gmane.linux.lfs.support

binutils - Pass 2 configuration issue

Void Abh — 2012-05-24T09:08:08

Hi All,

I have a user level understanding of linux and have done some application
programming using gcc.

I have just started out with LFS7.1 and it has been a smooth ride until 5.8
(Adjusting the toolchain).

The problem i am facing is with 5.9 :

While running the script: "binutils-2.22/configure", it failed with
config.log showing three errors:

*********************************************************************
configure:4009: x86_64-lfs-linux-gnu-gcc -B/tools/lib/ -V >&5
x86_64-lfs-linux-gnu-gcc: error: unrecognized option '-V'
x86_64-lfs-linux-gnu-gcc: fatal error: no input files
compilation terminated.
configure:4020: $? = 1
configure:4009: x86_64-lfs-linux-gnu-gcc -B/tools/lib/ -qversion >&5
x86_64-lfs-linux-gnu-gcc: error: unrecognized option '-qversion'
x86_64-lfs-linux-gnu-gcc: fatal error: no input files
compilation terminated.
configure:4020: $? = 1
configure:4040: checking for C compiler default output file name
configure:4062: x86_64-lfs-linux-gnu-gcc -B/tools/lib/    conftest.c  >&5
configure:4066: $? = 0
configure:4103: result: a.out
configure:4119: checking whether the C compiler works
configure:4128: ./a.out
../binutils-2.22/configure: line 4130: ./a.out: No such file or directory
configure:4132: $? = 127
configure:4139: error: in `/media/lin/sources/binutils-build':
configure:4143: error: cannot run C compiled programs.
If you meant to cross compile, use `--host'.
See `config.log' for more details.
***********************************************************************

Since man x86_64-lfs-linux-gnu-gcc didn't have -V  and -qversion options ,
i modified binutils-2.22/configure as follows:

#Original source:
for ac_option in --version -v -V -qversion; do

#Modified source
for ac_option in --version -v; do

With this change , i dont get the first two errors but the last error (not
finding ./a.out) still exists.
*********************************************************************
gcc version 4.6.2 (GCC)
configure:4020: $? = 0
configure:4040: checking for C compiler default output file name
configure:4062: x86_64-lfs-linux-gnu-gcc -B/tools/lib/    conftest.c  >&5
configure:4066: $? = 0
configure:4103: result: a.out
configure:4119: checking whether the C compiler works
configure:4128: ./a.out
../binutils-2.22/configure: line 4130: ./a.out: No such file or directory
configure:4132: $? = 127
configure:4139: error: in `/media/lin/sources/binutils-build':
configure:4143: error: cannot run C compiled programs.
If you meant to cross compile, use `--host'.
See `config.log' for more details.
*********************************************************************

My questions:
1) How come invalid options such as -V and -qversion have ended up in the
configure script ? Is there something that i have missed here ?
2) How come the same scripts worked during the first pass ?
3) How do i solve the last problem ? (./a.out)


I would be grateful for any help/ advise.

Regards

lfs7.1 cannot boot

Omar — 2012-05-24T07:07:34

Hi, all:
I finished all work of the LFS 7.1 book except the error when booting my LFS.

Firstly, I states my LFS 7.1.
I use VMware installed Ubuntu 10.04 on a virtual SCSI disk of 20G. Before beginning, I add another 8G virtual SCSI disk to VM and mount it in the Ubuntu manually. So the first disk with Ubuntu displays sda in /dev/ and the second which is mounted newly displays sdb in /dev/. Following the book I install LFS 7.1 on the sdb1 which is formatted with ext3 on the sdb and given only one partition. All pass with no error.
According to my disk in chapter 8.4.3 I run cmd as grub-install /dev/sdb and in chapter 8.4.4 I save grub.cfg as follows.

# Begin /boot/grub/grub.cfg
set default=0
set timeout=5

insmod ext2
set root=(hd1,1)

menuentry "GNU/Linux, Linux 3.2.6-lfs-7.1" {
linux /boot/vmlinuz-3.2.6-lfs-7.1 root=/dev/sdb1 ro
}

After a few works such as logout and unmount, I reboot the computer. But it starts up using Ubuntu again only appearing error checking of a moment.

