Backport windows fixes with INFINITY/NAN (#19390)

* Fixed `NAN`/`INFINITY` definitions to work on Windows 11 SDK.

* Regenerate stale files

php/ext/google/protobuf/php-upb.c

@@ -6256,9 +6256,35 @@ bool _upb_mapsorter_pushexts(_upb_mapsorter* s,
 
 // Must be last.
 
-const float kUpb_FltInfinity = INFINITY;
-const double kUpb_Infinity = INFINITY;
-const double kUpb_NaN = NAN;
+// The latest win32 SDKs have an invalid definition of NAN.
+// https://developercommunity.visualstudio.com/t/NAN-is-no-longer-compile-time-constant-i/10688907
+//
+// Unfortunately, the `0.0 / 0.0` workaround doesn't work in Clang under C23, so
+// try __builtin_nan first, if that exists.
+#ifdef _WIN32
+#ifdef __has_builtin
+#if __has_builtin(__builtin_nan)
+#define UPB_NAN __builtin_nan("0")
+#endif
+#if __has_builtin(__builtin_inf)
+#define UPB_INFINITY __builtin_inf()
+#endif
+#endif
+#ifndef UPB_NAN
+#define UPB_NAN 0.0 / 0.0
+#endif
+#ifndef UPB_INFINITY
+#define UPB_INFINITY 1.0 / 0.0
+#endif
+#else
+// For !_WIN32, assume math.h works.
+#define UPB_NAN NAN
+#define UPB_INFINITY INFINITY
+#endif
+
+const float kUpb_FltInfinity = UPB_INFINITY;
+const double kUpb_Infinity = UPB_INFINITY;
+const double kUpb_NaN = UPB_NAN;
 
 static const size_t overhead = sizeof(upb_Message_InternalData);
 

ruby/ext/google/protobuf_c/ruby-upb.c

@@ -5771,9 +5771,35 @@ bool _upb_mapsorter_pushexts(_upb_mapsorter* s,
 
 // Must be last.
 
-const float kUpb_FltInfinity = INFINITY;
-const double kUpb_Infinity = INFINITY;
-const double kUpb_NaN = NAN;
+// The latest win32 SDKs have an invalid definition of NAN.
+// https://developercommunity.visualstudio.com/t/NAN-is-no-longer-compile-time-constant-i/10688907
+//
+// Unfortunately, the `0.0 / 0.0` workaround doesn't work in Clang under C23, so
+// try __builtin_nan first, if that exists.
+#ifdef _WIN32
+#ifdef __has_builtin
+#if __has_builtin(__builtin_nan)
+#define UPB_NAN __builtin_nan("0")
+#endif
+#if __has_builtin(__builtin_inf)
+#define UPB_INFINITY __builtin_inf()
+#endif
+#endif
+#ifndef UPB_NAN
+#define UPB_NAN 0.0 / 0.0
+#endif
+#ifndef UPB_INFINITY
+#define UPB_INFINITY 1.0 / 0.0
+#endif
+#else
+// For !_WIN32, assume math.h works.
+#define UPB_NAN NAN
+#define UPB_INFINITY INFINITY
+#endif
+
+const float kUpb_FltInfinity = UPB_INFINITY;
+const double kUpb_Infinity = UPB_INFINITY;
+const double kUpb_NaN = UPB_NAN;
 
 static const size_t overhead = sizeof(upb_Message_InternalData);
 

upb/message/message.c

@@ -15,9 +15,35 @@
 // Must be last.
 #include "upb/port/def.inc"
 
-const float kUpb_FltInfinity = INFINITY;
-const double kUpb_Infinity = INFINITY;
-const double kUpb_NaN = NAN;
+// The latest win32 SDKs have an invalid definition of NAN.
+// https://developercommunity.visualstudio.com/t/NAN-is-no-longer-compile-time-constant-i/10688907
+//
+// Unfortunately, the `0.0 / 0.0` workaround doesn't work in Clang under C23, so
+// try __builtin_nan first, if that exists.
+#ifdef _WIN32
+#ifdef __has_builtin
+#if __has_builtin(__builtin_nan)
+#define UPB_NAN __builtin_nan("0")
+#endif
+#if __has_builtin(__builtin_inf)
+#define UPB_INFINITY __builtin_inf()
+#endif
+#endif
+#ifndef UPB_NAN
+#define UPB_NAN 0.0 / 0.0
+#endif
+#ifndef UPB_INFINITY
+#define UPB_INFINITY 1.0 / 0.0
+#endif
+#else
+// For !_WIN32, assume math.h works.
+#define UPB_NAN NAN
+#define UPB_INFINITY INFINITY
+#endif
+
+const float kUpb_FltInfinity = UPB_INFINITY;
+const double kUpb_Infinity = UPB_INFINITY;
+const double kUpb_NaN = UPB_NAN;
 
 static const size_t overhead = sizeof(upb_Message_InternalData);