Age | Commit message (Collapse) | Author | |
---|---|---|---|
2012-08-27 | a few libtool users are bogus, they confuse --export-symbols with the | Marc Espie | |
real option -export-symbols. Instead of blindly passing that one to the linker, error out beforehand, so that people know the problem IS libtool usage and not our program. (as exemplified by Todd Fries on ldns). | |||
2012-08-27 | vax does not get -rpath-link, but it also does not really need it. | Marc Espie | |
as found out by sebastia@ and used during 5.2 vax package build... no ill side-effects so far, so do this as a stopgap measure (matthieu@ reports that our libtool is not yet very good for static-only arches). | |||
2012-08-04 | zap nonsense code | Marc Espie | |
2012-07-18 | style | Marc Espie | |
2012-07-13 | tidbit | Marc Espie | |
2012-07-13 | minor cleanup | Marc Espie | |
2012-07-13 | move stuff around some more, do not load link parts unless we are actually | Marc Espie | |
linking. (and always load basic linker class when we're actually linking) Start making a proper "library stash class". | |||
2012-07-13 | fix shared library lookup... | Marc Espie | |
2012-07-12 | common code for deducing libraries. | Marc Espie | |
Still fucks up order somehow... | |||
2012-07-12 | okay, now the parser module is totally part of Mode/Link.pm, no reason | Marc Espie | |
for a separate file | |||
2012-07-12 | perlify: $_ is subject | Marc Espie | |
2012-07-12 | match what stupid gnu libtool is doing, namely -o is not a short option | Marc Espie | |
2012-07-12 | allow "long options" to be one letter, enforce with ! | Marc Espie | |
make sure -L dir will error out. "cute" gnu libtool: -x c will vanish from linker arguments, but only in library linking-mode. | |||
2012-07-12 | later | Marc Espie | |
2012-07-12 | match gnu-libtool: error out if -L dir is used instead of -Ldir | Marc Espie | |
2012-07-12 | deal with -pthread as a "once on" option. | Marc Espie | |
2012-07-12 | unwanted duplicate | Marc Espie | |
2012-07-12 | simplify yet a bit more | Marc Espie | |
2012-07-12 | move a bit of stuff around | Marc Espie | |
2012-07-11 | duh, idiot | Marc Espie | |
2012-07-11 | must check release is defined, it CAN be 0. | Marc Espie | |
2012-07-11 | try to make code vaguely sane | Marc Espie | |
2012-07-11 | simplify cache code a bit | Marc Espie | |
2012-07-11 | no need to trace libtool when executing an external command fails. | Marc Espie | |
otherwise I get bogus bug-reports from nitwits | |||
2012-07-11 | thinko | Marc Espie | |
2012-07-11 | fix bug for real. Turns out some libraries (xcb...) have NEEDED that go | Marc Espie | |
to standard ldconfig search dirs... Of course, ld complains because it doesn't have them in its standard search dir. BUT there's an option -rpath-link designed just for that (look up needed objects during compile and trust the end result to find them). So use that for ldconfig directories... so that xcb can link without needing to go have a look at the NEEDED list, but without encoding standard rpaths in the resulting binary... Also, put all the junk that needs -Wl together so we can add one single -Wl,-rpath,foo,-rpath-link,foo2 thingy... | |||
2012-07-11 | do the ldconfig search dir properly | Marc Espie | |
2012-07-11 | nit: bless object early so we can use methods on it | Marc Espie | |
2012-07-10 | quick fix ? | Marc Espie | |
2012-07-10 | unconfuse myself. | Marc Espie | |
having two methods that do completely different things with the same name (!) is evil... | |||
2012-07-10 | ouch, Rresolved and libsearchdirs no longer live in main, so make them | Marc Espie | |
part of $gp as they should be... Oh, and create an actual linker object for later. | |||
2012-07-10 | move some shit around to (more or less) where it belongs | Marc Espie | |
2012-07-10 | bye-bye Getopt::Long | Marc Espie | |
2012-07-10 | clean-up shit, pull some of LT::Util into ltconfig object | Marc Espie | |
make that a somewhat more "real" object... | |||
2012-07-10 | full support for -static -shared -prefer-pic -prefer-non-pic -fpie* | Marc Espie | |
in --mode=compile | |||
2012-07-10 | pass the option string along as a 4th parameter, in cases canonical is | Marc Espie | |
not all that matters (e.g., -fpie and friends) | |||
2012-07-10 | nit | Marc Espie | |
2012-07-10 | more config, sort stuff for easier spotting | Marc Espie | |
2012-07-09 | --config had better match more stupid shit. | Marc Espie | |
should fix devel/avr | |||
2012-07-09 | longarg: | Marc Espie | |
means longoption with a separate parameter, like for -Xcompiler (won't work as -Xcompiler=value) | |||
2012-07-09 | start ditching Getopt::Long, only LT::Mode::Link uses it, so make it | Marc Espie | |
autoload only when absolutely required. | |||
2012-07-09 | extend option parser a bit more: if we pass it a regexp, then we match | Marc Espie | |
that regexp. for permuted mode, expose the "future" arguments so we can tinker with them directly. Use that to support -Wc,* and -Xcompiler | |||
2012-07-09 | more compat shit for libtool1/libtool2 schizophrenia | Marc Espie | |
should fix print/pdflib found out by jasper@ | |||
2012-07-09 | use short accessor | Marc Espie | |
2012-07-09 | nits | Marc Espie | |
2012-07-09 | finally, also do accessors for short options. Turns out to be useful. | Marc Espie | |
fix the array accessor... | |||
2012-07-09 | a bit more scaffolding to distinguish between syntax errors during compilation | Marc Espie | |
and non existent files (which is normal here). | |||
2012-07-09 | avoid duplicate -o | Marc Espie | |
2012-07-09 | create mode objects thru factory always: uniform require. | Marc Espie | |
use getopt array mode for simpler test | |||
2012-07-09 | tweak auto-accessor for arrays to be more useful | Marc Espie | |