Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
From:   "Eric W. Biederman" <ebiederm@xmission.com>
To:     linux-kernel@vger.kernel.org
Cc:     Linus Torvalds <torvalds@linux-foundation.org>,
        Alexey Gladkov <legion@kernel.org>,
        Kyle Huey <me@kylehuey.com>, Oleg Nesterov <oleg@redhat.com>,
        Kees Cook <keescook@chromium.org>,
        Al Viro <viro@ZenIV.linux.org.uk>,
        Jens Axboe <axboe@kernel.dk>,
        "Eric W. Biederman" <ebiederm@xmission.com>
Date:   Wed,  9 Mar 2022 10:24:50 -0600
Message-Id: <20220309162454.123006-9-ebiederm@xmission.com>
In-Reply-To: <87o82gdlu9.fsf_-_@email.froward.int.ebiederm.org>
References: <87o82gdlu9.fsf_-_@email.froward.int.ebiederm.org>
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit
Subject: [PATCH 09/13] task_work: Decouple TIF_NOTIFY_SIGNAL and task_work
Precedence: bulk

There are a small handful of reasons besides pending signals that the
kernel might want to break out of interruptible sleeps.  The flag
TIF_NOTIFY_SIGNAL and the helpers that set and clear TIF_NOTIFY_SIGNAL
provide that the infrastructure for breaking out of interruptible
sleeps and entering the return to user space slow path for those
cases.

Expand tracehook_notify_signal inline in it's callers and remove it,
which makes clear that TIF_NOTIFY_SIGNAL and task_work are separate
concepts.

Update the comment on set_notify_signal to more accurately describe
it's purpose.

Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
---
 fs/io-wq.c                |  4 +++-
 fs/io_uring.c             |  4 +++-
 include/linux/tracehook.h | 15 ++-------------
 kernel/entry/kvm.c        |  7 +++++--
 4 files changed, 13 insertions(+), 17 deletions(-)

diff --git a/fs/io-wq.c b/fs/io-wq.c
index bb7f161bb19c..8b9147873c2c 100644
--- a/fs/io-wq.c
+++ b/fs/io-wq.c
@@ -515,7 +515,9 @@ static bool io_flush_signals(void)
 {
 	if (unlikely(test_thread_flag(TIF_NOTIFY_SIGNAL))) {
 		__set_current_state(TASK_RUNNING);
-		tracehook_notify_signal();
+		clear_notify_signal();
+		if (task_work_pending(current))
+			task_work_run();
 		return true;
 	}
 	return false;
diff --git a/fs/io_uring.c b/fs/io_uring.c
index e85261079a78..d5fbae1030f9 100644
--- a/fs/io_uring.c
+++ b/fs/io_uring.c
@@ -2592,7 +2592,9 @@ static inline bool io_run_task_work(void)
 {
 	if (test_thread_flag(TIF_NOTIFY_SIGNAL) || task_work_pending(current)) {
 		__set_current_state(TASK_RUNNING);
-		tracehook_notify_signal();
+		clear_notify_signal();
+		if (task_work_pending(current))
+			task_work_run();
 		return true;
 	}
 
diff --git a/include/linux/tracehook.h b/include/linux/tracehook.h
index b44a7820c468..e5d676e841e3 100644
--- a/include/linux/tracehook.h
+++ b/include/linux/tracehook.h
@@ -113,19 +113,8 @@ static inline void clear_notify_signal(void)
 }
 
 /*
- * called by exit_to_user_mode_loop() if ti_work & _TIF_NOTIFY_SIGNAL. This
- * is currently used by TWA_SIGNAL based task_work, which requires breaking
- * wait loops to ensure that task_work is noticed and run.
- */
-static inline void tracehook_notify_signal(void)
-{
-	clear_notify_signal();
-	if (task_work_pending(current))
-		task_work_run();
-}
-
-/*
- * Called when we have work to process from exit_to_user_mode_loop()
+ * Called to break out of interruptible wait loops, and enter the
+ * exit_to_user_mode_loop().
  */
 static inline void set_notify_signal(struct task_struct *task)
 {
diff --git a/kernel/entry/kvm.c b/kernel/entry/kvm.c
index cabf36a489e4..3ab5f98988c3 100644
--- a/kernel/entry/kvm.c
+++ b/kernel/entry/kvm.c
@@ -8,8 +8,11 @@ static int xfer_to_guest_mode_work(struct kvm_vcpu *vcpu, unsigned long ti_work)
 	do {
 		int ret;
 
-		if (ti_work & (_TIF_SIGPENDING | _TIF_NOTIFY_SIGNAL))
-			tracehook_notify_signal();
+		if (ti_work & (_TIF_SIGPENDING | _TIF_NOTIFY_SIGNAL)) {
+			clear_notify_signal();
+			if (task_work_pending(current))
+				task_work_run();
+		}
 
 		if (ti_work & _TIF_SIGPENDING) {
 			kvm_handle_signal_exit(vcpu);
-- 
2.29.2