I'm writting a Device Driver for a high bandwith acquisition board that can
do DMAs but without scater/gather
As the board as almost no internal fifo, I'm loosing datas if I have to
interrupt after each Page (4kB) Of transfert.
(The time needed to handle the interrupt is too big and the board internal
fifo explode)
=> How can I get LARGE DMA BUFFERS ? (256kB for example)
PS: the datas are coming in bursts, so if i use this kind of large DMA, I
have some time between DMAs as the bandwith is almost 0 every 256kB.
=> How can i remap this buffer to user space ? (Is that possible with rtl or
even linux ?)
(I don't want to copy datas as the CPU as to do a lot of work with these
datas, and I don't want to loose CPU time just for data copy...)
______________________________________________________________________________
message envoye depuis http://www.ifrance.com
emails (pop)-sites persos (espace illimite)-agenda-favoris (bookmarks)-forums
Ecoutez ce message par tel ! : 08 92 68 92 15 (france uniquement)
-- [rtl] ---
To unsubscribe:
echo "unsubscribe rtl" | mail [EMAIL PROTECTED] OR
echo "unsubscribe rtl <Your_email>" | mail [EMAIL PROTECTED]
---
For more information on Real-Time Linux see:
http://www.rtlinux.org/rtlinux/
