Re: AW: Re: [PATCH] S4bios for 2.5.52.

Re: AW: Re: [PATCH] S4bios for 2.5.52.

[Lyle Seaman]

> but that sounds like a unpluged block queue which get flushed every 5 seconds (or 30 seconds) by the bdflush daemon. There are two things which counters this, 1st: swsusp works on ide and there aren't such problems since aeons and 2nd Pavel knows that :-)
> Hence, apologies for this hip shot.

I'm not sure I understand what you're saying, Herbert.  

There was something broken in yield() or schedule() back around 2.5.20 - 
2.5.40 that made rw_swap_page_sync EXCRUTIATINGLY slow, but that's been fixed.

Still, writing pages synchronously, one at a time, is going to be slow no 
matter what you do.  *Especially* on slow laptop drives.

What I sent Nigel was a little patch to write them 20 at a time, which reduced 
my suspend times from nearly 60 seconds to something around 2 seconds.  I 
stopped at 20 because going beyond that was chasing diminishing returns and I 
didn't think that chewing up a huge chunk of stack was justified for a 
one-second reduction in suspend times.

I didn't send it on to the rest of the list because I figure Nigel's other 
changes will be coming along as well, and I didn't want to make merging too 
difficult for everyone.






-------------------------------------------------------
This SF.NET email is sponsored by: Geek Gift Procrastinating?
Get the perfect geek gift now!  Before the Holidays pass you by.
T H I N K G E E K . C O M      http://www.thinkgeek.com/sf/

RE: AW: Re: [PATCH] S4bios for 2.5.52.

[Nigel Cunningham]

Hmm. I didn't think that carefully about what was causing the write to
occur. Hence I unrolled I took your patch further and completely separated
the submitting of _all_ of the pages from the block_flushpage calls for all
of the pages (ie two loops). (I saw much more than a one second gain, even
on my faster desktop computer!) After reading this, I think I should look
sooner at what is going. I was going to anyway, but I was concentrating on
getting my whole port from 2.4 working first.

I'm not promising anything quickly - I have plenty of 'real' work at the
moment :>

Regards,

Nigel

> Still, writing pages synchronously, one at a time, is going
> to be slow no
> matter what you do.  *Especially* on slow laptop drives.
>
> What I sent Nigel was a little patch to write them 20 at a
> time, which reduced
> my suspend times from nearly 60 seconds to something around 2
> seconds.  I
> stopped at 20 because going beyond that was chasing
> diminishing returns and I
> didn't think that chewing up a huge chunk of stack was
> justified for a
> one-second reduction in suspend times.
>
> I didn't send it on to the rest of the list because I figure
> Nigel's other
> changes will be coming along as well, and I didn't want to
> make merging too
> difficult for everyone.



-------------------------------------------------------
This SF.NET email is sponsored by: Geek Gift Procrastinating?
Get the perfect geek gift now!  Before the Holidays pass you by.
T H I N K G E E K . C O M      http://www.thinkgeek.com/sf/

RE: Re: [PATCH] S4bios for 2.5.52.

[Nigel Cunningham]

[-- Attachment #1: Type: text/plain, Size: 1101 bytes --]

I don't think so, because AFAIK the bdflush daemon is in the refrigerator by
this point and it is consistently when a particular page number is submitted
that the first 'flush' occurs. I thought it might be that some threshhold is
reached. Oh - just in case there's anyone wondering, I'm not trying to write
to anything other than an IDE swap partition :>

Regards,

Nigel

> -----Original Message-----
> From: acpi-devel-admin-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
> [mailto:acpi-devel-admin-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org]On Behalf Of Herbert
> Nachtnebel
> Sent: Thursday, December 19, 2002 10:49 PM
> To: acpi-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
> Subject: AW: [ACPI] Re: [PATCH] S4bios for 2.5.52.
>
>
> Sorry to be stupid,
>
> but that sounds like a unpluged block queue which get flushed
> every 5 seconds (or 30 seconds) by the bdflush daemon. There
> are two things which counters this, 1st: swsusp works on ide
> and there aren't such problems since aeons and 2nd Pavel
> knows that :-)
> Hence, apologies for this hip shot.
>
> Herbert.
>

[-- Attachment #2: winmail.dat --]
[-- Type: application/ms-tnef, Size: 2364 bytes --]

AW: Re: [PATCH] S4bios for 2.5.52.

[Herbert Nachtnebel]

Sorry to be stupid, 

but that sounds like a unpluged block queue which get flushed every 5 seconds (or 30 seconds) by the bdflush daemon. There are two things which counters this, 1st: swsusp works on ide and there aren't such problems since aeons and 2nd Pavel knows that :-)
Hence, apologies for this hip shot.

Herbert.

-----Ursprüngliche Nachricht-----
Von: Nigel Cunningham [mailto:ncunningham-oKK1aGe2n869koe0gwxAeg@public.gmane.org]
Gesendet: Mittwoch, 18. Dezember 2002 20:38
An: 'Ducrot Bruno'; 'Kai Germaschewski'
Cc: 'Grover, Andrew'; linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org; 'Pavel Machek';
acpi-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
Betreff: RE: [ACPI] Re: [PATCH] S4bios for 2.5.52.


> Kai, if you see that S4OS is slower, it is probably a bug
> that have to be
> fixed in swsusp.

There does seem to be a problem with the rw_swap_page_sync routine. Perhaps
not on Pavel's machine (:>), but I've experienced painfully slow suspends
too. Performance was helped immensely by applying a concept that I'll credit
Lyle Seaman for - unrolling rw_swap_page_sync and submitting all the IO for
pages then looping again and waiting on each sync. I found with writing 7000
or so pages (a suspend from init S) that something gets triggered to force
the sync about every 3900 pages anyway.

Regards,

Nigel



-------------------------------------------------------
This SF.NET email is sponsored by: Order your Holiday Geek Presents Now!
Green Lasers, Hip Geek T-Shirts, Remote Control Tanks, Caffeinated Soap,
MP3 Players,  XBox Games,  Flying Saucers,  WebCams,  Smart Putty.
T H I N K G E E K . C O M       http://www.thinkgeek.com/sf/
_______________________________________________
Acpi-devel mailing list
Acpi-devel-5NWGOfrQmneRv+LV9MX5uipxlwaOVQ5f@public.gmane.org
https://lists.sourceforge.net/lists/listinfo/acpi-devel


-------------------------------------------------------
This SF.NET email is sponsored by: Geek Gift Procrastinating?
Get the perfect geek gift now!  Before the Holidays pass you by.
T H I N K G E E K . C O M      http://www.thinkgeek.com/sf/