Re: [Gm2] gm2 4.7.4 build failure

[Gaius Mulley]

Christoph Schlegel <address@hidden> writes:

> On Wed, 16 Mar 2016 11:26:16 +0000
> Gaius Mulley <address@hidden> wrote:
>
>> Steve Giess <address@hidden> writes:
>> 
>> > Dear Gaius,
>> >
>> > I downloaded 4.1.2, 4.7.4 & 5.2.0 - 15th of March versions.
>> >
>> > 4.1.2 built successfully on my Fedora 9, i686 system.
>> >
>> > However 4.7.4 failed on my usual Fedora 17, i686 system.
>> >
>> > Also 5.2.0 failed on my Fedora 23, x86_64 system.
>> >
>> > I've attached the fail results.
>> >
>> > Regards,
>> >
>> > Steve
>> 
>> Hi Steve,
>> 
>> thanks for the error report - yes I'm going to rewrite the tarball
>> generation script today - hopefully this will be fixed soon.  I've
>> also created 4.7.4 branch, 5.2.0 branch and 5.3.0 branch,
>
> Hello Gaius,
>
> I tried the latest tarball 5.2.0 and got:
>
> g++ -c -fpermissive   -g  -DIN_GCC    -fno-exceptions -fno-rtti 
> -fasynchronous-unwind-tables -W -Wall -Wno-narrowing -Wwrite-strings 
> -Wcast-qual  -fno-common  -DHAVE_CONFIG_H -I. -Igm2/gm2-libs-boot 
> -I../../gcc-5.2.0/gcc -I../../gcc-5.2.0/gcc/gm2/gm2-libs-boot 
> -I../../gcc-5.2.0/gcc/../include -I../../gcc-5.2.0/gcc/../libcpp/include  
> -I../../gcc-5.2.0/gcc/../libdecnumber 
> -I../../gcc-5.2.0/gcc/../libdecnumber/bid -I../libdecnumber 
> -I../../gcc-5.2.0/gcc/../libbacktrace   -I../../gcc-5.2.0/gcc/gm2 -Igm2 -I. 
> -Igm2/gm2-libs-boot -I. -Igm2/gm2-libs-boot -I../../gcc-5.2.0/gcc 
> -I../../gcc-5.2.0/gcc/gm2/gm2-libs-boot -I../../gcc-5.2.0/gcc/../include 
> -I../../gcc-5.2.0/gcc/../libcpp/include  
> -I../../gcc-5.2.0/gcc/../libdecnumber 
> -I../../gcc-5.2.0/gcc/../libdecnumber/bid -I../libdecnumber 
> -I../../gcc-5.2.0/gcc/../libbacktrace  gm2/gm2-libs-boot/Indexing.c -o 
> gm2/gm2-libs-boot/Indexing.o
> In file included from gm2/gm2-libs-boot/Indexing.c:24:0:
> gm2/gm2-libs-boot/Indexing.c: In function ‘void* 
> Indexing_InitIndex(unsigned int)’:
> gm2/gm2-libs-boot/Indexing.c:60:61: error: types may not be defined in 
> ‘sizeof’ expressions
>    i = reinterpret_cast<Indexing_Index>(Malloc(sizeof(struct {
>                                                              ^
> ../../gcc-5.2.0/gcc/gm2/p2c/p2c.h:105:54: note: in definition of macro 
> ‘Malloc’
>  #    define Malloc(n)   (ggc_internal_cleared_alloc (n))
>                                                       ^
> gm2/gm2-libs-boot/Indexing.c: In function ‘void Indexing_PutIndice(void*, 
> unsigned int, void*)’:
> gm2/gm2-libs-boot/Indexing.c:184:4: warning: pointer of type ‘void *’ 
> used in arithmetic [-Wpointer-arith]
>   b += oldSize;
>     ^
> gm2/gm2-libs-boot/Indexing.c:191:5: warning: pointer of type ‘void *’ 
> used in arithmetic [-Wpointer-arith]
>    b += (n - i->Low) * (sizeof(void *));
>      ^
> gm2/gm2-libs-boot/Indexing.c: In function ‘BOOLEAN Indexing_InBounds(void*, 
> unsigned int)’:
> gm2/gm2-libs-boot/Indexing.c:122:1: warning: control reaches end of non-void 
> function [-Wreturn-type]
>  }
>  ^
> gm2/gm2-libs-boot/Indexing.c: In function ‘unsigned int 
> Indexing_HighIndice(void*)’:
> gm2/gm2-libs-boot/Indexing.c:137:1: warning: control reaches end of non-void 
> function [-Wreturn-type]
>  }
>  ^
> gm2/gm2-libs-boot/Indexing.c: In function ‘unsigned int 
> Indexing_LowIndice(void*)’:
> gm2/gm2-libs-boot/Indexing.c:152:1: warning: control reaches end of non-void 
> function [-Wreturn-type]
>  }
>  ^
> ../../gcc-5.2.0/gcc/gm2/Make-lang.in:1032: recipe for target 
> 'gm2/gm2-libs-boot/Indexing.o' failed
> make[3]: *** [gm2/gm2-libs-boot/Indexing.o] Error 1
> make[3]: Leaving directory '/home/Christoph/gm2build/gcc'
> cp: cannot stat 'stage1/gm2/cc1gm2.exe': No such file or directory
> ../../gcc-5.2.0/gcc/gm2/Make-lang.in:575: recipe for target 'cc1gm2.exe' 
> failed
> make[2]: *** [cc1gm2.exe] Error 1
> make[2]: Leaving directory '/home/Christoph/gm2build/gcc'
> Makefile:4148: recipe for target 'all-gcc' failed
> make[1]: *** [all-gcc] Error 2
> make[1]: Leaving directory '/home/Christoph/gm2build'
> Makefile:885: recipe for target 'all' failed
> make: *** [all] Error 2
>
> Regards,
> Christoph

Hi Christoph,

thanks for testing this - yes the 5.2.0 is deep in development - I'm
still working on mc which once complete should fix the above bugs.
The good news is that some/many of the 5.2.0 changes have been made so I
think it should be speedy progress once this hurdle has been completed,

regards,
Gaius