[Bug bootstrap/49072] New: Error building the compiler

[Bug bootstrap/49072] New: Error building the compiler

[Jose.Pascual-Gutierr@1-ACT.com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49072

           Summary: Error building the compiler
           Product: gcc
           Version: 4.6.0
            Status: UNCONFIRMED
          Severity: critical
          Priority: P3
         Component: bootstrap
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: Jose.Pascual-Gutierr@1-ACT.com


I downloaded from one of the mirror sites the following tar balls:

gcc-4.6.0.tar.gz
gmp-4.3.2.tar.bz2
mpc-0.8.1.tar.gz
mpfr-2.4.2.tar.bz2

I decompressed and untarred the files, which resulted in four directories,
respectively: gcc-4.6.0, gmp-4.3.2, mpc-0.8.1, and mpfr-2.4.2. I then created
the directory ~/gcc4 and I moved gcc-4.6.0 in it. I then changed the names of
the directories gmp-4.3.2, mpc-0.8.1, and mpfr-2.4.2 to gmp, mpc, and mpfr,
respectively. These three directories, namely gmp, mpc, and mpfr were moved it
to ~/gcc4/gcc-4.6.0. In gcc4, I created the directory objdir. So, the directory
tree looks like the following:

|--gcc4
|  |--objdir
|  |--gcc-4.6.0
|  |  |--gmp
|  |  |--mpc
|  |  |--mpfr

I then proceeded to configure the installation. From ~/gcc4/objdir, I executed
“../gcc-4.6.0/configure”.
Once the Makefile was created, still from within ~/gcc4/objdir, I executed
“make bootstrap”. After about 24 hours of execution, the build failed with the
following error message:

../../../gcc-4.6.0/libgcc/../gcc/libgcc2.c:791:1: internal compiler error:
Floating point exception
Please submit a full bug report,
With preprocessed source if appropriate.
See http://gcc.gnu.org/bugs.html for instructinos.
make [3]: *** [_popcountsi2.o] Error 1
make [3]: Leaving directory
‘/home/Jose.Pascual-Gutierr/gcc4/objdir/i686-pc-cygwin/libgcc’
make [2]: *** [all-stage2-target-libgcc] Error 2
make [2]: Leaving directory ‘/home/Jose.Pascual-Gutierr/gcc4/objdir’
make [1]: *** [stage2-bubble] Error 2
make [1]: Leaving directory ‘/home/Jose.Pascual-Gutierr/gcc4/objdir’
make: *** [bootstrap] Error 2

All the help you can provide will be very appreciated.
Thank you very much,
José

[Bug bootstrap/49072] Error building the compiler

[redi at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49072

--- Comment #1 from Jonathan Wakely <redi at gcc dot gnu.org> 2011-05-19 22:30:41 UTC ---
(In reply to comment #0)
> I downloaded from one of the mirror sites the following tar balls:
> 
> gcc-4.6.0.tar.gz
> gmp-4.3.2.tar.bz2
> mpc-0.8.1.tar.gz
> mpfr-2.4.2.tar.bz2
> 
> I decompressed and untarred the files, which resulted in four directories,
> respectively: gcc-4.6.0, gmp-4.3.2, mpc-0.8.1, and mpfr-2.4.2. I then created
> the directory ~/gcc4 and I moved gcc-4.6.0 in it. I then changed the names of
> the directories gmp-4.3.2, mpc-0.8.1, and mpfr-2.4.2 to gmp, mpc, and mpfr,
> respectively. These three directories, namely gmp, mpc, and mpfr were moved it
> to ~/gcc4/gcc-4.6.0. In gcc4, I created the directory objdir. So, the directory
> tree looks like the following:
> 
> |--gcc4
> |  |--objdir
> |  |--gcc-4.6.0
> |  |  |--gmp
> |  |  |--mpc
> |  |  |--mpfr
> 
> I then proceeded to configure the installation. From ~/gcc4/objdir, I executed
> “../gcc-4.6.0/configure”.

