| Message ID | 1349878805-16352-3-git-send-email-coreyb@linux.vnet.ibm.com |
|---|---|
| State | New |
| Headers | show |
On 10/10/2012 08:20 AM, Corey Bryant wrote: > qmp_add_fd() gets an fd that was received over a socket with > SCM_RIGHTS and adds it to an fd set. This patch adds support > that will enable adding an fd that was inherited on the > command line to an fd set. > > This patch also prevents removal of an fd from an fd set during > initialization. This allows the fd to remain in the fd set after > probing of the image file. You're mixing code motion and semantic change in one patch (keyword "also" in your commit message; should the semantic change be moved to a separate patch (possibly squashed into 1/3)? > > Signed-off-by: Corey Bryant <coreyb@linux.vnet.ibm.com> > --- > > - if (has_fdset_id) { > - QLIST_FOREACH(mon_fdset, &mon_fdsets, next) { > - if (mon_fdset->id == fdset_id) { > - break; > - } > - } > - } Since you maintain mon_fdsets in sorted id order, should you optimize this loop to abort the QLIST_FOREACH early if fdset_id is less than mon_fdset->id? [I only noticed this because of code motion, so it is a pre-existing condition and therefore a separate patch, or another thing to squash into 1/3]
Am 10.10.2012 16:20, schrieb Corey Bryant: > qmp_add_fd() gets an fd that was received over a socket with > SCM_RIGHTS and adds it to an fd set. This patch adds support > that will enable adding an fd that was inherited on the > command line to an fd set. > > This patch also prevents removal of an fd from an fd set during > initialization. This allows the fd to remain in the fd set after > probing of the image file. "This patch also..." usually means that it should be split in two patches. Though in this case I'd vote for immediately dropping the second patch again: This makes the probing work with file descriptors using a hack for a certain situation (namely qemu startup) and leaves other cases (like hotplug) broken. I think it's kind of acceptable to leave it broken for both cases in this patch series, you can work around it by passing the 'format' option. If we do fix it, however, we should fix it consistently for all use cases. The best approach for this is probably to avoid opening the image file twice for probing; instead, we should first open the BlockDriverState for raw-posix, use that for probing and then put a second qcow2 or whatever BlockDriverState on top of the very same BDS without reopening it. The rest is code motion, except for: - if (mon_fdset->id == 0) { + if (!mon_fdset_cur) { It's not nice to hide such changes in code motion patches, but the change is obviously not wrong. The commit message should be changed to tell that this is (mostly) code motion. This doesn't really change anything semantically (except for the probing part). Kevin
On 10/10/2012 06:01 PM, Eric Blake wrote: > On 10/10/2012 08:20 AM, Corey Bryant wrote: >> qmp_add_fd() gets an fd that was received over a socket with >> SCM_RIGHTS and adds it to an fd set. This patch adds support >> that will enable adding an fd that was inherited on the >> command line to an fd set. >> >> This patch also prevents removal of an fd from an fd set during >> initialization. This allows the fd to remain in the fd set after >> probing of the image file. > > You're mixing code motion and semantic change in one patch (keyword > "also" in your commit message; should the semantic change be moved to a > separate patch (possibly squashed into 1/3)? > It should probably be another patch. I'll look into this for v3. >> >> Signed-off-by: Corey Bryant <coreyb@linux.vnet.ibm.com> >> --- > >> >> - if (has_fdset_id) { >> - QLIST_FOREACH(mon_fdset, &mon_fdsets, next) { >> - if (mon_fdset->id == fdset_id) { >> - break; >> - } >> - } >> - } > > Since you maintain mon_fdsets in sorted id order, should you optimize > this loop to abort the QLIST_FOREACH early if fdset_id is less than > mon_fdset->id? [I only noticed this because of code motion, so it is a > pre-existing condition and therefore a separate patch, or another thing > to squash into 1/3] > Sure that makes sense to optimize this.
