Hi All ,

1. We want to read 25-30K data from a memory using DMA and send
it back to USER-SPACE.

2. At present i am doing an dma_alloc_coherent of size 6K from
KERNEL and copying the data to user space buffer using copy_to_user
function call.

3. Can u guys please let me know what is the best mechanism
to transmit around 30K[Presently 7K] to user space .

The operation at present is slow rate , but we need to increase
the speed of transmission also.
[ Presently we have around 12MBytes/sec which we are aiming to
increase to around 25MBytes/sec]

-> So at high speed and large chunk what mechanism should i implement
-> I am not using x86 architecture. We are using OMAP processor
from Texas Instruments with ARM + LINUX
-> Also initial aim is to transfer large chunk , then the throughput
part

Thank You,
Ashish Mishra
[Banglore / India]

Hi All ,

Any input for the same .

The priority is to understand the best mechanism to transmit large data to user space .

Thank You,
Ashish Mishra
[Banglore / India]

MMAP can be used for higher size ....