Udma_ChPdmaPrms Struct Reference

Detailed Description

UDMA PDMA channel Static TR parameters.

Data Fields
uint32_t	elemSize
uint32_t	elemCnt
uint32_t	fifoCnt
uint8_t	burst
uint8_t	acc32
uint8_t	eol

Field Documentation

elemSize

uint32_t Udma_ChPdmaPrms::elemSize

[IN] Element size. This field specifies how much data is transferred in each write which is performed by the PDMA. This is the X static TR parameter of PDMA.

In case of MCAN TX/RX PDMA channel, this is not used and should be set to 0.

Refer Udma_PdmaElemSize for supported values.

elemCnt

uint32_t Udma_ChPdmaPrms::elemCnt

[IN] Element count. This field specifies how many elements to transfer each time a trigger is received on the PDMA channel. This is the Y static TR parameter of PDMA.

In case of MCAN PDMA channel, this represents the buffer size. In case of MCAN TX, this field specifies how many bytes should be written to an MCAN TX buffer. This field includes the 8 byte MCAN header on the initial packet fragment. The PDMA will break up the source packet into fragments of this buffer size, copying the 8 byte MCAN header for the initial fragment, and then skipping it for each additional fragment and thus reusing the header from the first fragment. A buffer size less than 16 is treated as 16, and a buffer size greater than 72 is treated as 72. In case of MCAN RX, this field specifies how many bytes should be read from an MCAN RX buffer. This field includes the 8 byte MCAN header on the initial packet fragment. A buffer size less than 16 is treated as 16, and a buffer size greater than 72 is treated as 72.

fifoCnt

uint32_t Udma_ChPdmaPrms::fifoCnt

[IN] FIFO count. This field specifies how many full FIFO operations comprise a complete packet. When the count has been reached, the PDMA will close the packet with an 'EOP' indication. If this parameter is set to 0, then no packet delineation is supplied by the PDMA and all framing is controlled via the UDMA TR.

This is the Z static TR parameter of PDMA. This is NA for TX and should be set to 0. In case of MCAN RX, this represents the buffer count. This field specifies how many MCAN RX buffers should be read before closing the CPPI packet with an 'EOP' indication. When this count is greater than 1, multiple MCAN RX buffers will be read into a single CPPI packet buffer. The 8 byte MCAN header will be skipped on subsequent MCAN buffer reads. Setting this field to NULL will suppress all packet delineation, and should be avoided.

burst

uint8_t Udma_ChPdmaPrms::burst

[IN] Burst mode. Enable VBUSP burst mode for this channel.

Since MCAN buffers are stored in linear memory, the burst mode for MCAN is a simple linear burst across the transfer window. The max burst size is set to the 72 byte size of the MCAN buffer. This will allow a full MCAN packet to be read out as a single burst.

acc32

uint8_t Udma_ChPdmaPrms::acc32

[IN] 32b access. Enable 32b access mode.

When set, enables 32-bit access mode. On a 32-bit PDMA, all accesses will have XCNT=4 to support legacy IP that is not fully VBUSP compliant. This bit is ignored if the PDMA VBUSP port is not 32 bits wide.

eol

uint8_t Udma_ChPdmaPrms::eol

[IN] EOL mode. Enable eol mode.

EOL mode. Normally, when the Z count of FIFO operations has been reached, the PDMA will close the packet with an 'EOP' indication. When this flag is set, the PDMA will instead trigger an EOL at the completion of Z.