Then I change the cmd in chapter 8.4.3 to grub-install /dev/sda and keep grub.cfg the same and reboot again. The computer gives errors as follows and stop starting up.

md: Autodetecting RAID arrays.
md: Scanned 0 and added 0 devices.
md: Autorun ...
md: ... autorun DONE.
Root-NFS: on NFS server address
VFS: Unable to mount root fs via NFS. trying floppy.
VFS: cannot open root device "sdb1" or unknown-block(2.0)
Please append a correct "root=" boot option; here are the available partitions:
0b00 1048575 sr0 driver: sr
Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(2.0)
Pid: 1. comm: swapper/0 Not tained 3.2.6 #1
Call trace:
...mount_block_root+0x141/0x1c9...mount_root...kernel_init...

I search this problem from lfs mail list and google, which says that compiling kernel needs some SCSI driver or changing hda to sda in the grub.cfg and etc. .I compile the kernel again with more drivers like SCSI. When I reboot again, it is the same.

Could anybody help me? Thanks in advance.

Omar--
http://linuxfromscratch.org/mailman/listinfo/lfs-support
FAQ: http://www.linuxfromscratch.org/lfs/faq.html
Unsubscribe: See the above information page

handle power button in LFS

xinglp — 2012-05-18T13:39:33

In order to use "power button" to shutdown or suspend my lfs.
I enabled the depreciate  /proc/acpi/event in kernel,
and install acpid to handle the button event.
It works well.

But is there any clear way to do that, such as by inittab or udev
or kernel callout userspace script ( /proc/sys/kernel/poweroff_cmd has
somethings )

I've looked into blfs, there's a lot more packages need to be
installed to enable this.(I'm not a X user,for now.)

Thanks.

LFS-7.1: 6.37. Automake-1.11.3 (TEST FAILURE)!

Yasser Zamani — 2012-05-14T18:26:10

Hi,

'make -j2 check' on 'LFS-7.1: 6.37.
        Automake-1.11.3 (http://www.linuxfromscratch.org/lfs/view/stable/chapter06/automake.html)' informs me for one failure with following red message :(

=====================================
1 of 842 tests failed
(96 tests were not run)
See tests/test-suite.log
Please report to bug-automake< at >gnu.org
=====================================
make[3]: *** [test-suite.log] Error 1
make[3]: Leaving directory `/sources/automake-1.11.3/tests'
make[2]: *** [check-TESTS] Error 2
make[2]: Leaving directory `/sources/automake-1.11.3/tests'
make[1]: *** [check-am] Error 2
make[1]: Leaving directory `/sources/automake-1.11.3/tests'
make: *** [check-recursive] Error 1

I also reviewed 'tests/test-suite.log' for word 'Error' and found following rows which are maybe related:

+ make
Making all in src
make[1]: Entering directory `/sources/automake-1.11.3/tests/depcomp10.dir/build/src'
source='../../src/foo.c' object='foo.o' libtool=no \
DEPDIR=.deps depmode=hp /bin/sh ../../depcomp \
gcc -DPACKAGE_NAME=\"depcomp10\" -DPACKAGE_TARNAME=\"depcomp10\" -DPACKAGE_VERSION=\"1.0\" -DPACKAGE_STRING=\"depcomp10\ 1.0\" -DPACKAGE_BUGREPORT=\"\" -DPACKAGE_URL=\"\" -DPACKAGE=\"depcomp10\" -DVERSION=\"1.0\" -I. -I../../src     -g -O2 -c -o foo.o ../../src/foo.c
cc1: error: unrecognized command line option '-M.deps/foo.TPo'
make[1]: *** [foo.o] Error 1
make[1]: Leaving directory `/sources/automake-1.11.3/tests/depcomp10.dir/build/src'
make: *** [all-recursive] Error 1
+ Exit 77

Is this failure expected or is a fatal error (and the process could not be continued before resolving it)?! Or do you think that eliminating '-j2' switch will resolve it (I should know before trying because it's test take 31 minutes to complete(i.e. 1 hour without -j2!))? 

Thanks in advance!

-Yasser
       -- 
http://linuxfromscratch.org/mailman/listinfo/lfs-support
FAQ: http://www.linuxfromscratch.org/lfs/faq.html
Unsubscribe: See the above information page

LFS-71.: Solution for automation of 'tar X', 'cd X' ... 'cd ..', 'rm -R X'.

