I did researched the mechanism of network drivers and CAN stack.:)
The ethernet driver get polled through the "void (*poll)(struct eth_drv_sc
*sc);" call back registered in its "eth_hwr_funs". Then Redboot can poll
ethernet driver through a hw-independent style.
But there is no such poll callback registed in "can_lowlevel_funs". Which
means when read a CAN device, it assumes the messages have been put into
the buffer(during interrupt or outside poll). Anyway, there is a "bool
(*getevent)(can_channel *priv, cyg_can_event *pevent, void *pdata);"
So I have to do a CYG_IO_SET_CONFIG_CAN_INPUT_DRAIN(added by myself) before
every read.
You should look at the CAN stack and see how to make it "think" you've
had an interrupt by calling some poll() entry point. That's how the
network drivers work - the same drivers work in either RedBoot or full
eCos with virtually no change.
--
------------------------------------------------------------
Gary Thomas | Consulting for the
MLB Associates | Embedded world
------------------------------------------------------------
_________________________________________________________________
享用世界上最大的电子邮件系统― MSN Hotmail。 http://www.hotmail.com
--
Before posting, please read the FAQ: http://ecos.sourceware.org/fom/ecos
and search the list archive: http://ecos.sourceware.org/ml/ecos-discuss
