost routine! This must be a bug.

System:
Linux x86_64 openSuse 11.0
Kernel 2.6.27.23-0.1-default #1 SMP 2009-05-26 17:02:05 -0400 x86_64 x86_64 x86_64 GNU/Linux
NVidia driver: 185.18.14
CUDA 2.2
CPU: core 2 quad Q9550 @ 2.83GHz
GPU: 2x ASUS 285GTX
Motherboard: ASUS Striker II Extreme (NVIDIA nForce 790i Ultra SLI)
RAM: 8GB DDR3