Is there anything more preventing it to be merged to trunk?
The patches (rebased to recent trunk) are attached. (Actually only the first one is updated due to merge conflict in ChangeLog files.)
>Wed, 20 Jan 2016, 1:41 +03:00 from Ivan Maidanski < [hidden email] >:
>I've checked objc tests - same as without the patches.
>Actually objc uses only GC_malloc/realloc and GC typed API - these are not affected.
>> Sun, 17 Jan 2016, 1:55 +01:00 from Matthias Klose < [hidden email] >:
>> On 16.01.2016 13:19, Ivan Maidanski wrote:
>> > Hello,
>> > Proposed 4 commits are code refactoring of boehm-gc (and libjava/boehm.cc, accordingly) to match API of recent BDWGC master ( https://github.com/ivmai/bdwgc ). This should simplify optional replacement of internal boehm-gcc with the external one.
>> > All gcj tests pass.
>> > Some of the benefits of using mainline boehm gc:
>> > * has no compiler warning
>> > * more targets supported (e.g. arm64)
>> > * no trouble with back-porting gc patches Similar thing was proposed for the Mono community - https://github.com/mono/mono/pull/2247 (the patches were accepted recently)
>> > References:
>> >  The proposed patches in GCC unofficial git mirror fork - https://github.com/ivmai/gcc/tree/match-bdwgc-api >>
>> for completeness, please check an ObjC enabled build configured with