From: Luiz Capitulino <lcapitulino@redhat.com>
To: Paolo Bonzini <pbonzini@redhat.com>
Cc: kwolf@redhat.com, famz@redhat.com, qemu-devel@nongnu.org,
stefanha@redhat.com, kraxel@redhat.com
Subject: Re: [Qemu-devel] [PATCH 5/6] monitor: protect outbuf with mutex
Date: Tue, 10 Jun 2014 10:10:18 -0400 [thread overview]
Message-ID: <20140610101018.34c62092@redhat.com> (raw)
In-Reply-To: <1401813551-6667-6-git-send-email-pbonzini@redhat.com>
On Tue, 3 Jun 2014 18:39:10 +0200
Paolo Bonzini <pbonzini@redhat.com> wrote:
> This lets the block layer emit QMP events from outside the I/O thread.
To be honest I'm starting to forget monitor code, but this looks correct
to me. I have only one comment below that doesn't prevent me from adding:
Reviewed-by: Luiz Capitulino <lcapitulino@redhat.com>
>
> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
> ---
> monitor.c | 35 ++++++++++++++++++++++++++++-------
> 1 file changed, 28 insertions(+), 7 deletions(-)
>
> diff --git a/monitor.c b/monitor.c
> index 342e83b..ebc66fb 100644
> --- a/monitor.c
> +++ b/monitor.c
> @@ -191,8 +191,11 @@ struct Monitor {
> int flags;
> int suspend_cnt;
> bool skip_flush;
> +
> + QemuMutex out_lock;
> QString *outbuf;
> - guint watch;
> + guint out_watch;
> +
> ReadLineState *rs;
> MonitorControl *mc;
> CPUState *mon_cpu;
> @@ -265,17 +268,22 @@ int monitor_read_password(Monitor *mon, ReadLineFunc *readline_func,
> }
> }
>
> +static void do_monitor_flush(Monitor *mon);
> +
> static gboolean monitor_unblocked(GIOChannel *chan, GIOCondition cond,
> void *opaque)
> {
> Monitor *mon = opaque;
>
> - mon->watch = 0;
> + qemu_mutex_lock(&mon->out_lock);
> + mon->out_watch = 0;
> monitor_flush(mon);
> + qemu_mutex_unlock(&mon->out_lock);
> return FALSE;
> }
>
> -void monitor_flush(Monitor *mon)
> +/* Called with mon->out_lock held. */
> +static void do_monitor_flush(Monitor *mon)
> {
> int rc;
> size_t len;
> @@ -302,18 +310,26 @@ void monitor_flush(Monitor *mon)
> QDECREF(mon->outbuf);
> mon->outbuf = tmp;
> }
> - if (mon->watch == 0) {
> - mon->watch = qemu_chr_fe_add_watch(mon->chr, G_IO_OUT,
> - monitor_unblocked, mon);
> + if (mon->out_watch == 0) {
> + mon->out_watch = qemu_chr_fe_add_watch(mon->chr, G_IO_OUT,
> + monitor_unblocked, mon);
> }
> }
> }
>
> +void monitor_flush(Monitor *mon)
> +{
> + qemu_mutex_lock(&mon->out_lock);
> + do_monitor_flush(mon);
> + qemu_mutex_unlock(&mon->out_lock);
> +}
> +
> /* flush at every end of line */
> static void monitor_puts(Monitor *mon, const char *str)
> {
> char c;
>
> + qemu_mutex_lock(&mon->out_lock);
> for(;;) {
> c = *str++;
> if (c == '\0')
> @@ -323,9 +339,10 @@ static void monitor_puts(Monitor *mon, const char *str)
> }
> qstring_append_chr(mon->outbuf, c);
> if (c == '\n') {
> - monitor_flush(mon);
> + do_monitor_flush(mon);
> }
> }
> + qemu_mutex_unlock(&mon->out_lock);
> }
>
> void monitor_vprintf(Monitor *mon, const char *fmt, va_list ap)
> @@ -690,6 +707,7 @@ static void handle_user_command(Monitor *mon, const char *cmdline);
> static void monitor_data_init(Monitor *mon)
> {
> memset(mon, 0, sizeof(Monitor));
> + qemu_mutex_init(&mon->out_lock);
> mon->outbuf = qstring_new();
> /* Use *mon_cmds by default. */
> mon->cmd_table = mon_cmds;
> @@ -698,6 +716,7 @@ static void monitor_data_init(Monitor *mon)
> static void monitor_data_destroy(Monitor *mon)
> {
> QDECREF(mon->outbuf);
> + qemu_mutex_destroy(&mon->out_lock);
> }
>
> char *qmp_human_monitor_command(const char *command_line, bool has_cpu_index,
> @@ -725,11 +744,13 @@ char *qmp_human_monitor_command(const char *command_line, bool has_cpu_index,
> handle_user_command(&hmp, command_line);
> cur_mon = old_mon;
>
> + qemu_mutex_lock(&hmp.out_lock);
> if (qstring_get_length(hmp.outbuf) > 0) {
> output = g_strdup(qstring_get_str(hmp.outbuf));
> } else {
> output = g_strdup("");
> }
> + qemu_mutex_unlock(&hmp.out_lock);
Are you sure we need to lock/unlock in this function? hmp is allocated
in the stack.
>
> out:
> monitor_data_destroy(&hmp);
next prev parent reply other threads:[~2014-06-10 14:10 UTC|newest]
Thread overview: 18+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-06-03 16:39 [Qemu-devel] [PATCH 0/5] qemu-char/monitor: make monitor_puts thread safe Paolo Bonzini
2014-06-03 16:39 ` [Qemu-devel] [PATCH 1/6] qemu-char: introduce qemu_chr_alloc Paolo Bonzini
2014-06-11 6:28 ` Fam Zheng
2014-06-03 16:39 ` [Qemu-devel] [PATCH 2/6] qemu-char: do not call chr_write directly Paolo Bonzini
2014-06-11 6:30 ` Fam Zheng
2014-06-03 16:39 ` [Qemu-devel] [PATCH 3/6] qemu-char: move pty_chr_update_read_handler around Paolo Bonzini
2014-06-11 6:32 ` Fam Zheng
2014-06-03 16:39 ` [Qemu-devel] [PATCH 4/6] qemu-char: make writes thread-safe Paolo Bonzini
2014-06-11 6:59 ` Fam Zheng
2014-06-11 8:16 ` Paolo Bonzini
2014-06-03 16:39 ` [Qemu-devel] [PATCH 5/6] monitor: protect outbuf with mutex Paolo Bonzini
2014-06-10 14:10 ` Luiz Capitulino [this message]
2014-06-10 14:24 ` Paolo Bonzini
2014-06-10 14:28 ` Luiz Capitulino
2014-06-03 16:39 ` [Qemu-devel] [PATCH 6/6] monitor: protect event emission Paolo Bonzini
2014-06-10 13:33 ` Luiz Capitulino
2014-06-27 9:43 ` [Qemu-devel] [PATCH 0/5] qemu-char/monitor: make monitor_puts thread safe Stefan Hajnoczi
2014-06-27 12:33 ` Luiz Capitulino
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=20140610101018.34c62092@redhat.com \
--to=lcapitulino@redhat.com \
--cc=famz@redhat.com \
--cc=kraxel@redhat.com \
--cc=kwolf@redhat.com \
--cc=pbonzini@redhat.com \
--cc=qemu-devel@nongnu.org \
--cc=stefanha@redhat.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.