Documentationspispi-summary.txt

 Chinese translated version of Documentationspispi-summary.txt

If you have any comment or update to the content, please contact the
original document maintainer directly.  However, if you have a problem
communicating in English you can also ask the Chinese maintainer for
help.  Contact the Chinese maintainer if this translation is outdated
or if there is a problem with the translation.

Chinese maintainer: majg <1519014266@qq.com>
---------------------------------------------------------------------
Documentationspispi-summary.txt 的中文翻译

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
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
译存在问题，请联系中文版维护者。

中文版维护者： 马建刚  majg <1519014266@qq.com>
中文版翻译者： 马建刚  majg <1519014266@qq.com>
中文版校译者： 马建刚  majg <1519014266@qq.com>

 

 

以下为正文
---------------------------------------------------------------------
  1 Overview of Linux kernel SPI support
-概述Linux内核SPI支持
  2 ====================================
  3
  4 02-Feb-2012
  5
  6 What is SPI?
  7 ------------
  8 The "Serial Peripheral Interface" (SPI) is a synchronous four wire serial
  9 link used to connect microcontrollers to sensors, memory, and peripherals.
 10 It's a simple "de facto" standard, not complicated enough to acquire a
 11 standardization body.  SPI uses a master/slave configuration.
-“串行外设接口（SPI）是一个同步的四线串行链路用于连接微控制器，传感器，
存储器和外设。这是一个简单的“事实”的标准，不够复杂，收购的标准化机构。
SPI采用了主/从配置。
 12
 13 The three signal wires hold a clock (SCK, often on the order of 10 MHz),
 14 and parallel data lines with "Master Out, Slave In" (MOSI) or "Master In,
 15 Slave Out" (MISO) signals.  (Other names are also used.)  There are four
 16 clocking modes through which data is exchanged; mode-0 and mode-3 are most
 17 commonly used.  Each clock cycle shifts data out and data in; the clock
 18 doesn't cycle except when there is a data bit to shift.  Not all data bits
 19 are used though; not every protocol uses those full duplex capabilities.
-这三个信号线保持一个时钟（SCK，通常的顺序为10 MHz），和“主输出，从输入（MOSI）或”主输入，
从输出（MISO）信号的并行数据线。 （也可用于其他名字）有四种时钟模式，通过该交换数据时，模
式0和模式3是最常用的。每个时钟周期将数据移出和数据，时钟不循环除非有一个数据位转移。不是
所有的数据位虽然不是每个协议使用这些全双工能力。
 20
 21 SPI masters use a fourth "chip select" line to activate a given SPI slave
 22 device, so those three signal wires may be connected to several chips
 23 in parallel.  All SPI slaves support chipselects; they are usually active
 24 low signals, labeled nCSx for slave 'x' (e.g. nCS0).  Some devices have
 25 other signals, often including an interrupt to the master.
-SPI主人使用第四个“片选”行，为了激活特定的SPI从属设备，所以这三个信号线，
可以连接到并行的几个芯片。所有SPI奴隶的支持chipselects，他们通常是活跃的
低信号，标记NCSX的奴隶'X'（例如NCS0）。有些设备有其他信号，通常包括一
个主中断。
 26
 27 Unlike serial busses like USB or SMBus, even low level protocols for
 28 SPI slave functions are usually not interoperable between vendors
 29 (except for commodities like SPI memory chips).
-不同于串行总线像USB或SMBus，即使是低级别的协议通常不为SPI从属功能
不同厂商之间的互操作（为SPI存储器芯片等商品除外）。
 30
 31   - SPI may be used for request/response style device protocols, as with
 32     touchscreen sensors and memory chips.
 33
 34   - It may also be used to stream data in either direction (half duplex),
 35     or both of them at the same time (full duplex).
 36
 37   - Some devices may use eight bit words.  Others may different word
 38     lengths, such as streams of 12-bit or 20-bit digital samples.
 39
 40   - Words are usually sent with their most significant bit (MSB) first,
 41     but sometimes the least significant bit (LSB) goes first instead.
 42
 43   - Sometimes SPI is used to daisy-chain devices, like shift registers.
 44
 45 In the same way, SPI slaves will only rarely support any kind of automatic
 46 discovery/enumeration protocol.  The tree of slave devices accessible from
 47 a given SPI master will normally be set up manually, with configuration
 48 tables.
-以同样的方式，SPI slaves很少会支持任何种类的自动发现/枚举协议。可以从一个给定的SPI主从设备的树通常会设置手动配置表。
 49
 50 SPI is only one of the names used by such four-wire protocols, and
 51 most controllers have no problem handling "MicroWire" (think of it as
 52 half-duplex SPI, for request/response protocols), SSP ("Synchronous
 53 Serial Protocol"), PSP ("Programmable Serial Protocol"), and other
 54 related protocols.
-SPI是唯一一个四线协议所使用的名称，和大多数控制器都没有问题处理“微丝”（认为它是半双工SPI，请求/响应协议），
SSP（同步串行协议“）PSP（“可编程串行议定书”的）​​，以及其他相关协议。
 55
 56 Some chips eliminate a signal line by combining MOSI and MISO, and
 57 limiting themselves to half-duplex at the hardware level.  In fact
 58 some SPI chips have this signal mode as a strapping option.  These
 59 can be accessed using the same programming interface as SPI, but of
 60 course they won't handle full duplex transfers.  You may find such
 61 chips described as using "three wire" signaling: SCK, data, nCSx.
 62 (That data line is sometimes called MOMI or SISO.)
-一些芯片消除了信号线相结合，MOSI和MISO，在硬件级别限制半双工。事实上，
一些的SPI芯片有这个信号模式作为一个魁梧的选项。这些都可以使用相同的编程
接口为SPI访问，但他们当然不会处理全双工传输。您可能会发现这样的芯片，使
用“三丝”的信号：数据线SCK，数据，NCSX（有时也被称为MOMI或SISO）。
 63
 64 Microcontrollers often support both master and slave sides of the SPI
 65 protocol.  This document (and Linux) currently only supports the master
 66 side of SPI interactions.
-微控制器通常支持主从双方的SPI协议。这个文件（和Linux）目前只支持SPI相互作用的主侧。
 67
 68
 69 Who uses it?  On what kinds of systems?
-谁使用它？什么样的系统？
 70 ---------------------------------------
 71 Linux developers using SPI are probably writing device drivers for embedded
 72 systems boards.  SPI is used to control external chips, and it is also a
 73 protocol supported by every MMC or SD memory card.  (The older "DataFlash"
 74 cards, predating MMC cards but using the same connectors and card shape,
 75 support only SPI.)  Some PC hardware uses SPI flash for BIOS code.
-Linux开发人员使用SPI很可能编写设备驱动程序，嵌入式系统板。SPI用于控制外部芯片，
并且它也是一个协议得到每MMC或SD存储卡。（旧的“数据闪存卡”，早于MMC卡，但使
用相同的连接器和卡的形状，只支持SPI）某些PC硬件的BIOS代码使用SPI闪存。
 76
 77 SPI slave chips range from digital/analog converters used for analog
 78 sensors and codecs, to memory, to peripherals like USB controllers
 79 or Ethernet adapters; and more.
-SPI从数字/模拟转换器芯片的范围从用于模拟传感器和编解码器，内存，外围设备，如USB控制器或以太网适配器等等。
 80
 81 Most systems using SPI will integrate a few devices on a mainboard.
 82 Some provide SPI links on expansion connectors; in cases where no
 83 dedicated SPI controller exists, GPIO pins can be used to create a
 84 low speed "bitbanging" adapter.  Very few systems will "hotplug" an SPI
 85 controller; the reasons to use SPI focus on low cost and simple operation,
 86 and if dynamic reconfiguration is important, USB will often be a more
 87 appropriate low-pincount peripheral bus.
-大多数系统使用SPI少数设备集成在主板上。一些提供SPI扩展连接器上的链接，
在没有专门的SPI控制器存在的情况下，GPIO引脚可以用来创建一个低速的“bitbanging”适配器。
很少有系统“热插拔”的SPI控制器;，使用SPI专注于成本低，操作简单的原因，
如果动态重新配置是很重要的，USB往往会是一个更合适的低引脚数外设总线。
 88
 89 Many microcontrollers that can run Linux integrate one or more I/O
 90 interfaces with SPI modes.  Given SPI support, they could use MMC or SD
 91 cards without needing a special purpose MMC/SD/SDIO controller.
-许多微控制器可以运行Linux集成一个或多个I / O接口与SPI模式。
由于SPI支持，他们可以使用MMC或SD卡，无需特殊目的MMC/ SD/ SDIO控制器。
 92
 93
 94 I'm confused.  What are these four SPI "clock modes"?
-我很困惑。这四个SPI时钟模式是什么“？
 95 -----------------------------------------------------
 96 It's easy to be confused here, and the vendor documentation you'll
 97 find isn't necessarily helpful.  The four modes combine two mode bits:
-这很容易被搞糊涂了，和供应商的文档，你会发现不一定有帮助。这四种模式结合了两种模式的位：
 98
 99  - CPOL indicates the initial clock polarity.  CPOL=0 means the
100    clock starts low, so the first (leading) edge is rising, and
101    the second (trailing) edge is falling.  CPOL=1 means the clock
102    starts high, so the first (leading) edge is falling.
-CPOL表示初始时钟极性。CPOL = 0意味着时钟开始比较低，所以第一（领先）的边缘呈上升趋势，
第二（后）沿下降。CPOL=1表示时钟开始高，所以第一（领先）的边缘下降。
103
104  - CPHA indicates the clock phase used to sample data; CPHA=0 says
105    sample on the leading edge, CPHA=1 means the trailing edge.
