On 2018-10-21 15:15 +0000, Debian Bug Tracking System wrote: Upstream confirms that the symbols come from:
Is the symbol referring to
void* operator new(size_t size) GC_DECL_NEW_THROW {
return GC_MALLOC_UNCOLLECTABLE(size);
}
(and corresponding 'operator delete')
in gc_cpp.cc?
And that those do use size_t parameters so the subst size_t notation
in the symbols files is correct and the arch-bits notation is wrong.
The attached patch works, but I think that going back to the subst
notation is actually more correct. Happy to prepare a patch for that if needed.
+diff -Nru libgc-7.6.4/debian/libgc1c2.symbols libgc-7.6.4/debian/libgc1c2.symb$
+--- libgc-7.6.4/debian/libgc1c2.symbols 2018-09-09 13:25:27.000000000 +$
++++ libgc-7.6.4/debian/libgc1c2.symbols 2018-09-20 21:28:53.000000000 +$
+@@ -690,14 +690,14 @@
+ (arch=kfreebsd-amd64 kfreebsd-i386)etext@Base 1:7.2d
+ libgccpp.so.1 libgc1c2 #MINVER#
+ _ZdaPv@Base 1:7.2d
+- (arch-bits=32)_ZdaPvj@Base 1:7.6.4
++ (arch-bits=32|arch=!arm64ilp32)_ZdaPvj@Base 1:7.6.4
+ (arch-bits=64)_ZdaPvm@Base 1:7.6.4
+ _ZdlPv@Base 1:7.2d
+- (arch-bits=32)_ZdlPvj@Base 1:7.6.4
++ (arch-bits=32|arch=!arm64ilp32)_ZdlPvj@Base 1:7.6.4
+ (arch-bits=64)_ZdlPvm@Base 1:7.6.4
+- (arch-bits=32)_Znaj@Base 1:7.6.4
++ (arch-bits=32|arch=!arm64ilp32)_Znaj@Base 1:7.6.4
+ (arch-bits=64)_Znam@Base 1:7.6.4
+- (arch-bits=32)_Znwj@Base 1:7.6.4
++ (arch-bits=32|arch=!arm64ilp32)_Znwj@Base 1:7.6.4
+ (arch-bits=64)_Znwm@Base 1:7.6.4
+ (arch=sparc sparc64)_etext@Base 1:7.2d
+ (arch=kfreebsd-amd64 kfreebsd-i386)etext@Base 1:7.2d
Wookey
--
Principal hats: Linaro, Debian, Wookware, ARM
http://wookware.org/
signature.asc
Description: PGP signature
