* Problems in Device Driver Development
@ 2003-10-16 8:45 Suresh Subramanian
2003-10-16 10:19 ` Tvrtko A. Uršulin
0 siblings, 1 reply; 2+ messages in thread
From: Suresh Subramanian @ 2003-10-16 8:45 UTC (permalink / raw)
To: linux-kernel
Hello everybody,
Me and my friends face the following problems while developing Device
Drivers.
Let me confess that we are all amateur programmers. I hope that i get a
reply for this from atleast one of the many people out there.
1. We are not able to use copy_to_user() a chunk of memory from 'write'
function of the device driver. The pointer in the user space is passed to
the kernel space through the 'ioctl' function of the device driver.
2. We are not able to send a signal using kill_fasync() from within a
kernel thread, to an user space process. The problem we face when we do
this is that we get an OOPS message. I tried hard to decode this message,
but in vain. So we came to the conclusion that i cannot send a signal from
within the kernel thread using kill_fasync to a process that is running in
the user space. Am i right in my conclusion, if not what should i do to
make it possible.
3. Similar to problem numbered 2, we face same problems in doing with
'tasklets'.
Thank you.
^ permalink raw reply [flat|nested] 2+ messages in thread
* Re: Problems in Device Driver Development
2003-10-16 8:45 Problems in Device Driver Development Suresh Subramanian
@ 2003-10-16 10:19 ` Tvrtko A. Uršulin
0 siblings, 0 replies; 2+ messages in thread
From: Tvrtko A. Uršulin @ 2003-10-16 10:19 UTC (permalink / raw)
To: Suresh Subramanian, linux-kernel
> 1. We are not able to use copy_to_user() a chunk of memory from 'write'
> function of the device driver. The pointer in the user space is passed to
> the kernel space through the 'ioctl' function of the device driver.
I am sure you get some return value from that function wich tells you what the
problem is...
> 2. We are not able to send a signal using kill_fasync() from within a
> kernel thread, to an user space process. The problem we face when we do
> this is that we get an OOPS message. I tried hard to decode this message,
> but in vain. So we came to the conclusion that i cannot send a signal from
> within the kernel thread using kill_fasync to a process that is running in
> the user space. Am i right in my conclusion, if not what should i do to
> make it possible.
>
> 3. Similar to problem numbered 2, we face same problems in doing with
> 'tasklets'.
kill_fasync() works just fine if used properly.
Try to find Linux Device Drivers 2nd edition online book, it can explain this
basic question to you.
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2003-10-16 10:17 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2003-10-16 8:45 Problems in Device Driver Development Suresh Subramanian
2003-10-16 10:19 ` Tvrtko A. Uršulin
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®