patch: make software suspend speedup in vmware virtual machine.

patch: make software suspend speedup in vmware virtual machine.

[Hu Gang]

[-- Attachment #1: Type: text/plain, Size: 641 bytes --]

Hello all:

With this patch 2.5.43 can resume only need ~5sec.
without this patch 2.5.43 also can resume, but need ~240sec.

This patch also can work in normal machine. But need more test.
--------------------------------------------------
--- arch/i386/kernel/suspend.c~old	Wed Oct 16 19:39:42 2002
+++ arch/i386/kernel/suspend.c	Wed Oct 16 19:38:21 2002
@@ -290,8 +290,8 @@
 		for (loop2=0; loop2 < PAGE_SIZE; loop2++) {
 			*(((char *)((pagedir_nosave+loop)->orig_address))+loop2) =
 				*(((char *)((pagedir_nosave+loop)->address))+loop2);
-			__flush_tlb();
 		}
+		__flush_tlb();
 	}
 
 	restore_processor_context();


-- 
		- Hu Gang

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]

Re: patch: make software suspend speedup in vmware virtual machine.

[Pavel Machek]

Hi!

> Hello all:
> 
> With this patch 2.5.43 can resume only need ~5sec.
> without this patch 2.5.43 also can resume, but need ~240sec.
> 
> This patch also can work in normal machine. But need more test.

You probably can move __flush_tlb() out of both loops, it is there
only to aid debugging. _But_ I'd like to keep it as is. __flush_tlb()
is there to prevent heisenbugs.
								Pavel

> --------------------------------------------------
> --- arch/i386/kernel/suspend.c~old	Wed Oct 16 19:39:42 2002
> +++ arch/i386/kernel/suspend.c	Wed Oct 16 19:38:21 2002
> @@ -290,8 +290,8 @@
>  		for (loop2=0; loop2 < PAGE_SIZE; loop2++) {
>  			*(((char *)((pagedir_nosave+loop)->orig_address))+loop2) =
>  				*(((char *)((pagedir_nosave+loop)->address))+loop2);
> -			__flush_tlb();
>  		}
> +		__flush_tlb();
>  	}
>  
>  	restore_processor_context();
> 
> 
> -- 
> 		- Hu Gang



-- 
I'm pavel@ucw.cz. "In my country we have almost anarchy and I don't care."
Panos Katsaloulis describing me w.r.t. patents at discuss@linmodems.org