From: "Dr. David Alan Gilbert" <dgilbert@redhat.com>
To: "Marc-André Lureau" <marcandre.lureau@gmail.com>
Cc: peterx@redhat.com, qemu-devel@nongnu.org, zhengchuan@huawei.com,
dovmurik@linux.vnet.ibm.com, hgcoin@gmail.com,
zhangjiachen.jaycee@bytedance.com, lvivier@redhat.com,
stefanha@redhat.com, vgoyal@redhat.com, jinyan12@huawei.com,
ann.zhuangyanying@huawei.com, quintela@redhat.com
Subject: Re: [PULL 23/26] monitor: Use LOCK_GUARD macros
Date: Tue, 7 Feb 2023 13:41:49 +0000 [thread overview]
Message-ID: <Y+JVHUrYY2rFbBes@work-vm> (raw)
In-Reply-To: <CAJ+F1C+6zPmeTjByKMaDZaFND5qr3Y4i9vJpyFJc2UFkgHWOjQ@mail.gmail.com>
* Marc-André Lureau (marcandre.lureau@gmail.com) wrote:
> Hi David
>
> On Fri, Sep 25, 2020 at 4:27 PM Dr. David Alan Gilbert (git)
> <dgilbert@redhat.com> wrote:
> >
> > From: "Dr. David Alan Gilbert" <dgilbert@redhat.com>
> >
> > Use the lock guard macros in monitor/misc.c - saves
> > a lot of unlocks in error paths, and the occasional goto.
> >
> > Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
> > Message-Id: <20200922095741.101911-1-dgilbert@redhat.com>
> > Reviewed-by: Markus Armbruster <armbru@redhat.com>
> > Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>
> > Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
> > ---
> > monitor/misc.c | 44 ++++++++++++++------------------------------
> > 1 file changed, 14 insertions(+), 30 deletions(-)
> >
> > diff --git a/monitor/misc.c b/monitor/misc.c
> > index 262f2bd951..6e0da0cb96 100644
> > --- a/monitor/misc.c
> > +++ b/monitor/misc.c
> > @@ -141,13 +141,13 @@ char *qmp_human_monitor_command(const char *command_line, bool has_cpu_index,
> > handle_hmp_command(&hmp, command_line);
> > cur_mon = old_mon;
> >
> > - qemu_mutex_lock(&hmp.common.mon_lock);
> > - if (qstring_get_length(hmp.common.outbuf) > 0) {
> > - output = g_strdup(qstring_get_str(hmp.common.outbuf));
> > - } else {
> > - output = g_strdup("");
> > + WITH_QEMU_LOCK_GUARD(&hmp.common.mon_lock) {
> > + if (qstring_get_length(hmp.common.outbuf) > 0) {
> > + output = g_strdup(qstring_get_str(hmp.common.outbuf));
> > + } else {
> > + output = g_strdup("");
> > + }
> > }
> > - qemu_mutex_unlock(&hmp.common.mon_lock);
> >
> > out:
> > monitor_data_destroy(&hmp.common);
> > @@ -1248,7 +1248,7 @@ void qmp_getfd(const char *fdname, Error **errp)
> > return;
> > }
> >
> > - qemu_mutex_lock(&cur_mon->mon_lock);
> > + QEMU_LOCK_GUARD(&cur_mon->mon_lock);
> > QLIST_FOREACH(monfd, &cur_mon->fds, next) {
> > if (strcmp(monfd->name, fdname) != 0) {
> > continue;
> > @@ -1256,7 +1256,6 @@ void qmp_getfd(const char *fdname, Error **errp)
> >
> > tmp_fd = monfd->fd;
> > monfd->fd = fd;
> > - qemu_mutex_unlock(&cur_mon->mon_lock);
> > /* Make sure close() is outside critical section */
> > close(tmp_fd);
>
> The comment doesn't look true after this patch, since the lock is
> released before returning. commit 9409fc05fe2 ("monitor: protect
> mon->fds with mon_lock") introduced the lock & comment. Apparently,
> you didn't convert qmp_closefd() to LOCK_GUARD, presumably because of
> the comment. But you did it in qmp_getfd. What's the reason for that
> though, avoiding syscalls while holding the lock or am I missing
> something?
No, I think you're right!
Yeh I think the reason was to avoid the (potentially blocking) syscall
with the lock held.
