softwre suspend 2 -- getting closer...
OK, I know I've been making lots of unhelpful noise on this topic
lately.
I feel I'm getting close to a working suspendable kernel.
On my IBM thinkpad600e, I have installed a 2.6.12 kernel, using the
debian sources and the experimental kernel-patch-suspend2 patches (so,
not the upstream source/patch combo, which I can't get to work). I am
compiling with gcc 4.0.2 (maybe this is a mistake, but I don't remember
how to set the compiler version...) and my .config and
grub/menu.lst are posted here:
http://www.racesci.org/.conf/config-2.6.12-suspend2-debian-01
http://www.racesci.org/.conf/menu.lst
[you'll note I have CONFIG_SUSPEND2_CHECK_RESUME_SAFE=y in my config
file. I understand, perhaps falsely, that's the origin of this
warning, but presumably not of the problem?
I am also using the hibernate-1.12 package from unstable.
The kernel boots fine, runs fine, suspends fine. But it will not
resume, and on every boot, I get a long message:
BIG FAT WARNING
Initrd is not properly configured for resuming...
and a bunch of stuff I didn't copy down, ending with a choice
"continue" [boot this kernel w/out resuming] or "reboot"[back to
bios]. If I continue, and I in fact have a hibernation image in my
swwap file, I get further warnings about erasing that image.
I've searched the archives & the FAQ but there isn't much info re:
this particular error. I imagine I've failed to set some switch
somewhere, or something isn't being echoed properly into "proc" -- but
I don't know what.
Also, on the same machine I have Bob Proulx's patched 2.6.8 kernel
installed, which I got here:
http://www.proulx.com/~bob/nc4000/
This kernel CAN find the suspend image it generates, but crashes when
it tries to read it -- from the error message it looks like acpi gets
upset; the acpi implementation is broken on my laptop, so I was hoping
a new kernel without acpi support would work -- but no such luck, at
least not yet.
Anyway, I would much appreciate whatever help folks can give, esp. in
terms of what specifically would generate this message, whether I
might be missing some configuration element, and whether I have
exhausted my options enough to file a bug.
Thanks,
Matt
--------------------------
.''`. Matt Price
: :' : Debian User
`. `'` & hemi-geek
`-
--------------------------
Reply to: