Eric Lee / smarc-ti-linux-kernel

Download as
Browse Code »

Commit 60fbf2bda140f27b0e9ab5b6d17342c9a5f9eacf

Authored by Linus Torvalds
2 parents 8cb652bb10 94f8cc0eea
Exists in master and in 13 other branches dlt-processor-sdk-linux-03.00.00.04, processor-sdk-linux-02.00.01, smarc-ti-linux-3.15.y, smarc-ti-lsk-linux-4.1.y, smarct3x-processor-sdk-04.01.00.06, smarct3x-processor-sdk-linux-02.00.01, smarct3x-processor-sdk-linux-03.00.00.04, smarct4x-800-processor-sdk-linux-02.00.01, smarct4x-processor-sdk-04.01.00.06, smarct4x-processor-sdk-linux-02.00.01, smarct4x-processor-sdk-linux-03.00.00.04, ti-linux-3.15.y, ti-lsk-linux-4.1.y

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

Showing 14 changed files Side-by-side Diff

Documentation/ja_JP/HOWTO
Diff comments   View file @ 60fbf2b
... ... @@ -315,7 +315,7 @@
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。
Documentation/ja_JP/stable_kernel_rules.txt
Diff comments   View file @ 60fbf2b
... ... @@ -50,16 +50,16 @@
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-
Documentation/stable_kernel_rules.txt
Diff comments   View file @ 60fbf2b
... ... @@ -39,7 +39,7 @@
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
Documentation/zh_CN/HOWTO
Diff comments   View file @ 60fbf2b
... ... @@ -237,7 +237,7 @@
237 237 如果没有2.6.x.y版本内核存在，那么最新的2.6.x版本内核就相当于是当前的稳定
238 238 版内核。
239 239  
240   -2.6.x.y版本由“稳定版”小组（邮件地址<stable@kernel.org>）维护，一般隔周发
  240 +2.6.x.y版本由“稳定版”小组（邮件地址<stable@vger.kernel.org>）维护，一般隔周发
241 241 布新版本。
242 242  
243 243 内核源码中的Documentation/stable_kernel_rules.txt文件具体描述了可被稳定
Documentation/zh_CN/io_ordering.txt
Diff comments   View file @ 60fbf2b
  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。
Documentation/zh_CN/stable_kernel_rules.txt
Diff comments   View file @ 60fbf2b
... ... @@ -42,7 +42,7 @@
42 42  
43 43 向稳定版代码树提交补丁的过程：
44 44  
45   - - 在确认了补丁符合以上的规则后，将补丁发送到stable@kernel.org。
  45 + - 在确认了补丁符合以上的规则后，将补丁发送到stable@vger.kernel.org。
46 46 - 如果补丁被接受到队列里，发送者会收到一个ACK回复，如果没有被接受，收
47 47 到的是NAK回复。回复需要几天的时间，这取决于开发者的时间安排。
48 48 - 被接受的补丁会被加到稳定版本队列里，等待其他开发者的审查。
... ... @@ -614,39 +614,6 @@
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);
... ... @@ -187,8 +187,8 @@
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  
drivers/base/topology.c
Diff comments   View file @ 60fbf2b
... ... @@ -39,8 +39,7 @@
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) || \
... ... @@ -48,14 +48,18 @@
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 */
... ... @@ -65,8 +69,10 @@
65 69 iattrs->ia_atime = iattrs->ia_mtime = iattrs->ia_ctime = CURRENT_TIME;
66 70  
67 71 simple_xattrs_init(&kn->iattr->xattrs);
68   -
69   - return kn->iattr;
  72 +out_unlock:
  73 + ret = kn->iattr;
  74 + mutex_unlock(&iattr_mutex);
  75 + return ret;
70 76 }
71 77  
72 78 static int __kernfs_setattr(struct kernfs_node *kn, const struct iattr *iattr)
... ... @@ -800,7 +800,10 @@
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 {
... ... @@ -453,96 +453,4 @@
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);
include/linux/device.h
Diff comments   View file @ 60fbf2b
... ... @@ -566,13 +566,7 @@
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);
571 569  
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   -
576 570 /* device resource management */
577 571 typedef void (*dr_release_t)(struct device *dev, void *res);
578 572 typedef int (*dr_match_t)(struct device *dev, void *res, void *match_data);
... ... @@ -932,10 +926,7 @@
932 926 extern struct device *__root_device_register(const char *name,
933 927 struct module *owner);
934 928  
935   -/*
936   - * This is a macro to avoid include problems with THIS_MODULE,
937   - * just as per what is done for device_schedule_callback() above.
938   - */
  929 +/* This is a macro to avoid include problems with THIS_MODULE */
939 930 #define root_device_register(name) \
940 931 __root_device_register(name, THIS_MODULE)
941 932  
include/linux/sysfs.h
Diff comments   View file @ 60fbf2b
... ... @@ -179,9 +179,6 @@
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,
... ... @@ -254,12 +251,6 @@
254 251 }
255 252  
256 253 #else /* CONFIG_SYSFS */
257   -
258   -static inline int sysfs_schedule_callback(struct kobject *kobj,
259   - void (*func)(void *), void *data, struct module *owner)
260   -{
261   - return -ENOSYS;
262   -}
263 254  
264 255 static inline int sysfs_create_dir_ns(struct kobject *kobj, const void *ns)
265 256 {