ELF binary type error during "make buildworld"
CB
ecliptica.ww at virgin.net
Sat Jan 5 01:28:09 GMT 2002
This is a multi-part message in MIME format.
------=_NextPart_000_0043_01C19588.3BE2E160
Content-Type: text/plain;
charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
Please see responses below in context - Thanks=20
----- Original Message -----=20
From: John Rochester=20
To: CB=20
Cc: freebsd-users at uk.freebsd.org=20
Sent: Saturday, January 05, 2002 1:03 AM
Subject: Re: ELF binary type error during "make buildworld"
CB wrote:
Following the advice of the Handbook #19.4 "Using make world", I =
initiated a script file and set the machine to work doing a "make =
buildworld". For over two hours all appeared to go fine until it bombed =
at:
cc -nostdinc -0 -pipe =
-I/usr/src/gnu/usr.bin/perl/perl/../../../../contrib/perl5 =
-I/usr/obj/usr/src/gnu/usr.bin/perl/perl =
-I/usr/obj/usr/src/i386/usr/include -c perlmain.c
ELF binary type not known. Use "brandelf" to brand it.
Abort trap
*** Error code 134
Stop in /usr/src/gnu/usr.bin/perl/perl.
*** Error code 1
Stop in /usr/src/gnu/usr.bin/perl.
*** Error code 1
Stop in /usr/src/gnu/usr.bin.
*** Error code 1
Stop in /usr/src/gnu.
*** Error code 1
Stop in /usr/src.
*** Error code 1
Stop in /usr/src.
# exit
This seems to be the compiler having a problem with the source file =
perlmain.c
This was actually the kernel complaining that it couldn't determine =
what system (FreeBSD/Linux/...) the program it was trying to run was =
compiled for. This is most likely due to a change in how the compiler =
writes this "branding" information. A little more context on the build =
would be useful - was this the first attempt at running a new cc that it =
had just built?
[ This was a first run through of "make buildworld" - the cc compiler =
is the one which was already in there with FreeBSD 4.0. I can dig back =
through the script file, it's over 6MBytes long but I can search for =
anything specific ]
Anyway, this has nothing to do with perlmain.c.
[ I had a look at perlmain.c and there didn't seem to be anything =
glaringly amiss ]
Does anybody else know if there are any problems with updating =
directly from 4.0-RELEASE to 4.5-PRERELEASE? I had a look at UPDATING, =
but couldn't see anything particularly relevant.
[ I browsed through UPDATING before doing the cvsup but nothing rang =
any bells for me when the error occurred ]
I located the file perlmain.c in directory =
/usr/obj/usr/src/gnu/usr.bin/perl/perl
and having checked the man page for "brandelf" tried running the =
command
brandelf -t FreeBSD perlmain.c
from within that directory.
It returns:
brandelf: file 'perlmain.c' is not in ELF format
The same response when I use brandelf with the -f command line =
switch (forces branding).
Is this a problem other people have encountered?
Am I doing something wrong?
Is there an easy fix?
Are other files likely to be similarly affected?
If I fix/replace this file will "make buildworld" insist on running =
through the whole performance again from the beginning or can it pick up =
where it left off?
Many thanks for any helpful guidance.
Charles
-- John Rochester Software Architect, Merus Software =
Ltd.Tel: +44 7870 174690 http://merus.co.uk
------=_NextPart_000_0043_01C19588.3BE2E160
Content-Type: text/html;
charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD><TITLE></TITLE>
<META content=3D"text/html; charset=3Diso-8859-1" =
http-equiv=3DContent-Type>
<META content=3D"MSHTML 5.00.2314.1000" name=3DGENERATOR></HEAD>
<BODY bgColor=3D#ffffff>
<DIV><FONT face=3DArial size=3D2>Please see responses below in context - =
Thanks=20
</FONT></DIV>
<BLOCKQUOTE=20
style=3D"BORDER-LEFT: #000000 2px solid; MARGIN-LEFT: 5px; MARGIN-RIGHT: =
0px; PADDING-LEFT: 5px; PADDING-RIGHT: 0px">
<DIV style=3D"FONT: 10pt arial">----- Original Message ----- </DIV>
<DIV=20
style=3D"BACKGROUND: #e4e4e4; FONT: 10pt arial; font-color: =
black"><B>From:</B>=20
<A href=3D"mailto:john at jrochester.org" =
title=3Djohn at jrochester.org>John=20
Rochester</A> </DIV>
<DIV style=3D"FONT: 10pt arial"><B>To:</B> <A=20
href=3D"mailto:ecliptica.ww at virgin.net" =
title=3Decliptica.ww at virgin.net>CB</A>=20
</DIV>
<DIV style=3D"FONT: 10pt arial"><B>Cc:</B> <A=20
href=3D"mailto:freebsd-users at uk.freebsd.org"=20
title=3Dfreebsd-users at uk.freebsd.org>freebsd-users at uk.freebsd.org</A> =
