Merge tag 'driver-core-5.3-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-core

Pull driver core fixes from Greg KH:
 "Here are two small fixes for some driver core issues that have been
  reported. There is also a kernfs "fix" here, which was then reverted
  because it was found to cause problems in linux-next.

  The driver core fixes both resolve reported issues, one with gpioint
  stuff that showed up in 5.3-rc1, and the other finally (and hopefully)
  resolves a very long standing race when removing glue directories.
  It's nice to get that issue finally resolved and the developers
  involved should be applauded for the persistence it took to get this
  patch finally accepted.

  All of these have been in linux-next for a while with no reported
  issues. Well, the one reported issue, hence the revert :)"

* tag 'driver-core-5.3-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-core:
  Revert "kernfs: fix memleak in kernel_ops_readdir()"
  kernfs: fix memleak in kernel_ops_readdir()
  driver core: Fix use-after-free and double free on glue directory
  driver core: platform: return -ENXIO for missing GpioInt

2 files changed, 59 insertions, 3 deletions

diff --git a/drivers/base/core.c b/drivers/base/core.c
index 636058bbf48a3d..1669d41fcddccd 100644
--- a/drivers/base/core.c
+++ b/drivers/base/core.c
@@ -1823,12 +1823,63 @@ static inline struct kobject *get_glue_dir(struct device *dev)
  */
 static void cleanup_glue_dir(struct device *dev, struct kobject *glue_dir)
 {
+	unsigned int ref;
+
 	/* see if we live in a "glue" directory */
 	if (!live_in_glue_dir(glue_dir, dev))
 		return;
 
 	mutex_lock(&gdp_mutex);
-	if (!kobject_has_children(glue_dir))
+	/**
+	 * There is a race condition between removing glue directory
+	 * and adding a new device under the glue directory.
+	 *
+	 * CPU1:                                         CPU2:
+	 *
+	 * device_add()
+	 *   get_device_parent()
+	 *     class_dir_create_and_add()
+	 *       kobject_add_internal()
+	 *         create_dir()    // create glue_dir
+	 *
+	 *                                               device_add()
+	 *                                                 get_device_parent()
+	 *                                                   kobject_get() // get glue_dir
+	 *
+	 * device_del()
+	 *   cleanup_glue_dir()
+	 *     kobject_del(glue_dir)
+	 *
+	 *                                               kobject_add()
+	 *                                                 kobject_add_internal()
+	 *                                                   create_dir() // in glue_dir
+	 *                                                     sysfs_create_dir_ns()
+	 *                                                       kernfs_create_dir_ns(sd)
+	 *
+	 *       sysfs_remove_dir() // glue_dir->sd=NULL
+	 *       sysfs_put()        // free glue_dir->sd
+	 *
+	 *                                                         // sd is freed
+	 *                                                         kernfs_new_node(sd)
+	 *                                                           kernfs_get(glue_dir)
+	 *                                                           kernfs_add_one()
+	 *                                                           kernfs_put()
+	 *
+	 * Before CPU1 remove last child device under glue dir, if CPU2 add
+	 * a new device under glue dir, the glue_dir kobject reference count
+	 * will be increase to 2 in kobject_get(k). And CPU2 has been called
+	 * kernfs_create_dir_ns(). Meanwhile, CPU1 call sysfs_remove_dir()
+	 * and sysfs_put(). This result in glue_dir->sd is freed.
+	 *
+	 * Then the CPU2 will see a stale "empty" but still potentially used
+	 * glue dir around in kernfs_new_node().
+	 *
+	 * In order to avoid this happening, we also should make sure that
+	 * kernfs_node for glue_dir is released in CPU1 only when refcount
+	 * for glue_dir kobj is 1.
+	 */
+	ref = kref_read(&glue_dir->kref);
+	if (!kobject_has_children(glue_dir) && !--ref)
 		kobject_del(glue_dir);
 	kobject_put(glue_dir);
 	mutex_unlock(&gdp_mutex);
diff --git a/drivers/base/platform.c b/drivers/base/platform.c
index 506a0175a5a781..ec974ba9c0c4ca 100644
--- a/drivers/base/platform.c
+++ b/drivers/base/platform.c
@@ -157,8 +157,13 @@ int platform_get_irq(struct platform_device *dev, unsigned int num)
 	 * the device will only expose one IRQ, and this fallback
 	 * allows a common code path across either kind of resource.
 	 */
-	if (num == 0 && has_acpi_companion(&dev->dev))
-		return acpi_dev_gpio_irq_get(ACPI_COMPANION(&dev->dev), num);
+	if (num == 0 && has_acpi_companion(&dev->dev)) {
+		int ret = acpi_dev_gpio_irq_get(ACPI_COMPANION(&dev->dev), num);
+
+		/* Our callers expect -ENXIO for missing IRQs. */
+		if (ret >= 0 || ret == -EPROBE_DEFER)
+			return ret;
+	}
 
 	return -ENXIO;
 #endif