Dave
> thanks
>
> > return;
> > @@ -1267,7 +1266,6 @@ void qmp_getfd(const char *fdname, Error **errp)
> > monfd->fd = fd;
> >
> > QLIST_INSERT_HEAD(&cur_mon->fds, monfd, next);
> > - qemu_mutex_unlock(&cur_mon->mon_lock);
> > }
> >
> > void qmp_closefd(const char *fdname, Error **errp)
> > @@ -1299,7 +1297,7 @@ int monitor_get_fd(Monitor *mon, const char *fdname, Error **errp)
> > {
> > mon_fd_t *monfd;
> >
> > - qemu_mutex_lock(&mon->mon_lock);
> > + QEMU_LOCK_GUARD(&mon->mon_lock);
> > QLIST_FOREACH(monfd, &mon->fds, next) {
> > int fd;
> >
> > @@ -1313,12 +1311,10 @@ int monitor_get_fd(Monitor *mon, const char *fdname, Error **errp)
> > QLIST_REMOVE(monfd, next);
> > g_free(monfd->name);
> > g_free(monfd);
> > - qemu_mutex_unlock(&mon->mon_lock);
> >
> > return fd;
> > }
> >
> > - qemu_mutex_unlock(&mon->mon_lock);
> > error_setg(errp, "File descriptor named '%s' has not been found", fdname);
> > return -1;
> > }
> > @@ -1350,11 +1346,10 @@ void monitor_fdsets_cleanup(void)
> > MonFdset *mon_fdset;
> > MonFdset *mon_fdset_next;
> >
> > - qemu_mutex_lock(&mon_fdsets_lock);
> > + QEMU_LOCK_GUARD(&mon_fdsets_lock);
> > QLIST_FOREACH_SAFE(mon_fdset, &mon_fdsets, next, mon_fdset_next) {
> > monitor_fdset_cleanup(mon_fdset);
> > }
> > - qemu_mutex_unlock(&mon_fdsets_lock);
> > }
> >
> > AddfdInfo *qmp_add_fd(bool has_fdset_id, int64_t fdset_id, bool has_opaque,
> > @@ -1389,7 +1384,7 @@ void qmp_remove_fd(int64_t fdset_id, bool has_fd, int64_t fd, Error **errp)
> > MonFdsetFd *mon_fdset_fd;
> > char fd_str[60];
> >
> > - qemu_mutex_lock(&mon_fdsets_lock);
> > + QEMU_LOCK_GUARD(&mon_fdsets_lock);
> > QLIST_FOREACH(mon_fdset, &mon_fdsets, next) {
> > if (mon_fdset->id != fdset_id) {
> > continue;
> > @@ -1409,12 +1404,10 @@ void qmp_remove_fd(int64_t fdset_id, bool has_fd, int64_t fd, Error **errp)
> > goto error;
> > }
> > monitor_fdset_cleanup(mon_fdset);
> > - qemu_mutex_unlock(&mon_fdsets_lock);
> > return;
> > }
> >
> > error:
> > - qemu_mutex_unlock(&mon_fdsets_lock);
> > if (has_fd) {
> > snprintf(fd_str, sizeof(fd_str), "fdset-id:%" PRId64 ", fd:%" PRId64,
> > fdset_id, fd);
> > @@ -1430,7 +1423,7 @@ FdsetInfoList *qmp_query_fdsets(Error **errp)
> > MonFdsetFd *mon_fdset_fd;
> > FdsetInfoList *fdset_list = NULL;
> >
> > - qemu_mutex_lock(&mon_fdsets_lock);
> > + QEMU_LOCK_GUARD(&mon_fdsets_lock);
> > QLIST_FOREACH(mon_fdset, &mon_fdsets, next) {
> > FdsetInfoList *fdset_info = g_malloc0(sizeof(*fdset_info));
> > FdsetFdInfoList *fdsetfd_list = NULL;
> > @@ -1460,7 +1453,6 @@ FdsetInfoList *qmp_query_fdsets(Error **errp)
> > fdset_info->next = fdset_list;
> > fdset_list = fdset_info;
> > }
> > - qemu_mutex_unlock(&mon_fdsets_lock);
> >
> > return fdset_list;
> > }
> > @@ -1554,7 +1546,7 @@ int monitor_fdset_dup_fd_add(int64_t fdset_id, int flags)
> > #else
> > MonFdset *mon_fdset;
> >
> > - qemu_mutex_lock(&mon_fdsets_lock);
> > + QEMU_LOCK_GUARD(&mon_fdsets_lock);
> > QLIST_FOREACH(mon_fdset, &mon_fdsets, next) {
> > MonFdsetFd *mon_fdset_fd;
> > MonFdsetFd *mon_fdset_fd_dup;
