From: Pavel Dovgalyuk <Pavel.Dovgaluk@ispras.ru>
To: qemu-devel@nongnu.org
Cc: kwolf@redhat.com, peter.maydell@linaro.org, war2jordan@live.com,
crosthwaite.peter@gmail.com, boost.lists@gmail.com,
quintela@redhat.com, ciro.santilli@gmail.com,
jasowang@redhat.com, mst@redhat.com, zuban32s@gmail.com,
armbru@redhat.com, maria.klimushenkova@ispras.ru,
dovgaluk@ispras.ru, kraxel@redhat.com, pavel.dovgaluk@ispras.ru,
thomas.dullien@googlemail.com, pbonzini@redhat.com,
mreitz@redhat.com, alex.bennee@linaro.org, dgilbert@redhat.com,
rth@twiddle.net
Subject: [Qemu-devel] [PATCH v6 20/25] replay: wake up vCPU when replaying
Date: Wed, 12 Sep 2018 11:19:45 +0300 [thread overview]
Message-ID: <20180912081945.3228.19776.stgit@pasha-VirtualBox> (raw)
In-Reply-To: <20180912081747.3228.21861.stgit@pasha-VirtualBox>
In record/replay icount mode vCPU thread and iothread synchronize
the execution using the checkpoints.
vCPU thread processes the virtual timers and iothread processes all others.
When iothread wants to wake up sleeping vCPU thread, it sends dummy queued
work. Therefore it could be the following sequence of the events in
record mode:
- IO: sending dummy work
- IO: processing timers
- CPU: wakeup
- CPU: clearing dummy work
- CPU: processing virtual timers
But due to the races in replay mode the sequence may change:
- IO: sending dummy work
- CPU: wakeup
- CPU: clearing dummy work
- CPU: sleeping again because nothing to do
- IO: Processing timers
- CPU: zzzz
In this case vCPU will not wake up, because dummy work is not to be set up
again.
This patch tries to wake up the vCPU when it sleeps and the icount warp
checkpoint isn't met. It means that vCPU has something to do, because
there are no other reasons of non-matching warp checkpoint.
Signed-off-by: Pavel Dovgalyuk <Pavel.Dovgaluk@ispras.ru>
--
v5: improve checking that vCPU is still sleeping
---
cpus.c | 31 +++++++++++++++++++++----------
include/sysemu/replay.h | 3 +++
replay/replay.c | 12 ++++++++++++
3 files changed, 36 insertions(+), 10 deletions(-)
diff --git a/cpus.c b/cpus.c
index f31a70a..b12a02f 100644
--- a/cpus.c
+++ b/cpus.c
@@ -576,18 +576,29 @@ void qemu_start_warp_timer(void)
return;
}
- /* warp clock deterministically in record/replay mode */
- if (!replay_checkpoint(CHECKPOINT_CLOCK_WARP_START)) {
- return;
- }
+ if (replay_mode != REPLAY_MODE_PLAY) {
+ if (!all_cpu_threads_idle()) {
+ return;
+ }
- if (!all_cpu_threads_idle()) {
- return;
- }
+ if (qtest_enabled()) {
+ /* When testing, qtest commands advance icount. */
+ return;
+ }
- if (qtest_enabled()) {
- /* When testing, qtest commands advance icount. */
- return;
+ replay_checkpoint(CHECKPOINT_CLOCK_WARP_START);
+ } else {
+ /* warp clock deterministically in record/replay mode */
+ if (!replay_checkpoint(CHECKPOINT_CLOCK_WARP_START)) {
+ /* vCPU is sleeping and warp can't be started.
+ It is probably a race condition: notification sent
+ to vCPU was processed in advance and vCPU went to sleep.
+ Therefore we have to wake it up for doing someting. */
+ if (replay_has_checkpoint()) {
+ qemu_clock_notify(QEMU_CLOCK_VIRTUAL);
+ }
+ return;
+ }
}
/* We want to use the earliest deadline from ALL vm_clocks */
diff --git a/include/sysemu/replay.h b/include/sysemu/replay.h
index c1646b0..3130e45 100644
--- a/include/sysemu/replay.h
+++ b/include/sysemu/replay.h
@@ -144,6 +144,9 @@ void replay_shutdown_request(ShutdownCause cause);
Returns 0 in PLAY mode if checkpoint was not found.
Returns 1 in all other cases. */
bool replay_checkpoint(ReplayCheckpoint checkpoint);
+/*! Used to determine that checkpoint is pending.