Yasser Zamani — 2012-05-14T17:18:01

Hi there,

As for each package you should decompress the package, enter decompressed folder, compile&install and finally remove decompressed folder, and also while you do not learn any new thing by doing these each time, so I would like to share my automation solution which does these just by getting a partial name of the package:

To create the main .sh file, issue following commands while you're logged in by chroot (section 6 of LFS-7.1) and also you're in '$LFS/sources' directory:

cat > t.sh << EOF
if [ "$1" == "" ]; then
    echo 'USAGE: t.bash [name]'
    return
fi
FNAME=$(ls $1* | grep tar.xz)
if [ "$FNAME" == "" ]; then
    FNAME=$(ls $1* | grep tar.gz)
    if [ "$FNAME" == "" ]; then
        FNAME=$(ls $1* | grep tar.bz2)
        if [ "$FNAME" == "" ]; then
            echo 'match not found!'
            return
        else
            TARPARAM='-jxf'
            EXT='.tar.bz2'
        fi
    else
        TARPARAM='-zxf'
        EXT='.tar.gz'
    fi
else
    TARPARAM='-Jxf'
    EXT='.tar.xz'
fi
DIR=$(echo $FNAME | sed "s/$EXT//g")
echo `rm -fR $DIR`
echo `ls $1*`
echo `tar $TARPARAM $FNAME`
rm -vfR ut.sh
echo -e 'cd ..\necho `rm -fR '$DIR'`\necho `ls '$1'*`\nrm -vfR ut.sh' >> ut.sh
chmod 777 ut.sh
cd $DIR
EOF
chmod 777 t.sh

Now, to use automation, act like following:
Issue following command to decompress and enter to decompressed folder by just specifying package partial name:source ./t.sh [package-partial-name] e.g. source ./t.sh autoco for autoconf-2.68 package.NOTE: package-partial-name should be long enough to distinct the package from others e.g. source ./t.sh bi is not suitable at all because can not distinct between binutils and bison!Do the book's instructions to configure, compile, test and install the package.Issue following command to undo the above instruction #1.source ../ut.sh.
Sicerely Yours,
Yasser.
       -- 
http://linuxfromscratch.org/mailman/listinfo/lfs-support
FAQ: http://www.linuxfromscratch.org/lfs/faq.html
Unsubscribe: See the above information page

6.2.3. Mounting Virtual Kernel File Systems question

Wayne Blaszczyk — 2012-05-13T05:58:43

Hi All,
I've recently updated my main workstation to LFS 7.1 (From LFS 6.7), and
noticed one of my auto build scripts is now failing.
I've tracked it down to the following command.
mount -vt tmpfs shm $LFS/dev/shm
It seems to me that this command is mounting onto /run/shm rather than
/mnt/lfs/dev/shm due to /mnt/lfs/dev/shm being a symlink to /run/shm.
Further more, when entering chroot environment, /dev/shm is now pointing
to a no existent /run/shm.
Am I'm doing something wrong or is this a bug?
Thanks,
Wayne.

Do you know why 2 CPUs act like a mirror in GCC test?

Yasser Zamani — 2012-05-12T16:41:19

Hi,

Sorry if it's off-topic; do you know why 2 CPUs act like a mirror while I'm running "make -k check" for testing GCC-4.6.2 (6.17's section of LFS-7.1)? it's not a problem but just I would like to know; I've attached an image which shows this while I was not running anything except GCC testing and Debian's System Monitor.

I think it'll be an interesting reason that causes 2 CPU's mirror action during all test process!

Thanks!

-Yasser
       -- 
http://linuxfromscratch.org/mailman/listinfo/lfs-support
FAQ: http://www.linuxfromscratch.org/lfs/faq.html
Unsubscribe: See the above information page

Problems starting the final LFS system

Lázaro Morales — 2012-05-11T19:30:01

Hello,

After complete the LFS book, when I restart the computer I get a few  
errors messages and the computer stuck after this:

     /lib/lsb/init-functions: line 590: /run/var/bootlog: No such file or  
directory
     /etc/rc.d/init.d/rc: line 217: /var/log/boot.log: Read-only file system
     /etc/rc.d/init.d/rc: line 220: /var/log/boot.log: Read-only file system
     [  301.811991] rc used greatest stack depth: 5556 byte left
     INIT: Id "2" respawning too fast: disable for 5 minutes
     INIT: Id "4" respawning too fast: disable for 5 minutes
     INIT: Id "5" respawning too fast: disable for 5 minutes
     INIT: Id "6" respawning too fast: disable for 5 minutes
     INIT: Id "1" respawning too fast: disable for 5 minutes
     INIT: Id "3" respawning too fast: disable for 5 minutes
     INIT: no more processes left in this runlevel

The LFS version is 7.1 and is running on VirtualBox with PAE/NX and VT-x,  
the processor is a dual-core E5300 and the virtual machine have 512MiB of  
RAM. Could be this a problem? LFS can be compiled on VirtualBox?

