All of lore.kernel.org
 help / color / mirror / Atom feed
From: Max Kellermann <max.kellermann@ionos.com>
To: dhowells@redhat.com, steved@redhat.com, viro@zeniv.linux.org.uk,
	netfs@lists.linux.dev, linux-kernel@vger.kernel.org
Cc: Max Kellermann <max.kellermann@ionos.com>, stable@vger.kernel.org
Subject: [PATCH] cachefiles: pass a poll key when waking the daemon
Date: Thu, 10 Sep 2026 20:22:00 +0200	[thread overview]
Message-ID: <20260910182200.1272489-1-max.kellermann@ionos.com> (raw)

Keyless wakeups bypass epoll's event-mask filtering.  With only
`EPOLLOUT` registered, any cachefiles_state_changed() call will wake
up the epoll_wait() system call, but will not return any events,
causing a busy loop.

Pass the relevant poll events with each wakeup so `epoll` can filter
out unwanted notifications.

Fixes: a5b3a80b899b ("CacheFiles: Provide read-and-reset release counters for cachefilesd")
Cc: stable@vger.kernel.org
Signed-off-by: Max Kellermann <max.kellermann@ionos.com>
---
Note: backports to older branches need to adjust ondemand.c as well
like this:
-       wake_up_all(&cache->daemon_pollwq);
+       cachefiles_wake_poll(cache, EPOLLIN);
---
 fs/cachefiles/cache.c    |  4 ++--
 fs/cachefiles/internal.h | 15 +++++++++++++--
 fs/cachefiles/namei.c    |  2 +-
 3 files changed, 16 insertions(+), 5 deletions(-)

diff --git a/fs/cachefiles/cache.c b/fs/cachefiles/cache.c
index 9fb06dc16520..08ca598618ac 100644
--- a/fs/cachefiles/cache.c
+++ b/fs/cachefiles/cache.c
@@ -255,7 +255,7 @@ int cachefiles_has_space(struct cachefiles_cache *cache,
 	    test_and_clear_bit(CACHEFILES_CULLING, &cache->flags)
 	    ) {
 		_debug("cease culling");
-		cachefiles_state_changed(cache);
+		cachefiles_state_changed(cache, EPOLLIN);
 	}
 
 	//_leave(" = 0");
@@ -275,7 +275,7 @@ int cachefiles_has_space(struct cachefiles_cache *cache,
 begin_cull:
 	if (!test_and_set_bit(CACHEFILES_CULLING, &cache->flags)) {
 		_debug("### CULL CACHE ###");
-		cachefiles_state_changed(cache);
+		cachefiles_state_changed(cache, EPOLLIN | EPOLLOUT);
 	}
 
 	_leave(" = %d", ret);
diff --git a/fs/cachefiles/internal.h b/fs/cachefiles/internal.h
index c93324e0f98c..b8a5ea4a5ced 100644
--- a/fs/cachefiles/internal.h
+++ b/fs/cachefiles/internal.h
@@ -15,6 +15,7 @@
 #include <linux/fscache-cache.h>
 #include <linux/cred.h>
 #include <linux/security.h>
+#include <linux/poll.h>
 
 #define CACHEFILES_DIO_BLOCK_SIZE 4096
 
@@ -117,13 +118,23 @@ struct cachefiles_object *cachefiles_cres_object(struct netfs_cache_resources *c
 	return fscache_cres_cookie(cres)->cache_priv;
 }
 
+/*
+ * Wake up the daemon polling on /dev/cachefiles.
+ */
+static inline void cachefiles_wake_poll(struct cachefiles_cache *cache,
+					__poll_t mask)
+{
+	__wake_up(&cache->daemon_pollwq, TASK_NORMAL, 0, poll_to_key(mask));
+}
+
 /*
  * note change of state for daemon
  */
-static inline void cachefiles_state_changed(struct cachefiles_cache *cache)
+static inline void cachefiles_state_changed(struct cachefiles_cache *cache,
+					    __poll_t mask)
 {
 	set_bit(CACHEFILES_STATE_CHANGED, &cache->flags);
-	wake_up_all(&cache->daemon_pollwq);
+	cachefiles_wake_poll(cache, mask);
 }
 
 /*
diff --git a/fs/cachefiles/namei.c b/fs/cachefiles/namei.c
index 88955249a1a6..52eb227f07e8 100644
--- a/fs/cachefiles/namei.c
+++ b/fs/cachefiles/namei.c
@@ -74,7 +74,7 @@ void cachefiles_unmark_inode_in_use(struct cachefiles_object *object,
 	if (!test_bit(CACHEFILES_OBJECT_USING_TMPFILE, &object->flags)) {
 		atomic_long_add(inode->i_blocks, &cache->b_released);
 		if (atomic_inc_return(&cache->f_released))
-			cachefiles_state_changed(cache);
+			cachefiles_state_changed(cache, EPOLLIN);
 	}
 }
 
-- 
2.47.3


                 reply	other threads:[~2026-09-10 18:22 UTC|newest]

Thread overview: [no followups] expand[flat|nested]  mbox.gz  Atom feed

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=20260910182200.1272489-1-max.kellermann@ionos.com \
    --to=max.kellermann@ionos.com \
    --cc=dhowells@redhat.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=netfs@lists.linux.dev \
    --cc=stable@vger.kernel.org \
    --cc=steved@redhat.com \
    --cc=viro@zeniv.linux.org.uk \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.