eventfd: reorganize the code to simplify new flags

PATCH In review Michael S. Tsirkin 2 participants 5 messages Aug 20, 2009 – Aug 27, 2009 1 message not in the archive

Subsystems (2)

VFS KVM

lore

Share & tools

lore

Michael S. Tsirkin

author

2009-08-20 15:57 UTC #1 lore raw

PATCH 1/2 eventfd: reorganize the code to simplify new flags

In reply to a message not in the archive

This slightly reorganizes the code in eventfd, encapsulating counter math in inline functions, so that it will be easier to add a new flag. No functional changes.

Signed-off-by Michael S. Tsirkin <[email protected]>

---
 fs/eventfd.c |   56 ++++++++++++++++++++++++++++++++++++++++++--------------
 1 files changed, 42 insertions(+), 14 deletions(-)

fs/eventfd.c +42 ~~-14~~

index 31d12de..347a0e0 100644
@@ -34,6 +34,37 @@ struct eventfd_ctx {
 	unsigned int flags;
 };
 
+
+static inline int eventfd_readable(struct eventfd_ctx *ctx)
+{
+	return ctx->count > 0;
+}
+
+static inline int eventfd_writeable(struct eventfd_ctx *ctx, u64 n)
+{
+	return ULLONG_MAX - n > ctx->count;
+}
+
+static inline int eventfd_overflow(struct eventfd_ctx *ctx, u64 cnt)
+{
+	return cnt == ULLONG_MAX;
+}
+
+static inline void eventfd_dowrite(struct eventfd_ctx *ctx, u64 ucnt)
+{
+	if (eventfd_writeable(ctx, ucnt))
+		ucnt = ULLONG_MAX - ctx->count;
+
+	ctx->count += ucnt;
+}
+
+static inline u64 eventfd_doread(struct eventfd_ctx *ctx)
+{
+	u64 ucnt = (ctx->flags & EFD_SEMAPHORE) ? 1 : ctx->count;
+	ctx->count -= ucnt;
+	return ucnt;
+}
+
 /**
  * eventfd_signal - Adds @n to the eventfd counter.
  * @ctx: [in] Pointer to the eventfd context.

@@ -57,9 +88,7 @@ int eventfd_signal(struct eventfd_ctx *ctx, int n)
 	if (n < 0)
 		return -EINVAL;
 	spin_lock_irqsave(&ctx->wqh.lock, flags);
-	if (ULLONG_MAX - ctx->count < n)
-		n = (int) (ULLONG_MAX - ctx->count);
-	ctx->count += n;
+	eventfd_dowrite(ctx, n);
 	if (waitqueue_active(&ctx->wqh))
 		wake_up_locked_poll(&ctx->wqh, POLLIN);
 	spin_unlock_irqrestore(&ctx->wqh.lock, flags);

@@ -119,11 +148,11 @@ static unsigned int eventfd_poll(struct file *file, poll_table *wait)
 	poll_wait(file, &ctx->wqh, wait);
 
 	spin_lock_irqsave(&ctx->wqh.lock, flags);
-	if (ctx->count > 0)
+	if (eventfd_readable(ctx))
 		events |= POLLIN;
-	if (ctx->count == ULLONG_MAX)
+	if (eventfd_overflow(ctx, ctx->count))
 		events |= POLLERR;
-	if (ULLONG_MAX - 1 > ctx->count)
+	if (eventfd_writeable(ctx, 1))
 		events |= POLLOUT;
 	spin_unlock_irqrestore(&ctx->wqh.lock, flags);

@@ -142,13 +171,13 @@ static ssize_t eventfd_read(struct file *file, char __user *buf, size_t count,
 		return -EINVAL;
 	spin_lock_irq(&ctx->wqh.lock);
 	res = -EAGAIN;
-	if (ctx->count > 0)
+	if (eventfd_readable(ctx))
 		res = sizeof(ucnt);
 	else if (!(file->f_flags & O_NONBLOCK)) {
 		__add_wait_queue(&ctx->wqh, &wait);
 		for (res = 0;;) {
 			set_current_state(TASK_INTERRUPTIBLE);
-			if (ctx->count > 0) {
+			if (eventfd_readable(ctx)) {
 				res = sizeof(ucnt);
 				break;
 			}

@@ -164,8 +193,7 @@ static ssize_t eventfd_read(struct file *file, char __user *buf, size_t count,
 		__set_current_state(TASK_RUNNING);
 	}
 	if (likely(res > 0)) {
-		ucnt = (ctx->flags & EFD_SEMAPHORE) ? 1 : ctx->count;
-		ctx->count -= ucnt;
+		ucnt = eventfd_doread(ctx);
 		if (waitqueue_active(&ctx->wqh))
 			wake_up_locked_poll(&ctx->wqh, POLLOUT);
 	}