Thanks very much,
Lázaro

Questions about locale settings

Lázaro Morales — 2012-05-11T13:44:07

Hello,

In section 7.13, the `/etc/profile` file could look like this?

# Begin /etc/profile
export LANG=en_GB.iso88591
# End /etc/profile

Need I specify some modifier?

Thanks,
Lázaro

.sh file for chapter 6 (all mounts in one)

Yasser Zamani — 2012-05-09T16:19:38

Hi there,

While processing chapter 6 of LFS-7.1 could be long, I would like to share my .sh file which prepares your environment to continue with chapter 6 after system reboot.

While you are logged in as root (via command su), issue following command to create a .sh file named go-lfs-6.sh:

cat > go-lfs-6.sh << EOF
LFS = /mnt/lfs
export LFS
mount -v <xxx> $LFS
mount -v --bind /dev $LFS/dev
mount -vt devpts devpts $LFS/dev/pts
mount -vt tmpfs shm $LFS/dev/shm
mount -vt proc proc $LFS/proc
mount -vt sysfs sysfs $LFS/sys
chroot "$LFS" /tools/bin/env -i HOME=/root TERM="xterm" PS1='\[\033[1;31m\]\u:\w$\[\033[0m\] ' PATH=/bin:/usr/bin:/sbin:/usr/sbin:/tools/bin /tools/bin/bash --login +h
EOF

NOTE: Replace <xxx> with your LFS partition e.g. /dev/sda7.
NOTE: chroot command has been rewrote to make the command prompt color to light red which can help you to separate command's results from each other by eye in bash!

Then every time you reboot, while you are logged in as root (via command su), just issue following command to prepare to continue with chapter 6:

bash ./go-lfs-6.sh

Sincerely Yours,
Yasser.
       -- 
http://linuxfromscratch.org/mailman/listinfo/lfs-support
FAQ: http://www.linuxfromscratch.org/lfs/faq.html
Unsubscribe: See the above information page

LFS Chapter 6.37. Automake-1.11.3 error

Andrew Kaunchy — 2012-05-09T13:07:00

Hello there!

This is my first building LFS by following the LFS-BOOK-7.1.
After long time work,more than 12 hours since the building process
started,a problem occured when I was making 'Automake-1.11.3' tests at
'Chapter 6.37. Automake-1.11.3.'The feedback given by the terminal is
as follows.

=====================================
1 of 842 tests failed
(96 tests were not run)
See tests/test-suite.log
Please report to bug-automake< at >gnu.org
=====================================
make[3]: *** [test-suite.log] Error 1
make[3]: Leaving directory `/sources/automake-1.11.3/tests'
make[2]: *** [check-TESTS] Error 2
make[2]: Leaving directory `/sources/automake-1.11.3/tests'
make[1]: *** [check-am] Error 2
make[1]: Leaving directory `/sources/automake-1.11.3/tests'
make: *** [check-recursive] Error 1

And I checked the make process,the FAIL info is:

FAIL: conffile-leading-dot.test

I have reinstalled the 'Automake-1.11.3' after this error occured
at the first time,and unfortunately,it failed again.

Can anyone tell me why this error occured and how to solve it?
Do I need to rebuild the LFS from zero again?

LFS 7.1 - 6.17. GCC-4.6.2

Lázaro Morales — 2012-05-08T12:12:37

Hello,

The GCC compilation process finished fine, but I get the next errors in  
the testing phase and then the test is stopped.

                          === libgomp Summary ===
    [...]
    make[2]: Target `check-target' not remade because of errors.
    make[2]: Leaving directory `/sources/gcc-build'
    make[1]: *** [do-check] Error 2
    make[1]: Target `check' not remade because of errors

I checked the log in  
http://www.linuxfromscratch.org/lfs/build-logs/7.1/core2duo/test-logs/080-gcc  
and this occur and the test continue without problems. Is there some  
reason of why did not continue the test? What do you recommend me at this  
point? Should I compile again or test the suite again?

Thanks!
Lázaro

wget-list update

Eleanore Boyd — 2012-05-08T04:14:28

I noticed that the wget-list file has outdated links for man-pages and 
zlib. For some reason, man7.org doesn't seem to host downloads anymore 
at all, and zlib got an update from the version in the book. Huzzah for 
google, at least.

http://www.kernel.org/pub/linux/docs/man-pages/man-pages-3.35.tar.gz
http://sourceforge.net/projects/libpng/files/zlib/1.2.6/zlib-1.2.6.tar.bz2/download

I don't know how to specify a mirror from Sourceforge, but I would think 
that would redirect properly.