> > @@ -1569,7 +1561,6 @@ int monitor_fdset_dup_fd_add(int64_t fdset_id, int flags)
> > QLIST_FOREACH(mon_fdset_fd, &mon_fdset->fds, next) {
> > mon_fd_flags = fcntl(mon_fdset_fd->fd, F_GETFL);
> > if (mon_fd_flags == -1) {
> > - qemu_mutex_unlock(&mon_fdsets_lock);
> > return -1;
> > }
> >
> > @@ -1580,25 +1571,21 @@ int monitor_fdset_dup_fd_add(int64_t fdset_id, int flags)
> > }
> >
> > if (fd == -1) {
> > - qemu_mutex_unlock(&mon_fdsets_lock);
> > errno = EACCES;
> > return -1;
> > }
> >
> > dup_fd = qemu_dup_flags(fd, flags);
> > if (dup_fd == -1) {
> > - qemu_mutex_unlock(&mon_fdsets_lock);
> > return -1;
> > }
> >
> > mon_fdset_fd_dup = g_malloc0(sizeof(*mon_fdset_fd_dup));
> > mon_fdset_fd_dup->fd = dup_fd;
> > QLIST_INSERT_HEAD(&mon_fdset->dup_fds, mon_fdset_fd_dup, next);
> > - qemu_mutex_unlock(&mon_fdsets_lock);
> > return dup_fd;
> > }
> >
> > - qemu_mutex_unlock(&mon_fdsets_lock);
> > errno = ENOENT;
> > return -1;
> > #endif
> > @@ -1609,7 +1596,7 @@ static int64_t monitor_fdset_dup_fd_find_remove(int dup_fd, bool remove)
> > MonFdset *mon_fdset;
> > MonFdsetFd *mon_fdset_fd_dup;
> >
> > - qemu_mutex_lock(&mon_fdsets_lock);
> > + QEMU_LOCK_GUARD(&mon_fdsets_lock);
> > QLIST_FOREACH(mon_fdset, &mon_fdsets, next) {
> > QLIST_FOREACH(mon_fdset_fd_dup, &mon_fdset->dup_fds, next) {
> > if (mon_fdset_fd_dup->fd == dup_fd) {
> > @@ -1619,17 +1606,14 @@ static int64_t monitor_fdset_dup_fd_find_remove(int dup_fd, bool remove)
> > if (QLIST_EMPTY(&mon_fdset->dup_fds)) {
> > monitor_fdset_cleanup(mon_fdset);
> > }
> > - goto err;
> > + return -1;
> > } else {
> > - qemu_mutex_unlock(&mon_fdsets_lock);
> > return mon_fdset->id;
> > }
> > }
> > }
> > }
> >
> > -err:
> > - qemu_mutex_unlock(&mon_fdsets_lock);
> > return -1;
> > }
> >
> > --
> > 2.26.2
> >
> >
>
>
> --
> Marc-André Lureau
>
--
Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK
next prev parent reply other threads:[~2023-02-07 13:42 UTC|newest]
Thread overview: 35+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-09-25 12:06 [PULL 00/26] migration and friends queue Dr. David Alan Gilbert (git)
2020-09-25 12:06 ` [PULL 01/26] migration: Properly destroy variables on incoming side Dr. David Alan Gilbert (git)
2020-09-25 12:06 ` [PULL 02/26] migration: Rework migrate_send_rp_req_pages() function Dr. David Alan Gilbert (git)
2020-09-25 12:06 ` [PULL 03/26] migration/dirtyrate: setup up query-dirtyrate framwork Dr. David Alan Gilbert (git)
2020-09-25 12:06 ` [PULL 04/26] migration/dirtyrate: add DirtyRateStatus to denote calculation status Dr. David Alan Gilbert (git)
2020-09-25 12:06 ` [PULL 05/26] migration/dirtyrate: Add RamblockDirtyInfo to store sampled page info Dr. David Alan Gilbert (git)
2020-09-25 12:06 ` [PULL 06/26] migration/dirtyrate: Add dirtyrate statistics series functions Dr. David Alan Gilbert (git)
2020-09-25 12:06 ` [PULL 07/26] migration/dirtyrate: move RAMBLOCK_FOREACH_MIGRATABLE into ram.h Dr. David Alan Gilbert (git)
2020-09-25 12:06 ` [PULL 08/26] migration/dirtyrate: Record hash results for each sampled page Dr. David Alan Gilbert (git)
