[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
htags stops with "cannot fork"
From: |
Dan Shearer |
Subject: |
htags stops with "cannot fork" |
Date: |
Wed, 28 Jan 2004 18:21:33 +1030 |
User-agent: |
Mutt |
Hello global team,
I did a search of source indexing software and I like the look of what
gtags can do when all symbols are referenced (which I notice in the
public instances of kernel trees on the net is not the case. Perhaps
this is a sign of things to come, read on :-)
I am attempting to create some HTML crossreferences of User Mode Linux
kernels. I have a well-specced IA32 machine for generating these (2GB
RAM, 80Gb disk, good CPU) but htags is dying on me. Unfortunately it
does this only after many hours work!
As my first test I unpacked the Linux kernel 2.4.23, applied apatches
and then ran gtags. No problems, and several Gb of files created. Then I
ran
htags --caution -aFnsvt 'Browse the UML 2.4.23-um2 tree'
and it spent a long time indexing nearly a million different symbols (I
gathered this by glancing at the output from time to time.) The stage
'Creating temporary database' took a long time, but it did finish.
Finally I got this:
:
:
[2466/11089] converting arch/sparc64/kernel/signal.c
[2467/11089] converting arch/sparc64/kernel/signal32.c
[2468/11089] converting arch/sparc64/kernel/smp.c
htags: cannot fork.
I can raise process limits and things, but my question is why should
htags have been trying to fork lots of processes? And why did it not
fall over before when doing what is presumably quite similar operations
with the same large dataset in order to generate the files in the first
place that are now being converted?
All advice appreciated,
--
Dan Shearer
address@hidden
----- End forwarded message -----
--
Dan Shearer
address@hidden
- htags stops with "cannot fork",
Dan Shearer <=