Hello James,
James Vahn wrote to Jan Wagemakers :
 JV> ~/asm$ gcc tijd.s -o tijd
 JV> For what it's worth, that ran without a segfault here (kernel 2.0.30,
 JV> gcc 2.7.2.1, libc5.4.33).
Well, I was using a Slackware-distribution here, but because Slackware
was/is difficult to upgrade, I have installed Debian here. And indeed, now
there is no segfault :-)
But still, it doesn't work like I expected :-/
/* (/home/jan/assembler/time/tijd.s)
        .globl main
main:
        pushl $currenttime      # push adres of currenttime on the stack
        call time               # call time
        addl $4,%esp
        movl currenttime,%eax   # test-rout. print value in currenttime
        pushl %eax
        pushl $.printdec
        call printf
        addl $8,%esp
        ret
.printdec:
        .string "%d"
currenttime:
        .byte 0,0,0,0
.END
*/
gcc tijd.s -o tijd
This example prints always the value zero on the screen instead of second
since 01Jan1970.
/* (/home/jan/assembler/time/tijd2.S)
        .globl main
main:
        pushl $0                # push zero on the stack
        call time               # call time
        addl $4,%esp
        movl %eax,currenttime   # push value in currenttime on stack
        movl currenttime,%eax   # test rout. print value in currenttime
        pushl %eax
        pushl $.printdec
        call printf
        addl $8,%esp
        ret
.printdec:
        .string "%d"
currenttime:
        .byte 0,0,0,0
.END
*/
gcc tijd2.s -o tijd2
And this example gives a segfault. This segfault disappears when I erase
the line with "movl %eax,currenttime". But I don't know why?
Anyone an idea?
Regards                                - Jan Wagemakers - 
 
o_o Life is short and love is always over in the morning (The Sisters of 
Mercy) 
--- Terminate 5.00/Pro /Linux.DosEmu.JanW-DOS(PTS)
---------------