Elly

mounting via .bashrc as root (2 suggestions)

Yasser Zamani — 2012-05-07T17:20:14

Hi there,

I have two suggestions for the book which I hope will be helpful:

As in every computer boot up and logging in as lfs user I had to run 'mount /dev/sda7 /mnt/lfs' as root, I added (rewrited) this to .bashrc file as following (rewriting LFS-7.1 4.4. Setting Up the Environment section):

cat > ~/.bashrc << "EOF"
set +h
umask 022
LFS=/mnt/lfs
LC_ALL=POSIX
LFS_TGT=$(uname -m)-lfs-linux-gnu
PATH=/tools/bin:/bin:/usr/bin
export LFS LC_ALL LFS_TGT PATH
mount /dev/sda7 $LFS
EOF

However I have two issues with this:
I know I should use 'sudo mount' in Debian but what command will work in all distros? because book's reader can be on any distro.The mount command should not run if it's already mounted. how it can be verified before mounting, via Linux commands?

Another suggestion is doing LFS-7.1 4.4 steps for root too in beginning of part III because after this part, all commands should be run as root but $LFS is not declared in root. so I issued following command while I was logged in as root and now I don't have to declare $LFS every time I login to root:
NOTE: I commented that lines which I was not sure about them if they should run in root mode too.

cat > ~/.bashrc << "EOF"
# set +h
# umask 022
LFS=/mnt/lfs
# LC_ALL=POSIX
# LFS_TGT=$(uname -m)-lfs-linux-gnu
# PATH=/tools/bin:/bin:/usr/bin
export LFS
# LC_ALL LFS_TGT PATH
mount /dev/sda7 $LFS
EOF

What's your idea? Are these useful and meaningful?

Thanks in advance!

Sincerely Yours,
Yasser.       -- 
http://linuxfromscratch.org/mailman/listinfo/lfs-support
FAQ: http://www.linuxfromscratch.org/lfs/faq.html
Unsubscribe: See the above information page

LFS 7.1 - 6.9. Glibc-2.14.1

Lázaro Morales — 2012-05-07T13:19:02

Hello,

After running the test suit, I get the next errors:

   # make -k check 2>&1 | tee glibc-check-log
   # grep Error glibc-check-log
   make[2]: *** [/sources/glibc-build/stdio-common/bug22.out] Error 1
   make[1]: *** [stdio-common/tests] Error 2
   make[2]: *** [/sources/glibc-build/posix/bug-regex32.out] Error 1
   make[2]: [/sources/glibc-build/posix/annexc.out] Error 1 (ignored)
   make[1]: *** [posix/tests] Error 2
   make: *** [check] Error 2

Can I continue without problems?

Thanks!
Lázaro

Automated LFS Program (Ta-Kenset)

TW — 2012-05-05T22:36:24

Hello,

I have been working on automating the Linux From Scratch build (and general builds) and 
came up with a program that you may find useful (hopefully).  I called it Ta-Kenset (takenset) and 
it can be retrieved at http://www.takenset.us.

I've built several LFS systems with this program, but I've only tried it from a fresh install of 
Debian, so I'm interested if it's not working on your distribution (or any errors, in general).  

Also, for now, I'm running this from home off of the line I get from Time Warner, so my apologies 
for any slow downloading.  Please let me know if you're having trouble getting into the site.

Take care and I hope this program helps your LFS builds.  And, thank you for your help over the past 
couple of years.

Shoul I remember (keep in mind) all steps?!

Yasser Zamani — 2012-05-05T15:20:22

Hi there,

At first thank you very much for this useful site and book!

Currently I'm doing jobs step-by-step with no getting any errors; however, sometimes the steps are very specific e.g. GCC pass 2 steps. GCC needs that patch, fixincludes should be suppressed, -fomit-frame-pointer should be removed and etc. should I really remember and keep in mind these?! how book writer discovered these? because we just do steps and if we try to e.g. use another version then we are not sure about patches and switches, right?

One more thing; sometimes I know why to run the script but I don't know how script works exactly. the main example is sed. I know it edits streams to replace or remove something but it's command in book is complicated for me at this time to understand. Is it essential to understand how it exactly works?