-CPHA表示时钟短语用于样本数据; CPHA = 0表示样品上的领先优势，CPHA= 1表示下降沿。
106
107    Since the signal needs to stablize before it's sampled, CPHA=0
108    implies that its data is written half a clock before the first
109    clock edge.  The chipselect may have made it become available.
-由于信号需要稳定下来之前，它的采样，CPHA= 0意味着它的数据写入第一个时钟沿前半个时钟。片选可能已经成为可用的。
110
111 Chip specs won't always say "uses SPI mode X" in as many words,
112 but their timing diagrams will make the CPOL and CPHA modes clear.
-芯片规格并不总是会说：“采用SPI模式X”在尽可能多的单词，但他们的时序图CPOL和CPHA模式的清晰。
113
114 In the SPI mode number, CPOL is the high order bit and CPHA is the
115 low order bit.  So when a chip's timing diagram shows the clock
116 starting low (CPOL=0) and data stabilized for sampling during the
117 trailing clock edge (CPHA=1), that's SPI mode 1.
-在SPI模式下号，CPOL是高位，和CPHA是低位位。所以，当一个芯片的时序图显示时钟
起点低（CPOL = 0）和采样数据稳定在后时钟沿（CPHA =1），这是SPI模式1。
118
119 Note that the clock mode is relevant as soon as the chipselect goes
120 active.  So the master must set the clock to inactive before selecting
121 a slave, and the slave can tell the chosen polarity by sampling the
122 clock level when its select line goes active.  That's why many devices
123 support for example both modes 0 and 3:  they don't care about polarity,
124 and alway clock data in/out on rising clock edges.
-请注意，时钟模式是尽快片选goes活跃相关。因此，主机必须设置时钟为无效，
然后再选择一个奴隶，奴隶可以告诉由采样时钟的水平时，其选线变为主动选择
的极性。这就是为什么很多设备支持，例如模式0和3：他们不关心极性，常常时
钟数据输入/输出时钟的上升沿。
125
126
127 How do these driver programming interfaces work?
-这些驱动程序的编程接口是怎样工作的？
128 ------------------------------------------------
129 The <linux/spi/spi.h> header file includes kerneldoc, as does the
130 main source code, and you should certainly read that chapter of the
131 kernel API document.  This is just an overview, so you get the big
132 picture before those details.
-头文件<linux/spi/spi.h>包括kerneldoc的，做的主要源代码，你当然应该读
这一章的内核API文档。这只是一个概述，所以你获得大画面之前，这些细节。
133
134 SPI requests always go into I/O queues.  Requests for a given SPI device
135 are always executed in FIFO order, and complete asynchronously through
136 completion callbacks.  There are also some simple synchronous wrappers
137 for those calls, including ones for common transaction types like writing
138 a command and then reading its response.
-SPI请求总是进入I / O队列。对于一个给定的SPI器件的请求总是执行先进先出的顺序，
并通过完成回调异步完成。也有一些简单的同步这些调用的包装，包括常见的交易类型，
如写一个命令，然后读取其响应。
139
140 There are two types of SPI driver, here called:
-有两种类型的SPI驱动程序，在这里被称为：
141
142   Controller drivers ... controllers may be built in to System-On-Chip
143         processors, and often support both Master and Slave roles.
144         These drivers touch hardware registers and may use DMA.
145         Or they can be PIO bitbangers, needing just GPIO pins.
-控制器驱动程序...控制器可建在片上系统处理器，并经常同时支持Master和Slave的角色。
这些驱动程序触摸硬件寄存器，并且可以使用DMA。或者，他们可以PIO bitbangers，只需要GPIO引脚。
146
147   Protocol drivers ... these pass messages through the controller
148         driver to communicate with a Slave or Master device on the
149         other side of an SPI link.
-协议驱动程序...这些通消息通过控制器驱动程序沟通与奴隶或主设备在另一侧的一个SPI链接。
150
151 So for example one protocol driver might talk to the MTD layer to export
152 data to filesystems stored on SPI flash like DataFlash; and others might
153 control audio interfaces, present touchscreen sensors as input interfaces,
154 or monitor temperature and voltage levels during industrial processing.
155 And those might all be sharing the same controller driver.
-因此，例如一个协议驱动程序可能谈论的MTD层将数据导出到存储在SPI闪存喜欢的DataFlash的文件系统;
和其他工业加工过程中控制音频接口，目前触摸屏传感器输入接口，或显示器温度和电压等级。
和那些可能都可以共享相同的控制器驱动程序。
156
157 A "struct spi_device" encapsulates the master-side interface between
158 those two types of driver.  At this writing, Linux has no slave side
159 programming interface.
-A“struct spi_device的”这两种类型的驱动程序之间的主端接口封装。在撰写本文时，Linux有没有从属端的编程接口。
160
161 There is a minimal core of SPI programming interfaces, focussing on
162 using the driver model to connect controller and protocol drivers using
163 device tables provided by board specific initialization code.  SPI
164 shows up in sysfs in several locations:
-SPI编程接口有一个最小的核心，侧重于使用驱动模型连接控制器和协议驱动程序，
使用设备提供的表板特定的初始化代码。SPI显示在sysfs中在几个地点：
165
166    /sys/devices/.../CTLR ... physical node for a given SPI controller
167
168    /sys/devices/.../CTLR/spiB.C ... spi_device on bus "B",
169         chipselect C, accessed through CTLR.
