From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:43674) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1fuV3b-0005As-US for qemu-devel@nongnu.org; Mon, 27 Aug 2018 23:52:45 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1fuV3Y-0006My-E5 for qemu-devel@nongnu.org; Mon, 27 Aug 2018 23:52:43 -0400 Received: from mx3-rdu2.redhat.com ([66.187.233.73]:57784 helo=mx1.redhat.com) by eggs.gnu.org with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.71) (envelope-from ) id 1fuV3Y-0006Mi-7s for qemu-devel@nongnu.org; Mon, 27 Aug 2018 23:52:40 -0400 Received: from smtp.corp.redhat.com (int-mx05.intmail.prod.int.rdu2.redhat.com [10.11.54.5]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mx1.redhat.com (Postfix) with ESMTPS id 8B81940241EE for ; Tue, 28 Aug 2018 03:52:39 +0000 (UTC) Date: Tue, 28 Aug 2018 11:52:32 +0800 From: Peter Xu Message-ID: <20180828035232.GD4446@xz-x1> References: <20180827222322.26009-1-marcandre.lureau@redhat.com> <20180827222322.26009-7-marcandre.lureau@redhat.com> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Disposition: inline In-Reply-To: <20180827222322.26009-7-marcandre.lureau@redhat.com> Content-Transfer-Encoding: quoted-printable Subject: Re: [Qemu-devel] [PATCH 6/9] terminal3270: do not use backend timer sources List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: =?utf-8?Q?Marc-Andr=C3=A9?= Lureau Cc: qemu-devel@nongnu.org, pbonzini@redhat.com, berrange@redhat.com, armbru@redhat.com On Tue, Aug 28, 2018 at 12:23:19AM +0200, Marc-Andr=C3=A9 Lureau wrote: > terminal3270 is uses the front-end side of the chardev. It shouldn't > create sources from backend side context. Fwiw, send_timing_mark_cb > calls qemu_chr_fe_write_all() which should be thread safe. >=20 > This partially reverts changes from commit > 2c716ba1506769c9be2caa02f0f6d6e7c00f4304. I don't fully understand too on why "It shouldn't create sources from backend side context". Could you explain a bit? If you don't want the backend gcontext to be exposed to the frontend, IMHO the simplest thing is just to use the NULL gcontext in qemu_chr_timeout_add_ms() which is a one-liner change. I don't see much point on re-using the GSource tags since IMHO GSource pointer is always superior to the old tag system. >=20 > CC: Peter Xu > Signed-off-by: Marc-Andr=C3=A9 Lureau > --- > hw/char/terminal3270.c | 15 ++++++--------- > 1 file changed, 6 insertions(+), 9 deletions(-) >=20 > diff --git a/hw/char/terminal3270.c b/hw/char/terminal3270.c > index e9c45e55b1..35b079d5c4 100644 > --- a/hw/char/terminal3270.c > +++ b/hw/char/terminal3270.c > @@ -31,7 +31,7 @@ typedef struct Terminal3270 { > uint8_t outv[OUTPUT_BUFFER_SIZE]; > int in_len; > bool handshake_done; > - GSource *timer_src; > + guint timer_tag; > } Terminal3270; > =20 > #define TYPE_TERMINAL_3270 "x-terminal3270" > @@ -47,10 +47,9 @@ static int terminal_can_read(void *opaque) > =20 > static void terminal_timer_cancel(Terminal3270 *t) > { > - if (t->timer_src) { > - g_source_destroy(t->timer_src); > - g_source_unref(t->timer_src); > - t->timer_src =3D NULL; > + if (t->timer_tag) { > + g_source_remove(t->timer_tag); > + t->timer_tag =3D 0; > } > } > =20 > @@ -100,8 +99,7 @@ static void terminal_read(void *opaque, const uint8_= t *buf, int size) > assert(size <=3D (INPUT_BUFFER_SIZE - t->in_len)); > =20 > terminal_timer_cancel(t); > - t->timer_src =3D qemu_chr_timeout_add_ms(t->chr.chr, 600 * 1000, > - send_timing_mark_cb, t); > + t->timer_tag =3D g_timeout_add_seconds(600, send_timing_mark_cb, t= ); > memcpy(&t->inv[t->in_len], buf, size); > t->in_len +=3D size; > if (t->in_len < 2) { > @@ -160,8 +158,7 @@ static void chr_event(void *opaque, int event) > * char-socket.c. Once qemu receives the terminal-type of the > * client, mark handshake done and trigger everything rolling = again. > */ > - t->timer_src =3D qemu_chr_timeout_add_ms(t->chr.chr, 600 * 100= 0, > - send_timing_mark_cb, t)= ; > + t->timer_tag =3D g_timeout_add_seconds(600, send_timing_mark_c= b, t); > break; > case CHR_EVENT_CLOSED: > sch->curr_status.scsw.dstat =3D SCSW_DSTAT_DEVICE_END; > --=20 > 2.18.0.547.g1d89318c48 >=20 Regards, --=20 Peter Xu