On 10/11/2012 07:25 AM, Kevin Wolf wrote: > Am 10.10.2012 16:20, schrieb Corey Bryant: >> qmp_add_fd() gets an fd that was received over a socket with >> SCM_RIGHTS and adds it to an fd set. This patch adds support >> that will enable adding an fd that was inherited on the >> command line to an fd set. >> >> This patch also prevents removal of an fd from an fd set during >> initialization. This allows the fd to remain in the fd set after >> probing of the image file. > > "This patch also..." usually means that it should be split in two > patches. Though in this case I'd vote for immediately dropping the > second patch again: This makes the probing work with file descriptors > using a hack for a certain situation (namely qemu startup) and leaves > other cases (like hotplug) broken. I don't think hotplug is broken. In that case the fd will only be removed from the fd set if the following is true: (mon_fdset_fd->removed || (QLIST_EMPTY(&mon_fdset->dup_fds) && mon_refcount == 0)) We can ignore the removed part for now. What's important here is that if there are no dup_fd references and there is at least one monitor connected, an fd will *not* be removed. The problem with the command line -add-fd is that there is no equivalent to mon_refcount. So I was using the check for runstate_is_running() as a (somewhat) equivalent to mon_refcount. In other words, if an fd is added to an fd set via the command line and it is not referenced by another command line option (ie. -drive or -blockdev), then clean it up after QEMU initialization. > > I think it's kind of acceptable to leave it broken for both cases in > this patch series, you can work around it by passing the 'format' > option. If we do fix it, however, we should fix it consistently for all > use cases. The best approach for this is probably to avoid opening the > image file twice for probing; instead, we should first open the > BlockDriverState for raw-posix, use that for probing and then put a > second qcow2 or whatever BlockDriverState on top of the very same BDS > without reopening it. > > > The rest is code motion, except for: > > - if (mon_fdset->id == 0) { > + if (!mon_fdset_cur) { > > It's not nice to hide such changes in code motion patches, but the > change is obviously not wrong. Sorry about that. I didn't mean to hide it. This probably makes more sense in patch 1. > > The commit message should be changed to tell that this is (mostly) code > motion. This doesn't really change anything semantically (except for the > probing part). Ok
Am 11.10.2012 17:04, schrieb Corey Bryant: > > > On 10/11/2012 07:25 AM, Kevin Wolf wrote: >> Am 10.10.2012 16:20, schrieb Corey Bryant: >>> qmp_add_fd() gets an fd that was received over a socket with >>> SCM_RIGHTS and adds it to an fd set. This patch adds support >>> that will enable adding an fd that was inherited on the >>> command line to an fd set. >>> >>> This patch also prevents removal of an fd from an fd set during >>> initialization. This allows the fd to remain in the fd set after >>> probing of the image file. >> >> "This patch also..." usually means that it should be split in two >> patches. Though in this case I'd vote for immediately dropping the >> second patch again: This makes the probing work with file descriptors >> using a hack for a certain situation (namely qemu startup) and leaves >> other cases (like hotplug) broken. > > I don't think hotplug is broken. In that case the fd will only be > removed from the fd set if the following is true: > > (mon_fdset_fd->removed || (QLIST_EMPTY(&mon_fdset->dup_fds) && > mon_refcount == 0)) > > We can ignore the removed part for now. What's important here is that > if there are no dup_fd references and there is at least one monitor > connected, an fd will *not* be removed. Ah yes, that's the part I missed. Then your approach of special-casing the command line is probably okay, though I'd still want to change the probing mechanism to avoid the reopen. Seems I need to find a better excuse to make someone do it. Meh. ;-) Kevin
diff --git a/monitor.c b/monitor.c index e53e733..4226249 100644 --- a/monitor.c +++ b/monitor.c @@ -2111,8 +2111,9 @@ static void monitor_fdset_cleanup(MonFdset *mon_fdset) MonFdsetFd *mon_fdset_fd_next; QLIST_FOREACH_SAFE(mon_fdset_fd, &mon_fdset->fds, next, mon_fdset_fd_next) { - if (mon_fdset_fd->removed || - (QLIST_EMPTY(&mon_fdset->dup_fds) && mon_refcount == 0)) { + if ((mon_fdset_fd->removed || + (QLIST_EMPTY(&mon_fdset->dup_fds) && mon_refcount == 0)) && + runstate_is_running()) { close(mon_fdset_fd->fd); g_free(mon_fdset_fd->opaque); QLIST_REMOVE(mon_fdset_fd, next); @@ -2141,9 +2142,6 @@ AddfdInfo *qmp_add_fd(bool has_fdset_id, int64_t fdset_id, bool has_opaque, { int fd; Monitor *mon = cur_mon; - MonFdset *mon_fdset = NULL; - MonFdsetFd *mon_fdset_fd; - AddfdInfo *fdinfo; fd = qemu_chr_fe_get_msgfd(mon->chr); if (fd == -1) { @@ -2151,68 +2149,7 @@ AddfdInfo *qmp_add_fd(bool has_fdset_id, int64_t fdset_id, bool has_opaque, goto error; } - if (has_fdset_id) { - QLIST_FOREACH(mon_fdset, &mon_fdsets, next) { - if (mon_fdset->id == fdset_id) { - break; - } - } - } - - if (mon_fdset == NULL) { - int64_t fdset_id_prev = -1; - MonFdset *mon_fdset_cur = QLIST_FIRST(&mon_fdsets); - - if (has_fdset_id) { - /* Use specified fdset ID */ - QLIST_FOREACH(mon_fdset, &mon_fdsets, next) { - mon_fdset_cur = mon_fdset; - if (fdset_id < mon_fdset_cur->id) { - break; - } - } - } else { - /* Use first available fdset ID */ - QLIST_FOREACH(mon_fdset, &mon_fdsets, next) { - mon_fdset_cur = mon_fdset; - if (fdset_id_prev == mon_fdset_cur->id - 1) { - fdset_id_prev = mon_fdset_cur->id; - continue; - } - break; - } - } - - mon_fdset = g_malloc0(sizeof(*mon_fdset)); - if (has_fdset_id) { - mon_fdset->id = fdset_id; - } else { - mon_fdset->id = fdset_id_prev + 1; - } - - /* The fdset list is ordered by fdset ID */ - if (mon_fdset->id == 0) { - QLIST_INSERT_HEAD(&mon_fdsets, mon_fdset, next); - } else if (mon_fdset->id < mon_fdset_cur->id) { - QLIST_INSERT_BEFORE(mon_fdset_cur, mon_fdset, next); - } else { - QLIST_INSERT_AFTER(mon_fdset_cur, mon_fdset, next); - } - } - - mon_fdset_fd = g_malloc0(sizeof(*mon_fdset_fd)); - mon_fdset_fd->fd = fd; - mon_fdset_fd->removed = false; - if (has_opaque) { - mon_fdset_fd->opaque = g_strdup(opaque); - } - QLIST_INSERT_HEAD(&mon_fdset->fds, mon_fdset_fd, next); - - fdinfo = g_malloc0(sizeof(*fdinfo)); - fdinfo->fdset_id = mon_fdset->id; - fdinfo->fd = mon_fdset_fd->fd; - - return fdinfo; + return monitor_fdset_add_fd(fd, has_fdset_id, fdset_id, has_opaque, opaque); error: if (fd != -1) { @@ -2298,6 +2235,77 @@ FdsetInfoList *qmp_query_fdsets(Error **errp) return fdset_list; } +AddfdInfo *monitor_fdset_add_fd(int fd, bool has_fdset_id, int64_t fdset_id, + bool has_opaque, const char *opaque) +{ + MonFdset *mon_fdset = NULL; + MonFdsetFd *mon_fdset_fd; + AddfdInfo *fdinfo; + + if (has_fdset_id) { + QLIST_FOREACH(mon_fdset, &mon_fdsets, next) { + if (mon_fdset->id == fdset_id) { + break; + } + } + } + + if (mon_fdset == NULL) { + int64_t fdset_id_prev = -1; + MonFdset *mon_fdset_cur = QLIST_FIRST(&mon_fdsets); + + if (has_fdset_id) { + /* Use specified fdset ID */ + QLIST_FOREACH(mon_fdset, &mon_fdsets, next) { + mon_fdset_cur = mon_fdset; + if (fdset_id < mon_fdset_cur->id) { + break; + } + } + } else { + /* Use first available fdset ID */ + QLIST_FOREACH(mon_fdset, &mon_fdsets, next) { + mon_fdset_cur = mon_fdset; + if (fdset_id_prev == mon_fdset_cur->id - 1) { + fdset_id_prev = mon_fdset_cur->id; + continue; + } + break; + } + } + + mon_fdset = g_malloc0(sizeof(*mon_fdset)); + if (has_fdset_id) { + mon_fdset->id = fdset_id; + } else { + mon_fdset->id = fdset_id_prev + 1; + } + + /* The fdset list is ordered by fdset ID */ + if (!mon_fdset_cur) { + QLIST_INSERT_HEAD(&mon_fdsets, mon_fdset, next); + } else if (mon_fdset->id < mon_fdset_cur->id) { + QLIST_INSERT_BEFORE(mon_fdset_cur, mon_fdset, next); + } else { + QLIST_INSERT_AFTER(mon_fdset_cur, mon_fdset, next); + } + } + + mon_fdset_fd = g_malloc0(sizeof(*mon_fdset_fd)); + mon_fdset_fd->fd = fd; + mon_fdset_fd->removed = false; + if (has_opaque) { + mon_fdset_fd->opaque = g_strdup(opaque); + } + QLIST_INSERT_HEAD(&mon_fdset->fds, mon_fdset_fd, next); + + fdinfo = g_malloc0(sizeof(*fdinfo)); + fdinfo->fdset_id = mon_fdset->id; + fdinfo->fd = mon_fdset_fd->fd; + + return fdinfo; +} + int monitor_fdset_get_fd(int64_t fdset_id, int flags) { #ifndef _WIN32 diff --git a/monitor.h b/monitor.h index b6e7d95..78133d2 100644 --- a/monitor.h +++ b/monitor.h @@ -90,6 +90,8 @@ int qmp_qom_set(Monitor *mon, const QDict *qdict, QObject **ret); int qmp_qom_get(Monitor *mon, const QDict *qdict, QObject **ret); +AddfdInfo *monitor_fdset_add_fd(int fd, bool has_fdset_id, int64_t fdset_id, + bool has_opaque, const char *opaque); int monitor_fdset_get_fd(int64_t fdset_id, int flags); int monitor_fdset_dup_fd_add(int64_t fdset_id, int dup_fd); int monitor_fdset_dup_fd_remove(int dup_fd);
qmp_add_fd() gets an fd that was received over a socket with SCM_RIGHTS and adds it to an fd set. This patch adds support that will enable adding an fd that was inherited on the command line to an fd set. This patch also prevents removal of an fd from an fd set during initialization. This allows the fd to remain in the fd set after probing of the image file. Signed-off-by: Corey Bryant <coreyb@linux.vnet.ibm.com> --- v2: - Removed Error** parameter from monitor_fdset_add_fd() monitor.c | 142 +++++++++++++++++++++++++++++++++----------------------------- monitor.h | 2 + 2 files changed, 77 insertions(+), 67 deletions(-)