Linux kernel mirror (for testing)
git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
kernel
os
linux
Merge tag 'driver-core-3.15-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-core
Pull driver core fixes from Greg KH:
"Here are some driver core fixes for 3.15-rc2. Also in here are some
documentation updates, as well as an API removal that had to wait for
after -rc1 due to the cleanups coming into you from multiple developer
trees (this one and the PPC tree.)
All have been in linux next successfully"
* tag 'driver-core-3.15-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-core:
drivers/base/dd.c incorrect pr_debug() parameters
Documentation: Update stable address in Chinese and Japanese translations
topology: Fix compilation warning when not in SMP
Chinese: add translation of io_ordering.txt
stable_kernel_rules: spelling/word usage
sysfs, driver-core: remove unused {sysfs|device}_schedule_callback_owner()
kernfs: protect lazy kernfs_iattrs allocation with mutex
fs: Don't return 0 from get_anon_bdev
+92
-160
+1
-1
Documentation/ja_JP/HOWTO
+1
-1
Documentation/ja_JP/HOWTO
···
315
315
もし、3.x.y カーネルが存在しない場合には、番号が一番大きい 3.x が
316
316
最新の安定版カーネルです。
317
317
318
-
3.x.y は "stable" チーム <stable@kernel.org> でメンテされており、必
318
+
3.x.y は "stable" チーム <stable@vger.kernel.org> でメンテされており、必
319
319
要に応じてリリースされます。通常のリリース期間は 2週間毎ですが、差し迫っ
320
320
た問題がなければもう少し長くなることもあります。セキュリティ関連の問題
321
321
の場合はこれに対してだいたいの場合、すぐにリリースがされます。
+3
-3
Documentation/ja_JP/stable_kernel_rules.txt
+3
-3
Documentation/ja_JP/stable_kernel_rules.txt
···
50
50
51
51
-stable ツリーにパッチを送付する手続き-
52
52
53
-
- 上記の規則に従っているかを確認した後に、stable@kernel.org にパッチ
53
+
- 上記の規則に従っているかを確認した後に、stable@vger.kernel.org にパッチ
54
54
を送る。
55
55
- 送信者はパッチがキューに受け付けられた際には ACK を、却下された場合
56
56
には NAK を受け取る。この反応は開発者たちのスケジュールによって、数
57
57
日かかる場合がある。
58
58
- もし受け取られたら、パッチは他の開発者たちと関連するサブシステムの
59
59
メンテナーによるレビューのために -stable キューに追加される。
60
-
- パッチに stable@kernel.org のアドレスが付加されているときには、それ
60
+
- パッチに stable@vger.kernel.org のアドレスが付加されているときには、それ
61
61
が Linus のツリーに入る時に自動的に stable チームに email される。
62
-
- セキュリティパッチはこのエイリアス (stable@kernel.org) に送られるべ
62
+
- セキュリティパッチはこのエイリアス (stable@vger.kernel.org) に送られるべ
63
63
きではなく、代わりに security@kernel.org のアドレスに送られる。
64
64
65
65
レビューサイクル-
+1
-1
Documentation/stable_kernel_rules.txt
+1
-1
Documentation/stable_kernel_rules.txt
···
39
39
the stable tree without anything else needing to be done by the author
40
40
or subsystem maintainer.
41
41
- If the patch requires other patches as prerequisites which can be
42
-
cherry-picked than this can be specified in the following format in
42
+
cherry-picked, then this can be specified in the following format in
43
43
the sign-off area:
44
44
45
45
Cc: <stable@vger.kernel.org> # 3.3.x: a1f84a3: sched: Check for idle
+1
-1
Documentation/zh_CN/HOWTO
+1
-1
Documentation/zh_CN/HOWTO
+67
Documentation/zh_CN/io_ordering.txt
+67
Documentation/zh_CN/io_ordering.txt
···
1
+
Chinese translated version of Documentation/io_orderings.txt
2
+
3
+
If you have any comment or update to the content, please contact the
4
+
original document maintainer directly. However, if you have a problem
5
+
communicating in English you can also ask the Chinese maintainer for
6
+
help. Contact the Chinese maintainer if this translation is outdated
7
+
or if there is a problem with the translation.
