Re: replace a legacy proprietary RTOS (Real-Time executive) with NetBSD!

["Thirupathiah Annapureddy" <writetothiru%gmail.com@localhost>]

Hi,

Thanks for your reply.

As I mentioned in the beginning, our goal is to replace a proprietary
RTOS with NetBSD with no/subtle changes to the original application.
It is possible that the original application is accessing the devices
in an ad-hoc manner as the RTOS lacks its own I/O subystem. So we are
thinking to port the application in several steps, the inital step
being porting the whole original application into a single user space
process. Later into multiple-user space processes and few parts to the
kernel space. While the kind of devices that the application accesses
range from simple serial device (UART) to the complex hw accelerators
for multimedia.

Thanks in advance,
Thiru

On Mon, Jan 12, 2009 at 6:55 PM, Antti Kantee <pooka%cs.hut.fi@localhost> wrote:
> On Mon Jan 12 2009 at 18:36:55 +0100, Thirupathiah Annapureddy wrote:
>> Hi,
>>
>> Thanks again for your quick reply.
>>
>> Regarding fielding the interrupts in user space, my question was
>> whether it is possible to write user mode device drivers. In the case
>> of user-mode drivers, the interrupts need to be handled at the user
>> space level. Also device address space need to be mapped into the user
>> space to control/configure the device.
>
> What kind of drivers do you have in mind?  I have been thinking
> about making it possible to run the same device driver code in user-
> or kernelspace for quite a while now, but I haven't had any motivation
> apart from "would be nice" so far.  Your goal seems to be a special case
> of my goal.
>
> We already support file systems and the tcp/ip stack in userspace, so
> device drivers would not be very far off, although of course they have
> their own specific subproblems.
>