This is all ok so far, which puts you well ahead of almost all people who
report problems building gcc!

> Once the Makefile was created, still from within ~/gcc4/objdir, I executed
> “make bootstrap”.

Just "make" is all you need, "make bootstrap" hasn't been needed for some time
now.


> After about 24 hours of execution, the build failed with the
> following error message:
> 
> ../../../gcc-4.6.0/libgcc/../gcc/libgcc2.c:791:1: internal compiler error:
> Floating point exception
> Please submit a full bug report,
> With preprocessed source if appropriate.
> See http://gcc.gnu.org/bugs.html for instructinos.
> make [3]: *** [_popcountsi2.o] Error 1
> make [3]: Leaving directory
> ‘/home/Jose.Pascual-Gutierr/gcc4/objdir/i686-pc-cygwin/libgcc’
> make [2]: *** [all-stage2-target-libgcc] Error 2
> make [2]: Leaving directory ‘/home/Jose.Pascual-Gutierr/gcc4/objdir’
> make [1]: *** [stage2-bubble] Error 2
> make [1]: Leaving directory ‘/home/Jose.Pascual-Gutierr/gcc4/objdir’
> make: *** [bootstrap] Error 2

What if you type "make" again?  Does it fail at the same place?

It might not be relevant, but although it's the minimum required version, MPFR
2.4.2 is a bit old now.  I suspect most people building GCC 4.6 are using a
newer version of MPFR.  Does it help if you repeat the whole process using MPFR
3.0.1 instead?

[Bug bootstrap/49072] Error building the compiler

[Jose.Pascual-Gutierr@1-ACT.com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49072

--- Comment #2 from Jose <Jose.Pascual-Gutierr@1-ACT.com> 2011-05-20 13:47:38 UTC ---
Thank you very much, Jonathan, for your comments.

I tried as you suggested. Instead of using mpfr-2.4.2, I downloaded the latest
version, namely, mpfr-3.0.1. I also tried with mpc-0.8.1 and mpc-0.9. The same
file tree structure and directory names were maintained as in my original
message. Also, I simply did a "make" and not a "make bootstrap". The building
failed in all cases:

make [5]: *** [acos.lo] Error 1
make [5]: Leavimng directory
'/home/Jose.Pascual-Gutierr/gcc_compileres/objdir/mpc/src'
Making all in tests
make [5]: Entering directory
''/home/Jose.Pascual-Gutierr/gcc_compileres/objdir/mpc/tests'
make [5]: Nothing to be done for 'all'.
make [5]: Leavimng directory
'/home/Jose.Pascual-Gutierr/gcc_compileres/objdir/mpc/tests'
Making all in doc
make [5]: Entering directory
''/home/Jose.Pascual-Gutierr/gcc_compileres/objdir/mpc/doc'
make [5]: Nothing to be done for 'all'.
make [5]: Leaving directory
''/home/Jose.Pascual-Gutierr/gcc_compileres/objdir/mpc/doc'
make [5]: Entering directory
''/home/Jose.Pascual-Gutierr/gcc_compileres/objdir/mpc'
make [5]: Leaving directory
''/home/Jose.Pascual-Gutierr/gcc_compileres/objdir/mpc'
make [4]: *** [all-recursive] Error 1
make [4]: Leaving directory
''/home/Jose.Pascual-Gutierr/gcc_compileres/objdir/mpc'
make [3]: *** [all] Error 2
make [3]: Leaving directory
''/home/Jose.Pascual-Gutierr/gcc_compileres/objdir/mpc'
make [2]: *** [all-stage1-mpc] Error 2
make [2]: Leaving directory ''/home/Jose.Pascual-Gutierr/gcc_compileres/objdir'
make [1]: *** [stage1-bubble] Error 2
make [1]: Leaving directory ''/home/Jose.Pascual-Gutierr/gcc_compileres/objdir'
make: *** [all] Error 2


Thank you very much again for your further comments,
Jose

[Bug bootstrap/49072] Error building the compiler

[Jose.Pascual-Gutierr@1-ACT.com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49072

