On Tuesday 30 September 2008 05:57:06 Eric Blake wrote: > According to Keith Marshall on 9/29/2008 5:25 PM: > > On this occasion, the `in place' failure is particularly nasty; > > the `all' primary goal proceeds normally, until it attempts to > > satisfy the `all-am' secondary goal, at which point it starts > > *installing* the partially built package -- very nasty, because > > it wasn't asked to install anything yet, but it is identifying > > the `install' goal as the only candidate to satisfy the > > `$(srcdir)/INSTALL' prerequisite -- and, because `install' itself > > depends on the only partially completed `all' goal, this recurses > > indefinitely. > > Sounds like the case-insensitive make behavior is still biting us, > then (which we previously determined only afflicts in-place > builds); I guess Stepan's patch didn't work as expected [1],[2]. > Could you please track down the config.log snippets related to > checking for case-insensitive make (ie. what is MAKE_CASE_SENSITIVE > set to in the Makefile.in?)
configure:2902: checking whether make is case sensitive configure:2918: result: no ... ac_cv_prog_make_make_case=no ... MAKE_CASE_SENSITIVE_FALSE='' MAKE_CASE_SENSITIVE_TRUE='#' > Also, does this patch help? > ... Yes; it resolves the issue I have observed, allowing the in-place build to complete successfully, (both `make all' and `make install'). Thanks. > Perhaps it is worth filing this as a bug of GNU make's > case-insensitive mode, for confusing phony targets of a different > case than real targets, but only when they occur in the same > directory. Maybe. I'll post one, when I can find the time. Another option may be to avoid any explicit mention of the ambiguous `INSTALL' goal, (as a file, it could equally well be called `install', and it would still be the same file); rather have it created as a side effect of a new unambiguous goal, (e.g. `INSTALL.stamp'). Regards, Keith.
