Date: Fri, 20 Jun 1997 13:10:09 +0930 (CST) From: Michael Smith <msmith@atrad.adelaide.edu.au> To: mark@quickweb.com (Mark Mayo) Cc: msmith@atrad.adelaide.edu.au, mark@quickweb.com, freebsd-hackers@FreeBSD.ORG Subject: Re: make world error in RELENG_2_2 Message-ID: <199706200340.NAA01931@genesis.atrad.adelaide.edu.au> In-Reply-To: <19970619215732.54070@vinyl.quickweb.com> from Mark Mayo at "Jun 19, 97 09:57:32 pm"
next in thread | previous in thread | raw e-mail | index | archive | help
Mark Mayo stands accused of saying:
>
> Yeah. However, I still don't understand why my header files are screwed.
> Here's what I'm confused about:
> 'make world' fairly early in the game does a 'make includes'. Based on this
> I assumed that I didn't have to manually do a 'make includes' myself.
> Result: doing a 'make includes' while the old /usr/include still existed
> didn't get me a clean build. After following the advice of Jordan and Mike,
> I did
> cd /usr/include
> rm -rf *
> cd /usr/src
> make includes
>
> Result:
>
> cd /usr/src/include; install -C -o bin -g bin -m 444 a.out.h ar.h assert.h
> bitstring.h ctype.h db.h dirent.h disktab.h err.h f2c.h fnmatch.h fstab.h fts.h
> glob.h grp.h strhash.h histedit.h kvm.h limits.h link.h locale.h malloc.h
> memory.h mpool.h ndbm.h netdb.h nl_types.h nlist.h paths.h pthread.h
> pthread_np.h pwd.h ranlib.h regex.h regexp.h resolv.h rune.h runetype.h
> setjmp.h sgtty.h signal.h stab.h stddef.h stdio.h stdlib.h string.h strings.h
> struct.h sysexits.h tar.h time.h timers.h ttyent.h unistd.h utime.h utmp.h
> vis.h /usr/include
> cd /usr/src/include/arpa; install -C -o bin -g bin -m 444 ftp.h inet.h
> nameser.h telnet.h tftp.h /usr/include/arpa
> usage: install [-CcDps] [-f flags] [-g group] [-m mode] [-o owner] file1 file2
> install [-CcDps] [-f flags] [-g group] [-m mode] [-o owner] file1 ...
> fileN directory
> install -d [-g group] [-m mode] [-o owner] directory ...
> *** Error code 64
>
Hmm. Looks like 'make includes' doesn't do the mtree stuff first.
Can you read Makefiles? Do you understand the basic idea behind them?
If not, now is the time to learn. I agree that install(1) should be
giving you a helpful error message, and it's not.
> I'll be adding a section entitles "What to do when make world fails", that
> will go throough some of the common problems that might occur during
> dramatic reorganizations of the source tree, and how to get around them.
Here is a checklist you might want to include somewhere near the top
where impatient people will find it. This isn't my work; if the
author wants to be identified I'm sure they'll stand up and say so :
1. If your CPU is overclocked, shut up and don't bother us again.
2. Make absolutely sure your entire source tree is 100% up-to-date,
using CVSup or CTM.
3. If you cheated in step 2, in any way whatsoever, shut up and
don't bother us again.
4. Execute these commands as root:
rm -rf /usr/obj
mkdir /usr/obj
find /usr/src -name obj -type l | xargs rm
make world
5. If you didn't type the commands in step 4 precisely as directed,
shut up and don't bother us again.
6. If it still fails, wait 24 hours. Then repeat steps 1-5
again one time.
7. If you didn't wait 24 hours and then repeat steps 1-5 exactly
as directed, shut up and don't bother us again.
8. If it still fails in exactly the same way as the first time, tell
us about it.
Ok, so they may want toning down a bit for general consumption...
--
]] Mike Smith, Software Engineer msmith@gsoft.com.au [[
]] Genesis Software genesis@gsoft.com.au [[
]] High-speed data acquisition and (GSM mobile) 0411-222-496 [[
]] realtime instrument control. (ph) +61-8-8267-3493 [[
]] Unix hardware collector. "Where are your PEZ?" The Tick [[
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199706200340.NAA01931>