--- Comment #3 from Jose <Jose.Pascual-Gutierr@1-ACT.com> 2011-05-20 21:22:03 UTC ---
Another error encountered on a second try:

xgcc: internal compiler error: Segmentation fault (program cc1)

Thank you

[Bug bootstrap/49072] Error building the compiler

[hp at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49072

Hans-Peter Nilsson <hp at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |hp at gcc dot gnu.org

--- Comment #4 from Hans-Peter Nilsson <hp at gcc dot gnu.org> 2011-06-09 03:44:56 UTC ---
(In reply to comment #3)
> Another error encountered on a second try:
> 
> xgcc: internal compiler error: Segmentation fault (program cc1)

You mean every time you run "make" you get an error for a different file?

If so, your system hardware is flaky.  Bad memory, bad cooling, or something
like that.  See <http://www.bitwizard.nl/sig11/> (which is getting old, but
still true for the important parts).

[Bug bootstrap/49072] Error building the compiler

[franck.z.bugzilla at orange dot fr]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49072

Franck Z <franck.z.bugzilla at orange dot fr> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |franck.z.bugzilla at orange
                   |                            |dot fr

--- Comment #5 from Franck Z <franck.z.bugzilla at orange dot fr> 2011-06-21 10:31:45 UTC ---
(In reply to comment #4)

During my first attempts to compile gcc, I also encountered problem with
in-tree compilation of GMP/MPFR/MPC, up to the point that I used the compiled
(but not up-to-date) version provided with the Cygwin distribution.

It took then place with the trunk.

Amont the weird things I observed stood out:
- a file "gmpxx.h" could be found and stopped compilation (with or without
--enable-xx flag in gcc's configure script)
- the architecture recognized by gmp's configure script was different when this
script was launched by me ("pentium4-"-something from what I recall...) or by
gcc's makefile. When with gcc's makefile, gmp opted for a cross-compilation for
i686-pc-cygwin to none-pc-cygwin - although my gcc build was native.

About the architecture point, I can't tell much. But about the gmpxx.h file, I
suspect that it is a mostly silent bug and that a file
/usr/local/include/gmpxx.h provided by Cygwin is used if present.
Apparently, gmp's scripts create a gmp.h file when configure is run, but the
gmpxx.h wrapper is copied only when a command "make install" is issued, so that
gcc multi-stage compilation process may not have it, and this as soon as mpfr
(or mpc, I can't remember) needs to link to gmp. The failure telling that
gmpxx.h was missing occured during a configure script in mpfr (or mpc, again).

When the unavailability of gmpxx.h was by-passed by a local installation of gmp
from Cygwin, I also got a division by zero error or such later in the build
process.

I'm trying to confirm it with 4.6.0 release. I've added a "choke me" line at
the beginning of /usr/local/include/gmpxx.h. Right now, the gcc "make" command
is running (it's entered "configuring stage 1 in ./gmp"). It may take time
before I can get an answer.

Hope this piece of information is useful. :-)
I can't propose a patch for this if it confirms, for want of skills in
makefiles as big as gcc's.

[Bug bootstrap/49072] Error building the compiler

[redi at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49072

--- Comment #6 from Jonathan Wakely <redi at gcc dot gnu.org> 2011-06-21 10:51:55 UTC ---
(In reply to comment #5)
> Apparently, gmp's scripts create a gmp.h file when configure is run, but the
> gmpxx.h wrapper is copied only when a command "make install" is issued, so that
> gcc multi-stage compilation process may not have it, and this as soon as mpfr
> (or mpc, I can't remember) needs to link to gmp. The failure telling that
> gmpxx.h was missing occured during a configure script in mpfr (or mpc, again).

gmpxx.h is part of the GMP source distribution, it's not generated

> Hope this piece of information is useful. :-)

Not really, if you don't say which versions of GCC and GMP/MFPR/MPC you're
using, and your exact configure command, then it's not reproducable or
verifiable.