170
171    /sys/bus/spi/devices/spiB.C ... symlink to that physical
172         .../CTLR/spiB.C device
173
174    /sys/devices/.../CTLR/spiB.C/modalias ... identifies the driver
175         that should be used with this device (for hotplug/coldplug)
176
177    /sys/bus/spi/drivers/D ... driver for one or more spi*.* devices
178
179    /sys/class/spi_master/spiB ... symlink (or actual device node) to
180         a logical node which could hold class related state for the
181         controller managing bus "B".  All spiB.* devices share one
182         physical SPI bus segment, with SCLK, MOSI, and MISO.
183
184 Note that the actual location of the controller's class state depends
185 on whether you enabled CONFIG_SYSFS_DEPRECATED or not.  At this time,
186 the only class-specific state is the bus number ("B" in "spiB"), so
187 those /sys/class entries are only useful to quickly identify busses.
-需要注意的是控制器的阶级的国家的实际位置取决于您是否启用或不CONFIG_SYSFS_DEPRECATED。
这时，唯一的类特定状态的总线号（“B”在SPIB“），让那些/ sys / class中的条目是唯一有用的快速识别总线。
188
189
190 How does board-specific init code declare SPI devices?
-板特定的初始化代码如何申报SPI器件？
191 ------------------------------------------------------
192 Linux needs several kinds of information to properly configure SPI devices.
193 That information is normally provided by board-specific code, even for
194 chips that do support some of automated discovery/enumeration.
-Linux需要正确配置SPI设备的多种信息。信息一般是由董事会特定的代码，甚至芯片，支持一些自动发现/枚举。
195
196 DECLARE CONTROLLERS
-DECLARE控制器
197
198 The first kind of information is a list of what SPI controllers exist.
199 For System-on-Chip (SOC) based boards, these will usually be platform
200 devices, and the controller may need some platform_data in order to
201 operate properly.  The "struct platform_device" will include resources
202 like the physical address of the controller's first register and its IRQ.
-第一类信息是SPI控制器存在的列表。对于片上系统（SOC）为基础的电路板，
这些通常是平台设备，控制器可能需要一些的platform_data的，为了正常运行。
“结构平台”将包括控制器的寄存器的物理地址和IRQ等资源。
203
204 Platforms will often abstract the "register SPI controller" operation,
205 maybe coupling it with code to initialize pin configurations, so that
206 the arch/.../mach-*/board-*.c files for several boards can all share the
207 same basic controller setup code.  This is because most SOCs have several
208 SPI-capable controllers, and only the ones actually usable on a given
209 board should normally be set up and registered.
-平台往往将抽象的“寄存器SPI控制器”操作，也许耦合代码初始化引脚配置，
使的 arch/.../mach-*/board-*.c文件几板都可以共享相同的基本控制器设置代码。
这是因为大多数SoC有几个的SPI功能控制器，只有那些实际可用在给定的电路板通常应设立和登记。
210
211 So for example arch/.../mach-*/board-*.c files might have code like:
-例如 arch/.../mach-*/board-*.c 文件可能有这样的代码：
212
213         #include <mach/spi.h>   /* for mysoc_spi_data */
214
215         /* if your mach-* infrastructure doesn't support kernels that can
216          * run on multiple boards, pdata wouldn't benefit from "__init".
217          */
218         static struct mysoc_spi_data __initdata pdata = { ... };
219
220         static __init board_init(void)
221         {
222                 ...
223                 /* this board only uses SPI controller #2 */
224                 mysoc_register_spi(2, &pdata);
225                 ...
226         }
227
228 And SOC-specific utility code might look something like:
-并且SOC特定的实用程序代码看起来可能是这样的：
229
230         #include <mach/spi.h>
231
232         static struct platform_device spi2 = { ... };
233
234         void mysoc_register_spi(unsigned n, struct mysoc_spi_data *pdata)
235         {
236                 struct mysoc_spi_data *pdata2;
237
238                 pdata2 = kmalloc(sizeof *pdata2, GFP_KERNEL);
239                 *pdata2 = pdata;
240                 ...
241                 if (n == 2) {
242                         spi2->dev.platform_data = pdata2;
243                         register_platform_device(&spi2);
244
245                         /* also: set up pin modes so the spi2 signals are
246                          * visible on the relevant pins ... bootloaders on
247                          * production boards may already have done this, but
248                          * developer boards will often need Linux to do it.
249                          */
250                 }
251                 ...
252         }
253
254 Notice how the platform_data for boards may be different, even if the
255 same SOC controller is used.  For example, on one board SPI might use
256 an external clock, where another derives the SPI clock from current
257 settings of some master clock.
