Subsystems (2)
From: Colonel <[email protected]>
To: [email protected]
CC: [email protected]
In-reply-to: <[email protected]>
(message from Marcelo Tosatti on Mon, 23 Apr 2001 20:31:12 -0300 (BRT))
Subject: Re: 2.4.4-pre6 : THANKS! very snappy here [nt]
Reply-to: [email protected]
In-Reply-To: <[email protected]>
Date: Mon, 23 Apr 2001 20:31:12 -0300 (BRT)
From: Marcelo Tosatti <[email protected]>
X-Sender: [email protected]
Just as curiosity --- did it got faster in pre5 ?
Quoted 1 line · Colonel
but since you read it, system seems like it's running twice as fast
Actually, it made not be pre6. I patched the kernel per some thread on LKML a few days before, but had not had the opportunity to reboot and run it until pre6. The patch was :
mm/page_alloc.c renamed from mm/page_alloc.c.~1~ +0 -1
@@ -454,7 +454,6 @@if (gfp_mask & __GFP_WAIT) { memory_pressure++;try_to_free_pages (gfp_mask); -wakeup_bdflush (0 ); goto try_again; } }
It would make more sense that the above change was responsible for the improvement.
PATCH 1/1 remove wakeup_bdflush(0) from __alloc_pages()
Linus,
Can you please apply the following patch ?
It avoids allocators from waking up bdflush all the time even when it does not have any job to do (ie no more than 30% (default) of dirty buffers).
Thanks
Quoted 42 lines · Colonel
From: Colonel <[email protected]>
To: [email protected]
CC: [email protected]
In-reply-to: <[email protected]>(message from Marcelo Tosatti on Mon, 23 Apr 2001 20:31:12 -0300 (BRT))Subject: Re: 2.4.4-pre6 : THANKS! very snappy here [nt]
Reply-to: [email protected]
References: <[email protected]>Date: Mon, 23 Apr 2001 20:31:12 -0300 (BRT)
From: Marcelo Tosatti <[email protected]>
X-Sender: [email protected]Just as curiosity --- did it got faster in pre5 ?
Quoted 1 line · Colonel
but since you read it, system seems like it's running twice as fast
Actually, it made not be pre6. I patched the kernel per some thread on LKML a few days before, but had not had the opportunity to reboot and run it until pre6. The patch was :
mm/page_alloc.c renamed from mm/page_alloc.c.~1~ +0
-1@@ -454,7 +454,6 @@if (gfp_mask & __GFP_WAIT) { memory_pressure++;try_to_free_pages (gfp_mask); -wakeup_bdflush (0 ); goto try_again; } }It would make more sense that the above change was responsible for the improvement.