Thread overview | |||||
---|---|---|---|---|---|
|
May 29, 2013 [phobos] MODEL inference | ||||
---|---|---|---|---|
| ||||
I'm seeing this is phobos/posix.mak: MODEL:=default ifneq (default,$(MODEL)) MODEL_FLAG:=-m$(MODEL) endif I find it unpleasant that we have directories generated/xxx/yyy/{32,64,default} of which default is identical to another. Is there a simple and robust technique to detect the default model on a given platform? Thanks, Andrei _______________________________________________ phobos mailing list phobos@puremagic.com http://lists.puremagic.com/mailman/listinfo/phobos |
May 29, 2013 Re: [phobos] MODEL inference | ||||
---|---|---|---|---|
| ||||
Posted in reply to Andrei Alexandrescu | On Wednesday, May 29, 2013 23:30:12 Andrei Alexandrescu wrote: > I'm seeing this is phobos/posix.mak: > > MODEL:=default > ifneq (default,$(MODEL)) > MODEL_FLAG:=-m$(MODEL) > endif > > I find it unpleasant that we have directories generated/xxx/yyy/{32,64,default} of which default is identical to another. > > Is there a simple and robust technique to detect the default model on a given platform? My first thought would be to compile a program and see what size size_t was, but I don't know feasible that is from inside of make. There's probably some way to get make to tell you, but I'm not very well versed in make. Timothee Cour has expressed interest in replacing all of the makefiles with D programs that do the build wth the idea that it would be easier to manage the various architectures and OSes that way. And if we did that, then all you'd have to do would be to check the size of size_t within the build script with size_t.sizeof. - Jonathan M Davis _______________________________________________ phobos mailing list phobos@puremagic.com http://lists.puremagic.com/mailman/listinfo/phobos |
May 30, 2013 Re: [phobos] MODEL inference | ||||
---|---|---|---|---|
| ||||
Posted in reply to Jonathan M Davis Attachments:
| coincidence, just sent out an email regarding this just today: http://forum.dlang.org/post/mailman.390.1369900211.13711.digitalmars-d@puremagic.com "D based build system to replace current makefile" Yes, no point in making the makefile more complicated, doing it in D is much easier/cleaner. On Wed, May 29, 2013 at 9:04 PM, Jonathan M Davis <jmdavisProg@gmx.com>wrote: > On Wednesday, May 29, 2013 23:30:12 Andrei Alexandrescu wrote: > > I'm seeing this is phobos/posix.mak: > > > > MODEL:=default > > ifneq (default,$(MODEL)) > > MODEL_FLAG:=-m$(MODEL) > > endif > > > > I find it unpleasant that we have directories generated/xxx/yyy/{32,64,default} of which default is identical to > another. > > > > Is there a simple and robust technique to detect the default model on a given platform? > > My first thought would be to compile a program and see what size size_t > was, > but I don't know feasible that is from inside of make. There's probably > some > way to get make to tell you, but I'm not very well versed in make. > > Timothee Cour has expressed interest in replacing all of the makefiles > with D > programs that do the build wth the idea that it would be easier to manage > the > various architectures and OSes that way. And if we did that, then all you'd > have to do would be to check the size of size_t within the build script > with > size_t.sizeof. > > - Jonathan M Davis > _______________________________________________ > phobos mailing list > phobos@puremagic.com > http://lists.puremagic.com/mailman/listinfo/phobos > |
Copyright © 1999-2021 by the D Language Foundation