</DIV>
<DIV style=3D"FONT: 10pt arial"><B>Sent:</B> Saturday, January 05, =
2002 1:03=20
AM</DIV>
<DIV style=3D"FONT: 10pt arial"><B>Subject:</B> Re: ELF binary type =
error during=20
"make buildworld"</DIV>
<DIV><BR></DIV>CB wrote:<BR>
<BLOCKQUOTE cite=3D"mid:000e01c1957a$93befa80$fa58fd3e at newroad.org"=20
type=3D"cite"><META content=3D"MSHTML 5.00.2314.1000" =
name=3DGENERATOR>
<STYLE></STYLE>
<DIV> <FONT face=3DArial size=3D2>Following the advice of the =
Handbook=20
#19.4 "Using make world", I initiated a script file and set the =
machine to=20
work doing a "make buildworld". For over two hours all appeared to =
go fine=20
until it bombed at:</FONT></DIV>
<DIV> </DIV>
<DIV><FONT face=3DArial size=3D2>cc -nostdinc -0 -pipe=20
-I/usr/src/gnu/usr.bin/perl/perl/../../../../contrib/perl5=20
-I/usr/obj/usr/src/gnu/usr.bin/perl/perl =
-I/usr/obj/usr/src/i386/usr/include=20
-c perlmain.c<BR>ELF binary type not known. Use "brandelf" to brand=20
it.<BR>Abort trap<BR>*** Error code 134<BR>Stop in=20
/usr/src/gnu/usr.bin/perl/perl.<BR>*** Error code 1<BR>Stop in=20
/usr/src/gnu/usr.bin/perl.<BR>*** Error code 1<BR>Stop in=20
/usr/src/gnu/usr.bin.<BR>*** Error code 1<BR>Stop in =
/usr/src/gnu.<BR>***=20
Error code 1<BR>Stop in /usr/src.<BR>*** Error code 1<BR>Stop in=20
/usr/src.<BR># exit</FONT></DIV>
<DIV> </DIV>
<DIV><FONT face=3DArial size=3D2>This seems to be the compiler =
having a problem=20
with the source file perlmain.c</FONT></DIV></BLOCKQUOTE>
<DIV><TT>This was actually the kernel complaining that it couldn't =
determine=20
what system (FreeBSD/Linux/...) the program it was trying to run was =
compiled=20
for. This is most likely due to a change in how the compiler =
writes this=20
"branding" information. A little more context on the build would =
be=20
useful - was this the first attempt at running a new cc that it had =
just=20
built?<BR></TT></DIV>
<DIV><FONT face=3DArial><TT>[ This was a first run through of "make =
buildworld"=20
- the cc compiler is the one which was already in there with FreeBSD =
4.0. I=20
can dig back through the script file, it's over 6MBytes long but I can =
search=20
for anything specific ]</TT></FONT></DIV>
<DIV><TT></TT> </DIV>
<DIV><TT><BR>Anyway, this has nothing to do with =
perlmain.c.<BR></TT></DIV>
<DIV><FONT face=3DArial><TT>[ I had a look at perlmain.c and there=20
didn't seem to be anything glaringly amiss ]</TT></FONT></DIV>
<DIV><TT><BR>Does anybody else know if there are any problems with =
updating=20
directly from 4.0-RELEASE to 4.5-PRERELEASE? I had a look at =
UPDATING,=20
but couldn't see anything particularly relevant.<BR></TT><BR><FONT =
face=3DArial=20
size=3D2>[ I browsed through UPDATING before doing the cvsup but =
nothing rang=20
any bells for me when the error occurred ]</FONT></DIV>
<BLOCKQUOTE cite=3D"mid:000e01c1957a$93befa80$fa58fd3e at newroad.org"=20
type=3D"cite"><DIV> </DIV>
<DIV><FONT face=3DArial size=3D2>I located the file perlmain.c in =
directory=20
/usr/obj/usr/src/gnu/usr.bin/perl/perl<BR>and having checked the man =
page=20
for "brandelf" tried running the command<BR>brandelf -t FreeBSD=20
perlmain.c<BR>from within that directory.</FONT></DIV>
<DIV> </DIV>
<DIV><FONT face=3DArial size=3D2>It returns:<BR>brandelf: file =
'perlmain.c' is=20
not in ELF format</FONT></DIV>
<DIV> </DIV>
<DIV><FONT face=3DArial size=3D2>The same response when I use =
brandelf with the=20
-f command line switch (forces branding).</FONT></DIV>
<DIV> </DIV>
<DIV><FONT face=3DArial size=3D2>Is this a problem other people have =
encountered?<BR>Am I doing something wrong?<BR>Is there an easy =
fix?<BR>Are=20
other files likely to be similarly affected?<BR>If I fix/replace =
this file=20
will "make buildworld" insist on running through the whole =
performance again=20
from the beginning or can it pick up where it left off?</FONT></DIV>
<DIV> </DIV>
<DIV><FONT face=3DArial size=3D2>Many thanks for any helpful=20
guidance.<BR>Charles</FONT></DIV>
<DIV> </DIV></BLOCKQUOTE><PRE class=3Dmoz-signature =
cols=3D"$mailwrapcol">-- <BR>John Rochester Software =
Architect, Merus Software Ltd.<BR>Tel: +44 7870 174690 <A =
class=3Dmoz-txt-link-freetext =
href=3D"http://merus.co.uk">http://merus.co.uk</A><BR></PRE><BR></BLOCKQU=
OTE></BODY></HTML>
------=_NextPart_000_0043_01C19588.3BE2E160--
More information about the Ukfreebsd
mailing list