-请注意，在板的platform_data可能不同，即使相同的SOC控制器使用。
例如，在一个板上SPI可能会使用外部时钟，另一个派生的SPI时钟从一些主时钟的当前设置。
258
259
260 DECLARE SLAVE DEVICES
-DECLARE从设备
261
262 The second kind of information is a list of what SPI slave devices exist
263 on the target board, often with some board-specific data needed for the
264 driver to work correctly.
-第二类信息是什么SPI从器件在目标板上存在，往往与一些特定板的驱动程序才能正常工作所需要的数据的列表。
265
266 Normally your arch/.../mach-*/board-*.c files would provide a small table
267 listing the SPI devices on each board.  (This would typically be only a
268 small handful.)  That might look like:
-通常情况下，你的arch/.../mach-*/board-*.c文件将列出每块电路板上的SPI设备提供一块
小的空间。（这通常是只有少数），可能看起来像：
269
270         static struct ads7846_platform_data ads_info = {
271                 .vref_delay_usecs       = 100,
272                 .x_plate_ohms           = 580,
273                 .y_plate_ohms           = 410,
274         };
275
276         static struct spi_board_info spi_board_info[] __initdata = {
277         {
278                 .modalias       = "ads7846",
279                 .platform_data  = &ads_info,
280                 .mode           = SPI_MODE_0,
281                 .irq            = GPIO_IRQ(31),
282                 .max_speed_hz   = 120000 /* max sample rate at 3V */ * 16,
283                 .bus_num        = 1,
284                 .chip_select    = 0,
285         },
286         };
287
288 Again, notice how board-specific information is provided; each chip may need
289 several types.  This example shows generic constraints like the fastest SPI
290 clock to allow (a function of board voltage in this case) or how an IRQ pin
291 is wired, plus chip-specific constraints like an important delay that's
292 changed by the capacitance at one pin.
-再次，注意特定板提供的信息，每个芯片可能需要几种类型。这个例子显示了通用的限制，
例如速度最快的SPI时钟允许（在这种情况下，一个功能板上电压）或IRQ引脚有线如何，
再加上芯片的具体限制，例如在一个引脚的电容变化的一个重要的延迟。
293
294 (There's also "controller_data", information that may be useful to the
295 controller driver.  An example would be peripheral-specific DMA tuning
296 data or chipselect callbacks.  This is stored in spi_device later.)
297
298 The board_info should provide enough information to let the system work
299 without the chip's driver being loaded.  The most troublesome aspect of
300 that is likely the SPI_CS_HIGH bit in the spi_device.mode field, since
301 sharing a bus with a device that interprets chipselect "backwards" is
302 not possible until the infrastructure knows how to deselect it.
-board_info应该提供足够的信息，让系统芯片的驱动程序被加载的情况下工作。
最麻烦的方面，可能是的SPI_CS_HIGH位在spi_device.mode领域，因为片选"backwards" 是不可能的，
直到基础设施知道如何取消它解释设备共享的总线。
303
304 Then your board initialization code would register that table with the SPI
305 infrastructure, so that it's available later when the SPI master controller
306 driver is registered:
-板初始化代码将注册该表与SPI基础设施，因此，它是可注册后，SPI主控制器驱动程序：
307
308         spi_register_board_info(spi_board_info, ARRAY_SIZE(spi_board_info));
309
310 Like with other static board-specific setup, you won't unregister those.
-与其他的静态设立特定板一样，你不会注销。
311
312 The widely used "card" style computers bundle memory, cpu, and little else
313 onto a card that's maybe just thirty square centimeters.  On such systems,
314 your arch/.../mach-.../board-*.c file would primarily provide information
315 about the devices on the mainboard into which such a card is plugged.  That
316 certainly includes SPI devices hooked up through the card connectors!
-被广泛使用的“一卡通”式电脑捆绑内存，CPU，以及其他小到一张卡片，也许只是30平方厘米​​。
在这样的系统中arch/.../mach-.../board-*.c 文件将主要提供信息到这样的卡插在主板上的设备。
这当然也包括SPI器件通过卡连接器连接！
317
318
319 NON-STATIC CONFIGURATIONS
-非静态的配置
320
321 Developer boards often play by different rules than product boards, and one
322 example is the potential need to hotplug SPI devices and/or controllers.
-开发板往往能起到比产品板不同的规则，一个例子是SPI器件和/或控制器热插拔的潜在需求。
323
324 For those cases you might need to use spi_busnum_to_master() to look
325 up the spi bus master, and will likely need spi_new_device() to provide the
326 board info based on the board that was hotplugged.  Of course, you'd later
327 call at least spi_unregister_device() when that board is removed.
-对于那些情况下，你可能需要使用spi_busnum_to_master（）来查找SPI总线主，
可能会需要spi_new_device（）向董事会提供信息的基础上热插拔的板。当然，
你以后调用至少spi_unregister_device()该板删除。
328
329 When Linux includes support for MMC/SD/SDIO/DataFlash cards through SPI, those
330 configurations will also be dynamic.  Fortunately, such devices all support
331 basic device identification probes, so they should hotplug normally.
