Please respond directly since I'm not on this mailing list.
I have 2 intertwined problems that my initial web research has failed to
reveal help. I recently upgraded machines and the new one has 1GB RAM. If I
build a 2.4.0pre10 (or 8 or 9, I haven't tried earlier) kernel and chose the
1GB memory setting then only 900504 K is detected (but everything runs
stably). If I chose the 4GB memory setting then the full 1 G is detected but
I get oops. I can reliably force an oops by mounting a samba drive and then
accessing it (via ls for example).
So, is this a known issue? Should I do an oops analysis? What can I do to
Also 2 items of note. The kernel that comes RetHat 6.2 detects all of the
RAM and is stable. Related to this, although not that important, I also
noticed that via this RedHad kernel, hdparm shows memory access (not disk)
of over 200 MB/s. On my 2.4 kernels this is about 120MB/s. Any ideas why?
Second, it is a dual PIII system so is an SMP kernel, if that makes a
Any help would be greatly appreciated.
On Mon, 6 Nov 2000, Rainer Mager wrote:
> I have 2 intertwined problems that my initial web research has failed to
> reveal help. I recently upgraded machines and the new one has 1GB RAM. If I
> build a 2.4.0pre10 (or 8 or 9, I haven't tried earlier) kernel and chose the
there is no such thing yet. 2.4.0 has not yet approached the "preX" cycle
but is currently in "textX" cycle and each testX internally contains lots
of "preY", of course. So, we have kernels test9,test10-pre1,test10 etc.
I assume you are talking about test10 kernel as it is the latest and there
is never any reason to run a development kernel except the latest.
> So, is this a known issue? Should I do an oops analysis? What can I do to
> fix this?
of course you should do an oops analysis. Without it, your email is not