From: Mo D. <md...@cy...> - 2001-01-06 23:27:50
|
On Sun, 7 Jan 2001, =?iso-8859-1?q?Danny=20Smith?= wrote: > Hi > > > > --- gcc/configure.in.orig Sat Jan 6 10:47:04 2001 > > +++ gcc/configure.in Sat Jan 6 10:47:33 2001 > > @@ -1553,9 +1553,9 @@ > > fi > > exeext=.exe > > case $machine in > > - *mingw32msv*) > > + *mingw32msv* | *mingw32*) > > ;; > > - *minwg32crt* | *mingw32*) > > + *mingw32crt*) > > tm_file="${tm_file} i386/crtdll.h" > > ;; > > esac > > --- gcc/configure.orig Sat Jan 6 10:47:07 2001 > > +++ gcc/configure Sat Jan 6 10:47:52 2001 > > @@ -3897,9 +3897,9 @@ > > fi > > exeext=.exe > > case $machine in > > - *mingw32msv*) > > + *mingw32msv* | *mingw32*) > > ;; > > - *minwg32crt* | *mingw32*) > > + *mingw32crt*) > > tm_file="${tm_file} i386/crtdll.h" > > ;; > > esac > > Hello. I have been wondering about that id for mingw. At first, the triple was: i386-mingw32-gcc When the crt vs msvcrt issue came up, this triple was suggested so that one would tell them apart. i386-mingw32msvc-gcc I am wondering if the "msvc" bit is really the most clear name we could be using here. It seems like it might be more clear if "msvcrt" was used instead. I think "msvc" could be confusing since someone might think that stood for the MS VC++ compiler. Why don't we use IDs like this? i386-pc-mingw32-msvcrt-gcc i386-pc-mingw32-crt-gcc That might better fit the GNU model of: $machine-$system-$release-$version What do you think? Mo DeJong Red Hat Inc |