-当Linux包括通过SPI支持MMC/ SD/ SDIO/数据闪存卡，这些配置也将是动态的。
幸运的是，这样的设备都支持基本的设备识别探头，所以他们应该正常热插拔。
332
333
334 How do I write an "SPI Protocol Driver"?
-我怎样写“SPI协议驱动程序”？
335 ----------------------------------------
336 Most SPI drivers are currently kernel drivers, but there's also support
337 for userspace drivers.  Here we talk only about kernel drivers.
-目前大多数的SPI驱动程序内核驱动程序，但也有用户空间的驱动器的支持。在这里，我们只谈内核驱动程序。
338
339 SPI protocol drivers somewhat resemble platform device drivers:
-SPI协议驱动程序有点像平台的设备驱动程序：
340
341         static struct spi_driver CHIP_driver = {
342                 .driver = {
343                         .name           = "CHIP",
344                         .owner          = THIS_MODULE,
345                 },
346
347                 .probe          = CHIP_probe,
348                 .remove         = CHIP_remove,
349                 .suspend        = CHIP_suspend,
350                 .resume         = CHIP_resume,
351         };
352
353 The driver core will automatically attempt to bind this driver to any SPI
354 device whose board_info gave a modalias of "CHIP".  Your probe() code
355 might look like this unless you're creating a device which is managing
356 a bus (appearing under /sys/class/spi_master).
-驱动核心会自动尝试此驱动程序绑定任何SPI设备其board_info MODALIAS给了“CHIP”。
probe()的代码看起来像这样，除非您要创建一个设备是管理一个总线（出现在 /sys/class/spi_master）。
357
358         static int CHIP_probe(struct spi_device *spi)
359         {
360                 struct CHIP                     *chip;
361                 struct CHIP_platform_data       *pdata;
362
363                 /* assuming the driver requires board-specific data: */
364                 pdata = &spi->dev.platform_data;
365                 if (!pdata)
366                         return -ENODEV;
367
368                 /* get memory for driver's per-chip state */
369                 chip = kzalloc(sizeof *chip, GFP_KERNEL);
370                 if (!chip)
371                         return -ENOMEM;
372                 spi_set_drvdata(spi, chip);
373
374                 ... etc
375                 return 0;
376         }
377
378 As soon as it enters probe(), the driver may issue I/O requests to
379 the SPI device using "struct spi_message".  When remove() returns,
380 or after probe() fails, the driver guarantees that it won't submit
381 any more such messages.
-只要它进入probe()，司机可能发出I/ O请求的SPI设备使用"struct spi_message"。
当 remove() 返回时，或probe()失败后，司机保证它不会提交任何更多的此类信息。
382
383   - An spi_message is a sequence of protocol operations, executed
384     as one atomic sequence.  SPI driver controls include:
385
386       + when bidirectional reads and writes start ... by how its
387         sequence of spi_transfer requests is arranged;
388
389       + which I/O buffers are used ... each spi_transfer wraps a
390         buffer for each transfer direction, supporting full duplex
391         (two pointers, maybe the same one in both cases) and half
392         duplex (one pointer is NULL) transfers;
393
394       + optionally defining short delays after transfers ... using
395         the spi_transfer.delay_usecs setting (this delay can be the
396         only protocol effect, if the buffer length is zero);
397
398       + whether the chipselect becomes inactive after a transfer and
399         any delay ... by using the spi_transfer.cs_change flag;
400
401       + hinting whether the next message is likely to go to this same
402         device ... using the spi_transfer.cs_change flag on the last
403         transfer in that atomic group, and potentially saving costs
404         for chip deselect and select operations.
405
406   - Follow standard kernel rules, and provide DMA-safe buffers in
407     your messages.  That way controller drivers using DMA aren't forced
408     to make extra copies unless the hardware requires it (e.g. working
409     around hardware errata that force the use of bounce buffering).
410
411     If standard dma_map_single() handling of these buffers is inappropriate,
412     you can use spi_message.is_dma_mapped to tell the controller driver
413     that you've already provided the relevant DMA addresses.
414
415   - The basic I/O primitive is spi_async().  Async requests may be
416     issued in any context (irq handler, task, etc) and completion
417     is reported using a callback provided with the message.
418     After any detected error, the chip is deselected and processing
419     of that spi_message is aborted.
420
421   - There are also synchronous wrappers like spi_sync(), and wrappers
422     like spi_read(), spi_write(), and spi_write_then_read().  These
423     may be issued only in contexts that may sleep, and they're all
424     clean (and small, and "optional") layers over spi_async().
425
426   - The spi_write_then_read() call, and convenience wrappers around
427     it, should only be used with small amounts of data where the
428     cost of an extra copy may be ignored.  It's designed to support
429     common RPC-style requests, such as writing an eight bit command
430     and reading a sixteen bit response -- spi_w8r16() being one its
431     wrappers, doing exactly that.
