Line discipline

[Okash Khawaja]

Also, the exit function speakup_test_cleanup() needs to contain code from
your module's exit function, if it is used as a kernel module, rather than
being built into kernel image.

On Thu, Dec 15, 2016 at 10:17 PM, Okash Khawaja <okash.khawaja at gmail.com>
wrote:

> Hi,
>
> I quickly put together speakup-test.c and compiled it into the kernel
> binary. Loading the kernel outputs test messages to ttyS0 and logs the
> event in kernel log. "got tty ffff8fcfe74b0000" message appears at time
> 3.258357. Both speakup-test.c and dmesg.log are attached. N_SPEAKUP is 26
> in /include/uapi/linux/tty.h.
>
> Also tested this as kernel module which seemed to work fine too. That
> required EXPORT_SYMBOL(tty_set_ldisc); in /drivers/tty/tty_ldisc.c. Will be
> able to spend more time over the weekend.
>
> Thanks!
> Okash
>
> On Sun, Dec 11, 2016 at 10:33 PM, Okash Khawaja <okash.khawaja at gmail.com>
> wrote:
>
>> Hi Samuel,
>>
>> That's great. Let me get back on this soon as I can.
>>
>> Thanks,
>> Okash
>>
>> > On 11 Dec 2016, at 20:22, Samuel Thibault <samuel.thibault at ens-lyon.org>
>> wrote:
>> >
>> > Hello,
>> >
>> > Okash Khawaja, on Sun 27 Nov 2016 20:41:50 +0000, wrote:
>> >> Thanks for explanation. I have not been able to work on this recently
>> - pulled
>> >> away for something urgent. As soon as it's over, I'll be back on it.
>> Probably
>> >> third week of December.
>> >
>> > I've taken some time to have a look, I have attached the result:
>> >
>> > - tty_NULL is a patch against the kernel to make tty functions accept
>> >  being called with filp == NULL
>> > - mymodule.tgz is a dumb module which opens ttyS0 (major 4 minor 64),
>> >  and writes to it through the tty write operation.
>> >
>> > So this, combined with your work on the line discipline, should get
>> > something working, and opening other serial ports (e.g. ttyUSB0) is a
>> > matter of changing the major/minor pair (e.g. 188, 0).
>> >
>> > Samuel
>> > <mymodule.tgz>
>> > <tty_NULL>
>>
>
>