8
+
9
+
Chinese maintainer: Lin Yongting <linyongting@gmail.com>
10
+
---------------------------------------------------------------------
11
+
Documentation/io_ordering.txt 的中文翻译
12
+
13
+
如果想评论或更新本文的内容，请直接联系原文档的维护者。如果你使用英文
14
+
交流有困难的话，也可以向中文版维护者求助。如果本翻译更新不及时或者翻
15
+
译存在问题，请联系中文版维护者。
16
+
17
+
中文版维护者： 林永听 Lin Yongting <linyongting@gmail.com>
18
+
中文版翻译者： 林永听 Lin Yongting <linyongting@gmail.com>
19
+
中文版校译者： 林永听 Lin Yongting <linyongting@gmail.com>
20
+
21
+
22
+
以下为正文
23
+
---------------------------------------------------------------------
24
+
25
+
在某些平台上，所谓的内存映射I/O是弱顺序。在这些平台上，驱动开发者有责任
26
+
保证I/O内存映射地址的写操作按程序图意的顺序达到设备。通常读取一个“安全”
27
+
设备寄存器或桥寄存器，触发IO芯片清刷未处理的写操作到达设备后才处理读操作，
28
+
而达到保证目的。驱动程序通常在spinlock保护的临界区退出之前使用这种技术。
29
+
这也可以保证后面的写操作只在前面的写操作之后到达设备（这非常类似于内存
30
+
屏障操作，mb()，不过仅适用于I/O）。
31
+
32
+
假设一个设备驱动程的具体例子：
33
+
34
+
...
35
+
CPU A: spin_lock_irqsave(&dev_lock, flags)
36
+
CPU A: val = readl(my_status);
37
+
CPU A: ...
38
+
CPU A: writel(newval, ring_ptr);
39
+
CPU A: spin_unlock_irqrestore(&dev_lock, flags)
40
+
...
41
+
CPU B: spin_lock_irqsave(&dev_lock, flags)
42
+
CPU B: val = readl(my_status);
43
+
CPU B: ...
44
+
CPU B: writel(newval2, ring_ptr);
45
+
CPU B: spin_unlock_irqrestore(&dev_lock, flags)
46
+
...
47
+
48
+
上述例子中，设备可能会先接收到newval2的值，然后接收到newval的值，问题就
49
+
发生了。不过很容易通过下面方法来修复：
50
+
51
+
...
52
+
CPU A: spin_lock_irqsave(&dev_lock, flags)
53
+
CPU A: val = readl(my_status);
54
+
CPU A: ...
55
+
CPU A: writel(newval, ring_ptr);
56
+
CPU A: (void)readl(safe_register); /* 配置寄存器？*/
57
+
CPU A: spin_unlock_irqrestore(&dev_lock, flags)
58
+
...
59
+
CPU B: spin_lock_irqsave(&dev_lock, flags)
60
+
CPU B: val = readl(my_status);
61
+
CPU B: ...
62
+
CPU B: writel(newval2, ring_ptr);
63
+
CPU B: (void)readl(safe_register); /* 配置寄存器？*/
64
+
CPU B: spin_unlock_irqrestore(&dev_lock, flags)
65
+
66
+
在解决方案中，读取safe_register寄存器，触发IO芯片清刷未处理的写操作，
67
+
再处理后面的读操作，防止引发数据不一致问题。
+1
-1
Documentation/zh_CN/stable_kernel_rules.txt
+1
-1
Documentation/zh_CN/stable_kernel_rules.txt
-33
drivers/base/core.c
-33
drivers/base/core.c
···
614
614
}
615
615
EXPORT_SYMBOL_GPL(device_remove_bin_file);
616
616
617
-
/**
618
-
* device_schedule_callback_owner - helper to schedule a callback for a device
619
-
* @dev: device.
620
-
* @func: callback function to invoke later.
621
-
* @owner: module owning the callback routine
622
-
*
623
-
* Attribute methods must not unregister themselves or their parent device
624
-
* (which would amount to the same thing). Attempts to do so will deadlock,
625
-
* since unregistration is mutually exclusive with driver callbacks.
626
-
*
627
-
* Instead methods can call this routine, which will attempt to allocate
628
-
* and schedule a workqueue request to call back @func with @dev as its
629
-
* argument in the workqueue's process context. @dev will be pinned until
630
-
* @func returns.
631
-
*
632
-
* This routine is usually called via the inline device_schedule_callback(),
633
-
* which automatically sets @owner to THIS_MODULE.
634
-
*
635
-
* Returns 0 if the request was submitted, -ENOMEM if storage could not
636
-
* be allocated, -ENODEV if a reference to @owner isn't available.
637
-
*
638
-
* NOTE: This routine won't work if CONFIG_SYSFS isn't set! It uses an
639
-
* underlying sysfs routine (since it is intended for use by attribute
640
-
* methods), and if sysfs isn't available you'll get nothing but -ENOSYS.
