theraven added inline comments.
================
Comment at:
lldb/source/Plugins/LanguageRuntime/ObjC/GNUstepObjCRuntime/GNUstepObjCRuntime.cpp:42
+ LanguageType language) {
+ if (language != eLanguageTypeObjC)
+ return nullptr;
----------------
I'm not familiar with lldb internals, will this exclude ObjC++?
================
Comment at:
lldb/source/Plugins/LanguageRuntime/ObjC/GNUstepObjCRuntime/GNUstepObjCRuntime.cpp:47
+ const llvm::Triple &TT = process->GetTarget().GetArchitecture().GetTriple();
+ if (TT.getVendor() == llvm::Triple::VendorType::Apple)
+ return nullptr;
----------------
I never finished the v2 ABI support for Mach-O, but v1 works on Apple platforms
and v2 will eventually. We definitely shouldn't be the default here, but it
would be nice not to prevent it working on macOS.
================
Comment at:
lldb/source/Plugins/LanguageRuntime/ObjC/GNUstepObjCRuntime/GNUstepObjCRuntime.cpp:131
+ ExecutionContext &exe_ctx) {
+ // TODO: This function is supposed to check whether an ObjC selector is
+ // present for an object. Might be implemented similar as in the Apple V2
----------------
I'm not sure how this is meant to work. I'd expect this to call one of the
runtime's introspection functions. I believe the Apple version has variants of
these that run without acquiring locks, which can be used in the debugger. We
can add these quite easily if necessary.
================
Comment at: lldb/test/Shell/Expr/objc-gnustep-print.m:68
+//
+// MEMBERS_OUTSIDE: (lldb) p t->_int
+// MEMBERS_OUTSIDE: (int) $0 = 0
----------------
It's not clear from this test if it's correctly computing the offsets. I
suspect that it isn't, since I don't see any reference to the ivar offset
variables.
Repository:
rG LLVM Github Monorepo
CHANGES SINCE LAST ACTION
https://reviews.llvm.org/D146154/new/
https://reviews.llvm.org/D146154
_______________________________________________
lldb-commits mailing list
lldb-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
