📄 xdma_multi.h
字号:
/* $Id: xdma_multi.h,v 1.2 2006/06/02 21:43:44 meinelte Exp $ *//******************************************************************************** XILINX IS PROVIDING THIS DESIGN, CODE, OR INFORMATION "AS IS"* AS A COURTESY TO YOU, SOLELY FOR USE IN DEVELOPING PROGRAMS AND* SOLUTIONS FOR XILINX DEVICES. BY PROVIDING THIS DESIGN, CODE,* OR INFORMATION AS ONE POSSIBLE IMPLEMENTATION OF THIS FEATURE,* APPLICATION OR STANDARD, XILINX IS MAKING NO REPRESENTATION* THAT THIS IMPLEMENTATION IS FREE FROM ANY CLAIMS OF INFRINGEMENT,* AND YOU ARE RESPONSIBLE FOR OBTAINING ANY RIGHTS YOU MAY REQUIRE* FOR YOUR IMPLEMENTATION. XILINX EXPRESSLY DISCLAIMS ANY* WARRANTY WHATSOEVER WITH RESPECT TO THE ADEQUACY OF THE* IMPLEMENTATION, INCLUDING BUT NOT LIMITED TO ANY WARRANTIES OR* REPRESENTATIONS THAT THIS IMPLEMENTATION IS FREE FROM CLAIMS OF* INFRINGEMENT, IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS* FOR A PARTICULAR PURPOSE.** (c) Copyright 2003-2004 Xilinx Inc.* All rights reserved.*******************************************************************************//*****************************************************************************//**** @file xdma_multi.h** <b>Description</b>** This file contains the multichannel DMA implementation. This component* supports a channelized DMA design in which each device can have it's own* dedicated multichannel DMA, as opposed to a centralized DMA design.* A device which uses DMA typically contains at least two channels of DMA,* one for sending data and the other for receiving data.** This component is designed to be used as a basic building block for* designing a device driver. It provides register accesses such that all* DMA processing can be maintained easier, but the device driver designer* must still understand all the details of the multichannel DMA.** The multichannel DMA allows a CPU to minimize the CPU interaction required* to move data between a memory and a device. The CPU requests the DMA hardware* to perform a DMA operation and typically continues performing other* processing until the DMA operation completes. DMA could be considered a* primitive form of multiprocessing such that caching and address translation* can be an issue.** <b>Scatter Gather Operations</b>** The multichannel DMA may support scatter gather operations. A scatter gather* operation automates the DMA operation such that multiple buffers can be* sent or received with minimal software interaction with the hardware. Buffer* descriptors, contained in the XBufDescriptor component, are used by the* scatter gather operations of the multichannel DMA to describe the buffers to be* processed.** <b>Scatter Gather List Operations</b>** A scatter gather list may be supported by each channel of DMA. The scatter* gather list allows buffer descriptors to be put into the list by a device* driver which requires scatter gather. The hardware processes the buffer* descriptors which are contained in the list and modifies the buffer* descriptors to reflect the status of the DMA operations. The device driver* is notified by interrupt that specific DMA events occur including scatter* gather events. The device driver removes the completed buffer descriptors* from the scatter gather list to evaluate the status of each DMA operation.** The scatter gather list is created and buffer descriptors are inserted into* the list. Buffer descriptors are never removed from the list after it's* creation such that a put operation copies from a temporary buffer descriptor* to a buffer descriptor in the list. Get operations don't copy from the list* to a temporary, but return a pointer to the buffer descriptor in the list.* A buffer descriptor in the list may be locked to prevent it from being* overwritten by a put operation. This allows the device driver to get a* descriptor from a scatter gather list and prevent it from being overwritten* until the buffer associated with the buffer descriptor has been processed.** <b>Typical Scatter Gather Processing</b>** The following steps illustrate the typical processing to use the* scatter gather features of a multichannel DMA.** 1. Create a scatter gather list for the each channel of DMA which puts empty* buffer descriptors into the list.<br>* 2. Create buffer descriptors which describe the buffers to be filled with* receive data or the buffers which contain data to be sent.<br>* 3. Put buffer descriptors into the multichannel DMA scatter list such that scatter* gather operations are requested.<br>* 4. Commit the buffer descriptors in the list such that they are ready to be* used by the multichannel DMA hardware.<br>* 5. Start the scatter gather operations of the multichannel DMA.<br>* 6. Process any interrupts which occur as a result of the scatter gather* operations or poll the each channel of DMA to determine the status.** <b>Interrupts</b>** Each channel of the DMA has the ability to generate an interrupt. This component* does not perform processing for the interrupt as this processing is typically* tightly coupled with the device which is using the multichannel DMA. It is the* responsibility of the caller of DMA functions to manage the interrupt* including connecting to the interrupt and enabling/disabling the interrupt.** <b>Critical Sections</b>** It is the responsibility of the device driver designer to use critical* sections as necessary when calling functions of the multichannel DMA. This* component does not protect critical sections and it does access registers using* read-modify-write operations. Calls to DMA functions from a main thread* and from an interrupt context could produce unpredictable behavior such that* the caller must provide the appropriate protection.