From: Alex Bennée <alex.bennee@linaro.org>
Date: Fri, 30 Sep 2016 21:31:01 +0000 (+0100)
Subject: linux-user/syscall: extend lock around cpu-list
X-Git-Tag: v2.9.0-rc2~371^2~21
X-Git-Url: https://git.proxmox.com/?a=commitdiff_plain;h=dd1f63493adbbb06fa16ed15f8fc16584f55ee81;p=mirror_qemu.git

linux-user/syscall: extend lock around cpu-list

There is a potential race if several threads exit at once. To serialise
the exits extend the lock above the initial checking of the CPU list.

Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20160930213106.20186-11-alex.bennee@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---

diff --git a/linux-user/syscall.c b/linux-user/syscall.c
index 0815f30965..fa559be47f 100644
--- a/linux-user/syscall.c
+++ b/linux-user/syscall.c
@@ -7476,13 +7476,16 @@ abi_long do_syscall(void *cpu_env, int num, abi_long arg1,
             break;
         }
 
+        cpu_list_lock();
+
         if (CPU_NEXT(first_cpu)) {
             TaskState *ts;
 
-            cpu_list_lock();
             /* Remove the CPU from the list.  */
             QTAILQ_REMOVE(&cpus, cpu, node);
+
             cpu_list_unlock();
+
             ts = cpu->opaque;
             if (ts->child_tidptr) {
                 put_user_u32(0, ts->child_tidptr);
@@ -7495,6 +7498,8 @@ abi_long do_syscall(void *cpu_env, int num, abi_long arg1,
             rcu_unregister_thread();
             pthread_exit(NULL);
         }
+
+        cpu_list_unlock();
 #ifdef TARGET_GPROF
         _mcleanup();
 #endif