2020-09-25 12:06 ` [PULL 09/26] migration/dirtyrate: Compare page hash results for recorded " Dr. David Alan Gilbert (git)
2020-09-25 12:06 ` [PULL 10/26] migration/dirtyrate: skip sampling ramblock with size below MIN_RAMBLOCK_SIZE Dr. David Alan Gilbert (git)
2020-09-25 12:06 ` [PULL 11/26] migration/dirtyrate: Implement set_sample_page_period() and is_sample_period_valid() Dr. David Alan Gilbert (git)
2020-09-25 12:06 ` [PULL 12/26] migration/dirtyrate: Implement calculate_dirtyrate() function Dr. David Alan Gilbert (git)
2020-09-25 12:06 ` [PULL 13/26] migration/dirtyrate: Implement qmp_cal_dirty_rate()/qmp_get_dirty_rate() function Dr. David Alan Gilbert (git)
2020-09-25 12:06 ` [PULL 14/26] migration/dirtyrate: Add trace_calls to make it easier to debug Dr. David Alan Gilbert (git)
2020-09-25 12:06 ` [PULL 15/26] migration: Truncate state file in xen-save-devices-state Dr. David Alan Gilbert (git)
2020-09-25 12:06 ` [PULL 16/26] migration: increase max-bandwidth to 128 MiB/s (1 Gib/s) Dr. David Alan Gilbert (git)
2020-09-25 12:06 ` [PULL 17/26] migration/tls: save hostname into MigrationState Dr. David Alan Gilbert (git)
2020-09-25 12:06 ` [PULL 18/26] migration/tls: extract migration_tls_client_create for common-use Dr. David Alan Gilbert (git)
2020-09-25 12:06 ` [PULL 19/26] migration/tls: add tls_hostname into MultiFDSendParams Dr. David Alan Gilbert (git)
2020-09-25 12:06 ` [PULL 20/26] migration/tls: extract cleanup function for common-use Dr. David Alan Gilbert (git)
2020-09-25 12:06 ` [PULL 21/26] migration/tls: add support for multifd tls-handshake Dr. David Alan Gilbert (git)
2020-09-25 12:06 ` [PULL 22/26] migration/tls: add trace points for multifd-tls Dr. David Alan Gilbert (git)
2020-09-25 12:06 ` [PULL 23/26] monitor: Use LOCK_GUARD macros Dr. David Alan Gilbert (git)
2023-02-07 13:26 ` Marc-André Lureau
2023-02-07 13:41 ` Dr. David Alan Gilbert [this message]
2023-02-07 14:06 ` Marc-André Lureau
2020-09-25 12:06 ` [PULL 24/26] virtiofsd: document cache=auto default Dr. David Alan Gilbert (git)
2020-09-25 12:06 ` [PULL 25/26] virtiofsd: Used glib "shared" thread pool Dr. David Alan Gilbert (git)
2020-09-25 12:06 ` [PULL 26/26] virtiofsd: Add -o allow_direct_io|no_allow_direct_io options Dr. David Alan Gilbert (git)
2020-09-29 21:53 ` Vivek Goyal
2020-09-30 2:14 ` [External] " Jiachen Zhang
2020-09-25 12:35 ` [PULL 00/26] migration and friends queue no-reply
2020-09-25 14:31 ` Dr. David Alan Gilbert
2020-09-25 16:22 ` Peter Maydell
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=Y+JVHUrYY2rFbBes@work-vm \
--to=dgilbert@redhat.com \
--cc=ann.zhuangyanying@huawei.com \
--cc=dovmurik@linux.vnet.ibm.com \
--cc=hgcoin@gmail.com \
--cc=jinyan12@huawei.com \
--cc=lvivier@redhat.com \
--cc=marcandre.lureau@gmail.com \
--cc=peterx@redhat.com \
--cc=qemu-devel@nongnu.org \
--cc=quintela@redhat.com \
--cc=stefanha@redhat.com \
--cc=vgoyal@redhat.com \
--cc=zhangjiachen.jaycee@bytedance.com \
--cc=zhengchuan@huawei.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).