Nigel Henry wrote:
On Friday 12 January 2007 03:03, Jim Cornette wrote:
How do I find out if the kernels installed are the type that Kam refers to?
You might try running the below.
cat /boot/config* |grep SMP
I get this output on FC6.
CONFIG_SMP=y
# CONFIG_X86_BIGSMP is not set
CONFIG_SUSPEND_SMP=y
CONFIG_X86_FIND_SMP_CONFIG=y
CONFIG_X86_SMP=y
CONFIG_SMP=y
# CONFIG_X86_BIGSMP is not set
CONFIG_SUSPEND_SMP=y
CONFIG_X86_FIND_SMP_CONFIG=y
CONFIG_X86_SMP=y
This is not a professional approach, (no doubt) you might try
cat /boot/config-2.6.18-1.2869.fc6 |grep SMP
depending on your kernel version.
So I guess that you would append each of the versions listed below after
/boot/config.
According to synaptic the six kernels I have installed are:
2.6.15-1.2054_FC5 (this one shutsdown completely)
2.6.17-1.2174_FC5
2.6.17-1.2187_FC5
2.6.18-1.2200_FC5
2.6.18-1.2239_FC5
and
2.6.18-1.2257_FC5
I dont have any of the smp ones installed, as I only have a single processor.
None of the ones on the first list would shutdown completely as of recently,
apart from kernel 2.6.15, (which was the original one from the install
cdroms) . The others were shutting down completely, until some, as of yet
illusive upgraded package that appears to have changed the behaviour of how
later kernels than 2.6.15 deal with acpi.
Tims suggestion to use acpi=force has worked for me on FC5, but has not solved
Aarons problems on FC6.
Maybe someone with knowledge of what changes can be made to the kernel
config file or whatever can chime in. Unfortunately, I have no idea. I
just know adding acpi=on in my past systems was needed since acpi was
configured off or on at different times during the severn beta cycle.
I said on my reply to Aaron, I'm going to do a fresh install of FC5 over one
of my FC3 installs, on the same machine with the problems. I will install all
the kernels that are on the existing FC5 on this machine, then do the updates
a few at a time. Hopefully I have most of the updates
in /var/cache/apt/archives amongst the 3 installs I have of FC5 at present.
If I can avoid hours of dialup downloading of updates it will put me in a
better frame of mind, but am determined to find out which of these darned
packages has messed with acpi.
Good luck on your quest to pinpoint the origins of trouble with some
computers after upgrades were applied.
I'll be back, maybe admitting that I'm defeated, but I'll be back.
I'm interested in what is causing these errors myself. I fortunately do
not have those problems now with the present kernels. There was a bug
awhile back that caused problems with complete shutdown of the system.
It was corrected but I have no idea as to what fixed the problem. Maybe
there is a patch that was applied and now was removed, I don't know.
sorry I'm not able to give more knowledgeable information to you though.
See you with the verdict. :-)
Jim
Nigel.
--
"The trouble with doing something right the first time is that nobody
appreciates how difficult it was."
-- Walt West
