cpet/openbsd-ports
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
66,993 Commits 1 Branch 0 Tags 554 MiB
c92312ea9e
Commit Graph

80 Commits

Author SHA1 Message Date
robert
d75e3af324 use a different method to force using the bundled mcs.exe for bootstrapping 
to avoid random build failures in fake
 2011-06-05 15:16:48 +00:00
robert
16f1c2b8eb Make mono happy with vmmap. Thanks to Zoltan Varga from Mono for looking 
into this issue and coming up with the fix:
Fix NOMAP32BIT support in specific trampolines.
 2011-06-04 11:17:08 +00:00
ajacoutot
6f3853d4f2 And once again, fix the DLLMAP_FILES regex. 2011-05-03 10:40:59 +00:00
robert
5413f6acc9 update to 2.10.2 2011-04-27 08:50:15 +00:00
ajacoutot
0f6542179b Use the converters/libiconv MODULE. 2011-03-15 10:02:59 +00:00
robert
ec1135afbf remove old patches 2011-03-02 08:47:24 +00:00
robert
e8f3634c5b update to 2.10.1 2011-03-02 08:37:51 +00:00
ajacoutot
4d44cd5446 Better regex, with robert@ 2011-01-16 16:44:27 +00:00
robert
3a126321a6 update to 2.8.2 2011-01-14 21:14:12 +00:00
robert
3cea55a8c8 update to 2.8.1 2010-11-18 22:40:16 +00:00
espie
88d20077a4 new depends 2010-11-17 08:05:12 +00:00
espie
5e243a5257 convert WANTLIB on geo/lang 2010-11-07 19:12:42 +00:00
sthen
d15e1ce121 Reorder the ifdefs so this works both with and without kvm_getprocs. 
It didn't build on -current before, and no change on systems with
kvm_getprocs removed, so no need to bump.

Thanks naddy@ for report + build test.
 2010-10-27 14:45:26 +00:00
sthen
48233a5e3f adjust ifdefs so the kinfo_proc2 support is used. ok robert@ 2010-10-26 17:30:38 +00:00
espie
88dd25abf4 USE_GROFF=Yes 2010-10-18 19:20:41 +00:00
robert
13e1377fd8 ${_SYSTRACE_CMD} is not needed here 2010-10-16 13:21:20 +00:00
robert
e653dbf40a Update to mono-2.8; 
there might be some breakage but this is the good time to do the update
 2010-10-15 16:21:07 +00:00
ajacoutot
9bfe9801e2 Sync WANTLIB. 2010-10-01 05:23:52 +00:00
jasper
79370ddcd2 - revert patches after subst_cmd was ran sometime ago, so we have ${LIBTOOL} again 2010-09-28 18:33:06 +00:00
ajacoutot
d25bd8856b Remove powerpc from mono supported arches. 
It builds but there is no support for sigcontext and mono crashes right on
startup. No need to loose time building this...

input and ok robert@
 2010-09-16 18:53:29 +00:00
robert
be2e0b37ac - fix exception handling on amd64 
- fix i386 builds
- remove a patchfile that is not needed anymore
 2010-08-23 13:18:50 +00:00
robert
50c5451a44 add a missing patchfile missed by either me or cvs 2010-08-22 16:59:04 +00:00
robert
19c8a21871 Update to 2.6.7; remove patches that were merged backed to upstream 2010-08-18 10:51:55 +00:00
espie
792fe50e7c MODMONO_ONLY_FOR_ARCHS 
zap () subshells
 2010-07-05 09:13:05 +00:00
ajacoutot
038dca8ad6 ONLY_FOR_ARCHS is already part of the mono MODULE. 2010-07-05 06:16:58 +00:00
naddy
db2480f1f2 fix SO_PEERCRED fallout 2010-07-03 16:58:11 +00:00
robert
b225ba5909 Do not define GNU_SOURCE on OpenBSD otherwise it will pick up some 
GNU API which will break badly on OpenBSD.
 2010-06-28 18:15:24 +00:00
ajacoutot
fc908f7fe4 Bump after recent mono MODULE change. 
"sure" robert@
 2010-05-15 09:40:17 +00:00
ajacoutot
1b229dd600 Fix substitution for DLLMAP_FILES. 
This makes sure we catch all the stupid so.X stuffs.

ok robert@
 2010-05-15 09:37:48 +00:00
robert
c391b61a05 remove the gmake workaround 2010-05-10 20:17:31 +00:00
robert
cfa8e4468c add some missing patches that cvs missed .. bad cvs 2010-04-28 17:06:26 +00:00
robert
5d24fb6f79 Update to mono 2.6.4; 
(a lot of patches are in upstream now)
 2010-04-27 12:39:49 +00:00
robert
dbd0eb5db0 Correctly calculate GC_stackbottom on OpenBSD and set the threadstack size 
to a high value just like other BSD -s.
 2010-04-16 11:59:45 +00:00
robert
ec43dd560d Use KERN_BOOTTIME to get the uptime instead of making up a value 
because we don't have /proc.
 2010-04-08 16:40:04 +00:00
robert
0f11488b05 Disable shared handles because they are sometimes working, sometimes not. 2010-04-08 14:58:10 +00:00
ajacoutot
6ea207d134 Add MODMONO_{BUILD,RUN}_DEPENDS so other ports can use it. 
ok robert@
 2010-04-08 13:36:13 +00:00
robert
281e83e581 Fix kinfo_proc2 code and add support for more mono stuff using sysctl to 
get process informations. Remove the mkbundle.cs patch and replace it
with one that uses configure.in so the libgc dependency ends up in the
pkg-config file. Move disabled features to configure.in from our Makefile
 2010-04-08 12:12:33 +00:00
robert
2ea6d63a22 Move lang/mono dependency to mono.port.mk 2010-04-05 10:39:35 +00:00
robert
ecb79bdd63 add ONLY_FOR_ARCHS here 2010-04-05 01:15:00 +00:00
robert
702fa474ed Enable shared handles and add USE_NANT to mono.port.mk to use NAnt 
to build stuff
 2010-04-05 01:02:57 +00:00
robert
40cfd19ee0 - use pthread_stackseg_np to get the stack boundaries on OpenBSD 
- remove some scripts that we don't use from BASH_SCRIPTS
- ignore one of the regression tests and continue if it fails
 2010-04-01 21:53:51 +00:00
robert
38713339c6 resurrect powerpc support 2010-03-29 10:32:06 +00:00
robert
f390996a68 mkdbundle has to link to libgc. 2010-03-28 15:25:15 +00:00
robert
be3e2f5ed5 Use autoconf 2.63 and remove the SIZEOF_VOID_P definition from the CFLAGS 
because new m4 in base fixes the SIZEOV_VOID_P -> SIZEOF_VOID__ stupidity,
so make sure you have up-to-date m4 before building the port.
 2010-03-27 12:37:15 +00:00
robert
f99e840e0c Remove the handles.c patch because that's not needed anymore because 
Zoltan Varga from the mono team helped me to debug the problem we
were having. So the correct problem description is that on OpenBSD,
ctx can be NULL if we are interrupting poll().
The upstream diff also fixes the issue where we are interrupting
unmanaged code.
 2010-03-27 12:06:26 +00:00
robert
d68bbb33d2 retry 5 times then skip 2010-03-26 17:53:26 +00:00
robert
87c1dfbfef Sometimes we have the mutex is still in use by some other thread so 
let's wait until it gets unlocked, this makes the random build errors
go away.. again.
 2010-03-26 17:01:02 +00:00
robert
35962ca14a Remove the previous diff to mini.h because that was totally wrong, i made 
it up before having my morning coffee.
So it turns out that sometimes the context on openbsd can be 0. This will
lead to problems. I suspect that this is because of our fantastic pthread
library. With the following diff we just skip doing some JIT stuff if the
context is 0 because the thread probably exited before.
Anyways this way everything *seems* to work fine and almost all regression
tests are passing now except for 2 out of 380.
Pepole understanding pthreads are welcome to look at it and come up with
something that is not a tripe XXX hack.
 2010-03-26 13:13:47 +00:00
robert
fdc60f1997 take maintainership 2010-03-26 10:41:55 +00:00
robert
e7939895f2 Use the correct GET_CONTEXT macro otherwise we will get a 0 context 
and that will freak everything out and crash.
Define EOVERFLOW based on our headers and remove the Makefile hack
that we don't need anymore.
 2010-03-26 10:40:44 +00:00