Hi, Just wanted to do a follow-up on async mcp251x driver which I posted sometime ago: beside "Gianluca" is anyone else tested/checked that driver ? Any feedback ?
Regards, -- Fawad Lateef On Wed, Jul 28, 2010 at 5:12 PM, Fawad Lateef <[email protected]> wrote: > Hi, > > On Wed, Jul 28, 2010 at 5:05 PM, gianluca > <[email protected]> wrote: >> Hello list! >> >> Just to let to everyone know the spi_async() driver for mcp2515 written by >> Fawad Lateef works good. >> > > Thanks for testing and posting results on list. Just wanted to _clear_ > that I _rewrote_ tx/rx path to use spi_async while switching to old > synchronous functionality if some error occurs then switch back to > asynchronous when error is handled. So its _not_ completely written by > me :) > >> Fawad Lateef wrote: >>> >>> Good, Really nice to hear this. >>> >>> Can you do me a favour ? and Post your reply again while replying to >>> ALL, so that others on mailing-list can see the results too ? >>> >>> -- Fawad Lateef >>> >>> >>> >>> On Wed, Jul 28, 2010 at 4:48 PM, gianluca >>> <[email protected]> wrote: >>>> >>>> Fawad Lateef wrote: >>>>> >>>>> I am sending mcp251x.c file to you directly. Please find it attached. >>>>> >>>>>>> Hoping to get feedback and test results soon. >>>>>> >>>>>> Sure! as soon as the driver is compiled, I will try with my torture >>>>>> test >>>>>> just to see how is going on!! >>>> >>>> Ok, here the results: >>>> >>>> It works! Our busload is one canpacket every 2-3 msec, with the Qt >>>> Embedded >>>> application running sometime at 30/40% of CPU Load (Samsung S3C2442 >>>> running >>>> at 300Mhz). >>>> Our torture test is on receiving: our S3C board sends a start command to >>>> a >>>> special node on the bus, and this node starts to send a packet with ID >>>> counting up (from 0x00 to 0x7ff) and with a payload (32bit-4 bytes) with >>>> a >>>> progressive number (counter). >>>> >>>> Our application after sending the start command, it waits (recv()) the >>>> packets incoming. Every received-packet it checks the previous payload, >>>> and >>>> if it is match the correct counter, dump all data on the screen (serial >>>> console running at 115.2K). >>>> If some packet is missing (due to ID behaviour priority and/or packet >>>> lost >>>> due to speed of the isr interrupt) it prints a special line on the >>>> screen. >>>> At the end of transfer it will check the rx_over_errors. >>>> If it is ZERO it displays a smile on the screen, otherwise it display a >>>> sort >>>> of BAD sign... >>>> >>>> Now everything is working as expected. >>>> >>>> Thank you very much. >>>> >>>> Ciao >>>> -- >>>> ,,, >>>> (o o) >>>> ======oOO==(_)==OOo====== >>>> >>>> Gianluca Renzi >>>> R&D >>>> phone: +39.0542.609120 >>>> fax: +39.0542.609212 >>>> >>>> .oooO Oooo. >>>> ======( )==( )======= >>>> \ ( ) / >>>> \_) (_/ >>>> >>> >>> >>> >> > > Regards, > > Fawad Lateef > _______________________________________________ Socketcan-core mailing list [email protected] https://lists.berlios.de/mailman/listinfo/socketcan-core
