Giulio Benetti

> can you change the target branch to 'develop', please? Done

Unfortunately I haven't take into account 2 things: 1) __APPLE__ doesn't support SYMVER at all 2) FUSE_SYMVER can use symver or assembly symver directive depending on HAVE_SYMVER_ATTRIBUTE So here I...

Hi @aakefbs, > @giuliobenetti Why did you close it? I'm not sure if your patch is right. I agree it should test for HAVE_SYMVER_ATTRIBUTE, but if this is not defined,...

Then, what about in fuse_lowlevel.h ``` #if (!defined(__UCLIBC__) && !defined(__APPLE__)) int fuse_parse_cmdline(struct fuse_args *args, struct fuse_cmdline_opts *opts); #else #if FUSE_USE_VERSION < FUSE_MAKE_VERSION(3, 12) int fuse_parse_cmdline_30(struct fuse_args *args, struct fuse_cmdline_opts *opts);...

I've updated and build-tested succesfully with: - glibc - uclibc - musl on Buildroot.

> @giuliobenetti Could you test this patch? > > ``` > diff --git a/meson.build b/meson.build > index 3cef64fd..0695cff7 100644 > --- a/meson.build > +++ b/meson.build > @@ -100,6 +100,13 @@...

> I fear there is another problem with your patch - anything compiled against a previous libfuse won't be ABI compatible, as the fuse_parse_cmdline() symbol will not be defined. Oh,...

> @giuliobenetti Thanks for your help to solve this. I think I know what we need to do - the `fuse_parse_cmdline` compat symbol needs to go into a separated .c...

> I have two patches here https://github.com/aakefbs/libfuse/commits/ulibc > > One is to fix the fuse_parse_cmdline issue and the other is to test **APPLE** and **ULIBC**. The test fails with a...

> Now my question is how this was ever successfully compiling on ulibc? Honestly I don't know since I don't have a test case :-/ Anyway Buildroot's Autobuilders should catch...