From bochs-dev  Tue Jun 21 19:51:03 2005
From: Russell Miller <rmiller () duskglow ! com>
Date: Tue, 21 Jun 2005 19:51:03 +0000
To: bochs-dev
Subject: Re: [Bochs-developers] Keyboard interception question
Message-Id: <200506211251.03945.rmiller () duskglow ! com>
X-MARC-Message: https://marc.info/?l=bochs-dev&m=111938336728905
MIME-Version: 1
Content-Type: multipart/mixed; boundary="--Boundary-00=_n+GuC1H9zipjZwe"

--Boundary-00=_n+GuC1H9zipjZwe
Content-Type: text/plain;
  charset="utf-8"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

On Tuesday 21 June 2005 12:30, Daniel Gryniewicz wrote:

> possible.  The problem is that DOS drivers access the hardware directly
> by invoking BIOS function mapped into specific locations in memory.
> These BIOS functions directly poke the harware.  That kind of hardware
> access is not possible in protected mode in userspace, only in kernel
> space.  You'd have to write a kernel level driver to directly emulate
> the DOS hardware interface, and then export that to userpsace, and then
> hook the output of the bochs BIOS to it.  My guess is that would require
> enough information to write a real driver for Linux for the keyboard.
>
Regarding the access to the hardware not being possible in protected mode in 
userspace, I respectfully say BS. :)  Because I've done it, and I even have a 
little program to do it.  All you have to do is have permission to 
access /dev/port.  I wrote the proof of concept last night.

Basically, you read from position 0x64 and 0x60 in /dev/port, and write to 
0x60.

Attached is the sample program that works under linux, just to prove it.

So now that that is out of the way, how do I hook the output of the BOCHS bios 
to the userspace program?

Thanks.

--Russell


-- 

Russell Miller - rmiller@duskglow.com - Agoura, CA

--Boundary-00=_n+GuC1H9zipjZwe
Content-Type: text/x-csrc;
  charset="utf-8";
  name="prg.c"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
	filename="prg.c"


#include <stdio.h>
#include <errno.h>
#include <stdlib.h>
#include <fcntl.h>

#define CMD_E9		0xE9
#define ENABLE_KBD	0xF4
#define DISABLE_KBD	0xF5
#define ECHO_KBD	0xEE
#define RST_SELFTST_KBD 0xFF

void wait_for_kbd(int fd) {

	char data;

	do {
		lseek(fd, 0x64, 0);
		read(fd, &data, 1);
	} while ((data & 2) == 2);

}
void write_kbd(int fd, unsigned char data) {

	fprintf(stderr, "writing keyboard\n");
	wait_for_kbd(fd);
	lseek(fd, 0x60, 0);
	write(fd, &data, 1);
}

void read_kbd(int fd, unsigned char *data) {

	lseek(fd, 0x60, 0);
	read(fd, data, 1);

}
void enable_kbd(int fd) {

	fprintf(stderr, "enabling keyboard\n");
	write_kbd(fd, 0xf4);

}

void disable_kbd(int fd) {

	fprintf(stderr, "disabling keyboard\n");
	write_kbd(fd, 0xf5);
}

int main(int argc, char **argv) {

	int fd;
	int tries = 0;
	unsigned char data;
	

	if ((fd = open("/dev/port", O_RDWR)) < 0) {
		fprintf(stderr, "Cannot open /dev/port: %s\n", strerror(errno));
		exit(1);
	}

/*	disable_kbd(fd);*/

	data = RST_SELFTST_KBD;
	write_kbd(fd, data);

	data = 0;
	fprintf(stderr, "reading keyboard\n");
	do {
		read_kbd(fd, &data);	
		if (tries == 200000) {
			break;
		} 	
		tries++;
	} while (data != ECHO_KBD);

/*	enable_kbd(fd);*/

	if (data == 0xaa) {
		printf("test succeeded!\n");
	} else {
		printf("test failed! (data is %d)\n", data);
	}
}

--Boundary-00=_n+GuC1H9zipjZwe--


-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies
from IBM. Find simple to follow Roadmaps, straightforward articles,
informative Webcasts and more! Get everything you need to get up to
speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
_______________________________________________
bochs-developers mailing list
bochs-developers@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bochs-developers