From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:59056) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1WIF5d-0001Gr-FH for qemu-devel@nongnu.org; Tue, 25 Feb 2014 05:18:21 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1WIF5Z-0001Co-CX for qemu-devel@nongnu.org; Tue, 25 Feb 2014 05:18:17 -0500 Received: from mx3-phx2.redhat.com ([209.132.183.24]:47341) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1WIF5Y-0001Cf-TF for qemu-devel@nongnu.org; Tue, 25 Feb 2014 05:18:13 -0500 Date: Tue, 25 Feb 2014 05:18:00 -0500 (EST) From: Gal Hammer Message-ID: <1499128156.9265161.1393323480208.JavaMail.zimbra@redhat.com> In-Reply-To: <530B309F.8050906@redhat.com> References: <1393237002-16980-1-git-send-email-ghammer@redhat.com> <1393237002-16980-3-git-send-email-ghammer@redhat.com> <530B1FC9.3080507@redhat.com> <530B309F.8050906@redhat.com> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 7bit Subject: Re: [Qemu-devel] [PATCH 2/2] qtest: fix a "!chr->fd_in_tag" assertion error in qtest. Reply-To: Gal Hammer List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Paolo Bonzini , qemu-devel@nongnu.org Cc: amit shah , peter maydell , anthony@codemonkey.ws Hi, After reading the code again. It make sense to leave the call to remove_fd_in_watch() because I changed the code so the chr_update_read_handler function is called only when the front end is opened. Submitting a new version of the patch. Gal. ----- Original Message ----- From: "Gal Hammer" To: "Paolo Bonzini" , qemu-devel@nongnu.org Cc: "amit shah" , "peter maydell" , anthony@codemonkey.ws Sent: Monday, February 24, 2014 1:44:31 PM Subject: Re: [Qemu-devel] [PATCH 2/2] qtest: fix a "!chr->fd_in_tag" assertion error in qtest. On 24/02/2014 12:32, Paolo Bonzini wrote: > Il 24/02/2014 11:16, Gal Hammer ha scritto: >> Replacement of the default chardev handlers now requires a call to >> release the current handlers. >> >> Signed-off-by: Gal Hammer >> --- >> monitor.c | 1 + >> qtest.c | 2 ++ >> 2 files changed, 3 insertions(+) >> >> diff --git a/monitor.c b/monitor.c >> index de90fba..db52e7f 100644 >> --- a/monitor.c >> +++ b/monitor.c >> @@ -5024,6 +5024,7 @@ void monitor_init(CharDriverState *chr, int flags) >> if (monitor_ctrl_mode(mon)) { >> mon->mc = g_malloc0(sizeof(MonitorControl)); >> /* Control mode requires special handlers */ >> + qemu_chr_add_handlers(chr, NULL, NULL, NULL, NULL); >> qemu_chr_add_handlers(chr, monitor_can_read, >> monitor_control_read, >> monitor_control_event, mon); >> qemu_chr_fe_set_echo(chr, true); >> diff --git a/qtest.c b/qtest.c >> index ae941d6..a5682ee 100644 >> --- a/qtest.c >> +++ b/qtest.c >> @@ -519,6 +519,8 @@ void qtest_init(const char *qtest_chrdev, const >> char *qtest_log, Error **errp) >> return; >> } >> >> + /* Replace the default tcp's handlers with qtest's handlers. */ >> + qemu_chr_add_handlers(chr, NULL, NULL, NULL, NULL); >> qemu_chr_add_handlers(chr, qtest_can_read, qtest_read, >> qtest_event, chr); >> qemu_chr_fe_set_echo(chr, true); >> >> > > Why can't qemu_chr_add_handlers call remove_fd_in_watch when handlers > are replaced, before setting the new ones? > > Paolo It used to be like that (see patch 1/2 for the change). However, I removed it after I saw that the function remove_fd_in_watch() is always called when setting all the handlers to NULL (qemu-char.c +3872) so a second call in every chr_update_read_handler callback seems redundant. As far as I can tell only the qtest change the read handlers and not using the default ones. Gal.