641
-
*/
642
-
int device_schedule_callback_owner(struct device *dev,
643
-
void (*func)(struct device *), struct module *owner)
644
-
{
645
-
return sysfs_schedule_callback(&dev->kobj,
646
-
(void (*)(void *)) func, dev, owner);
647
-
}
648
-
EXPORT_SYMBOL_GPL(device_schedule_callback_owner);
649
-
650
617
static void klist_children_get(struct klist_node *n)
651
618
{
652
619
struct device_private *p = to_device_private_parent(n);
+2
-2
drivers/base/dd.c
+2
-2
drivers/base/dd.c
···
187
187
return;
188
188
}
189
189
190
-
pr_debug("driver: '%s': %s: bound to device '%s'\n", dev_name(dev),
191
-
__func__, dev->driver->name);
190
+
pr_debug("driver: '%s': %s: bound to device '%s'\n", dev->driver->name,
191
+
__func__, dev_name(dev));
192
192
193
193
klist_add_tail(&dev->p->knode_driver, &dev->driver->p->klist_devices);
194
194
+1
-2
drivers/base/topology.c
+1
-2
drivers/base/topology.c
···
39
39
static ssize_t show_##name(struct device *dev, \
40
40
struct device_attribute *attr, char *buf) \
41
41
{ \
42
-
unsigned int cpu = dev->id; \
43
-
return sprintf(buf, "%d\n", topology_##name(cpu)); \
42
+
return sprintf(buf, "%d\n", topology_##name(dev->id)); \
44
43
}
45
44
46
45
#if defined(topology_thread_cpumask) || defined(topology_core_cpumask) || \
+10
-4
fs/kernfs/inode.c
+10
-4
fs/kernfs/inode.c
···
48
48
49
49
static struct kernfs_iattrs *kernfs_iattrs(struct kernfs_node *kn)
50
50
{
51
+
static DEFINE_MUTEX(iattr_mutex);
52
+
struct kernfs_iattrs *ret;
51
53
struct iattr *iattrs;
52
54
55
+
mutex_lock(&iattr_mutex);
56
+
53
57
if (kn->iattr)
54
-
return kn->iattr;
58
+
goto out_unlock;
55
59
56
60
kn->iattr = kzalloc(sizeof(struct kernfs_iattrs), GFP_KERNEL);
57
61
if (!kn->iattr)
58
-
return NULL;
62
+
goto out_unlock;
59
63
iattrs = &kn->iattr->ia_iattr;
60
64
61
65
/* assign default attributes */
···
69
65
iattrs->ia_atime = iattrs->ia_mtime = iattrs->ia_ctime = CURRENT_TIME;
70
66
71
67
simple_xattrs_init(&kn->iattr->xattrs);
72
-
73
-
return kn->iattr;
68
+
out_unlock:
69
+
ret = kn->iattr;
70
+
mutex_unlock(&iattr_mutex);
71
+
return ret;
74
72
}
75
73
76
74
static int __kernfs_setattr(struct kernfs_node *kn, const struct iattr *iattr)
+4
-1
fs/super.c
+4
-1
fs/super.c
···
800
800
801
801
static DEFINE_IDA(unnamed_dev_ida);
802
802
static DEFINE_SPINLOCK(unnamed_dev_lock);/* protects the above */
803
-
static int unnamed_dev_start = 0; /* don't bother trying below it */
803
+
/* Many userspace utilities consider an FSID of 0 invalid.
804
+
* Always return at least 1 from get_anon_bdev.
805
+
*/
806
+
static int unnamed_dev_start = 1;
804
807
805
808
int get_anon_bdev(dev_t *p)
806
809
{
-92
fs/sysfs/file.c
-92
fs/sysfs/file.c
···
453
453
kernfs_remove_by_name(kobj->sd, attr->attr.name);
454
454
}
455
455
EXPORT_SYMBOL_GPL(sysfs_remove_bin_file);
456
-
457
-
struct sysfs_schedule_callback_struct {
458
-
struct list_head workq_list;
459
-
struct kobject *kobj;
460
-
void (*func)(void *);
461
-
void *data;
462
-
struct module *owner;
463
-
struct work_struct work;
464
-
};
465
-
466
-
static struct workqueue_struct *sysfs_workqueue;
467
-
static DEFINE_MUTEX(sysfs_workq_mutex);
468
-
static LIST_HEAD(sysfs_workq);
469
-
static void sysfs_schedule_callback_work(struct work_struct *work)
470
-
{
471
-
struct sysfs_schedule_callback_struct *ss = container_of(work,
472
-
struct sysfs_schedule_callback_struct, work);
473
-
474
-
(ss->func)(ss->data);
475
-
kobject_put(ss->kobj);
476
-
module_put(ss->owner);
477
-
mutex_lock(&sysfs_workq_mutex);
478
-
list_del(&ss->workq_list);
479
-
mutex_unlock(&sysfs_workq_mutex);
480
-
kfree(ss);
481
-
}
482
-
483
-
/**
484
-
* sysfs_schedule_callback - helper to schedule a callback for a kobject
485
-
* @kobj: object we're acting for.
486
-
* @func: callback function to invoke later.
487
-
* @data: argument to pass to @func.
