Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

CMOS memory

4 views
Skip to first unread message

Zhane

unread,
Dec 7, 2009, 10:30:07 AM12/7/09
to
are there any unused space in the CMOS memory that I can use for byte
storage n retrieval?

Alexei A. Frounze

unread,
Dec 7, 2009, 11:25:49 AM12/7/09
to
On Dec 7, 7:30 am, Zhane <zhan...@gmail.com> wrote:
> are there any unused space in the CMOS memory that I can use for byte
> storage n retrieval?

Maybe, but I'm not sure how reliable and useful that is. I don't think
there's any location left for such purposes. Usually, only the BIOS
uses that. I think you could try using the century byte.

Alex

wolfgang kern

unread,
Dec 8, 2009, 5:57:51 AM12/8/09
to

Zhane asked:

> are there any unused space in the CMOS memory that I can use for byte
> storage n retrieval?

RBIL reports a few as unused(reserved) and due to b7 is usually the
NMI-gate, 80..FFh will mirror 00..7Fh anyway.

Some chipset-RTC have also a 256 byte set available via port pairs
72/73h or 74/76h (! could already be gated to hardware !).

Because BIOS-extensions and OS may already use them all, I wouldn't
recommend to alter them from user-land.

My OS uses a single HD-sector to store configuration-bits, and it's
faster to RD/WR a RAM-copy from HD than to use the awful slow CMOS.

__
wolfgang

perhaps not up-to-date anymore:
copied from RBIL (once added notes marked with *):

PORT 0070-007F - CMOS RAM/RTC (REAL TIME CLOCK)
Note: the real-time clock may be either a discrete MC146814, MC146818, or
an emulation thereof built into the motherboard chipset
SeeAlso: PORT 00A0h"XT"

0070 -W CMOS RAM index register port (ISA, EISA)
bit 7 = 1 NMI disabled from reaching CPU
= 0 NMI enabled

bit 6-0 CMOS RAM index
(64 bytes in early systems, now usually 128 bytes)
Note: any write to PORT 0070h should be followed by an action to
PORT 0071h or the RTC will be left in an unknown state.

0071 RW CMOS RAM data port (ISA, EISA) (see #P0409)
(Table P0409)
Values for Real-Time Clock register number (see also CMOS.LST):
00h-0Dh clock registers
0Eh diagnostics status byte
0Fh shutdown status byte
10h diskette drive type for A: and B:
11h reserved / IBM fixed disk / setup options
12h fixed disk drive type for drive 0 and drive 1
13h reserved / AMI Extended CMOS setup (AMI Hi-Flex BIOS)
14h equipment byte
15h LSB of system base memory in Kb
16h MSB of system base memory in Kb
17h LSB of total extended memory in Kb
18h MSB of total extended memory in Kb
19h drive C extension byte
1Ah drive D extension byte
1Bh-2Dh reserved [* more drives?]
20h-27h commonly used for first user-configurable drive type
2Eh CMOS MSB checksum over 10-2D
2Fh CMOS LSB checksum over 10-2D
30h LSB of extended memory found above 1Mb at POST
31h MSB of extended memory found above 1Mb at POST
32h date century in BCD
33h information flags
34h-3Fh reserved [* windoze seem to use this]
35h-3Ch commonly used for second user-configurable drive type
3Dh-3Eh word to 82335 MCR memory config register at [22] (Phoenix)
*40h...7Fh vendor specific (not present/mirror 00..3Fh on older chips)
-eof


Zhane

unread,
Dec 8, 2009, 7:55:11 AM12/8/09
to
>
> Some chipset-RTC have also a 256 byte set available via port pairs
> 72/73h or 74/76h (! could already be gated to hardware !).

what are these for traditionally?

wolfgang kern

unread,
Dec 9, 2009, 6:08:57 AM12/9/09
to

Zhane asked:

>> Some chipset-RTC have also a 256 byte set available via port pairs
>> 72/73h or 74/76h (! could already be gated to hardware !).

> what are these for traditionally?

on old machines they may have not existed at all.
even not standard, 72/73h is found as CMOS-extension in various chipsets.
74/76h is mentioned for IBM(hw-config), I never had one, so can't tell.

Modern chipsets often contain much more than what standard BIOS/OS
would be able to detect and use. So I check the hardware manuals
rather than the Plug&Play-story, and just run a try&fail on a few
particular standarised legacy addons (ie: LAN,Sound) during install.

__
wolfgang


0 new messages