@@ -188,17 +216,17 @@ static ssize_t eventfd_write(struct file *file, const char __user *buf, size_t c
 		return -EINVAL;
 	if (copy_from_user(&ucnt, buf, sizeof(ucnt)))
 		return -EFAULT;
-	if (ucnt == ULLONG_MAX)
+	if (eventfd_overflow(ctx, ucnt))
 		return -EINVAL;
 	spin_lock_irq(&ctx->wqh.lock);
 	res = -EAGAIN;
-	if (ULLONG_MAX - ctx->count > ucnt)
+	if (eventfd_writeable(ctx, ucnt))
 		res = sizeof(ucnt);
 	else if (!(file->f_flags & O_NONBLOCK)) {
 		__add_wait_queue(&ctx->wqh, &wait);
 		for (res = 0;;) {
 			set_current_state(TASK_INTERRUPTIBLE);
-			if (ULLONG_MAX - ctx->count > ucnt) {
+			if (eventfd_writeable(ctx, ucnt)) {
 				res = sizeof(ucnt);
 				break;
 			}

@@ -214,7 +242,7 @@ static ssize_t eventfd_write(struct file *file, const char __user *buf, size_t c
 		__set_current_state(TASK_RUNNING);
 	}
 	if (likely(res > 0)) {
-		ctx->count += ucnt;
+		eventfd_dowrite(ctx, ucnt);
 		if (waitqueue_active(&ctx->wqh))
 			wake_up_locked_poll(&ctx->wqh, POLLIN);
 	}

Signature · 1 line

1.6.2.5

Michael S. Tsirkin

author

2009-08-20 15:57 UTC #5 lore raw

PATCH 2/2 eventfd: EFD_STATE flag

In reply to a message not in the archive

This implements a new EFD_STATE flag for eventfd.
This changes eventfd behaviour in the following way:
- write simply stores the value written, and is always non-blocking
- read unblocks when the value written changes, and returns the value written

Motivation: we'd like to use eventfd in qemu to pass interrupts from (emulated or assigned) devices to guest. For level interrupts, the counter supported currently by eventfd is not a good match: we really need to set interrupt to a level, typically 0 or 1, and give the guest ability to see the last value written.

Acked-by Avi Kivity <[email protected]>

Signed-off-by Michael S. Tsirkin <[email protected]>

---
 fs/eventfd.c            |   41 ++++++++++++++++++++++++++++++++++-------
 include/linux/eventfd.h |    3 ++-
 2 files changed, 36 insertions(+), 8 deletions(-)

fs/eventfd.c +34 -7

index 347a0e0..7b279e3 100644
@@ -31,37 +31,59 @@ struct eventfd_ctx {
 	 * issue a wakeup.
 	 */
 	__u64 count;
+	/*
+	 * When EF_STATE flag is set, eventfd behaves differently:
+	 * value written gets stored in "count", read will copy
+	 * "count" to "state".
+	 */
+	__u64 state;
 	unsigned int flags;
 };
 
 
 static inline int eventfd_readable(struct eventfd_ctx *ctx)
 {
-	return ctx->count > 0;
+	if (ctx->flags & EFD_STATE)
+		return ctx->state != ctx->count;
+	else
+		return ctx->count > 0;
 }
 
 static inline int eventfd_writeable(struct eventfd_ctx *ctx, u64 n)
 {
-	return ULLONG_MAX - n > ctx->count;
+	if (ctx->flags & EFD_STATE)
+		return 1;
+	else
+		return ULLONG_MAX - n > ctx->count;
 }
 
 static inline int eventfd_overflow(struct eventfd_ctx *ctx, u64 cnt)
 {
-	return cnt == ULLONG_MAX;
+	if (ctx->flags & EFD_STATE)
+		return 0;
+	else
+		return cnt == ULLONG_MAX;
 }
 
 static inline void eventfd_dowrite(struct eventfd_ctx *ctx, u64 ucnt)
 {
-	if (eventfd_writeable(ctx, ucnt))
-		ucnt = ULLONG_MAX - ctx->count;
+	if (ctx->flags & EFD_STATE)
+		ctx->count = ucnt;
+	else {
+		if (ULLONG_MAX - ctx->count < ucnt)
+			ucnt = ULLONG_MAX - ctx->count;
 
-	ctx->count += ucnt;
+		ctx->count += ucnt;
+	}
 }
 
 static inline u64 eventfd_doread(struct eventfd_ctx *ctx)
 {
 	u64 ucnt = (ctx->flags & EFD_SEMAPHORE) ? 1 : ctx->count;
-	ctx->count -= ucnt;
+	if (ctx->flags & EFD_STATE)
+		ctx->state = ucnt;
+	else
+		ctx->count -= ucnt;
 	return ucnt;
 }

