Re: Re: compiling gcc 5.x from the scratch: issues

Re: Re: compiling gcc 5.x from the scratch: issues

[Michael Mishourovsky]

Dear Sirs, following your suggestions I checked again  https://gcc.gnu.org/wiki/InstallingGCC.

Key point was here: https://gcc.gnu.org/install/download.html (last passage)
"Likewise the GMP, MPFR and MPC libraries can be automatically built together with GCC. Unpack the GMP, MPFR and/or MPC source distributions in the directory containing the GCC sources and rename their directories to gmp, mpfr and mpc, respectively (or use symbolic links with the same name)."

Actually my assumption was that if I run download_prerequisites from /contrib, it will download everything and put it at right places; but actually it does not follow statement above (it does not create gmp, mpfr and mpc folder in main gcc package directory)
So I did it manually and copied files there.

Then, from a folder I am going to have gcc build, I run ../gcc_5_2_0_release/configure --prefix=/gcc_520obj --enable-languages=c,c++ --disablemultilib --disable-bootstrap --disable-shared --enable-shared=static 
Then invoked make

When it was completed, I made install using the following: make DESTDIR=$HOME/gcc_tmp/ install
Another problem was that I don't have sudo rights, so I cannot install gcc into standard folders; but installation into some other folder works for me.
And lastly, appended PATH & LD_LIBRARY_PATH.

In conclusion, key problem was in putting GMP, OMC and MPFR source files in right directories before build is stated. Other stuff is more or less trivial.
If you feel OK with that, this information might be added to guidelines (at least FAQ)

Thank you for your motivational advice, have a good day,
Best regards, Mikhail Mishurovskiy


------- Original Message -------
Sender : Jonathan Wakely<jwakely.gcc@gmail.com>
Date : Sep 09, 2015 18:21 (GMT+09:00)
Title : Re: compiling gcc 5.x from the scratch: issues

On 9 September 2015 at 09:50, Markus Trippelsdorf wrote:
> On 2015.09.09 at 08:36 +0000, Michael Mishourovsky wrote:
>>  At my work I would like to have recent gcc installed but i have no
>>  sudo rights to update the current gcc (its 4.4.7, and OS is redhat
>>  linux).
>>
>>  So I checked out latest version of gcc via svn, and following
>>  guidelines given at gcc webpage, tried to download prerequisites,
>>  configure gcc to be compiled into a separate folder and to start
>>  make. But got a lot of errors including errors in test files (that
>>  are easy to fix), errors during linkage (seems to be almost final
>>  step); eventually failed to  build it up. As it is supposed that this
>>  SW is extensively tested, could you tell me: whether someone tried to
>>  build latest gcc (say, gcc 5.1 or 5.2) from scratch and was succeed
>>  in this (if he/she follows the guidelines given at the gcc web-site?
>>  If possible, could you check -whether it is still possible to compile
>>  gcc or not and update guidelines, if something has been changed when
>>  time passes by?
>
> Please make sure that you use a build directory outside the gcc source
> tree.
> You may simply follow: https://gcc.gnu.org/install/
>
> If the issue still persists, please post the exact errors that you
> encounter.

Also https://gcc.gnu.org/wiki/InstallingGCC

Please don't cross-post to both the gcc and gcc-help lists, we have
separate lists for a reason not just so you have additional places to
email to. You're asking for help building GCC so your email is
off-topic on the gcc list.

Re: Re: compiling gcc 5.x from the scratch: issues

[Markus Trippelsdorf]

On 2015.09.10 at 05:36 +0000, Michael Mishourovsky wrote:
> Dear Sirs, following your suggestions I checked again  https://gcc.gnu.org/wiki/InstallingGCC.
> 
> Key point was here: https://gcc.gnu.org/install/download.html (last
> passage) "Likewise the GMP, MPFR and MPC libraries can be
> automatically built together with GCC. Unpack the GMP, MPFR and/or MPC
> source distributions in the directory containing the GCC sources and
> rename their directories to gmp, mpfr and mpc, respectively (or use
> symbolic links with the same name)."
> 
> Actually my assumption was that if I run download_prerequisites from
> /contrib, it will download everything and put it at right places; but
> actually it does not follow statement above (it does not create gmp,
> mpfr and mpc folder in main gcc package directory) So I did it
> manually and copied files there.

