Rockbox mail archive

• Next message: Blue Chip: "Re: another 8MB upgrade success story"
• Previous message: Justin Heiner: "Re: another 8MB upgrade success story"
• In reply to: Daniel Stenberg: "Re: another 8MB upgrade success story"
• Next in thread: Mike Holden: "Re: another 8MB upgrade success story"
• Reply: Mike Holden: "Re: another 8MB upgrade success story"
• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]

On Wed, 19 Mar 2003 19:02:10 +0100 (CET) Daniel Stenberg <daniel_at_haxx.se>
wrote:

> On Wed, 19 Mar 2003, Justin Heiner wrote:
>
> > Couldn't you do a quick write/read test on some part of the memory over
> > 4mb? Or would that cause too many problems.
>
> That would be the easy part. Detecting that the unit has 8MB ram should be
> easy enough (doing something like you describe).
>
> Then what? Various buffers should use different sizes. How do you achieve
> that with no dynamic memory allocation?

minimal introspection in the code: find the address for the 'old' buffers and
point them to a virtual stack at the back of the new space, or reserves which
are the 8mb equivalents.

loosing space in 1-2mb to gain 8 may be worth it.

at this level, hacks in machinecode to re-write addresses in the instruction
stream held in memory is fine. painful but ...

-George
>
> --
> Daniel Stenberg -- http://rockbox.haxx.se/ -- http://daniel.haxx.se/

-- 
George Michaelson       |  APNIC
Email: ggm_at_apnic.net    |  PO Box 2131 Milton QLD 4064
Phone: +61 7 3367 0490  |  Australia
  Fax: +61 7 3367 0482  |  http://www.apnic.net

• Next message: Blue Chip: "Re: another 8MB upgrade success story"
• Previous message: Justin Heiner: "Re: another 8MB upgrade success story"
• In reply to: Daniel Stenberg: "Re: another 8MB upgrade success story"
• Next in thread: Mike Holden: "Re: another 8MB upgrade success story"
• Reply: Mike Holden: "Re: another 8MB upgrade success story"
• Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]