It also seems to be completely unrelated to this bug report, which is about
segfaults (almost certainly due to faulty hardware) not incorrectly-detected
system properties during configuration.

[Bug bootstrap/49072] Error building the compiler

[franck.z.bugzilla at orange dot fr]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49072

--- Comment #7 from Franck Z <franck.z.bugzilla at orange dot fr> 2011-06-22 09:21:41 UTC ---
In reply to comment #6)
> > Hope this piece of information is useful. :-)
> Not really, if you don't say which versions of GCC and GMP/MFPR/MPC you're
> using, and your exact configure command, then it's not reproducable or
> verifiable.

My Cygwin is recent (one week old).
I have the same source tree for gcc, gcc/gmp, gcc/mpc and gcc/mpfr as the one
advised here (4.6.0, 4.3.2, 0.9 and 3.0.1).

Along with Cygwin comes:
- libgmp, libgmp3, libgmpxx4 and libgmp-devel 4.3.1-3
- libmpc1 and libmpc-devel 0.8-1
- libmpfr1 and libmpfr-devel 2.4.1-4
I edited /usr/local/include/gmpxx.h to added "choke me" in order to verify if
it's used by gcc's "make" despite gmp's source being in my source tree.

gcc output is in an "objdir" directory separate from my gcc source directory.
Configure script (after a "make distclean" command in "objdir"), with "objdir"
as my present working directory :
/"path to my source"/configure --enable-cx

then:
make

That's the configuration I'm testing now ("make" isn't finished). The first
build failures I refer too were done with pretty much the same configuration,
but for a trunk version of gcc and a one-two months old Cygwin. As I've lost
track of this precise configuration, I'm trying again with the configuration
described above.

> It also seems to be completely unrelated to this bug report, which is about
> segfaults (almost certainly due to faulty hardware) not incorrectly-detected
> system properties during configuration.

I won't fight on this point, ;-).
If my bug shows up again, I'll submit a specific bug report, so as not to
pollute this one.

Thank you.

[Bug bootstrap/49072] Error building the compiler

[franck.z.bugzilla at orange dot fr]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49072

--- Comment #8 from Franck Z <franck.z.bugzilla at orange dot fr> 2011-06-22 11:07:52 UTC ---
(In reply to comment #7)

Failure. But with a tag passed to libtool.
My Cygwin version of libtool is 2.4-1.

Took place in objdir/gmp/mpn.
With the command:
/bin/sh ../libtool --mode=compile gcc -DHAVE_CONFIG_H -I.
-I/cygdrive/c/gcc-4.6.0/gmp/mpn -I.. -D__GMP_WITHIN_GMP
-I/cygdrive/c/gcc-4.6.0/gmp -DOPERATION_`echo fib_table | sed 's/_$//'` 
-DNO_ASM -g -fkeep-inline-functions -c -o fib_table.lo fib_table.c

messages sent:
libtool: compile: unable to infer tagged configuration
libtool: compile: specify a tag with `--tag'

backtraces from "make"s rewinding from /home/Défaut/objdir/gmp/mpn (make[5]) to
/home/Défaut/objdir (make):
[fib_table.lo] Error 1 (in mpn)
[all-recursive] Error 1 (in gmp)
[all] Error 2 (in gmp)
[all-stage1-gmp] Error 2 (in objdir)
[stage1-bubble] Error 2 (in objdir)
[all] Error 2

Should I file in a new bug for this one with is different?

[Bug bootstrap/49072] Error building the compiler

[redi at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49072

Jonathan Wakely <redi at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Severity|critical                    |normal

--- Comment #9 from Jonathan Wakely <redi at gcc dot gnu.org> 2011-06-22 11:19:32 UTC ---
(In reply to comment #8)
> Should I file in a new bug for this one with is different?

Yes please, it's completely unrelated.  Please provide all the information
requested at http://gcc.gnu.org/bugs/

[Bug bootstrap/49072] Error building the compiler

[pinskia at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49072

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
         Resolution|---                         |INVALID

--- Comment #10 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
No feedback in 2 years so closing.