[lug] Odd mixing of SIMM modules causing boot failure

[Warren Sanders]

As stated, this weekend I wanted to upgrade my P200MMX 48MB 72 pin DRAM (2
8MB and 2 16MB FPM).  I replaced the 8MB modules with a couple 16MB *EDO*.
The system recognized the configuration, but my Red Hat 7.2 using GRUB came
in black and white... and then it had a kernel panic reporting the
"root=/dev/hda3" label was messed up or not found bootable.

I then assumed that my recent kernel upgrade had been corrupt and only
survived one boot.  Needless to say, I had already axed the other installed
kernels.  I made a similar mistake before by removing kernels before
rebooting, but this wasn't the case.  In the past I had removed the drive
and mounted it on another system to get the kernel reinstalled.  I was about
to attempt this again when I decided to boot with this drive and noticed it
was a color GRUB screen and knew right away that it was going to work now.

So placing it back in the original troubled system, I had the b/w Grub
screen again.  What did I change before this happened?  Well I never would
have thought mixing FPM with EDO would cause such a SNAFU!  Placing the old
8MB FPM modules back proved it correct.

Now I have heard many DON'Ts on using SIMMS and SDRAM but not much on FPM
and EDO.  Regardless, my system did recognize it and Linux didn't report
memory related problems; just boot partition problems!

Any thoughts?

Cheers,
-- 
Warren Sanders
http://MontanaLinux.Org