Re: psdatabase file in /boot
>>| psdatabase is responsible for the WCHAN column of the output of "ps
>>| -l", so your system will boot. Without psdatabase, you will get
>>| numbers in this column instead of routine names.
>What if I recompiled my kernel, changing some kernel config options
>in the process, but failed to run `psupdate'?
Then you'd get strange information in WCHAN, for a start. I don't
have the source for ps handy, but I expect that it hits a hash table
created from psdatabase, and you see GIGO. I doubt, knowing the
coding standards, that ps assumes it'll hit a hash entry, but it's
possible... I'll have to get the source and check it out.
> I am faced with a strange and (to me) worrisome situation in which
>I can do the debian.rules build of librl-2.0.3-2 if and only if I
>insert a call to `strace' at one pont in the file readline/Makefile,
>such that the sole difference between the altered Makefile and the
>original is as follows:
>< $(DLL_TOOLS_PATH)/mkimage -f -l librl -v $(DLL_VERSION) -a $(DLL_LOADADDR) \
>> strace $(DLL_TOOLS_PATH)/mkimage -f -l librl -v $(DLL_VERSION) -a $(DLL_LOADADDR) \
> Any possibility of a connection between this fact and the fact that
>up until a few minutes ago I had indeed been using a kernel that I had
>recompiled without updating the file `/boot/psdatabase'?
I see no possible connection. What happens if you don't insert the
'strace' in there? It looks like a context-related syntax error. I
suggest that you post to the list a context diff
('diff --context=5 oldfile newfile'), and also add in the same rule,
before that one, lines to the effect of:
cat > vars.expnd <<\xyzzy
:$(DLL_TOOLS_PATH):$(DLL_VERSION):$(DLL_LOADADDR):
xyzzy
so that vars.expnd will contain the expanded variables.
--
http://www.wp.com/joelh --- Joel Ray Holveck --- joelh@gnu.ai.mit.edu
Fourth law of computing:
Anything that can go wro
.signature: segmentation violation -- core dumped
Reply to: