qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: "Alex Bennée" <alex.bennee@linaro.org>
To: mttcg@listserver.greensocs.com, qemu-devel@nongnu.org,
	fred.konrad@greensocs.com, a.rigo@virtualopensystems.com,
	cota@braap.org, bobby.prani@gmail.com, nikunj@linux.vnet.ibm.com
Cc: mark.burton@greensocs.com, pbonzini@redhat.com,
	jan.kiszka@siemens.com, serge.fdrv@gmail.com, rth@twiddle.net,
	peter.maydell@linaro.org, claudio.fontana@huawei.com,
	"Alex Bennée" <alex.bennee@linaro.org>,
	"Peter Crosthwaite" <crosthwaite.peter@gmail.com>
Subject: [Qemu-devel] [RFC v4 18/28] tcg: remove global exit_request
Date: Thu, 11 Aug 2016 16:24:14 +0100	[thread overview]
Message-ID: <1470929064-4092-19-git-send-email-alex.bennee@linaro.org> (raw)
In-Reply-To: <1470929064-4092-1-git-send-email-alex.bennee@linaro.org>

There are now only two uses of the global exit_request left.

The first ensures we exit the run_loop when we first start to process
pending work and in the kick handler. This is just as easily done by
setting the first_cpu->exit_request flag.

The second use is in the round robin kick routine. The global
exit_request ensured every vCPU would set its local exit_request and
cause a full exit of the loop. Now the iothread isn't being held while
running we can just rely on the kick handler to push us out as intended.

We lightly re-factor the main vCPU thread to ensure cpu->exit_requests
cause us to exit the main loop and process any IO requests that might
come along.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>

---
v4
  - moved to after iothread unlocking patch
  - needed to remove kick exit_request as well.
  - remove extraneous cpu->exit_request check
  - remove stray exit_request setting
  - remove needless atomic operation
---
 cpu-exec-common.c       |  1 -
 cpu-exec.c              |  9 ++-------
 cpus.c                  | 18 ++++++++++--------
 include/exec/exec-all.h |  1 -
 4 files changed, 12 insertions(+), 17 deletions(-)

diff --git a/cpu-exec-common.c b/cpu-exec-common.c
index e220ba7..e29cf6d 100644
--- a/cpu-exec-common.c
+++ b/cpu-exec-common.c
@@ -23,7 +23,6 @@
 #include "exec/exec-all.h"
 #include "exec/memory-internal.h"
 
-bool exit_request;
 int tcg_pending_threads;
 
 /* exit the current TB, but without causing any exception to be raised */
diff --git a/cpu-exec.c b/cpu-exec.c
index 317ad8e..d2e8b9e 100644
--- a/cpu-exec.c
+++ b/cpu-exec.c
@@ -531,9 +531,8 @@ static inline void cpu_loop_exec_tb(CPUState *cpu, TranslationBlock *tb,
         /* Something asked us to stop executing
          * chained TBs; just continue round the main
          * loop. Whatever requested the exit will also
-         * have set something else (eg exit_request or
-         * interrupt_request) which we will handle
-         * next time around the loop.  But we need to
+         * have set something else (eg interrupt_request) which we
+         * will handle next time around the loop.  But we need to
          * ensure the tcg_exit_req read in generated code
          * comes before the next read of cpu->exit_request
          * or cpu->interrupt_request.
@@ -589,10 +588,6 @@ int cpu_exec(CPUState *cpu)
 
     rcu_read_lock();
 
-    if (unlikely(atomic_mb_read(&exit_request))) {
-        cpu->exit_request = 1;
-    }
-
     cc->cpu_exec_enter(cpu);
 
     /* Calculate difference between guest clock and host clock.
diff --git a/cpus.c b/cpus.c
index 5ed6dc0..8a40a08 100644
--- a/cpus.c
+++ b/cpus.c
@@ -1215,7 +1215,6 @@ static CPUState *tcg_current_rr_cpu;
 static void qemu_cpu_kick_rr_cpu(void)
 {
     CPUState *cpu;
-    atomic_mb_set(&exit_request, 1);
     do {
         cpu = atomic_mb_read(&tcg_current_rr_cpu);
         if (cpu) {
@@ -1265,11 +1264,11 @@ static void *qemu_tcg_cpu_thread_fn(void *arg)
         timer_mod(kick_timer, qemu_tcg_next_kick());
     }
 
-    /* process any pending work */
-    atomic_mb_set(&exit_request, 1);
-
     cpu = first_cpu;
 
+    /* process any pending work */
+    cpu->exit_request = 1;
+
     while (1) {
         /* Account partial waits to QEMU_CLOCK_VIRTUAL.  */
         qemu_account_warp_timer();
@@ -1278,7 +1277,7 @@ static void *qemu_tcg_cpu_thread_fn(void *arg)
             cpu = first_cpu;
         }
 
-        for (; cpu != NULL && !exit_request; cpu = CPU_NEXT(cpu)) {
+        while (cpu && !cpu->exit_request) {
             atomic_mb_set(&tcg_current_rr_cpu, cpu);
 
             qemu_clock_enable(QEMU_CLOCK_VIRTUAL,
@@ -1300,12 +1299,15 @@ static void *qemu_tcg_cpu_thread_fn(void *arg)
                 break;
             }
 
-        } /* for cpu.. */
+            cpu = CPU_NEXT(cpu);
+        } /* while (cpu && !cpu->exit_request).. */
+
         /* Does not need atomic_mb_set because a spurious wakeup is okay.  */
         atomic_set(&tcg_current_rr_cpu, NULL);
 
-        /* Pairs with smp_wmb in qemu_cpu_kick.  */
-        atomic_mb_set(&exit_request, 0);
+        if (cpu && cpu->exit_request) {
+            atomic_mb_set(&cpu->exit_request, 0);
+        }
 
         handle_icount_deadline();
 
diff --git a/include/exec/exec-all.h b/include/exec/exec-all.h
index 3547dfa..27d1a24 100644
--- a/include/exec/exec-all.h
+++ b/include/exec/exec-all.h
@@ -409,7 +409,6 @@ extern int singlestep;
 
 /* cpu-exec.c, accessed with atomic_mb_read/atomic_mb_set */
 extern int tcg_pending_threads;
-extern bool exit_request;
 
 /**
  * qemu_work_cond - condition to wait for CPU work items completion
-- 
2.7.4

  parent reply	other threads:[~2016-08-11 15:24 UTC|newest]

Thread overview: 64+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-08-11 15:23 [Qemu-devel] [RFC v4 00/28] Base enabling patches for MTTCG Alex Bennée
2016-08-11 15:23 ` [Qemu-devel] [RFC v4 01/28] cpus: make all_vcpus_paused() return bool Alex Bennée
2016-08-11 15:23 ` [Qemu-devel] [RFC v4 02/28] translate_all: DEBUG_FLUSH -> DEBUG_TB_FLUSH Alex Bennée
2016-08-11 15:23 ` [Qemu-devel] [RFC v4 03/28] translate-all: add DEBUG_LOCKING asserts Alex Bennée
2016-08-11 15:24 ` [Qemu-devel] [RFC v4 04/28] cpu-exec: include cpu_index in CPU_LOG_EXEC messages Alex Bennée
2016-09-07  2:21   ` Richard Henderson
2016-08-11 15:24 ` [Qemu-devel] [RFC v4 05/28] docs: new design document multi-thread-tcg.txt (DRAFTING) Alex Bennée
2016-08-11 15:24 ` [Qemu-devel] [RFC v4 06/28] tcg: comment on which functions have to be called with tb_lock held Alex Bennée
2016-09-07  2:30   ` Richard Henderson
2016-08-11 15:24 ` [Qemu-devel] [RFC v4 07/28] linux-user/elfload: ensure mmap_lock() held while setting up Alex Bennée
2016-09-07  2:34   ` Richard Henderson
2016-08-11 15:24 ` [Qemu-devel] [RFC v4 08/28] translate-all: Add assert_(memory|tb)_lock annotations Alex Bennée
2016-09-07  2:41   ` Richard Henderson
2016-09-07  7:08     ` Alex Bennée
2016-08-11 15:24 ` [Qemu-devel] [RFC v4 09/28] tcg: protect TBContext with tb_lock Alex Bennée
2016-09-07  2:48   ` Richard Henderson
2016-08-11 15:24 ` [Qemu-devel] [RFC v4 10/28] target-arm/arm-powerctl: wake up sleeping CPUs Alex Bennée
2016-08-11 15:24 ` [Qemu-devel] [RFC v4 11/28] tcg: move tcg_exec_all and helpers above thread fn Alex Bennée
2016-09-07  2:53   ` Richard Henderson
2016-08-11 15:24 ` [Qemu-devel] [RFC v4 12/28] tcg: cpus rm tcg_exec_all() Alex Bennée
2016-08-11 15:24 ` [Qemu-devel] [RFC v4 13/28] tcg: add options for enabling MTTCG Alex Bennée
2016-09-07  3:06   ` Richard Henderson
2016-08-11 15:24 ` [Qemu-devel] [RFC v4 14/28] tcg: add kick timer for single-threaded vCPU emulation Alex Bennée
2016-09-07  3:25   ` Richard Henderson
2016-09-07  5:40     ` Paolo Bonzini
2016-09-07 10:15       ` Alex Bennée
2016-09-07 10:19     ` Alex Bennée
2016-08-11 15:24 ` [Qemu-devel] [RFC v4 15/28] tcg: rename tcg_current_cpu to tcg_current_rr_cpu Alex Bennée
2016-09-07  3:34   ` Richard Henderson
2016-08-11 15:24 ` [Qemu-devel] [RFC v4 16/28] tcg: drop global lock during TCG code execution Alex Bennée
2016-09-07  4:03   ` Richard Henderson
2016-09-07  5:43     ` Paolo Bonzini
2016-09-07  6:43       ` Richard Henderson
2016-09-07 15:15         ` Paolo Bonzini
2016-08-11 15:24 ` [Qemu-devel] [RFC v4 17/28] cpus: re-factor out handle_icount_deadline Alex Bennée
2016-09-07  4:06   ` Richard Henderson
2016-08-11 15:24 ` Alex Bennée [this message]
2016-09-07  4:11   ` [Qemu-devel] [RFC v4 18/28] tcg: remove global exit_request Richard Henderson
2016-08-11 15:24 ` [Qemu-devel] [RFC v4 19/28] tcg: move locking for tb_invalidate_phys_page_range up Alex Bennée
2016-09-27 15:56   ` Paolo Bonzini
2016-08-11 15:24 ` [Qemu-devel] [RFC v4 20/28] cpus: tweak sleeping and safe_work rules for MTTCG Alex Bennée
2016-09-07  4:22   ` Richard Henderson
2016-09-07 10:05   ` Paolo Bonzini
2016-08-11 15:24 ` [Qemu-devel] [RFC v4 21/28] tcg: enable tb_lock() for SoftMMU Alex Bennée
2016-09-07  4:26   ` Richard Henderson
2016-09-27 16:16   ` Paolo Bonzini
2016-08-11 15:24 ` [Qemu-devel] [RFC v4 22/28] tcg: enable thread-per-vCPU Alex Bennée
2016-08-11 15:24 ` [Qemu-devel] [RFC v4 23/28] atomic: introduce cmpxchg_bool Alex Bennée
2016-09-08  0:12   ` Richard Henderson
2016-08-11 15:24 ` [Qemu-devel] [RFC v4 24/28] cputlb: add assert_cpu_is_self checks Alex Bennée
2016-09-08 17:19   ` Richard Henderson
2016-08-11 15:24 ` [Qemu-devel] [RFC v4 25/28] cputlb: introduce tlb_flush_* async work Alex Bennée
2016-09-07 10:08   ` Paolo Bonzini
2016-09-08 17:23   ` Richard Henderson
2016-08-11 15:24 ` [Qemu-devel] [RFC v4 26/28] cputlb: tweak qemu_ram_addr_from_host_nofail reporting Alex Bennée
2016-09-08 17:24   ` Richard Henderson
2016-08-11 15:24 ` [Qemu-devel] [RFC v4 27/28] cputlb: make tlb_reset_dirty safe for MTTCG Alex Bennée
2016-09-08 17:34   ` Richard Henderson
2016-08-11 15:24 ` [Qemu-devel] [RFC v4 28/28] cputlb: make tlb_flush_by_mmuidx " Alex Bennée
2016-09-07 10:09   ` Paolo Bonzini
2016-09-08 17:54   ` Richard Henderson
2016-08-11 17:22 ` [Qemu-devel] [RFC v4 00/28] Base enabling patches " Alex Bennée
2016-08-12  8:02   ` Alex Bennée
2016-09-06  9:24 ` Alex Bennée

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=1470929064-4092-19-git-send-email-alex.bennee@linaro.org \
    --to=alex.bennee@linaro.org \
    --cc=a.rigo@virtualopensystems.com \
    --cc=bobby.prani@gmail.com \
    --cc=claudio.fontana@huawei.com \
    --cc=cota@braap.org \
    --cc=crosthwaite.peter@gmail.com \
    --cc=fred.konrad@greensocs.com \
    --cc=jan.kiszka@siemens.com \
    --cc=mark.burton@greensocs.com \
    --cc=mttcg@listserver.greensocs.com \
    --cc=nikunj@linux.vnet.ibm.com \
    --cc=pbonzini@redhat.com \
    --cc=peter.maydell@linaro.org \
    --cc=qemu-devel@nongnu.org \
    --cc=rth@twiddle.net \
    --cc=serge.fdrv@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).