Package: kernel-patch-badram
Version: 2.6.12.0-1
Severity: important
I just added 2GB of RAM to my system (for a total of 3GB), and memtest86 
reports that I have a bad region of memory.  The badram pattern suggested 
by memtest86 is 0x5c8e400c,0xfffffffc.  I do plan on exchanging these bad
DIMMs for good ones, but I can't do that for a couple of weeks, so I
thought I'd just work around the problem for now.

After patching my kernel and rebooting, I checked the Memory line in the 
dmesg output as suggested by the documentation.  It reported 0k space used 
by badram.  I figured since the bad region was only four bytes, perhaps it 
was just rounding down.  To be sure, I changed the mask to 0xfffff000, to 
mark 4KiB, a full page, as bad.  But the kernel still reported 0k badram.  
I tried using a mask of 0xffff0000 to mark 64KiB, 16 full pages, as bad.  
Still 0k badram, but I did notice another change:  The reported available 
memory decreased by 16MiB!  Looking back at my previous attempts, and also 
comparing to the results when I used no badram parameter, I found that the 
reported available memory did not change until I use the larger mask.

It appears that either badram is not working or, perhaps, that the memory 
display calculated by the badram patch is incorrect.  I'm not sure how to 
tell which is the case, hence my decision to mark this bug "important", on 
the grounds that the patch may not be excluding my bad memory from the pool 
available for allocation.

Here's a list of the results I got with various badram configs:

no badram option
Memory: 3114472k/3145408k available (1616k kernel code, 29756k reserved,
728k data, 172k init, 2227904k highmem, 0k BadRAM)

Kernel command line: root=/dev/hda1 ro badram= 0x5c8e400c,0xfffffffc
Memory: 3114472k/3145408k available (1616k kernel code, 29756k reserved,
728k data, 172k init, 2227904k highmem, 0k BadRAM)

Kernel command line: root=/dev/hda1 ro badram=0x5c8e400c,0xfffff000
Memory: 3114472k/3145408k available (1616k kernel code, 29756k reserved,
728k data, 172k init, 2227904k highmem, 0k BadRAM)

Kernel command line: root=/dev/hda1 ro badram=0x5c8e400c,0xffff8000
Memory: 3114472k/3145408k available (1616k kernel code, 29756k reserved,
728k data, 172k init, 2227904k highmem, 0k BadRAM)

Kernel command line: root=/dev/hda1 ro badram=0x5c8e400c,0xffff0000
Memory: 3098088k/3145408k available (1616k kernel code, 29756k reserved,
728k data, 172k init, 2227904k highmem, 0k BadRAM

My kernel config is just a copy of the 2.6.11 AMD K8 package config,
except with badram turned on, obviously.

Thanks,

Shawn

-- System Information:
Debian Release: testing/unstable
  APT prefers unstable
  APT policy: (500, 'unstable'), (500, 'testing')
Architecture: i386 (i686)
Shell:  /bin/sh linked to /bin/bash
Kernel: Linux 2.6.11badram
Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8)

Versions of packages kernel-patch-badram depends on:
ii  bash                          3.0-16     The GNU Bourne Again SHell
ii  grep-dctrl                    2.6.7      Grep Debian package information
ii  patch                         2.5.9-2    Apply a diff file to an original

kernel-patch-badram recommends no packages.

-- no debconf information


-- 
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]

Reply via email to