This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.
Index Nav: | [Date Index] [Subject Index] [Author Index] [Thread Index] | |
---|---|---|
Message Nav: | [Date Prev] [Date Next] | [Thread Prev] [Thread Next] |
Other format: | [Raw text] |
On 07/30/18 19:48, Joseph Myers wrote:
I whole-heatedly agree with that principle. [FWIW, I've taken a similar approach with keeping the cross and native compilation code paths as close as possible. This is why I'm building the libraries separately in the first place.]On the contrary, I think an important principle here is that non-multilib and multilib builds follow the same code paths as far as possible, with the multilib variables just set to trivial values (modulo osdirname) in the case of a non-multilib build - a non-multilib build should be building libraries exactly the same, with the same logic and the same variable settings, as the default multilib in a multilib build.
That said, it is my tentative understanding that the point of having config-ml is to cordon-off all the necessarily-multilib-specific logic so it doesn't pollute everything else. When that script isn't run, I think the Makefiles already contain default "trivial values" for capitalized MULTI* variables (which are the only ones actually used by the build itself), yielding precisely that deduplication of code paths we both want.
John
Index Nav: | [Date Index] [Subject Index] [Author Index] [Thread Index] | |
---|---|---|
Message Nav: | [Date Prev] [Date Next] | [Thread Prev] [Thread Next] |