432
433 Some drivers may need to modify spi_device characteristics like the
434 transfer mode, wordsize, or clock rate.  This is done with spi_setup(),
435 which would normally be called from probe() before the first I/O is
436 done to the device.  However, that can also be called at any time
437 that no message is pending for that device.
-一些驱动程序可能需要，修改spi_device特点像传输模式，WORDSIZE，或时钟速率。
这是通过spi_setup（），这通常被称为probe（）之前，先完成I/ O设备。然而，这也
可以被称为没有消息等待用于该装置在任何时间。
438
439 While "spi_device" would be the bottom boundary of the driver, the
440 upper boundaries might include sysfs (especially for sensor readings),
441 the input layer, ALSA, networking, MTD, the character device framework,
442 or other Linux subsystems.
-虽然“spi_device”的驱动程序将在底部边界，上边界可能包括sysfs中（尤其是传感器读数），
输入层，ALSA，网络，MTD，字符设备的框架，或其他的Linux子系统。
443
444 Note that there are two types of memory your driver must manage as part
445 of interacting with SPI devices.
-请注意，有两种类型的内存，你的驱动必须管理与SPI设备交互的一部分。
446
447   - I/O buffers use the usual Linux rules, and must be DMA-safe.
448     You'd normally allocate them from the heap or free page pool.
449     Don't use the stack, or anything that's declared "static".
450
451   - The spi_message and spi_transfer metadata used to glue those
452     I/O buffers into a group of protocol transactions.  These can
453     be allocated anywhere it's convenient, including as part of
454     other allocate-once driver data structures.  Zero-init these.
455
456 If you like, spi_message_alloc() and spi_message_free() convenience
457 routines are available to allocate and zero-initialize an spi_message
458 with several transfers.
-如果你喜欢，spi_message_alloc（）和spi_message_free（）例程可分配和零初始化多个传输spi_message。
459
460
461 How do I write an "SPI Master Controller Driver"?
-我怎样写一个“SPI主控制器驱动程序”？
462 -------------------------------------------------
463 An SPI controller will probably be registered on the platform_bus; write
464 a driver to bind to the device, whichever bus is involved.
-SPI控制器可能会被登记在平台上总线;编写驱动程序绑定到设备，总线参与。
465
466 The main task of this type of driver is to provide an "spi_master".
467 Use spi_alloc_master() to allocate the master, and spi_master_get_devdata()
468 to get the driver-private data allocated for that device.
-这种类型的驱动器的主要任务是提供一个“spi_master”。使用spi_alloc_master（）分配，
和spi_master_get_devdata（）分配给该设备的驱动程序专用数据。
469
470         struct spi_master       *master;
471         struct CONTROLLER       *c;
472
473         master = spi_alloc_master(dev, sizeof *c);
474         if (!master)
475                 return -ENODEV;
476
477         c = spi_master_get_devdata(master);
478
479 The driver will initialize the fields of that spi_master, including the
480 bus number (maybe the same as the platform device ID) and three methods
481 used to interact with the SPI core and SPI protocol drivers.  It will
482 also initialize its own internal state.  (See below about bus numbering
483 and those methods.)
-该驱动器将初始化字段这spi_master，包括总线数（可能是相同的平台的移动设备ID）
和三种方法用来与SPI核心和SPI协议驱动程序进行交互。它还将其自己的内部状态初始化。
（总线编号和那些方法见下文）。
484
485 After you initialize the spi_master, then use spi_register_master() to
486 publish it to the rest of the system. At that time, device nodes for the
487 controller and any predeclared spi devices will be made available, and
488 the driver model core will take care of binding them to drivers.
-经过初始化的spi_master，然后使用spi_register_master“（）将它发布到系统其余部分。
控制器和任何预先声明的SPI设备的设备节点将被提供，并会处理它们绑定到驱动器的驱动程序模型的核心。
489
490 If you need to remove your SPI controller driver, spi_unregister_master()
491 will reverse the effect of spi_register_master().
-如果您需要删除您的SPI控制器驱动程序，spi_unregister_master（）将扭转spi_register_master（）的效果。
492
493
494 BUS NUMBERING
-总线编号
495
496 Bus numbering is important, since that's how Linux identifies a given
497 SPI bus (shared SCK, MOSI, MISO).  Valid bus numbers start at zero.  On
498 SOC systems, the bus numbers should match the numbers defined by the chip
499 manufacturer.  For example, hardware controller SPI2 would be bus number 2,
500 and spi_board_info for devices connected to it would use that number.
-总线编号是很重要的，因为这是Linux如何确定一个给定的SPI总线（共享SCK，MOSI，MISO）。
有效的总线数从零开始。在SOC系统，总线数字应该匹配的芯片制造商定义的数字。例如，
硬件控制器SPI2将总线数字2，spi_board_info连接到它的设备将使用该号码。
501
502 If you don't have such hardware-assigned bus number, and for some reason
503 you can't just assign them, then provide a negative bus number.  That will
504 then be replaced by a dynamically assigned number. You'd then need to treat
505 this as a non-static configuration (see above).