+ Does not proceed to the next event in the log. */
+bool replay_has_checkpoint(void);
/* Asynchronous events queue */
diff --git a/replay/replay.c b/replay/replay.c
index a106211..2145686 100644
--- a/replay/replay.c
+++ b/replay/replay.c
@@ -241,6 +241,18 @@ out:
return res;
}
+bool replay_has_checkpoint(void)
+{
+ bool res = false;
+ if (replay_mode == REPLAY_MODE_PLAY) {
+ g_assert(replay_mutex_locked());
+ replay_account_executed_instructions();
+ res = EVENT_CHECKPOINT <= replay_state.data_kind
+ && replay_state.data_kind <= EVENT_CHECKPOINT_LAST;
+ }
+ return res;
+}
+
static void replay_enable(const char *fname, int mode)
{
const char *fmode = NULL;
next prev parent reply other threads:[~2018-09-12 8:19 UTC|newest]
Thread overview: 43+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-09-12 8:17 [Qemu-devel] [PATCH v6 00/25] Fixing record/replay and adding reverse debugging Pavel Dovgalyuk
2018-09-12 8:17 ` [Qemu-devel] [PATCH v6 01/25] block: implement bdrv_snapshot_goto for blkreplay Pavel Dovgalyuk
2018-09-12 8:18 ` [Qemu-devel] [PATCH v6 02/25] replay: disable default snapshot for record/replay Pavel Dovgalyuk
2018-09-12 8:18 ` [Qemu-devel] [PATCH v6 03/25] replay: update docs for record/replay with block devices Pavel Dovgalyuk
2018-09-12 8:18 ` [Qemu-devel] [PATCH v6 04/25] replay: don't drain/flush bdrv queue while RR is working Pavel Dovgalyuk
2018-09-12 8:18 ` [Qemu-devel] [PATCH v6 05/25] replay: finish record/replay before closing the disks Pavel Dovgalyuk
2018-09-12 8:18 ` [Qemu-devel] [PATCH v6 06/25] qcow2: introduce icount field for snapshots Pavel Dovgalyuk
2018-09-12 8:18 ` [Qemu-devel] [PATCH v6 07/25] migration: " Pavel Dovgalyuk
2018-09-12 8:18 ` [Qemu-devel] [PATCH v6 08/25] replay: provide and accessor for rr filename Pavel Dovgalyuk
2018-09-12 8:18 ` [Qemu-devel] [PATCH v6 09/25] replay: introduce info hmp/qmp command Pavel Dovgalyuk
2018-09-12 8:18 ` [Qemu-devel] [PATCH v6 10/25] replay: introduce breakpoint at the specified step Pavel Dovgalyuk
2018-09-12 8:18 ` [Qemu-devel] [PATCH v6 11/25] replay: implement replay-seek command to proceed to the desired step Pavel Dovgalyuk
2018-09-12 8:18 ` [Qemu-devel] [PATCH v6 12/25] replay: flush events when exiting Pavel Dovgalyuk
2018-09-12 8:19 ` [Qemu-devel] [PATCH v6 13/25] replay: refine replay-time module Pavel Dovgalyuk
2018-09-12 8:19 ` [Qemu-devel] [PATCH v6 14/25] translator: fix breakpoint processing Pavel Dovgalyuk
2018-09-12 8:19 ` [Qemu-devel] [PATCH v6 15/25] replay: flush rr queue before loading the vmstate Pavel Dovgalyuk
2018-09-12 8:19 ` [Qemu-devel] [PATCH v6 16/25] gdbstub: add reverse step support in replay mode Pavel Dovgalyuk
2018-09-12 8:19 ` [Qemu-devel] [PATCH v6 17/25] gdbstub: add reverse continue " Pavel Dovgalyuk
2018-09-12 8:19 ` [Qemu-devel] [PATCH v6 18/25] replay: describe reverse debugging in docs/replay.txt Pavel Dovgalyuk
2018-09-12 8:19 ` [Qemu-devel] [PATCH v6 19/25] replay: allow loading any snapshots before recording Pavel Dovgalyuk
2018-09-12 8:19 ` Pavel Dovgalyuk [this message]
2018-09-13 10:12 ` [Qemu-devel] [PATCH v6 20/25] replay: wake up vCPU when replaying Paolo Bonzini
2018-09-13 11:06 ` Pavel Dovgalyuk
2018-09-12 8:19 ` [Qemu-devel] [PATCH v6 21/25] replay: replay BH for IDE trim operation Pavel Dovgalyuk
2018-09-13 17:05 ` John Snow
2018-09-14 5:48 ` Pavel Dovgalyuk
2018-09-14 7:27 ` Pavel Dovgalyuk
2018-09-14 14:17 ` John Snow
2018-09-14 14:00 ` John Snow
2018-09-14 15:19 ` Paolo Bonzini
2018-09-14 15:59 ` Peter Maydell
2018-09-12 8:19 ` [Qemu-devel] [PATCH v6 22/25] replay: add BH oneshot event for block layer Pavel Dovgalyuk
2018-09-13 10:22 ` Paolo Bonzini
2018-09-13 11:13 ` Pavel Dovgalyuk
2018-09-13 13:45 ` Paolo Bonzini
2018-09-14 6:17 ` Pavel Dovgalyuk
2018-09-12 8:20 ` [Qemu-devel] [PATCH v6 23/25] timer: introduce new virtual clock Pavel Dovgalyuk
2018-09-12 8:20 ` [Qemu-devel] [PATCH v6 24/25] slirp: fix ipv6 timers Pavel Dovgalyuk
2018-09-12 8:45 ` Samuel Thibault
2018-09-12 8:20 ` [Qemu-devel] [PATCH v6 25/25] ui: fix virtual timers Pavel Dovgalyuk
2018-09-13 10:27 ` [Qemu-devel] [PATCH v6 00/25] Fixing record/replay and adding reverse debugging Paolo Bonzini
2018-09-13 13:40 ` Pavel Dovgalyuk
2018-09-13 13:46 ` Paolo Bonzini
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=20180912081945.3228.19776.stgit@pasha-VirtualBox \
--to=pavel.dovgaluk@ispras.ru \
--cc=alex.bennee@linaro.org \
--cc=armbru@redhat.com \
--cc=boost.lists@gmail.com \
--cc=ciro.santilli@gmail.com \
--cc=crosthwaite.peter@gmail.com \
--cc=dgilbert@redhat.com \
--cc=dovgaluk@ispras.ru \
--cc=jasowang@redhat.com \
--cc=kraxel@redhat.com \
--cc=kwolf@redhat.com \
--cc=maria.klimushenkova@ispras.ru \
--cc=mreitz@redhat.com \
--cc=mst@redhat.com \
--cc=pbonzini@redhat.com \
--cc=peter.maydell@linaro.org \
--cc=qemu-devel@nongnu.org \
--cc=quintela@redhat.com \
--cc=rth@twiddle.net \
--cc=thomas.dullien@googlemail.com \
--cc=war2jordan@live.com \
--cc=zuban32s@gmail.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).