[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Qemu-stable] [PATCH 45/56] chardev: fix CHR_EVENT_OPENED events for mux
From: |
Michael Roth |
Subject: |
[Qemu-stable] [PATCH 45/56] chardev: fix CHR_EVENT_OPENED events for mux chardevs |
Date: |
Tue, 13 Aug 2013 10:11:09 -0500 |
As of bd5c51ee6c4f1c79cae5ad2516d711a27b4ea8ec, chardevs no longer use
bottom-halves to issue CHR_EVENT_OPENED events. To maintain past
semantics, we instead defer the CHR_EVENT_OPENED events toward the end
of chardev initialization.
For muxes, this isn't good enough, since a range of FEs must be able
to attach to the mux prior to any CHR_EVENT_OPENED being issued, else
each FE will immediately print it's initial output (prompts, banners,
etc.) just prior to us switching to the next FE as part of
initialization.
The is new and confusing behavior for users, as they'll see output for
things like the HMP monitor, even though their the current mux focus
may be a guest serial port with potentially no output.
We fix this by further deferring CHR_EVENT_OPENED events for FEs
associated with muxes until after machine init by flagging mux chardevs
with 'explicit_be_open', which suppresses emission of CHR_EVENT_OPENED
events until we explicitly set the mux as opened later.
Currently, we must defer till after machine init since we potentially
associate FEs with muxes as part of realize (for instance,
serial_isa_realizefn).
Signed-off-by: Michael Roth <address@hidden>
Message-id: address@hidden
Cc: address@hidden
Signed-off-by: Anthony Liguori <address@hidden>
(cherry picked from commit 7b7ab18d0b9769b5f39e663fa55caed461b1202e)
Signed-off-by: Michael Roth <address@hidden>
---
include/sysemu/char.h | 1 +
qemu-char.c | 50 +++++++++++++++++++++++++++++++++++++++++++++++++
2 files changed, 51 insertions(+)
diff --git a/include/sysemu/char.h b/include/sysemu/char.h
index 066c216..ac2aaaf 100644
--- a/include/sysemu/char.h
+++ b/include/sysemu/char.h
@@ -77,6 +77,7 @@ struct CharDriverState {
int explicit_fe_open;
int explicit_be_open;
int avail_connections;
+ int is_mux;
QemuOpts *opts;
QTAILQ_ENTRY(CharDriverState) next;
};
diff --git a/qemu-char.c b/qemu-char.c
index c16aea1..dbb574d 100644
--- a/qemu-char.c
+++ b/qemu-char.c
@@ -478,6 +478,46 @@ static void mux_chr_update_read_handler(CharDriverState
*chr)
mux_chr_send_event(d, d->focus, CHR_EVENT_MUX_IN);
}
+static bool muxes_realized;
+
+/**
+ * Called after processing of default and command-line-specified
+ * chardevs to deliver CHR_EVENT_OPENED events to any FEs attached
+ * to a mux chardev. This is done here to ensure that
+ * output/prompts/banners are only displayed for the FE that has
+ * focus when initial command-line processing/machine init is
+ * completed.
+ *
+ * After this point, any new FE attached to any new or existing
+ * mux will receive CHR_EVENT_OPENED notifications for the BE
+ * immediately.
