Comm ports on Pc386
Angelo Fraietta
angelo at hunterlink.net.au
Thu Jul 5 04:47:26 UTC 2001
Thank you very much. I can see what is going on from this. My original problem stemmed from
the fact that
#define CONFIGURE_LIBIO_MAXIMUM_FILE_DESCRIPTORS 20
had to be defined (the version of the manual that I have says it should default at 20)
The program below ran fine from my pentium machine but did nothing on my 386dx, so removed the
calls to the shell and removed
#define CONFIGURE_MEMORY_OVERHEAD 16384
#define CONFIGURE_EXTRA_TASK_STACKS
(CONFIGURE_MAXIMUM_TASKS*4*RTEMS_MINIMUM_STACK_SIZE)
as these caused my machine to hang(I don't actually need a shell).
I also modified the code to put a fflush after a certain number of chars (thnaks for informing
me that the output was buffered)
I can now move on with my project, so I would like to thank you again.
Warmest regards
Angelo Fraietta
"Correo Fernando-ruiz (E-mail)" wrote:
> Hi Angelo.
>
> Trying touch a minimum the config.c well compiled and executed is this.
>
> /*
> * Simple test program -- simplified version of sample test hello.
> *
> * COPYRIGHT (c) 1989-1999.
> * On-Line Applications Research Corporation (OAR).
> *
> * The license and distribution terms for this file may be
> * found in the file LICENSE in this distribution or at
> * http://www.OARcorp.com/rtems/license.html.
> *
> * $Id: main.c,v 1.3 2000/06/12 15:00:12 joel Exp $
> */
>
> #define TEST_INIT
>
> #include <bsp.h>
>
> void test_main( void );
>
> rtems_task Init(
> rtems_task_argument ignored
> )
> {
> test_main();
>
> rtems_task_delete(RTEMS_SELF);
>
> }
>
> /* configuration information */
>
> #include <rtems.h>
> #include <tty_drv.h>
> #include <console.h>
> #include <clockdrv.h>
> #include <timerdrv.h>
>
> rtems_driver_address_table Device_drivers[] = {
> CONSOLE_DRIVER_TABLE_ENTRY,
> TTY1_DRIVER_TABLE_ENTRY,
> TTY2_DRIVER_TABLE_ENTRY,
> CLOCK_DRIVER_TABLE_ENTRY,
> {NULL,NULL,NULL,NULL,NULL,NULL},
> };
>
> #define CONFIGURE_HAS_OWN_DEVICE_DRIVER_TABLE
> #define CONFIGURE_RTEMS_INIT_TASKS_TABLE
> #define CONFIGURE_MAXIMUM_TASKS 6
>
> #define CONFIGURE_LIBIO_MAXIMUM_FILE_DESCRIPTORS 20
>
> #define CONFIGURE_RTEMS_INIT_TASKS_TABLE
> #define CONFIGURE_USE_IMFS_AS_BASE_FILESYSTEM
>
> /* You needs this in order to start anything wrong when the task_create
> (JOEL?)*/
>
> #define CONFIGURE_MEMORY_OVERHEAD 16384
> #define CONFIGURE_EXTRA_TASK_STACKS
> (CONFIGURE_MAXIMUM_TASKS*4*RTEMS_MINIMUM_STACK_SIZE)
>
> #define CONFIGURE_INIT
>
> #include <confdefs.h>
> #include <stdio.h>
>
> rtems_id Task_id;
> rtems_name Task_name;
>
> rtems_task Test_task( rtems_task_argument unused);
> void test_main()
> /* Whether you have a high priority task without
> rtems_wait(delay/event/sem/queue/etc)
> you can hangup the system
> (Try to change to realtime any task in the Windows2000/NT task list*/
> {
> rtems_status_code status;
> Task_name = rtems_build_name( 'T', 'A', '1', ' ' );
> status = rtems_task_create(
> Task_name, 100, RTEMS_MINIMUM_STACK_SIZE * 2, RTEMS_DEFAULT_MODES,
> RTEMS_DEFAULT_ATTRIBUTES, &Task_id
> );
> status = rtems_task_start( Task_id, Test_task, 1 );
>
> /*
> Problems in the device driver when the port is reopened/shared
> Test_task uses also /dev/ttyS1 (Perhaps no problem only a little mistake)
> shell_init("sh1",16384,100,"/dev/ttyS1" ,B9600|CS8,TRUE);
> */
> shell_init("sh2",16384,100,"/dev/ttyS2" ,B9600|CS8,TRUE);
> shell_init("shc",16384,100,"/dev/console",B9600|CS8,TRUE);
>
> }
>
> const char* tty_ports [] =
> {
> "/dev/ttyS1",
> "/dev/ttyS2"
> };
>
> rtems_task Test_task(
> rtems_task_argument unused
> )
> {
>
> FILE* fp = fopen (tty_ports [0], "r+"); /* B9600|CS0 termios default*/
>
> if (!fp)
> {
> printf ("Unable to open %s", tty_ports [0]);
> }
> else
> {
> while (1)
> {
> if (!fputc('t', fp)) /* Not the best use */
> {
> printf ("unable to write to port");
> }
> rtems_task_wake_after(10); /* A while */
> }
> }
> }
>
> /* end of file */
>
> The 'tttt..' stream is showed every bufer size number of t's.
>
> The fputc uses I/O buffered. You need raw I/O better(like the shell).
>
> browse the sources and see rtems/c/src/libmisc/shell/shell.c
>
> There you can see how a termios port is changed to raw mode and the buffer
> is eliminated.
> Also you can see how the bauds are changed. (My source was
> rtems/c/src/tests/libtests/termios)
>
> in shell_scanline I do a readkey() and putch() using two fildes of same
> device.
>
> PS:
> When login... (On a minicom/hyperterminal client or on /dev/console)
> The user 'root' is the superuser and 'rtems' is the normal user.
> The password is not validated (crypt.c missing) by the moment.
>
> Write a /etc/motd and /etc/issue and you can have a more friendly logiun
> process.
>
> browse the sources and see rtems/c/src/libmisc/shell/cmds.c in order add
> your commands.
>
> Enjoy it.
>
> Do you want another kind of serial device mode?
>
> Fernando RUIZ CASAS
> home: correo at fernando-ruiz.com
> work: fernando.ruiz at ctv.es
>
> -----Mensaje original-----
> De: Angelo Fraietta [mailto:angelo at hunterlink.net.au]
> Enviado el: martes, 03 de julio de 2001 22:50
> Para: fernando.ruiz at ctv.es
> CC: Correo Fernando-ruiz (E-mail) (E-mail)
> Asunto: Re: Comm ports on Pc386
>
> Find attached the source and makefile. Thank you very much for taking the
> time to look at these
> Warmest regards
> Angelo Fraietta
> "Fernando RUIZ CASAS (E-mail)" wrote:
> Can you send me all the source test application in order to rebuild in my
> environment?Regards
> Fernando RUIZ CASAS
>
> home:correo at fernando-ruiz.com
> work:fernando.ruiz at ctv.es
>
> ...
--
Angelo Fraietta
PO Box 859
Hamilton NSW 2303
Home Page
http://users.hunterlink.net.au/~maaaf/
There are those who seek knowledge for the sake of knowledge - that is CURIOSITY
There are those who seek knowledge to be known by others - that is VANITY
There are those who seek knowledge in order to serve - that is LOVE
Bernard of Clairvaux (1090 - 1153)
More information about the users
mailing list