Graham Murray <[EMAIL PROTECTED]> writes:
> I suspect that the cause of this is IOCTL changes in the new driver.
Sorry to follow up my own posting, but I have been looking some
more. An strace on diald shows that it does look like a problems with
the Ioctl. The following 2 lines seem to be the ones of interest. They
are repeated many times in the trace.
ioctl(5, PPPIOCGUNIT, 0x8061e38) = -1 EINVAL (Invalid argument)
ioctl(5, 0x5494, 0x8061e38) = -1 EINVAL (Invalid argument)
I think that these map to the start of the routine ppp_set_addrs()
where diald is trying to get the interface number which pppd has
opened.
I have also checked in the kernel sources and the definition for
PPPIOCGUNIT_2_2_0 in diald matches that for PPPIOCGUNIT in both the
kernel and pppd sources. Looking at the handlers for these Ioctls in
the kernel, they cannot return EINVAL, so it looks as though the
Ioctl calls from diald are not reaching the kernel pppd driver.
Could it therefore be that it now necessary to send the Ioctls via
some other mechanism than via the "data" tty?
-
To unsubscribe from this list: send the line "unsubscribe linux-diald" in
the body of a message to [EMAIL PROTECTED]
