On Tue, Nov 30, 2021 at 09:18:10PM +0100, Andreas F. Borchert wrote: > Hi, > > config.guess guesses Solaris 11 to run on a 32-bit platform > despite Solaris 11 no longer supporting any 32-bit platform. > > See the following code at lines 434 to 445: > > | SUN_ARCH=i386 > | # If there is a compiler, see if it is configured for 64-bit objects. > | # Note that the Sun cc does not turn __LP64__ into 1 like gcc does. > | # This test works for both compilers. > | if test "$CC_FOR_BUILD" != no_compiler_found; then > | if (echo '#ifdef __amd64'; echo IS_64BIT_ARCH; echo '#endif') | \ > | (CCOPTS="" $CC_FOR_BUILD -E - 2>/dev/null) | \ > | grep IS_64BIT_ARCH >/dev/null > | then > | SUN_ARCH=x86_64 > | fi > | fi > > If "cc" is installed, i.e. the Oracle Studio compiler, this one > is chosen for $CC_FOR_BUILD. This compiler, the gcc provided > by Oracle and also gcc bootstrapped from sources on that platform > with a default configuration will by default generate 32-bit > binaries -- even on a 64-bit platform. And __amd64 will not > be defined for compilations targeting a 32-bit platform. This > is different from the corresponding behaviour on GNU/Linux systems > where the local platform is targeted by default. > > Thus, as long as you do not add "-m64" or if you have a custom-built > gcc which defaults to 64 bit, you will get 32-bit binaries on Solaris > despite living on a 64-bit platform. > > You can either omit this test for Solaris 11 (as this operating system > will not run on any 32-bit platform) or adapt the test by adding the > "-m64" flag. This this will work properly for Solaris 10 as well (the > last Solaris release that supported x86 32-bit platforms). > > Here is a patch that simply adds the "-m64" flag: > > ------------------------------------------------------------------------------- > *** config.guess Tue Nov 30 20:42:03 2021 > --- config.guess-fixed Tue Nov 30 20:44:54 2021 > *************** > *** 438,442 **** > if test "$CC_FOR_BUILD" != no_compiler_found; then > if (echo '#ifdef __amd64'; echo IS_64BIT_ARCH; echo '#endif') | \ > ! (CCOPTS="" $CC_FOR_BUILD -E - 2>/dev/null) | \ > grep IS_64BIT_ARCH >/dev/null > then > --- 438,442 ---- > if test "$CC_FOR_BUILD" != no_compiler_found; then > if (echo '#ifdef __amd64'; echo IS_64BIT_ARCH; echo '#endif') | \ > ! (CCOPTS="" $CC_FOR_BUILD -m64 -E - 2>/dev/null) | \ > grep IS_64BIT_ARCH >/dev/null > then > -------------------------------------------------------------------------------
I've applied this, although it looks really strange and I cannot tell whether it breaks something in Solaris land or not, let's see if somebody starts complaining. > This is related to following discussion: > https://github.com/latex3/luaotfload/issues/203 > > BTW, I wonder why config.guess does not distinguish between > SPARCv8 (32-bit platform) and SPARCv9 and other SPARC-based > 64-bit platforms. config.guess always generates > "sparc-sun-solaris2.x" for these cases. > > And I also wonder why config.sub supports "solaris3" alias > "sunos6". This is pure fantasy. There will be no "solaris3" > as we are already arrived at Solaris 11. (There was a jump > from Solaris 2.6 to Solaris 7 where "uname -r" just moved from > 5.6 to 5.7. We are now at 5.11 and nobody expects a move towards > 6.x and, even if this would happen, this would not be named > Solaris 3. It seems to be made long before my tenure, cannot shed any light on these issues, sorry. -- ldv
