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

Re: Etch becoming slower than Sarge?

Andrei Popescu wrote:
On Fri, 05 Jan 2007 17:45:40 +0000
andy <geek_show@dsl.pipex.com> wrote:

Andrei Popescu wrote:
On Fri, 05 Jan 2007 16:18:27 +0000
andy <geek_show@dsl.pipex.com> wrote:

Although the man page will identify as a bug that the info in
dmidecode can be completely wrong (!), this does show that I have
1024MB of RAM - i.e. 1GB. So, somewhere betwixt BIOS and fired up
Etch I have lost recognition for 250MB-odd of memory.
Thanks for your thoughts. On a note that might be related, I have
just watched an *.avi video from the hard-drive using Totem and was
drawing 100% CPU, with 42% of memory in cache. Perhaps this is
normal ?

My only strange experiences with Etch thus far are:
* not having the full 1 GB stick recognised,
* the heavy use of CPU,
* the overflow into swap, and
* the occasional hard freeze when burning DVDs after converting
them from non-DVD formats

Perhaps these issues are related? Could I have a bad stick of RAM?
Is it software? Will the Knoppix and memtest86 CDs you mention
help sort out which is which?

Thanks for your thoughts.

You still didn't answer any questions about the video card. Maybe
you have an on-board chipset with shared mem?

Sorry Andrei, that e-mail must have slipped past my radar. The
graphics card is 128MB ATI Radeon X200 Express Graphics built-in. I
think that it is a 32-bit and that memory is prefetchable
[size=256MB]. If this means that it shares memory from the RAM chip,
then yes, this would account for the missing RAM.

AFAIK prefetchable means something else, but if the chip is built-in it
might use shared memory. Just look through the bios settings, the
amount of shared memory is usually configurable.

OK - next reboot I'll look through the BIOS settings. I would be looking to reduce the amount of shared memory, but by how much? I am assuming that disabling or reducing it to 0 would negatively impact on performance?

Thanks for your input.


Reply to: