On 16 October 2015 at 12:26, Peter Maydell <[email protected]> wrote: > On 32-bit ARM the test-ivshmem program got stuck somehow (it had been running > for an hour when I killed it). It didn't write anything to the log, > I'm afraid.
What is happening here is that we are looping infinitely in mktempshmem() because shm_open() returns -1 with errno ENOSYS, and there's no code in the loop that stops the loop on anything except shm_open succeeding, or even prints anything out about shm_open failing. I think this is failing for me because my system's chroot doesn't have /dev/shm mounted. It would be nice if we could at a minimum handle this reasonably gracefully... thanks -- PMM
