From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Jeff Dickens" To: help-gcc@gnu.org Subject: Re: building gcc 2.95.2 - genattrtab.c - INT_MAX undeclared - help! Date: Mon, 06 Dec 1999 08:08:00 -0000 Message-id: <82gmqg$55p$0@dosa.alt.net> References: <828pnp$oe6$0@dosa.alt.net> <828t9h$6d1$0@dosa.alt.net> X-SW-Source: 1999-12/msg00079.html Hey, fwiw, I just defined HAVE_LIMITS_H as 1 in gcc/auto-host.h, and away it went. "make bootstrap" completed without error. Jeff Dickens wrote in message news: 828t9h$6d1$0@dosa.alt.net ... > genattrtab.c includes system.h. > > In system.h, I find this: > > #if HAVE_LIMITS_H > # include > #endif > > Who decides if "HAVE_LIMITS_H" should be defined ? And why would this > decision be no ? > > Also,genattrtab.c includes sys/resource.h. > > sys/resource.h includes sys/time.h which includes sys/types.h, which does > contain INT_MAX, but only inside > "#if defined(_KERNEL)". > > > > Jeff Dickens wrote in message > news: 828pnp$oe6$0@dosa.alt.net ... > > I'm building gcc 2.95.2 on NCR System V MP-RAS V3.02. > > > > I'm using the NCR c compiler to build gcc: > > NCR High Performance C Compiler R3.0c > > > > I soft-linked in all the binutils components, and building those seemed to > > go fine. > > > > Then when it went on to compiling gcc, it failed on genattrtab.c: > > > > > cc -c -DIN_GCC -g -DHAVE_CONFIG_H -I. -I../../gcc-2.95.2/gcc -I../. > > ./gc > > c-2.95.2/gcc/config -I../../gcc-2.95.2/gcc/../include > > ../../gcc-2.95.2/gcc/genat > > trtab.c > > . > > . > > E "../../gcc-2.95.2/gcc/genattrtab.c",L4733/C21(#237): INT_MAX: > Identifier > > is u > > ndeclared. > > . > > . > > 1 user error 192 warnings > > make[2]: *** [genattrtab.o] Error 1 > > make[2]: Leaving directory `/build1/ersbuild/src/gcc-2.95.2-build/gcc' > > make[1]: *** [bootstrap] Error 2 > > make[1]: Leaving directory `/build1/ersbuild/src/gcc-2.95.2-build/gcc' > > make: *** [bootstrap] Error 2 > > > > (I removed the 192 warnings to improve readability) > > > > > > INT_MAX is defined in /usr/include/limits.h. Why would't it find that ? > > > > thanks... > > > > > > > > > > > > > > > > From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Jeff Dickens" To: help-gcc@gnu.org Subject: Re: building gcc 2.95.2 - genattrtab.c - INT_MAX undeclared - help! Date: Fri, 31 Dec 1999 22:24:00 -0000 Message-ID: <82gmqg$55p$0@dosa.alt.net> References: <828pnp$oe6$0@dosa.alt.net> <828t9h$6d1$0@dosa.alt.net> X-SW-Source: 1999-12n/msg00079.html Message-ID: <19991231222400.LMh6bwbg4WXusOYmG4D8ThJ2ccK3qqFF9S90S7BVYOE@z> Hey, fwiw, I just defined HAVE_LIMITS_H as 1 in gcc/auto-host.h, and away it went. "make bootstrap" completed without error. Jeff Dickens wrote in message news: 828t9h$6d1$0@dosa.alt.net ... > genattrtab.c includes system.h. > > In system.h, I find this: > > #if HAVE_LIMITS_H > # include > #endif > > Who decides if "HAVE_LIMITS_H" should be defined ? And why would this > decision be no ? > > Also,genattrtab.c includes sys/resource.h. > > sys/resource.h includes sys/time.h which includes sys/types.h, which does > contain INT_MAX, but only inside > "#if defined(_KERNEL)". > > > > Jeff Dickens wrote in message > news: 828pnp$oe6$0@dosa.alt.net ... > > I'm building gcc 2.95.2 on NCR System V MP-RAS V3.02. > > > > I'm using the NCR c compiler to build gcc: > > NCR High Performance C Compiler R3.0c > > > > I soft-linked in all the binutils components, and building those seemed to > > go fine. > > > > Then when it went on to compiling gcc, it failed on genattrtab.c: > > > > > cc -c -DIN_GCC -g -DHAVE_CONFIG_H -I. -I../../gcc-2.95.2/gcc -I../. > > ./gc > > c-2.95.2/gcc/config -I../../gcc-2.95.2/gcc/../include > > ../../gcc-2.95.2/gcc/genat > > trtab.c > > . > > . > > E "../../gcc-2.95.2/gcc/genattrtab.c",L4733/C21(#237): INT_MAX: > Identifier > > is u > > ndeclared. > > . > > . > > 1 user error 192 warnings > > make[2]: *** [genattrtab.o] Error 1 > > make[2]: Leaving directory `/build1/ersbuild/src/gcc-2.95.2-build/gcc' > > make[1]: *** [bootstrap] Error 2 > > make[1]: Leaving directory `/build1/ersbuild/src/gcc-2.95.2-build/gcc' > > make: *** [bootstrap] Error 2 > > > > (I removed the 192 warnings to improve readability) > > > > > > INT_MAX is defined in /usr/include/limits.h. Why would't it find that ? > > > > thanks... > > > > > > > > > > > > > > > >