[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: mono on openbsd i386 current
On 3 Nov 2002, Marc Matteo wrote:
> Let me re-phrase: what did you have do to pkgconfig to "get linking with
> threads right"? :)
i downloaded 0.14 and built it from scratch. i modifed this line:
pkgconfig-0.14.0/glib-1.2.8/gthread/Makefile:
testgthread_LDADD = ../libglib.la libgthread.la
and appended -pthread. worked, it seems. included unit tests passed:
make check-TESTS
PASS: array-test
PASS: dirname-test
PASS: hash-test
PASS: list-test
PASS: node-test
PASS: relation-test
PASS: slist-test
PASS: string-test
PASS: strfunc-test
PASS: tree-test
PASS: type-test
=================== All 11 tests passed ===================
just a point of reference, maybe someone wants to update the port to 0.14?
the ./configure detected it needed -lc_r but didn't know how to get at it.
i think everything else here worked for pkgconfig-0.14.
current status of mono on openbsd-current (my test machine for this was a
k6-2/300) is that it builds ok but something doesn't seem to work right.
about 1% of all tests and code fails with errors, the rest fails to start
doing anything. hence, i think its a configuration issue (along the lines
of MONO_PATH) and haven't investigated it at all. it could also be
bytecode differences between intel and amd processors ... :/ it could just
be bytecode generation entirely.
for the cvs snapshot i had to delete the checks for glib-2.0 via
pkg-config (it looks for .pc files, which we dont have). the 0.16 release
didn't have this problem. if worse comes to worse either remove those
checks nd regen ./configure or add stub files in in the port.
like i said this is about as far as i want to dork around with mono on
openbsd. i want a c# compiler (i like c# more than java and c++) but i
lack the time/skills needed to bring mono up to being fully functional on
openbsd. i was fully surprised it compiled so cleanly. this was only to
provide a data point for people wishing to see mono on openbsd. it looks
promising is the conclusion.
___________________________
jose nazario, ph.d. jose@monkey.org
http://www.monkey.org/~jose/