488
-
* @owner: module owning the callback code
489
-
*
490
-
* sysfs attribute methods must not unregister themselves or their parent
491
-
* kobject (which would amount to the same thing). Attempts to do so will
492
-
* deadlock, since unregistration is mutually exclusive with driver
493
-
* callbacks.
494
-
*
495
-
* Instead methods can call this routine, which will attempt to allocate
496
-
* and schedule a workqueue request to call back @func with @data as its
497
-
* argument in the workqueue's process context. @kobj will be pinned
498
-
* until @func returns.
499
-
*
500
-
* Returns 0 if the request was submitted, -ENOMEM if storage could not
501
-
* be allocated, -ENODEV if a reference to @owner isn't available,
502
-
* -EAGAIN if a callback has already been scheduled for @kobj.
503
-
*/
504
-
int sysfs_schedule_callback(struct kobject *kobj, void (*func)(void *),
505
-
void *data, struct module *owner)
506
-
{
507
-
struct sysfs_schedule_callback_struct *ss, *tmp;
508
-
509
-
if (!try_module_get(owner))
510
-
return -ENODEV;
511
-
512
-
mutex_lock(&sysfs_workq_mutex);
513
-
list_for_each_entry_safe(ss, tmp, &sysfs_workq, workq_list)
514
-
if (ss->kobj == kobj) {
515
-
module_put(owner);
516
-
mutex_unlock(&sysfs_workq_mutex);
517
-
return -EAGAIN;
518
-
}
519
-
mutex_unlock(&sysfs_workq_mutex);
520
-
521
-
if (sysfs_workqueue == NULL) {
522
-
sysfs_workqueue = create_singlethread_workqueue("sysfsd");
523
-
if (sysfs_workqueue == NULL) {
524
-
module_put(owner);
525
-
return -ENOMEM;
526
-
}
527
-
}
528
-
529
-
ss = kmalloc(sizeof(*ss), GFP_KERNEL);
530
-
if (!ss) {
531
-
module_put(owner);
532
-
return -ENOMEM;
533
-
}
534
-
kobject_get(kobj);
535
-
ss->kobj = kobj;
536
-
ss->func = func;
537
-
ss->data = data;
538
-
ss->owner = owner;
539
-
INIT_WORK(&ss->work, sysfs_schedule_callback_work);
540
-
INIT_LIST_HEAD(&ss->workq_list);
541
-
mutex_lock(&sysfs_workq_mutex);
542
-
list_add_tail(&ss->workq_list, &sysfs_workq);
543
-
mutex_unlock(&sysfs_workq_mutex);
544
-
queue_work(sysfs_workqueue, &ss->work);
545
-
return 0;
546
-
}
547
-
EXPORT_SYMBOL_GPL(sysfs_schedule_callback);
+1
-10
include/linux/device.h
+1
-10
include/linux/device.h
···
566
566
const struct bin_attribute *attr);
567
567
extern void device_remove_bin_file(struct device *dev,
568
568
const struct bin_attribute *attr);
569
-
extern int device_schedule_callback_owner(struct device *dev,
570
-
void (*func)(struct device *dev), struct module *owner);
571
-
572
-
/* This is a macro to avoid include problems with THIS_MODULE */
573
-
#define device_schedule_callback(dev, func) \
574
-
device_schedule_callback_owner(dev, func, THIS_MODULE)
575
569
576
570
/* device resource management */
577
571
typedef void (*dr_release_t)(struct device *dev, void *res);
···
926
932
extern struct device *__root_device_register(const char *name,
927
933
struct module *owner);
928
934
929
-
/*
930
-
* This is a macro to avoid include problems with THIS_MODULE,
931
-
* just as per what is done for device_schedule_callback() above.
932
-
*/
935
+
/* This is a macro to avoid include problems with THIS_MODULE */
933
936
#define root_device_register(name) \
934
937
__root_device_register(name, THIS_MODULE)
935
938
-9
include/linux/sysfs.h
-9
include/linux/sysfs.h
···
179
179
180
180
#ifdef CONFIG_SYSFS
181
181
182
-
int sysfs_schedule_callback(struct kobject *kobj, void (*func)(void *),
183
-
void *data, struct module *owner);
184
-
185
182
int __must_check sysfs_create_dir_ns(struct kobject *kobj, const void *ns);
186
183
void sysfs_remove_dir(struct kobject *kobj);
187
184
int __must_check sysfs_rename_dir_ns(struct kobject *kobj, const char *new_name,
···
251
254
}
252
255
253
256
#else /* CONFIG_SYSFS */
254
-
255
-
static inline int sysfs_schedule_callback(struct kobject *kobj,
256
-
void (*func)(void *), void *data, struct module *owner)
257
-
{
258
-
return -ENOSYS;
259
-
}
260
257
261
258
static inline int sysfs_create_dir_ns(struct kobject *kobj, const void *ns)
262
259
{