| Electronic Components Datasheet Search |
|
MPC8540 Datasheet(PDF) 807 Page - Freescale Semiconductor, Inc |
|
|
|||||||||||||||||||||||||||||
MPC8540 Datasheet(HTML) 807 Page - Freescale Semiconductor, Inc |
|
807 / 1302 page ![]() Functional Description MPC8540 PowerQUICC III Integrated Host Processor Reference Manual, Rev. 1 Freescale Semiconductor 15-29 The DMA controller supports misaligned transfers for both the source and destination addresses. In order to maximize performance, the source and destination engines align the source and destination addresses to a 64-byte boundary. The DMA always reads/writes the maximum number of bytes for a given transfer as described by the capability inputs of the DMA controller except for globally coherent transactions that use the size of the cache coherence granule as described by the mode select input. Using 256 bytes over the RapidIO interface reduces packet overhead which translates to increased bandwidth utilization through the interface. The DMA controller supports bandwidth control, which prevents a channel from consuming all the data bandwidth in the controller. Each channel is allowed to consume the bandwith of the shared resources as specified by the bandwidth control value. After the channel uses its allotted bandwidth, the arbiter grants the next channel access to the shared resources. The arbitration is round robin between the channels. This feature is also used to implement the external control pause feature. If the external control start and pause are enabled in the MRn, the channel enters a paused state after transferring the data described in the bandwidth control. External control can restart the channel from a paused state. The DMA controller is designed to support RapidIO transaction types, including various priority level support. The DMA controller offers additional features from previous generations in which the reads can be mapped to globally coherent (IO_READ), non-coherent (NREAD), or maintenance reads. In addition, the writes can be mapped to coherent (flush with data) and non-coherent (NWRITE, NWRITE_R) writes, messages, and maintenance writes. The DMA programming model permits software to program each DMA engine independently to interrupt on completed segment, chain, or error. It also provides the capability for software to resume the DMA engine from a hardware halted condition by setting the channel continue bit, MRn[CC]. See Table 15-21 on page 15-36 for more complete descriptions of the channel states and state transitions. 15.4.1.1 Basic DMA Mode Transfer This mode is primarily included for backward compatibility with existing DMA controllers which use a simple programming model. This is the default mode out of reset. The different modes of operation under the basic mode are explained in the following sections. 15.4.1.1.1 Basic Direct Mode In basic direct mode, the DMA controller does not read descriptors from memory, but instead uses the current parameters programmed in the DMA registers to start the DMA transfer. Software is responsible for initializing SARn, SATRn, DARn, DATRn, and BCRn registers. The DMA transfer is started when MRn[CS] is set. Software is expected to program all the appropriate registers before setting MRn[CS] to a 1. The transfer is finished after all the bytes specified in the |
|
Link URL |
| Does ALLDATASHEET help your business so far? [ DONATE ] |
About Alldatasheet | Advertisement | Contact us | Privacy Policy | Link to Datasheet | Link Exchange | Manufacturer List All Rights Reserved©Alldatasheet.com |
| Russian : Alldatasheetru.com | Korean : Alldatasheet.co.kr | Spanish : Alldatasheet.es | French : Alldatasheet.fr | Italian : Alldatasheetit.com Portuguese : Alldatasheetpt.com | Polish : Alldatasheet.pl | Vietnamese : Alldatasheet.vn Indian : Alldatasheet.in | Mexican : Alldatasheet.com.mx | British : Alldatasheet.co.uk | New Zealand : Alldatasheet.co.nz |
|
Family Site : ic2ic.com |
icmetro.com |