makeinfo if no native makeinfo executable exists. Honor TEXINFO_REQD
when determining whether the native makeinfo can be used.
* Remove USE_MAKEINFO and replace it with USE_TOOLS+=makeinfo.
* Get rid of all the "split" argument deduction for makeinfo since
the PLIST module already handles varying numbers of split info files
correctly.
NOTE: Platforms that have "makeinfo" in the base system should check
that the makeinfo entries of pkgsrc/mk/tools.${OPSYS}.mk are
correct.
them completely, as they had been unnecessary. As ${MKDIR} already
includes the -p option, the ${TEST} is not necessary and has been
dropped, too. Fixes PR 32002.
rather than find. This behaves better if ${ALT_GCC} does not specify
a gcc executable.
2) Change to uniform handling of ABI baseline-pairs for NetBSD. There
must be agreement between the Makefile and libstdc++-v3/configure.host;
this change ensures consistency by introducing ABI_BASELINE_PAIR
variable used in both.
3) Bump revision number.
alpha--netbsd1.6.2:
1) Rename the baseline_symbols.txt files that define abi and implement
logic in Makefile that uses these according to architecture and
NetBSD version (1.x versus 2.x)
2) Add several patches that allow code to compile on alpha--netbsd1.6.2
3) Revise Makefile so that libiconv is found under NetBSD 1.6.2
a) The BUILD_ADA must be YES (strictly capitalized) for the
pre-configure target to work, which is inconsistent.
b) For me (using ALT_GCC), the directory .work/.gcc was not
present at pre-configure time, which made some "cd" fail.
2) Makefile cleanup as recommended by Lubomir.
3) pkglint cleanup
4) Increment PKGREVISION
2) Added include statment for zlib/buildlink3.mk to Makefile to go with
CONFIGURE_ARGS+=--with-system-zlib already there. Thank you
Georg Schwarz <geos@epost.de> for pointing out this ommision.
3) Incorporated logic from gcc3 that determines whether to add
binutils dependency or not.
4) Eliminated hardcoded target of netbsdelf2.0 as a step toward supporting
other architectures.