-如果你不具备这样的硬件分配的总线数量，出于某种原因，你不能只是将它们分配，
然后提供一个负总线编号。，将被替换为一个动态分配的编号。然后，你需要将此视
为一个非静态的配置（见上文）。
506
507
508 SPI MASTER METHODS
-SPI MASTER方法的
509
510     master->setup(struct spi_device *spi)
511         This sets up the device clock rate, SPI mode, and word sizes.
512         Drivers may change the defaults provided by board_info, and then
513         call spi_setup(spi) to invoke this routine.  It may sleep.
514
515         Unless each SPI slave has its own configuration registers, don't
516         change them right away ... otherwise drivers could corrupt I/O
517         that's in progress for other SPI devices.
518
519                 ** BUG ALERT:  for some reason the first version of
520                 ** many spi_master drivers seems to get this wrong.
521                 ** When you code setup(), ASSUME that the controller
522                 ** is actively processing transfers for another device.
523
524     master->cleanup(struct spi_device *spi)
525         Your controller driver may use spi_device.controller_state to hold
526         state it dynamically associates with that device.  If you do that,
527         be sure to provide the cleanup() method to free that state.
528
529     master->prepare_transfer_hardware(struct spi_master *master)
530         This will be called by the queue mechanism to signal to the driver
531         that a message is coming in soon, so the subsystem requests the
532         driver to prepare the transfer hardware by issuing this call.
533         This may sleep.
534
535     master->unprepare_transfer_hardware(struct spi_master *master)
536         This will be called by the queue mechanism to signal to the driver
537         that there are no more messages pending in the queue and it may
538         relax the hardware (e.g. by power management calls). This may sleep.
539
540     master->transfer_one_message(struct spi_master *master,
541                                  struct spi_message *mesg)
542         The subsystem calls the driver to transfer a single message while
543         queuing transfers that arrive in the meantime. When the driver is
544         finished with this message, it must call
545         spi_finalize_current_message() so the subsystem can issue the next
546         transfer. This may sleep.
547
548     DEPRECATED METHODS
549
550     master->transfer(struct spi_device *spi, struct spi_message *message)
551         This must not sleep. Its responsibility is arrange that the
552         transfer happens and its complete() callback is issued. The two
553         will normally happen later, after other transfers complete, and
554         if the controller is idle it will need to be kickstarted. This
555         method is not used on queued controllers and must be NULL if
556         transfer_one_message() and (un)prepare_transfer_hardware() are
557         implemented.
558
559
560 SPI MESSAGE QUEUE
-SPI消息队列
561
562 If you are happy with the standard queueing mechanism provided by the
563 SPI subsystem, just implement the queued methods specified above. Using
564 the message queue has the upside of centralizing a lot of code and
565 providing pure process-context execution of methods. The message queue
566 can also be elevated to realtime priority on high-priority SPI traffic.
-如果你是快乐由SPI子系统提供与标准的排队机制，只是实施上述指定排队的方法。
使用消息队列上攻集中了大量的代码，并提供纯进程上下文执行方法。消息队列，
也可以提升实时优先级，高优先级的SPI流量。
567
568 Unless the queueing mechanism in the SPI subsystem is selected, the bulk
569 of the driver will be managing the I/O queue fed by the now deprecated
570 function transfer().
-除非选择在SPI子系统的排队机制，将大块的驱动程序管理I / O队列喂养现在已经过时的功能transfer（）。
571
572 That queue could be purely conceptual.  For example, a driver used only
573 for low-frequency sensor access might be fine using synchronous PIO.
-该队列可能是纯粹的概念。例如，驱动程序仅用于低频传感器访问可能会使用同步PIO。
574
575 But the queue will probably be very real, using message->queue, PIO,
576 often DMA (especially if the root filesystem is in SPI flash), and
577 execution contexts like IRQ handlers, tasklets, or workqueues (such
578 as keventd).  Your driver can be as fancy, or as simple, as you need.
579 Such a transfer() method would normally just add the message to a
580 queue, and then start some asynchronous transfer engine (unless it's
581 already running).
-但队列将是非常现实的，使用消息队列，PIO，经常DMA（特别是如果根文件系统是在SPI闪存），
IRQ处理程序一样，tasklet的，或工作队列（如keventd的）的执行上下文。您的驱动程序，可以花哨，
或简单，因为你需要。这种 transfer（）方法通常只是一个队列中添加消息，然后开始一些异步传输引擎（除非它已经运行）。
582
583
584 THANKS TO
585 ---------
586 Contributors to Linux-SPI discussions include (in alphabetical order,
587 by last name):
588
589 David Brownell
590 Russell King
591 Dmitry Pervushin
592 Stephen Street
593 Mark Underwood
594 Andrew Victor
595 Vitaly Wool
596 Grant Likely
597 Mark Brown
598 Linus Walleij