package/libfuse3: fix build failure on Microblaze due to wrong symver attribute detection

Add a patch to fix symver wrong detection. As explained into the patch
pending upstream[1] __has attribute() gives false positive when an
attribute is not supported[2] and this leads to assuming symver attribute
is present but it's not. This consequentely leads to build failure. I.e.
Microblaze doesn't really support it since it doesn't include elfos.h in
config.gcc that determines symver to be available.

[1]: https://github.com/libfuse/libfuse/pull/620
[2]: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=101766

Fixes:
http://autobuild.buildroot.net/results/d6c/d6cfaf2aafaeda3c12d127f6a2d2e175b25e654f

Signed-off-by: Giulio Benetti <giulio.benetti@benettiengineering.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>

diff --git a/package/libfuse3/0001-meson.build-fix-wrong-.symver-detection.patch b/package/libfuse3/0001-meson.build-fix-wrong-.symver-detection.patch
new file mode 100644 (file)
index 0000000..da20c26
--- /dev/null
+++ b/package/libfuse3/0001-meson.build-fix-wrong-.symver-detection.patch
@@ -0,0 +1,61 @@
+From 3aba09a5c56e017746c5c1652dbc845f4db7374a Mon Sep 17 00:00:00 2001
+From: Giulio Benetti <giulio.benetti@benettiengineering.com>
+Date: Tue, 3 Aug 2021 23:39:46 +0200
+Subject: [PATCH] meson.build: fix wrong .symver detection
+
+As pointed here [1] __has_attribute() is broken for many attributes and
+if it doesn't support the specific attribute it returns true, so we
+can't really rely on that for this check. This lead to Buildroot
+libfuse3 build failure [2] where that shows up with:
+```
+error: symver is only supported on ELF platforms
+```
+Indeed Microblaze doesn't support ELF since it doesn't include elfos.h,
+but __has_attribute(symver) returns true.
+
+So let's substitute the #ifdef __has_attribute() with a stronger test on
+a function foo() with __attribute__((symver ("test@TEST"))).
+
+[1]: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=101766
+[2]: http://autobuild.buildroot.net/results/d6c/d6cfaf2aafaeda3c12d127f6a2d2e175b25e654f/build-end.log
+
+Signed-off-by: Giulio Benetti <giulio.benetti@benettiengineering.com>
+---
+ meson.build | 16 +++++++---------
+ 1 file changed, 7 insertions(+), 9 deletions(-)
+
+diff --git a/meson.build b/meson.build
+index b0250ee..a7281f3 100644
+--- a/meson.build
++++ b/meson.build
+@@ -96,20 +96,18 @@ endif
+ # to have __has_attribute, then they are too old to support symver)
+ # other compilers might not have __has_attribute, but in those cases
+ # it is safe for this check to fail and for us to fallback to the old _asm_
+-# method for symver
++# method for symver. Anyway the attributes not supported by __has_attribute()
++# unfortunately return true giving a false positive. So let's try to build
++# using __attribute__ ((symver )) and see the result.
+ code = '''
+-#if defined __has_attribute
+-# if !__has_attribute (symver)
+-# error symver attribute not supported
+-# endif
+-#else
+-#error __has_attribute not defined, assume we do not have symver
+-#endif
++__attribute__ ((symver ("test@TEST")))
++void foo(void) {
++}
+ 
+ int main(void) {
+     return 0;
+ }'''
+-if cc.compiles(code, args: [ '-O0', '-c'])
++if cc.compiles(code, args: [ '-O0', '-c', '-Werror'])
+      message('Compiler supports symver attribute')
+      add_project_arguments('-DHAVE_SYMVER_ATTRIBUTE', language: 'c')
+ else
+-- 
+2.25.1
+