------- Comment #14 from iains at gcc dot gnu dot org 2010-04-12 13:36 ------- Created an attachment (id=20365) --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=20365&action=view) sort out some nits with config/{*,}/darwin*.h and hack in a solution for dsymtuil
The dsymutils issue is not a GCC bug - but it is a problem that gcc darwin developers need to deal with in the short to medium term (possibly long-term for darwin<10) ... OK, this is a patch which is work in progress to clean up darwin*.h: (a) clear up some nits in the spacing of asm directives (b) make sure that we only need to maintain one LINK_COMMAND_SPEC - and ... now ... (c) provide a hack to get round the dsymtuils problem this is NOT a proposed patch to solve the problem - but it is a patch to allow things to proceed for developers without being inundated with error messages. I've made the dsymutils produce flat files for the output (.dwarf) which seem to work fine with the XCode gdb and avoid the "dSYM is a directory" config warnings. debug is working for fortran with this but ONLY if you put -save-temps on the C/L .. dunno why that is yet.. (-pipe doesn't solve it). -====- Perhaps someone could look at how this can be packaged so that we can use it for things like "regress"? -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=43254