Thanks in advance!
       -- 
http://linuxfromscratch.org/mailman/listinfo/lfs-support
FAQ: http://www.linuxfromscratch.org/lfs/faq.html
Unsubscribe: See the above information page

Chapter 5 questions

Scott Robertson — 2012-05-03T04:53:19

Hello,

Can someone explain what the directory structure is supposed to look like in Chapter 5?

1. Where does the Binutils folder/directory go? Where does the GCC folder go? That's all the further I have gotten so far.  I think I will probably want to know where all the folders go.   I created a separate partition for LFS, and put the folders there, is that right, or not?
Or is it supposed to go into the /mnt/lfs folder? Or somewhere else?


2.  Can someone also explain permissions?  Who is supposed to have permissions on what?  What permissions are supposed to be on the Binutils and GCC directories?  root?  lfs?  

And what are they supposed to be? read, write and execute?

3.  What directory am I supposed to be in when I execute the commands?  As I said, I have an LFS partition, so am I supposed to be in /media/LFS?  Or more like /mnt/lfs? Or somewhere else?

4. Also, what does the two dots and slash mean?    ../(some command)  

How does that factor into it?



Thank in advance, 

Scott

Bk7.1_Ch6.17 GCC-4.6.2 search order of 'include directories'

pete Spam-Avoider — 2012-04-29T00:01:37

    snipped 


 >  -- Bruce


Thanks Bruce and Fernando for replying so quickly.

Bruce, your explanation sounds reasonable, and, if its working on your system
then I'm sure it'll work for me too.

Kind of suprised there weren't more posts about this.

Thanks again.


pete
x2164

New failure in coreutils tests

Ken Moffat — 2012-04-28T23:30:17

 I'm building the current book with the newer grep, man-pages,
kernel-headers, and autofoo that Matt is working on.  This is as a
test of my hardware (looking good at the moment, but a few BLFS
packages have caused me ICEs recently, so not yet certain if it is
now ok - will reply to the earlier thread with the gory details when
I'm convinced :)

 The host system is LFS-7.1.  In coreutils it now fails in 'check
user' with
../build-aux/test-driver: line 95: 12348 Aborted "$< at >" > $log_file 2>&1
FAIL: test-getlogin

 which sets non-zero status and breaks my build at the end of 'check
user'.  The code is

# Test script is run here.
"$< at >" >$log_file 2>&1
estatus=$?
if test $enable_hard_errors = no && test $estatus -eq 99; then
  estatus=1
fi

 I'm assuming this is from gnulib-tests/test-getlogin.c.  Not seen
this failure in my builds earlier this week, perhaps it's a one-off
or perhaps something in the 3.3.4 headers has changed.

 Or, maybe it's just my hardware (again).

ĸen

Bk7.1_Ch6.17 GCC-4.6.2 search order of 'includedirectories'

pete Spam-Avoider — 2012-04-28T21:00:33

  Hi, 

  I'm building stable book 7.1 (Bk-7.1) using a stable 
  book 6.8 (Bk-6.8) system which meets all the
  "Host System Requirement" of Bk-7.1.

  I've followed the book to the letter. All commands have
  been taken from the text of Bk-7.1 by cutting and pasting text
  from a lynx browser.

  So far I've reached Bk-7.1_Ch-6.17, GCC-4.6.2, which has been
  compiled, tested and installed.  No unexpected failures
  occurred during 'make -k check'.  I've halted further
  progress past this chapter to see if I can get some input
  concerning the difference between the expected output of
  Bk-7.1_Ch-6.17's command:

     'grep -B4 '^ /usr/include' dummy.log'

  and the actual output I received.  A similar problem was
  described in a post to this list on Jan. 17, 2012 by Ronnie van Aarle
  with a response each from jasonpsage and emteeoh.
 
  The expected output is:

#include <...> search starts here:
 /usr/local/include
 /usr/lib/gcc/i686-pc-linux-gnu/4.6.2/include
 /usr/lib/gcc/i686-pc-linux-gnu/4.6.2/include-fixed
 /usr/include

  The output from my system is:

#include <...> search starts here:
 /usr/lib/gcc/i686-pc-linux-gnu/4.6.2/include
 /usr/local/include
 /usr/lib/gcc/i686-pc-linux-gnu/4.6.2/include-fixed
 /usr/include


  My concern is to have the correct 'include directory' search
  order.

  All of the other commands in Ch-6.17 for testing the installation of
  gcc-4.6.2 produced the example outputs as presented in the chapter.

  Am I looking at an 'include directory' search order problem?

  Thanks in advance.

  pete jordan
  x2164