Re: [patch] take 2 Re: PG_swap_entry bug in recent kernels

[kanoj@google.engr.sgi.com (Kanoj Sarcar)]

> 
> On Fri, 7 Apr 2000, Kanoj Sarcar wrote:
> 
> >[..] you should try stress
> >testing with swapdevice removal with a large number of runnable
> >processes.[..]
> 
> swapdevice removal during swapin activity is broken right now as far I can
> see. I'm trying to fix that stuff right now.

Be aware that I already have a patch for this. I have been meaning to 
clean it up against latest 2.3 and submit it to Linus ... FWIW, it 
has been broken since 2.2.

> 
> >Also, did you have a good reason to want to make lookup_swap_cache()
> >invoke find_get_page(), and not find_lock_page()? I coded some of the 
> 
> Using find_lock_page and then unlocking the page is meaningless. If you
> are going to unconditionally unlock the page then you shouldn't lock it in
> first place.

I will have to think a little bit about why the code does what it does
currently. I will let you know ...

Kanoj

> 
> Andrea
> 

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux.eu.org/Linux-MM/