While trying to build a native compiler, I came across a situation where one of the rules in CLIB2 was explicitly using ppc-amigaos-strip. Instead, by setting that as a variable - as is already done for CC, AR, LD and RANLIB - it allows the overriding of its value.
The request is to set a variable in the makefile:
STRIP=ppc-amigaos-strip
and use $(STRIP) instead.
===
Reason: Firstly, it is more consistent this way. Other BINs like CC, AR use variables. Secondly, when building ADTOOLS the locations of those binaries may not yet be on the path - in the ADTOOLS makefile there are things like:
CC=/path/to/Xcompiler/bin/ppc-amigaos-gcc
and then CC is fed into CLIB2. This does not happen for STRIP, so it causes a failure.
===
Request to merge into BETA10, which should finally propagate through to MASTER.
https://github.com/afxgroup/clib2/blob/28aa4eec565160d7fbe90cd534fc57542d921d65/GNUmakefile.os4#L205
While trying to build a native compiler, I came across a situation where one of the rules in CLIB2 was explicitly using ppc-amigaos-strip. Instead, by setting that as a variable - as is already done for CC, AR, LD and RANLIB - it allows the overriding of its value.
The request is to set a variable in the makefile:
STRIP=ppc-amigaos-strip
and use $(STRIP) instead.===
Reason: Firstly, it is more consistent this way. Other BINs like CC, AR use variables. Secondly, when building ADTOOLS the locations of those binaries may not yet be on the path - in the ADTOOLS makefile there are things like:
CC=/path/to/Xcompiler/bin/ppc-amigaos-gcc and then CC is fed into CLIB2. This does not happen for STRIP, so it causes a failure.
===
Request to merge into BETA10, which should finally propagate through to MASTER.