[patch] [sched] x86 idle thread should clear %fs, %gs

[patch] [sched] x86 idle thread should clear %fs, %gs

[Ingo Molnar]

the idle thread (on SMP, all idle threads) should clear the %fs and %gs
register, we forgot to clear these registers after the boot process. This
is especially important in kernels with the previous %fs/%gs patch
applied. (The patch below is against 2.5.3-pre3.)

i've checked and it apperas that the 0.01 Linux kernel source code has the
same problem, in boot/head.s the idle task loads the 0x10 selector, and
INIT_TASK's init TSS uses the 0x17 selector for %gs and never clears it.
:-)

while it's not an issue from the correctness point of view in the 0.01
kernel either, the TSS switching microcode probably exeutes slightly
faster if %gs is 0 for both tasks.

so it appears that this lowlevel x86 performance bug(?) is more than 11
years old! :-)

	Ingo

--- linux/arch/i386/kernel/setup.c.orig	Sun Jan 27 15:14:43 2002
+++ linux/arch/i386/kernel/setup.c	Sun Jan 27 16:01:34 2002
@@ -2803,9 +2803,10 @@
 	load_TR(nr);
 	load_LDT(&init_mm);

-	/*
-	 * Clear all 6 debug registers:
-	 */
+	/* Clear %fs and %gs. */
+	asm volatile ("xorl %eax, %eax; movl %eax, %fs; movl %eax, %gs");
+
+	/* Clear all 6 debug registers: */

 #define CD(register) __asm__("movl %0,%%db" #register ::"r"(0) );

Re: [patch] [sched] x86 idle thread should clear %fs, %gs

[Pavel Machek]

Hi!

> so it appears that this lowlevel x86 performance bug(?) is more than 11
> years old! :-)

Probably 386's were not optimized for %fs==%gs==0 case?

BTW does it really make a difference for the CPU?
									Pavel

> @@ -2803,9 +2803,10 @@
>  	load_TR(nr);
>  	load_LDT(&init_mm);
> 
> -	/*
> -	 * Clear all 6 debug registers:
> -	 */
> +	/* Clear %fs and %gs. */
> +	asm volatile ("xorl %eax, %eax; movl %eax, %fs; movl %eax, %gs");
> +
> +	/* Clear all 6 debug registers: */
> 
>  #define CD(register) __asm__("movl %0,%%db" #register ::"r"(0) );

-- 
(about SSSCA) "I don't say this lightly.  However, I really think that the U.S.
no longer is classifiable as a democracy, but rather as a plutocracy." --hpa

Re: [patch] [sched] x86 idle thread should clear %fs, %gs

[Ingo Molnar]

On Tue, 29 Jan 2002, Pavel Machek wrote:

> > so it appears that this lowlevel x86 performance bug(?) is more than 11
> > years old! :-)
>
> Probably 386's were not optimized for %fs==%gs==0 case?
>
> BTW does it really make a difference for the CPU?

whether %fs/%gs is 0 or not? With the current 2.5.3-pre6 task switching
code there definitely is a performance difference, because we optimize the
'both values are 0' case. With the original TSS-based task switching
microcode i suspect it made a difference as well. (switching between two
different %fs/%gs values should be slower.)

	Ingo