This is the mail archive of the
gcc-help@gcc.gnu.org
mailing list for the GCC project.
Re: disappointing build problems with gcc 3.2.2 and 3.3 on i686-pc-linux-gnu :(
- From: John Sincock <jss at au dot mensa dot org>
- To: Phil Edwards <phil at jaj dot com>
- Cc: gcc-help at gcc dot gnu dot org
- Date: Sun, 18 May 2003 23:33:12 +0930
- Subject: Re: disappointing build problems with gcc 3.2.2 and 3.3 on i686-pc-linux-gnu :(
- References: <20030517173817.3CAF1277FC@ns2.jaj.com> <20030518024738.GA13926@disaster.jaj.com>
- Reply-to: jss at au dot mensa dot org
Hi Phil,
thanks for your response
On Sunday 18 May 2003 12:17 pm, you wrote:
> > i?86-*-linux*
> > You will need binutils 2.9.1.0.15 or newer for exception handling to
> > work.
> > -------------------------------------------------------------------------
> > I hope that binutils bit doesnt apply to my system.
> You have 2.11. 2.11 is newer than 2.9. You should be fine there.
Ah thanks, sorry, i read the 2.9.1 as 2.91, so it thought my 2.11
might be out of date.
> > gcc-3.2.2]$ gcc -v: Reading specs from
> > /usr/local/lib/gcc-lib/i586-pc-linux-gnu/3.2.2/specs
> > Configured with: ./configure
> > Thread model: posix
> > gcc version 3.2.2
>
> Reread the install instructions, where we strongly recommend building in
> a different directory than the source dir.
I have been building in different dir than source dir. That snippet was from
someone elses build... I pasted in that section to show that someone
else with mandrake 8.2 has somehow managed to compile 3.2.2. Sorry
re confusion there.
> > That report contradicts the statement from
> > http://gcc.gnu.org/install/specific.html which says glibc 2.2.5 is
> > required for gcc 3.2.1 or greater.
> You need it for C++ at the very least.
This confuses me a bit.. :(
I did read: http://gcc.gnu.org/install/specific.html#*-*-linux-gnu
Where it says:
>Versions of libstdc++-v3 starting with 3.2.1 require bugfixes present in
>glibc 2.2.5 and later.
>More information is available in the libstdc++-v3 documentation
BUT!!! If you look at the list of successful builds at:
http://gcc.gnu.org/gcc-3.2/buildstat.html
There are a variety of successful builds with glibc 2.2.4
The test reports also show that they have built g++, and it
has passed almost all tests, with only a few unexpected failures.
eg these builds of 3.2.2 were ok
-------------------------------------------------------------------
It took a total of 11hours start to finish to build this on a
intel233mmx machine.
gcc-3.2.2]$ ./config.guess: i586-pc-linux-gnu
gcc-3.2.2]$ gcc -v: Reading specs from
/usr/local/lib/gcc-lib/i586-pc-linux-gnu/3.2.2/specs
Configured with: ./configure
Thread model: posix
gcc version 3.2.2
/etc/issue: Mandrake Linux release 8.2 (Bluebird) for i586
Kernel 2.4.18-6mdk on an i586
uname -a: Linux localhost.localdomain 2.4.18-6mdk #1 Fri Mar 15
02:59:08 CET 2002 i586 unknown
gcc-3.2.2]$ rpm -q glibc: glibc-2.2.4-26.2mdk
-----------------------------------------------------------------
...Reading specs from
/usr/local/lib/gcc-lib/i686-pc-linux-gnu/3.2.2/specs
Configured with: ./configure
Thread model: posix
gcc version 3.2.2
3. RedHat 8.0
4. uname -a
Linux 2.4.18-14 #1 Wed Sep 4 13:35:50 EDT 2002 i686
i686 i386 GNU/Linux
5. rpm -q glibc
glibc-2.2.93-5
--------------------------------------------------------------------
And there were quite a few others. I havent found any successful
builds with the exact same version of glibc as me though - which
is glibc-2.2.4-25mdk
I think I am about to give up trying to compile a new gcc for now.
Though i will have another try sometime soon, when i have upgraded
my glibc to 2.2.5 or so. I dont think i have enough enthusiasm to do
that this weekend.
I think ive found a few 3.2 rpms, which should install on my system as
well, so I may just install them next weekend, and then have another go
at compiing 3.3 with that, just to see how i go.
Arrrghhh defeated by a simple compile of gcc!! Nothing else, in gigabytes
of software compiled, has managed to humiliate me to this extent!!
Enlightenment 0.17 has probably come close... but at least i can blame
a failure to compile that mess on their demented software.... with gcc,
i think i will have to assume that I have messed up my system in some
way, or my different versions of gcc are getting in each others way,
or something.
Oh well,
and thanks again,
John