Re: [Qemu-devel] [PATCH 0/2][RFC?] Remove SIGUSR2 from posix-aio

2011-09-19 Thread Stefan Hajnoczi
On Mon, Sep 19, 2011 at 12:54 PM, Frediano Ziglio wrote: > Now that iothread is always compiled sending a signal seems only an > additional step. This patch also avoid writing to two pipe (one from signal > and one in qemu_service_io). > > Tested and works correctly with KVM enabled. Performances

[Qemu-devel] [PATCH 0/2][RFC?] Remove SIGUSR2 from posix-aio

2011-09-19 Thread Frediano Ziglio
Now that iothread is always compiled sending a signal seems only an additional step. This patch also avoid writing to two pipe (one from signal and one in qemu_service_io). Tested and works correctly with KVM enabled. Performances are only sligthly better (as I expected). strace output is more rea