@@ -337,6 +359,10 @@ SYSCALL_DEFINE2(eventfd2, unsigned int, count, int, flags)
 
 	if (flags & ~EFD_FLAGS_SET)
 		return -EINVAL;
+	/* State together with semaphore does not make sense. */
+	if ((flags & EFD_STATE) && (flags & EFD_SEMAPHORE))
+		return -EINVAL;
+
 
 	ctx = kmalloc(sizeof(*ctx), GFP_KERNEL);
 	if (!ctx)

@@ -344,6 +370,7 @@ SYSCALL_DEFINE2(eventfd2, unsigned int, count, int, flags)
 
 	kref_init(&ctx->kref);
 	init_waitqueue_head(&ctx->wqh);
+	ctx->state = count;
 	ctx->count = count;
 	ctx->flags = flags;

include/linux/eventfd.h +2 -1

index 3b85ba6..78ff649 100644
@@ -19,11 +19,12 @@
  * shared O_* flags.
  */
 #define EFD_SEMAPHORE (1 << 0)
+#define EFD_STATE (1 << 1)
 #define EFD_CLOEXEC O_CLOEXEC
 #define EFD_NONBLOCK O_NONBLOCK
 
 #define EFD_SHARED_FCNTL_FLAGS (O_CLOEXEC | O_NONBLOCK)
-#define EFD_FLAGS_SET (EFD_SHARED_FCNTL_FLAGS | EFD_SEMAPHORE)
+#define EFD_FLAGS_SET (EFD_SHARED_FCNTL_FLAGS | EFD_SEMAPHORE | EFD_STATE)
 
 #ifdef CONFIG_EVENTFD

Signature · 1 line

1.6.2.5

Paolo Bonzini

2009-08-27 11:48 UTC #2 lore raw

In reply to Michael S. Tsirkin

fs/eventfd.c struct eventfd_ctx { · line 58

Quoted 11 lines · Michael S. Tsirkin

fs/eventfd.c +11 -0

+static inline int eventfd_writeable(struct eventfd_ctx *ctx, u64 n)
+{
+	return ULLONG_MAX - n>  ctx->count;
+}
+
+static inline void eventfd_dowrite(struct eventfd_ctx *ctx, u64 ucnt)
+{
+	if (eventfd_writeable(ctx, ucnt))
+		ucnt = ULLONG_MAX - ctx->count;
+
+	ctx->count += ucnt;

In any case, this usage of eventfd_writeable is wrong: the code was like this:

- if (ULLONG_MAX - ctx->count < n)
- n = (int) (ULLONG_MAX - ctx->count);
- ctx->count += n;
+ eventfd_dowrite(ctx, n);

and so it should be !eventfd_writable. (This smelled when I was editing patch 2, so I went back and checked patch 1).

Paolo

Michael S. Tsirkin

author

2009-08-27 12:00 UTC #3 lore raw

Quoted 21 lines · Paolo Bonzini

fs/eventfd.c struct eventfd_ctx { · line 58

Quoted 11 lines · Michael S. Tsirkin

fs/eventfd.c +11 -0

+static inline int eventfd_writeable(struct eventfd_ctx *ctx, u64 n)
+{
+	return ULLONG_MAX - n>  ctx->count;
+}
+
+static inline void eventfd_dowrite(struct eventfd_ctx *ctx, u64 ucnt)
+{
+	if (eventfd_writeable(ctx, ucnt))
+		ucnt = ULLONG_MAX - ctx->count;
+
+	ctx->count += ucnt;

In any case, this usage of eventfd_writeable is wrong: the code was like this:

fs/eventfd.c +1 -3

-	if (ULLONG_MAX - ctx->count < n)
-		n = (int) (ULLONG_MAX - ctx->count);
-	ctx->count += n;
+	eventfd_dowrite(ctx, n);

Quoted 4 lines

and so it should be !eventfd_writable. (This smelled when I was editing patch 2, so I went back and checked patch 1).
Paolo

Right, patch 2 actually fixed this back. The best thing is to open-code it actually, since we handle it in a way that only applies to a counter, anyway. Will post a fixed v1.

Signature · 1 line

MST

Michael S. Tsirkin

author

2009-08-27 12:03 UTC #4 lore raw

In reply to Paolo Bonzini

Quoted 14 lines · Paolo Bonzini

fs/eventfd.c struct eventfd_ctx { · line 58

Quoted 11 lines · Michael S. Tsirkin

fs/eventfd.c +11 -0

+static inline int eventfd_writeable(struct eventfd_ctx *ctx, u64 n)
+{
+	return ULLONG_MAX - n>  ctx->count;
+}
+
+static inline void eventfd_dowrite(struct eventfd_ctx *ctx, u64 ucnt)
+{
+	if (eventfd_writeable(ctx, ucnt))
+		ucnt = ULLONG_MAX - ctx->count;
+
+	ctx->count += ucnt;

In any case, this usage of eventfd_writeable is wrong:

Fixed, thanks!