阅读:2399回复:5
EOT#有效后,9054的响应。
别告诉我去看DATASHEET,我看了N遍,觉得里面的问题很严重,一定是写错了。才不得已POST.
清详细说明,越详细越好。比如在BLAST#使能的情况下,.....在BLAST#不使能的情况下.......。谢谢了。 |
|
沙发#
发布于:2002-04-17 10:14
老大们,帮帮忙吧。
|
|
板凳#
发布于:2002-04-17 14:30
我看过很多遍了,觉得没什么问题.你说的问题在哪里?
|
|
|
地板#
发布于:2002-04-17 19:40
End of Transfer (EOT#) Input
The DMA EOT# Enable bit(s) (DMAMODE0[14] and/or DMAMODE1[14]) determines the number of Lwords to transfer after a DMA controller asserts EOT# input. EOT# input should be asserted only when the PCI 9054 owns a bus. If BLAST# output is not required for the last Lword of the DMA transfer (DMAMODE0[15]=1 and/or DMAMODE1[15]=1), the DMA controller releases the data bus and terminates DMA after it receives an external READY#. Or, the internal wait state counter decrements to 0 for the current Lword. If the DMA controller is currently bursting data that is not the last Data phase for the burst, BLAST# output is not asserted. If BLAST# output is required for last Lword of the DMA transfer (DMAMODE0[15]=0 and/or DMAMODE1[15]=0), the DMA controller transfers one or two Lwords, depending on the Local Bus width. If EOT# is asserted, the DMA controller completes the current Lword and one additional Lword (this allows BLAST# output to be asserted during the final Lword). If the DMA FIFO is full or empty after the Data phase in which EOT# is asserted, the second Lword is not transferred. EOT enable bit愣决定EOT有效之后的反应,这不是显然错了吗?我觉得应该改成The Fast/Slow Terminate Mode Select bit(s) (DMAMODE0[15] and/or DMAMODE1[15])来决定。 老大,是不是错了? |
|
地下室#
发布于:2002-04-18 09:42
好像是有些问题,老兄看的够仔细的,找PLX的技术支持联系一下.
|
|
|
5楼#
发布于:2002-04-19 11:56
没找过,如何找?谢谢。
|
|