Well, https://gcc.gnu.org/wiki/InstallingGCC says:

»Alternatively, after extracting the GCC source archive, simply run the
./contrib/download_prerequisites script in the GCC source directory.«

The »in the GCC source directory« part is important. If you run it from
anywhere else it will not work.

-- 
Markus

Re: Re: compiling gcc 5.x from the scratch: issues

[Markus Trippelsdorf]

On 2015.09.10 at 08:40 +0200, Markus Trippelsdorf wrote:
> On 2015.09.10 at 05:36 +0000, Michael Mishourovsky wrote:
> > Dear Sirs, following your suggestions I checked again  https://gcc.gnu.org/wiki/InstallingGCC.
> > 
> > Key point was here: https://gcc.gnu.org/install/download.html (last
> > passage) "Likewise the GMP, MPFR and MPC libraries can be
> > automatically built together with GCC. Unpack the GMP, MPFR and/or MPC
> > source distributions in the directory containing the GCC sources and
> > rename their directories to gmp, mpfr and mpc, respectively (or use
> > symbolic links with the same name)."
> > 
> > Actually my assumption was that if I run download_prerequisites from
> > /contrib, it will download everything and put it at right places; but
> > actually it does not follow statement above (it does not create gmp,
> > mpfr and mpc folder in main gcc package directory) So I did it
> > manually and copied files there.
> 
> Well, https://gcc.gnu.org/wiki/InstallingGCC says:
> 
> »Alternatively, after extracting the GCC source archive, simply run the
> ./contrib/download_prerequisites script in the GCC source directory.«
> 
> The »in the GCC source directory« part is important. If you run it from
> anywhere else it will not work.

So perhaps a patch like this one would help:

diff --git a/contrib/download_prerequisites b/contrib/download_prerequisites
index de0e7c41847d..44dc1238ff5d 100755
--- a/contrib/download_prerequisites
+++ b/contrib/download_prerequisites
@@ -24,6 +24,11 @@
 # be downloaded.
 GRAPHITE_LOOP_OPT=yes
 
+if  [ ! -s gcc/BASE-VER ] ; then
+	echo "You must run this script in the toplevel GCC source directory."
+	exit 1
+fi
+
 # Necessary to build GCC.
 MPFR=mpfr-2.4.2
 GMP=gmp-4.3.2

-- 
Markus

Re: compiling gcc 5.x from the scratch: issues

[Martin Sebor]

On 09/10/2015 12:49 AM, Markus Trippelsdorf wrote:
> On 2015.09.10 at 08:40 +0200, Markus Trippelsdorf wrote:
>> On 2015.09.10 at 05:36 +0000, Michael Mishourovsky wrote:
>>> Dear Sirs, following your suggestions I checked again  https://gcc.gnu.org/wiki/InstallingGCC.
>>>
>>> Key point was here: https://gcc.gnu.org/install/download.html (last
>>> passage) "Likewise the GMP, MPFR and MPC libraries can be
>>> automatically built together with GCC. Unpack the GMP, MPFR and/or MPC
>>> source distributions in the directory containing the GCC sources and
>>> rename their directories to gmp, mpfr and mpc, respectively (or use
>>> symbolic links with the same name)."
>>>
>>> Actually my assumption was that if I run download_prerequisites from
>>> /contrib, it will download everything and put it at right places; but
>>> actually it does not follow statement above (it does not create gmp,
>>> mpfr and mpc folder in main gcc package directory) So I did it
>>> manually and copied files there.
>>
>> Well, https://gcc.gnu.org/wiki/InstallingGCC says:
>>
>> »Alternatively, after extracting the GCC source archive, simply run the
>> ./contrib/download_prerequisites script in the GCC source directory.«
>>
>> The »in the GCC source directory« part is important. If you run it from
>> anywhere else it will not work.
>
> So perhaps a patch like this one would help:

FWIW, I think the more common source of problems for those new
to building GCC is not having all the right prerequisites. The
GCC configure script tries to detect whether the right versions
are available but does only a superficial job (checks their
version macros) which can lead to cryptic build failures down
the line when not all the right architectures are installed
(e.g., the i686 in addition to x86_64). It also proceeds when
MPC is not installed at all which then causes the build to fail.