+ */
+static void muxes_realize_done(Notifier *notifier, void *unused)
+{
+ CharDriverState *chr;
+
+ QTAILQ_FOREACH(chr, &chardevs, next) {
+ if (chr->is_mux) {
+ MuxDriver *d = chr->opaque;
+ int i;
+
+ /* send OPENED to all already-attached FEs */
+ for (i = 0; i < d->mux_cnt; i++) {
+ mux_chr_send_event(d, i, CHR_EVENT_OPENED);
+ }
+ /* mark mux as OPENED so any new FEs will immediately receive
+ * OPENED event
+ */
+ qemu_chr_be_generic_open(chr);
+ }
+ }
+ muxes_realized = true;
+}
+
+static Notifier muxes_realize_notify = {
+ .notify = muxes_realize_done,
+};
+
static CharDriverState *qemu_chr_open_mux(CharDriverState *drv)
{
CharDriverState *chr;
@@ -494,6 +534,11 @@ static CharDriverState *qemu_chr_open_mux(CharDriverState
*drv)
chr->chr_accept_input = mux_chr_accept_input;
/* Frontend guest-open / -close notification is not support with muxes */
chr->chr_set_fe_open = NULL;
+ /* only default to opened state if we've realized the initial
+ * set of muxes
+ */
+ chr->explicit_be_open = muxes_realized ? 0 : 1;
+ chr->is_mux = 1;
return chr;
}
@@ -3786,6 +3831,11 @@ static void register_types(void)
/* Bug-compatibility: */
register_char_driver_qapi("memory", CHARDEV_BACKEND_KIND_MEMORY,
qemu_chr_parse_ringbuf);
+ /* this must be done after machine init, since we register FEs with muxes
+ * as part of realize functions like serial_isa_realizefn when -nographic
+ * is specified
+ */
+ qemu_add_machine_init_done_notifier(&muxes_realize_notify);
}
type_init(register_types);
--
1.7.9.5
- [Qemu-stable] [PATCH 37/56] gtk: Fix compiler warning (GTK 3 deprecated function), (continued)
- [Qemu-stable] [PATCH 37/56] gtk: Fix compiler warning (GTK 3 deprecated function), Michael Roth, 2013/08/13
- [Qemu-stable] [PATCH 42/56] virtio-console: Use exitfn for virtserialport, too, Michael Roth, 2013/08/13
- [Qemu-stable] [PATCH 38/56] dataplane: refuse to start if device is already in use, Michael Roth, 2013/08/13
- [Qemu-stable] [PATCH 32/56] block: Add return value for bdrv_flush_all(), Michael Roth, 2013/08/13
- [Qemu-stable] [PATCH 39/56] xhci: handle USB_RET_IOERROR, Michael Roth, 2013/08/13
- [Qemu-stable] [PATCH 33/56] cpus: Add return value for vm_stop(), Michael Roth, 2013/08/13
- [Qemu-stable] [PATCH 40/56] qemu-char: Register ring buffer driver with correct name "ringbuf", Michael Roth, 2013/08/13
- [Qemu-stable] [PATCH 41/56] qapi: Rename ChardevBackend member "memory" to "ringbuf", Michael Roth, 2013/08/13
- [Qemu-stable] [PATCH 43/56] pci-bridge: update mappings for migration/restore, Michael Roth, 2013/08/13
- [Qemu-stable] [PATCH 34/56] cpus: Let vm_stop[_force_state]() always flush block devices, Michael Roth, 2013/08/13
- [Qemu-stable] [PATCH 45/56] chardev: fix CHR_EVENT_OPENED events for mux chardevs,
Michael Roth <=
- [Qemu-stable] [PATCH 54/56] dataplane: sync virtio.c and vring.c virtqueue state, Michael Roth, 2013/08/13
- [Qemu-stable] [PATCH 55/56] virtio: clear signalled_used_valid when switching from dataplane, Michael Roth, 2013/08/13
- [Qemu-stable] [PATCH 49/56] target-i386: Fix X86CPU error handling, Michael Roth, 2013/08/13
- [Qemu-stable] [PATCH 47/56] seccomp: add additional asynchronous I/O syscalls, Michael Roth, 2013/08/13
- [Qemu-stable] [PATCH 56/56] vhost: clear signalled_used_valid on vhost stop, Michael Roth, 2013/08/13
- [Qemu-stable] [PATCH 53/56] i82801b11: Fix i82801b11 PCI host bridge config space, Michael Roth, 2013/08/13
- [Qemu-stable] [PATCH 48/56] iov: handle EOF in iov_send_recv, Michael Roth, 2013/08/13
- [Qemu-stable] [PATCH 52/56] Bugfix for loading multiboot kernels, Michael Roth, 2013/08/13
- [Qemu-stable] [PATCH 51/56] semaphore: fix a hangup problem under load on NetBSD hosts., Michael Roth, 2013/08/13
- [Qemu-stable] [PATCH 50/56] ignore SIGPIPE in qemu-img and qemu-io, Michael Roth, 2013/08/13