v2: - added comments about the monitor lock being used to protect the file descriptor 'fd' - eofcb is not set anywhere, so also removing it; the if statement checking on it seems dead code I am replacing the last instances of close() I found with VIR_CLOSE() / VIR_FORCE_CLOSE respectively. The first part patches virsh, which I missed out on previously. The 2nd patch I had left out intentionally to look at it more carefully: The 'closed' variable could be easily removed since it wasn't used anywhere else. The possible race condition that could result from the filedescriptor being closed and not set to -1 (and possibly let us write into 'something' totally different if the fd was allocated by another thread) seems to be prevented by the qemuMonitorLock() already placed around the code that reads from or writes to the fd. So the change of this code as shown in the patch should not have any side-effects. Signed-off-by: Stefan Berger&lt;stefanb(a)us.ibm.com&gt; --- src/qemu/qemu_monitor.c | 32 +++++++++++--------------------- tools/virsh.c | 4 ++-- 2 files changed, 13 insertions(+), 23 deletions(-) Index: libvirt-acl/tools/virsh.c =================================================================== --- libvirt-acl.orig/tools/virsh.c +++ libvirt-acl/tools/virsh.c @@ -8994,12 +8994,12 @@ editWriteToTempFile (vshControl *ctl, co if (safewrite (fd, doc, strlen (doc)) == -1) { vshError(ctl, _("write: %s: failed to write to temporary file: %s"), ret, strerror(errno)); - close (fd); + VIR_FORCE_CLOSE(fd); unlink (ret); VIR_FREE(ret); return NULL; } - if (close (fd) == -1) { + if (VIR_CLOSE(fd)< 0) { vshError(ctl, _("close: %s: failed to write or close temporary file: %s"), ret, strerror(errno)); unlink (ret); Index: libvirt-acl/src/qemu/qemu_monitor.c =================================================================== --- libvirt-acl.orig/src/qemu/qemu_monitor.c +++ libvirt-acl/src/qemu/qemu_monitor.c @@ -44,7 +44,7 @@ #define DEBUG_RAW_IO 0 struct _qemuMonitor { - virMutex lock; + virMutex lock; /* also used to protect fd */ virCond notify; int refs; @@ -71,11 +71,6 @@ struct _qemuMonitor { * the next monitor msg */ int lastErrno; - /* If the monitor EOF callback is currently active (stops more commands being run) */ - unsigned eofcb: 1; - /* If the monitor is in process of shutting down */ - unsigned closed: 1; - unsigned json: 1; }; @@ -356,6 +351,7 @@ qemuMonitorIOProcess(qemuMonitorPtr mon) } +/* Call this function while holding the monitor lock. */ static int qemuMonitorIOWriteWithFD(qemuMonitorPtr mon, const char *data, @@ -400,7 +396,10 @@ qemuMonitorIOWriteWithFD(qemuMonitorPtr return ret; } -/* Called when the monitor is able to write data */ +/* + * Called when the monitor is able to write data + * Call this function while holding the monitor lock. + */ static int qemuMonitorIOWrite(qemuMonitorPtr mon) { @@ -433,6 +432,7 @@ qemuMonitorIOWrite(qemuMonitorPtr mon) /* * Called when the monitor has incoming data to read + * Call this function while holding the monitor lock. * * Returns -1 on error, or number of bytes read */ @@ -505,6 +505,7 @@ qemuMonitorIO(int watch, int fd, int eve qemuMonitorPtr mon = opaque; int quit = 0, failed = 0; + /* lock access to the monitor and protect fd */ qemuMonitorLock(mon); qemuMonitorRef(mon); #if DEBUG_IO @@ -692,17 +693,11 @@ void qemuMonitorClose(qemuMonitorPtr mon VIR_DEBUG("mon=%p", mon); qemuMonitorLock(mon); - if (!mon->closed) { + + if (mon->fd>= 0) { if (mon->watch) virEventRemoveHandle(mon->watch); - if (mon->fd != -1) - close(mon->fd); - /* NB: ordinarily one might immediately set mon->watch to -1 - * and mon->fd to -1, but there may be a callback active - * that is still relying on these fields being valid. So - * we merely close them, but not clear their values and - * use this explicit 'closed' flag to track this state */ - mon->closed = 1; + VIR_FORCE_CLOSE(mon->fd); } if (qemuMonitorUnref(mon)> 0) @@ -715,11 +710,6 @@ int qemuMonitorSend(qemuMonitorPtr mon, { int ret = -1; - if (mon->eofcb) { - msg->lastErrno = EIO; - return -1; - } - mon->msg = msg; qemuMonitorUpdateWatch(mon);