Martin

Re: compiling gcc 5.x from the scratch: issues

[Jonathan Wakely]

On 10 September 2015 at 17:23, Martin Sebor wrote:
> FWIW, I think the more common source of problems for those new
> to building GCC is not having all the right prerequisites. The
> GCC configure script tries to detect whether the right versions
> are available but does only a superficial job (checks their
> version macros) which can lead to cryptic build failures down
> the line when not all the right architectures are installed
> (e.g., the i686 in addition to x86_64). It also proceeds when
> MPC is not installed at all which then causes the build to fail.

Not having the right prerequisites certainly used to be a common
problem (not quite as common as having them but installing them
outside ld.so's search dirs and then not setting LD_LIBRARY_PATH).
However, since I've been on a crusade to stop people doing that by
pointing them at https://gcc.gnu.org/wiki/InstallingGCC and telling
them to use contrib/download_prerequisites it is *much* less common
than it used to be.

Markus, I think your patch wouldn't hurt, and might help people who
fail to read the instructions carefully, so I suggest sending it to
gcc-patches for review.

Re: compiling gcc 5.x from the scratch: issues

[Markus Trippelsdorf]

On 2015.09.10 at 19:53 +0100, Jonathan Wakely wrote:
> On 10 September 2015 at 17:23, Martin Sebor wrote:
> > FWIW, I think the more common source of problems for those new
> > to building GCC is not having all the right prerequisites. The
> > GCC configure script tries to detect whether the right versions
> > are available but does only a superficial job (checks their
> > version macros) which can lead to cryptic build failures down
> > the line when not all the right architectures are installed
> > (e.g., the i686 in addition to x86_64). It also proceeds when
> > MPC is not installed at all which then causes the build to fail.
> 
> Not having the right prerequisites certainly used to be a common
> problem (not quite as common as having them but installing them
> outside ld.so's search dirs and then not setting LD_LIBRARY_PATH).
> However, since I've been on a crusade to stop people doing that by
> pointing them at https://gcc.gnu.org/wiki/InstallingGCC and telling
> them to use contrib/download_prerequisites it is *much* less common
> than it used to be.
> 
> Markus, I think your patch wouldn't hurt, and might help people who
> fail to read the instructions carefully, so I suggest sending it to
> gcc-patches for review.

I did that already and Jeff has approved the patch. Will commit it
tomorrow.

-- 
Markus

Re: compiling gcc 5.x from the scratch: issues

[Jonathan Wakely]

On 10 September 2015 at 20:15, Markus Trippelsdorf wrote:
> On 2015.09.10 at 19:53 +0100, Jonathan Wakely wrote:
>> Markus, I think your patch wouldn't hurt, and might help people who
>> fail to read the instructions carefully, so I suggest sending it to
>> gcc-patches for review.
>
> I did that already and Jeff has approved the patch. Will commit it
> tomorrow.

Oh nice, thanks.

Re: Re: compiling gcc 5.x from the scratch: issues

[Michael Mishourovsky]

Dear All
thank you for so responsive support.
you are right - theoretically everything is documented; 
but practically a person who does not compile and build gcc every day might be confused with all these details.

Anyway your advice and update really helps, thanks.
Have a good weekend,

Best regards, Mikhail 

------- Original Message -------
Sender : Jonathan Wakely<jwakely.gcc@gmail.com>
Date : Sep 11, 2015 04:17 (GMT+09:00)
Title : Re: compiling gcc 5.x from the scratch: issues

On 10 September 2015 at 20:15, Markus Trippelsdorf wrote:
> On 2015.09.10 at 19:53 +0100, Jonathan Wakely wrote:
>> Markus, I think your patch wouldn't hurt, and might help people who
>> fail to read the instructions carefully, so I suggest sending it to
>> gcc-patches for review.
>
> I did that already and Jeff has approved the patch. Will commit it
> tomorrow.

Oh nice, thanks.