** <b>Address Translation</b>** All addresses of data structures which are passed to DMA functions must* be physical (real) addresses as opposed to logical (virtual) addresses.** <b>Caching</b>** The memory which is passed to the function which creates the scatter gather* list must not be cached such that buffer descriptors are non-cached. This* is necessary because the buffer descriptors are kept in a ring buffer and* not directly accessible to the caller of DMA functions.** The caller of DMA functions is responsible for ensuring that any data* buffers which are passed to the multichannel DMA are cache-line aligned if* necessary.** The caller of DMA functions is responsible for ensuring that any data* buffers which are passed to the multichannel DMA have been flushed from the* cache.** The caller of DMA functions is responsible for ensuring that the cache is* invalidated prior to using any data buffers which are the result of a DMA* operation.** <b>Memory Alignment</b>** The addresses of data buffers which are passed to DMA functions must be* 32 bit word aligned if the peripheral is on the OPB bus and must be* 64-bit word aligned if the peripheral is on the PLB bus since the DMA* hardware performs bus width word transfers. The alignment of the provided* buffers is not checked by the driver.** <b>Mutual Exclusion</b>** The functions of the multichannel DMA are not thread safe such that the caller* of all DMA functions is responsible for ensuring mutual exclusion for each* channel of DMA. Mutual exclusion across multiple instances of the* multichannel DMAs is not necessary.** <b>Asserts</b>** Asserts are used within all Xilinx drivers to enforce constraints on argument* values. Asserts can be turned off on a system-wide basis by defining, at compile* time, the NDEBUG identifier. By default, asserts are turned on and it is* recommended that application developers leave asserts on during development.** @note** Many of the provided functions which are register accessors don't provide* a lot of error detection. The caller is expected to understand the impact* of a function call based upon the current state of the multichannel DMA. This* is done to minimize the overhead in this component.* <br><br>* In the following diagram, arrows are used to illustrate addresses or pointers* such that this diagram is aimed for illustration purposes rather than a literal* implementation.* <br><br>* The user memory block is passed in and used by the device driver. The user* memory is partitioned into two parts with the 1st part being a table of* addresses (indexed by the channel number) which contains the address of each* channels registers. The 2nd part of the memory is a table of structures* (indexed by the channel number) which contains the scatter gather data for* each channel.* <pre>* MODIFICATION HISTORY:** Ver Who Date Changes* ----- ---- -------- -----------------------------------------------* 1.00a ecm 09/16/03 First release* 1.00a xd 10/27/04 Doxygenated for inclusion in API documentation* 1.00b ecm 10/31/05 Updated for the check sum offload changes.* </pre>** @internal** Diagram of memory allocation and usage*** XDmaMulti* Instance DMA Hardware* ------------------- -----------* | BaseAddress |-------------------------------->| |* | | User Memory | device |* |-------------------| ----------------- | registers |* | AddrTablePtr |-----| | | |-----------|* | | | | Channel | | |* |-------------------| | | Address | | channel |* | SgDataTablePtr |---| | | Table | |-->| 0 |* | | | | | ---------- | | | registers |* ------------------- | |---->|channel 0 |------| |-----------|* | | | address | | | |* | | |----------| | | channel |* | | |channel 1 |--------->| 1 |* | | | address | | | registers |* | | ---------- | -----------* | | |* | | Scatter |* | | Gather |* | | Data Table |* | | ---------- |* |------>| sg data | |* | |channel 0 | |* | |----------| |* | | sg data | |* | |channel 1 | |* | ---------- |* -----------------********************************************************************************/#ifndef XDMA_MULTI_H /* prevent circular inclusions */#define XDMA_MULTI_H /* by using protection macros */#ifdef __cplusplusextern "C" {#endif/***************************** Include Files *********************************/#include "xdma_multi_i.h" /* constants shared with buffer descriptor */#include "xbasic_types.h"#include "xstatus.h"#include "xbuf_descriptor.h"/************************** Constant Definitions *****************************/#define XDM_CHANNEL_BASE_OFFSET 0x10000 /**< channel base offset */#define XDM_CHANNEL_OFFSET 0x20 /**< interval spacing of DMA registers *//** @name DMA Control Register bit fields * * The following constants provide access to the bit fields of the DMA control * register (DMACR). * @{ */#define XDM_DMACR_SOURCE_INCR_MASK 0x80000000UL /**< increment source address */#define XDM_DMACR_DEST_INCR_MASK 0x40000000UL /**< increment dest address */#define XDM_DMACR_SOURCE_LOCAL_MASK 0x20000000UL /**< local source address */#define XDM_DMACR_DEST_LOCAL_MASK 0x10000000UL /**< local dest address */#define XDM_DMACR_SG_STOP_MASK 0x08000000UL /**< scatter gather stop */#define XDM_DMACR_LAST_BD_MASK XDM_CONTROL_LAST_BD_MASK /**< last buffer descriptor *//* @} *//** @name DMA Status Register bit fields * * The following constants provide access to the bit fields of the DMA status * register (DMASR). * @{ */#define XDM_DMASR_BUSY_MASK 0x80000000UL /**< Channel is busy */#define XDM_DMASR_BUS_ERROR_MASK 0x40000000UL /**< Bus error occurred */#define XDM_DMASR_BUS_TIMEOUT_MASK 0x20000000UL /**< Bus timeout occurred */#define XDM_DMASR_LAST_BD_MASK XDM_STATUS_LAST_BD_MASK /**< Last buffer descriptor */
⌨️ 快捷键说明
复制代码
Ctrl + C
搜索代码
Ctrl + F
全屏模式
F11
切换主题
Ctrl + Shift + D
显示快捷键
?
增大字号
Ctrl + =
减小字号
Ctrl + -