61 lines
2.3 KiB
ReStructuredText
61 lines
2.3 KiB
ReStructuredText
|
.. SPDX-License-Identifier: GPL-2.0
|
|||
|
|
|
|||
|
|
.. include:: ../disclaimer-zh_CN.rst
|
|||
|
|
|
|||
|
|
:Original: Documentation/driver-api/io_ordering.rst
|
|||
|
|
|
|||
|
|
:翻译:
|
|||
|
|
|
|||
|
|
林永听 Lin Yongting <linyongting@gmail.com>
|
|||
|
|
司延腾 Yanteng Si <siyanteng@loongson.cn>
|
|||
|
|
|
|||
|
|
:校译:
|
|||
|
|
|
|||
|
|
===========================
|
|||
|
|
对内存映射地址的I/O写入排序
|
|||
|
|
===========================
|
|||
|
|
|
|||
|
|
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
|
|||
|
|
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”
|
|||
|
|
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，
|
|||
|
|
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。
|
|||
|
|
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
|
|||
|
|
屏障操作，mb()，不过仅适用于I/O）。
|
|||
|
|
|
|||
|
|
假设一个设备驱动程的具体例子::
|
|||
|
|
|
|||
|
|
...
|
|||
|
|
CPU A: spin_lock_irqsave(&dev_lock, flags)
|
|||
|
|
CPU A: val = readl(my_status);
|
|||
|
|
CPU A: ...
|
|||
|
|
CPU A: writel(newval, ring_ptr);
|
|||
|
|
CPU A: spin_unlock_irqrestore(&dev_lock, flags)
|
|||
|
|
...
|
|||
|
|
CPU B: spin_lock_irqsave(&dev_lock, flags)
|
|||
|
|
CPU B: val = readl(my_status);
|
|||
|
|
CPU B: ...
|
|||
|
|
CPU B: writel(newval2, ring_ptr);
|
|||
|
|
CPU B: spin_unlock_irqrestore(&dev_lock, flags)
|
|||
|
|
...
|
|||
|
|
|
|||
|
|
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
|
|||
|
|
发生了。不过很容易通过下面方法来修复::
|
|||
|
|
|
|||
|
|
...
|
|||
|
|
CPU A: spin_lock_irqsave(&dev_lock, flags)
|
|||
|
|
CPU A: val = readl(my_status);
|
|||
|
|
CPU A: ...
|
|||
|
|
CPU A: writel(newval, ring_ptr);
|
|||
|
|
CPU A: (void)readl(safe_register); /* 配置寄存器？*/
|
|||
|
|
CPU A: spin_unlock_irqrestore(&dev_lock, flags)
|
|||
|
|
...
|
|||
|
|
CPU B: spin_lock_irqsave(&dev_lock, flags)
|
|||
|
|
CPU B: val = readl(my_status);
|
|||
|
|
CPU B: ...
|
|||
|
|
CPU B: writel(newval2, ring_ptr);
|
|||
|
|
CPU B: (void)readl(safe_register); /* 配置寄存器？*/
|
|||
|
|
CPU B: spin_unlock_irqrestore(&dev_lock, flags)
|
|||
|
|
|
|||
|
|
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，
|
|||
|
|
再处理后面的读操作，防止引发数据不一致问题。
|