[Lldb-commits] [lldb] [lldb] Handle STT_TLS symbol type in ELF parser (PR #178975)

[David Spickett via lldb-commits]

================
@@ -8,7 +8,7 @@
 
 
 class PlatformProcessCrashInfoTestCase(TestBase):
-    @expectedFailureAll(oslist=["windows", "linux", "freebsd", "netbsd"])
+    @expectedFailureAll(oslist=["windows", "freebsd", "netbsd"])
----------------
DavidSpickett wrote:

I mean can you build lldb using the changes in this PR and check that the test case works. This is about lldb's ability to parse the input properly rather than the system's, and as a result read variables that are thread local. 

(certainly good to know FreeBSD generally knows about the type though, would certainly have been a blocker if it didn't)

Let's leave NetBSD in the xfails then, worse case it unexpectedly passes.

As for Darwin, there is already a `skipIfDarwin`, and you're right that Darwin is mach-o. So there's no reason to re-enable it in this PR.

https://github.com/llvm/llvm-project/pull/178975