On 12/6/25 18:38, Uri Lublin wrote:
On Sat, Dec 6, 2025 at 10:54 AM Nicholas Vinson <[email protected]
<mailto:[email protected]>> wrote:
Disable the -Wmissing-braces warning as it suggests adding braces to
designated initializes in a manner that violates C++20 rules.
Signed-off-by: Nicholas Vinson <[email protected]
<mailto:[email protected]>>
---
m4/spice-compile-warnings.m4 | 1 +
1 file changed, 1 insertion(+)
diff --git a/m4/spice-compile-warnings.m4 b/m4/spice-compile-warnings.m4
index abe4a912..d2e1acc8 100644
--- a/m4/spice-compile-warnings.m4
+++ b/m4/spice-compile-warnings.m4
@@ -84,6 +84,7 @@ AC_DEFUN([SPICE_COMPILE_WARNINGS],[
dontwarn="$dontwarn -Wstrict-prototypes"
dontwarn="$dontwarn -Wsuggest-final-methods"
dontwarn="$dontwarn -Wsuggest-final-types"
+ dontwarn="$dontwarn -Wmissing-braces"
Probably better to (also?) remove "-Wmissing-braces" from the
list iterated by gl_manywarn_item.
Just before gl_manywarn_item there is a comment that says "List all gcc
warning categories". That comment is why I did not remove it from the
gl_manywarn_item list.
It seemed more logically consistent to me to place it in the list and
then use the dontwarn feature to disable the flag.
That said, I am amenable to either approach; I just would want to make
sure everyone is OK with the requested change.
Thanks,
Nicholas Vinson
I suggest applying the m4 patch first (and g++ fix second) such
that the build works after the first patch is applied.
Thanks,
Uri
# Get all possible GCC warnings
gl_MANYWARN_ALL_GCC([maybewarn])
--
2.52.0
