From: "coly li" <colyli@263.net>
To: "alfred hitch" <alfred.hitch@gmail.com>
Cc: grub-devel <grub-devel@gnu.org>
Subject: Re: Re: Fwd: memory probing
Date: Tue, 10 May 2005 16:50:51 +0800 [thread overview]
Message-ID: <20050510085058.DEF762FAC4@smtp.263.net> (raw)
alfred hitch:
Bios can provide us a memory map table, for it knows the initial address space layout, why it knows, because bios itself does the initial mapping for address space.
In a non-bios enviornment, we can't get the memory map table by calling bios service, so, we have to detect the size of memory by ourselves.
write and verify means: you read the content from location X, and change the value, then write back to location X; then re-read the content from X, if the value is what you write, the location X is valid; otherwise, maybe the location X is invalid.
as a trick, you need not verify the memory linearly, you can verify it in the power-of-two, only perform linearly dectecting after the prior detecting failed.
coly li
colyli@263.net
2005-05-10
======= 2005-05-10 16:33:25 original messages:=======
>thanks,
>but what did u mean write and verify,
>I thhot grub depends heavilly on bios calls for all this ..
>
>Alfred
>
>On 5/10/05, coly li <colyli@263.net> wrote:
>> alfred hitch:
>>
>> It's not so easy to get the proper answer. it depends on your hardware. for I'm not familiar with embedded enviornmnet, I can only give you some idea for x86 platform.
>>
>> you should know your memory address space layouts. which range is for other peripheral, which range is for real memory. different range use difference detecting scheme. for the range of real memory, the simplest method is "write and verify".
>>
>> but, for I'm not a embedded programmer, I can't give you more information for non-x86. FYI.
>>
>> coly li
>> colyli@263.net
>> 2005-05-10
>>
>> ======= 2005-05-10 14:36:51 original messages:=======
>>
>> >---------- Forwarded message ----------
>> >From: alfred hitch <alfred.hitch@gmail.com>
>> >Date: May 10, 2005 2:13 AM
>> >Subject: Re: memory probing
>> >To: The development of GRUB 2 <grub-devel@gnu.org>
>> >
>> >
>> >Hi,
>> >
>> >thanks for your response,
>> >actually what I am looking for is that does grub and all use bios
>> >calls to find out the memory size ?
>> >on my system there is no bios, ixdp425 based plattform,
>> >then can I somehow probe amount of memory and not use a #define'ed one ?
>> >
>> >Cheers,
>> >Alfred
>> >
>> >On 5/10/05, coly li <colyli@263.net> wrote:
>> >> alfred hitch:
>> >>
>> >> hi! I guess what you want to know more is the initialization for linux kernel.
>> >> I know a guy who wirtes a perfect text named "i386 linux boot howto", you can find out this text on www.tldp.org.
>> >> also, I sugest you read the linux boot protocol, you can find this text in linux kenrel source code. maybe the name is boot.txt.
>> >>
>> >> good luck;-)
>> >>
>> >> coly li
>> >> colyli@263.net
>> >> 2005-05-10
>> >>
>> >> ======= 2005-05-10 12:08:54 original messages:=======
>> >>
>> >> >Hi All,
>> >> >
>> >> >I am trying to understand working of bootloaders and have a question
>> >> >(which might be very elementary perhaps).
>> >> >
>> >> >I wanted to understand how does bootloaders probe memory installed on
>> >> >system and thus pass the correct mem option across to linux kernel
>> >> >say.
>> >> >
>> >> >Can someone please explain if this is dependant on bios necessarilly,
>> >> >as I am looking for doing something similar on ixdp425 plattform.
>> >> >
>> >> >Can u please point me to some relavant doc / code ..
>> >> >
>> >> >Cheers,
>> >> >Alfred
>> >> >
>> >> >
>> >> >_______________________________________________
>> >> >Grub-devel mailing list
>> >> >Grub-devel@gnu.org
>> >> >http://lists.gnu.org/mailman/listinfo/grub-devel
>> >>
>> >> = = = = = = = = = = = = = = = = = = = =
>> >>
>> >>
>> >> _______________________________________________
>> >> Grub-devel mailing list
>> >> Grub-devel@gnu.org
>> >> http://lists.gnu.org/mailman/listinfo/grub-devel
>> >>
>> >>
>> >>
>> >
>> >
>>
>> = = = = = = = = = = = = = = = = = = = =
>>
>>
= = = = = = = = = = = = = = = = = = = =
next reply other threads:[~2005-05-10 9:00 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2005-05-10 8:50 coly li [this message]
2005-05-10 9:24 ` Fwd: memory probing Marco Gerards
2005-05-10 10:23 ` alfred hitch
2005-05-10 12:13 ` Stefan Reinauer
2005-05-11 2:11 ` alfred hitch
2005-05-11 4:39 ` Douglas Wade Needham
2005-05-11 7:47 ` alfred hitch
2005-05-11 17:09 ` Douglas Wade Needham
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20050510085058.DEF762FAC4@smtp.263.net \
--to=colyli@263.net \
--cc=alfred.hitch@gmail.com \
--cc=grub-devel@gnu.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.