intel assembly register st?

[/bin/sh]

While I was doing some intel assembly code on kate text editor I by mistake wrote "st" and I saw that st was blue (I had set kate to blue for a register), this means st is a register. What is st used for, is it 8, 16, 32 or 64 bit or a segment register?

[nvteighen]

While I was doing some intel assembly code on kate text editor I by mistake wrote "st" and I saw that st was blue (I had set kate to blue for a register), this means st is a register. What is st used for, is it 8, 16, 32 or 64 bit or a segment register?

My ASM knowledge is very limited, but I never heard of any ST register... Take a look at: http://en.wikipedia.org/wiki/X86#x86_registers

Anyway, never deduce anything from syntax highlighting. Under common-lisp-mode + SLIME, Emacs marks anything starting with "with-" as if it was a keyword... But no matter how hard you try looking for "with-my-girlfriend" on the Common Lisp HyperSpec, you won't find it... or her... :P

(No, it's not a bug, it's a feature. There's a reason for it and it's actually handy, but it'll be a bit lengthy to explain here, apart from completely off-topic)

[/bin/sh]

I found that in your link ([http://en.wikipedia.org/wiki/X86#x86_registers) under 32-bit at line 9 it said:
"With the 80486 a floating-point processing unit (FPU) was added, with eight 80-bit wide registers: st(0) to st(7)."
So I guess st(0) to st(7) are registers.

[nvteighen]

I found that in your link ([http://en.wikipedia.org/wiki/X86#x86_registers) under 32-bit at line 9 it said:
"With the 80486 a floating-point processing unit (FPU) was added, with eight 80-bit wide registers: st(0) to st(7)."
So I guess st(0) to st(7) are registers.

Yay, I fail at reading! Hehe... So, yup, they are then :D

[Bachstelze]

The reason you have never heard of them is that they have been superseded by the mmx registers with the Pentium MMX, which themselves have been superseded by the xmm registers with the Pentium III. Nowadays, compilers will use the xmm registers (if the processor supports them) when compiling code that contains floating-point operations.


firas@av104151 ~ % cat test.c
float foo(float x)
{
return 2*x;
}
firas@av104151 ~ % gcc -std=c99 -pedantic -Wall -Wextra -m32 -S test.c
firas@av104151 ~ % cat test.s
.text
.globl _foo
_foo:
pushl %ebp
movl %esp, %ebp
subl $24, %esp
movss 8(%ebp), %xmm0
addss %xmm0, %xmm0
movss %xmm0, -12(%ebp)
flds -12(%ebp)
leave
ret
.subsections_via_symbols
firas@av104151 ~ % gcc -std=c99 -pedantic -Wall -Wextra -m32 -march=i386 -S test.c
firas@av104151 ~ % cat test.s
.text
.globl _foo
_foo:
pushl %ebp
movl %esp, %ebp
subl $8, %esp
flds 8(%ebp)
fadd %st(0), %st
leave
ret
.subsections_via_symbols