187 * 188 * Post a notification of an event into a set of watch queues and let the users 189 * know. However most of them have never been documented so they are not easily available to other developers.
Linux kernel kref code#
185 * The creds of the process that triggered the notification. In the Linux kernel code base there are many design patterns that have been found to be effective. */ 84 static const struct pipe_buf_operations watch_queue_pipe_buf_ops = ġ80 181 /** 182 * _post_watch_notification - Post an event notification 183 * The watch list to post the event to. 67 */ 68 bit = buf-> offset + buf-> len Ħ9 if (( bit & ( WATCH_QUEUE_NOTE_SIZE - 1)) = 0)ħ5 76 set_bit( bit, wqueue->notes_bitmap) ħ9 80 // No try_steal function => no stealing 81 #define watch_queue_pipe_buf_try_steal NULL 82 83 /* New data written to a pipe may be appended to a buffer with this type. OTOH, the note must've been more than zero size. 41 */ 42 static inline bool lock_wqueue(struct watch_queue *wqueue)ĥ1 52 static inline void unlock_wqueue(struct watch_queue *wqueue)ĥ6 57 static void watch_queue_pipe_buf_release(struct pipe_inode_info * pipe,Ħ0 struct watch_queue *wqueue = (struct watch_queue *) buf-> private Ħ3 64 /* We need to work out which note within the page this refers to, but 65 * the note might have been maximum size, so merely ANDing the offset 66 * off doesn't work. It can then take the lock, 39 * and check that the wqueue hasn't been destroyed, which in 40 * turn makes sure that the notification pipe still exists. 5 * Written by David Howells 6 * 7 * See Documentation/core-api/watch_queue.rst 8 */ 9 10 #define pr_fmt( fmt) "watchq: " fmt 11 #include Ģ9 30 MODULE_DESCRIPTION( "Watch queue") ģ4 #define WATCH_QUEUE_NOTES_PER_PAGE ( PAGE_SIZE / WATCH_QUEUE_NOTE_SIZE)ģ5 36 /* 37 * This must be called under the RCU read-lock, which makes 38 * sure that the wqueue still exists.
Linux kernel kref driver#
~ ~ ~ ~ ~ ~ ~ ~ġ // SPDX-License-Identifier: GPL-2.0 2 /* Watch queue and general notification mechanism, built on pipes 3 * 4 * Copyright (C) 2020 Red Hat, Inc. As the Linux kernel maintainer for the -stable branch, the staging subsystem, USB, driver core, debugfs, kref, kobject, and the sysfs kernel subsystems.
Linux kernel kref drivers#
Documentation arch block crypto drivers firmware fs include init ipc kernel lib mm net samples scripts security sound tools usr virt/ kvm. When the channel is freed, the coresponding kref_put GitHub - kref/linux-oxnas: linux kernel for oxnas kref / linux-oxnas Public master 5 branches 0 tags 415,001 commits Failed to load latest commit information. When you are done with a pointer, you must call krefput (): krefput (&data->refcount, datarelease) If.
Every time one of the device's channels is allocated to a client, Once you have an initialized kref, you must follow the following rules: If you make a non-temporary copy of a pointer, especially if it can be passed to another thread of execution, you must. + * device is released, the coresponding kref_put is done in the release A kref_get is done for each device registered. * class_device is released, the coresponding kref_put is done in the release A kref_get is done for each class_device registered. * Each device has a kref, which is initialized to 1 when the device is A struct kobject represents a kernel object, maybe a device or so, such as the things that. +++ -41,12 +41,12 the definition of dma_event_callback in dmaengine.h. A struct kref is an object that handles reference counting.
Part of the work to eliminate struct class_device.ĭrivers/dma/dmaengine.c | 43 ++++++++++++++++++++++-Ģ files changed, 24 insertions(+), 22 deletions(-) Messages sorted by: Ĭontent-Disposition: inline filename=dma.patchĬonvert from class_device to device for drivers/dma/dmaengine.
0 kommentar(er)
