This issue appears to be related to the amount of data that I am
attempting to pass in ioctl calls.
Question 1: I am confused about the true use of the _IOWR macro, is the
third argument the type of the data being passed or is it the type of
the pointer being used to pass the data eg
typedef struct
{
long one;
long two
int three;
} DMA;<==== this struct is 625 bytes long FYI
#define THECALL _IOWR(UNIQUEMAGIC, 0x0d, DMA)
OR
#define THECALL _IOWR(UNIQUEMAGIC, 0x0d, DMA*)
Question 2:
Is there a limit to the size of the data that I can pass?
The symptom is that ALL ioctl calls fail if I add a new case to the
ioctl handler switch statment. At this time this call is not even
happening, but calls that are initializing the device begin to fail once
I have added the code for this call. The failure mode is that the
copy_from_user statment fails to copy any data.
If I comment out the case statment reference to the struct in question,
things begin to work again.
ANY insight to this would be greatly appreciated.
In previous episodes:
I am writing a driver that is causing me fits. I can open the driver
without problem, but the first ioctl call causes an infinite number...or
until I hit reset, number of "cannot handle kernel paging request"
errors printed out. If I change the code slightly....like adding a
printk the error statment may change to invalid operand: 0000. I have
several prink statements that don't get printed out within the ioctl
function in the driver prior to the switch statment that will handle the
given request...this acts like when I insmod my module, the function
address don't get resolved correctly and that my function call gets lost.
Any thoughts and/or debugging suggestions would be most welcome.
TIA
David
--
Kernelnewbies: Help each other learn about the Linux kernel.
Archive: http://mail.nl.linux.org/kernelnewbies/
FAQ: http://kernelnewbies.org/faq/
