Dear Reinhard Meyer, dear Lei Wen,
In message 4CB2EA80.2090804@emk-elektronik.de you wrote:
Ok, I am also fine with not include the 512KiB restriction. So we comes to a conclusion that we still use v1 patch, but cut the 512KiB limitation?
Considering the comments that were given to that one, yes.
I would like to point out that, from the previous discussion, I actually think that a 512 KiB limit would be wrong. If the limitation is really in the alignment, i. e. when crossing a 512 KiB boundary, then the allowable transfer size depends on the start address, and 512 KiB is actually only the maximum possible value.
As mmc host limitation, the max number of block in one go should be limited to 65535, and the max buffer size should not excceed 512k bytes.
Better reads somehow like this: Since some hardware has a 16 bit block counter, split the multiple block write into a maximum of (2**16 - 1) blocks to be written at one go.
Agreed.
Best regards,
Wolfgang Denk