[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: upgraded to lenny/sid and ordered RAM



On 5/19/07, Finn Thain <fthain@telegraphics.com.au> wrote:


On Thu, 17 May 2007, Brian Morris wrote:

> actually I think a lot of people quoting each other and the entropy
> effects in that are less information. the reference above i gave the guy
> actually describes doing real tests.

Sure, real tests are good. It would be great if you could test your new
SIMMs in both 605 and 63x and let us know (once they arrive).

OK, I got them today and put 64MB in q605 and 128MB in q630.

I have another set of identifying numbers too if you like
(no preference to any particular corporation, apparently
though my order was redirected ) you could search for also

64T8E164-60HYN AFE 64MB
128T16E164-60K0J-MIC AFF 128MB

neither card is apple or any other major manufacturer branded
and the individual chips differ between cards, but all say
"Korea"

both cards are double banked (unless I am confuse, i take
that to mean they have two rows of chips on a side, that
is 16 chips on a side). the 64MB is single sided and the 128
is double sided.

I wasn't really planning to try them the other way around
because I expect that the 128 would work in 605 and i am
more interest in whether the 64 would work in a 630 with
a second slot -- that is the missing one on mine which is
more toward the center of the motherboard.  So could
you or someone else try one ???

Personally, I've tested a few hundred random SIMMs so far. See,

  http://www.telegraphics.com.au/~fthain/dram/index.html

... the idea of this page being to help identify SIMM capactity -- not
compatibility (almost all work in Q650).



I don't trust generalisations like "composite SIMMs don't work" or
"single-sided don't work" or "double-sided work". What works and what
doesn't work ought to be determined by the structure of the chips
themselves (e.g. 1M x 4 chips vs 4M x 1 chips) and the number of those
chips on the module (given correct refresh rate and EDO/FPM).

well you could be right. but i do believe that some slots
are physically one sided (only) which particular
possibly your second
slot on the 631. have you used any double sided card
in the second slot ?

if you have seen the Apple Memory Guide, although the
capacities are outdated, it does discuss the single/double
sided issue some.

> what do you have now. do you have any double / single sideds that you
> can switch around just to test that part of the theory (that slot 1 is
> double sided and slot 2 is single).

I have no 64 or 128 sticks to test. I will try to obtain some, but I'd be
more comfortable if I had the specs (chip count and chip part codes) of a
module that you or someone else had already tested!

I meant trying single sided 16 in first slot or double sided
32 in second slot. some these I pulled out of junk pile of
'97 pentium pc.

not just for yourself but for possibly others out there
who are contemplating these upgrades and wondering
about their memory. I have no two slotted machines to
test, but i would offer my spare cards if I thought they
could help someone. I do think for sure that 48  or 64 MB
for most users upgrading would help significantly.

But for people building or/and programming projects now
the real test ensues. my interest sort of between system
and user level. trying / retrying some compile/build  will
take me a little while.

Although i did start up aptitude and it seems much happier
now it can grab 50MB.

__________________________

In regards to the X stuff I am going to try respecting
the starting of new thread for that.

I am sort of in the middle and hoping others will get theirs
working soon and give us some more examples.





> 1) it refuses to believe i have more than 300k of video ram. I know i
>    have at least 640x480x2 bytes, upward of 512k... but it will only go
>    in 640x480x1 Byte not 640x480x2 or 800x600x1 as in macos.

Sounds like a similar problem to the one Raylynn reported with the
console... so maybe this is a fbdev problem not xorg...
there are three layers of fbdev. from xorg their fbdev and
their fbdevhw and then the fbdev from the kernel.

so far, if I boot in 640x480x16bit then the kernel admits i
do have 600k vRAM unfortunately xorg still refuses to
see it. If i don't see something pretty soon i will file bug
on xorg driver. maybe i will fly it by you guys though.



> 2) most apps that have both X and console versions won't start their X
>    versions. for instance xemacs -nw works but not xemacs, info and w3m
>    produce their text versions in the window you type but do not start
>    from the menu (on ppc both ways give you new windows). but
>    graphicsmajick xedit xman all work so i don't know what's up.
>
> 3) the cursor keys don't work in X.

I don't know why that would be... does xorg understand the various
different ADB keyboards? I'm guessing, but I'd suggest have a close look
at the xorg.conf and failing that try xev and xmodmap.

well there are new keyboard stuffs in xorg that produces a lot of
errors though they never mattered on ppc I may have
to dig through that. I will try some things and
both of yours' suggestions



-f




Reply to: