[lld] [LLD][COFF] Follow up comments on pr146610 (PR #147152)

[Martin Storsjö via llvm-commits]

================
@@ -207,10 +207,10 @@ void ArchiveFile::parse() {
   // Read the symbol table to construct Lazy objects.
   bool skipDllMain = false;
   for (const Archive::Symbol &sym : file->symbols()) {
-    // If the DllMain symbol was exported by mistake, skip importing it
-    // otherwise we might end up with a import thunk in the final binary which
-    // is wrong.
-    if (sym.getName() == "__imp_DllMain" || sym.getName() == "DllMain") {
+    // If an import library provides the DllMain symbol, skip importing it, as
+    // we should be using our own DllMain, not another DLL's DllMain.
+    if (sym.getName() == "__imp_DllMain" || sym.getName() == "DllMain" ||
+        sym.getName() == "_DllMain") {
----------------
mstorsjo wrote:

We shouldn't be testing two different syntaxes, we should only be testing the right one for each architecture. (Otherwise we'd be matching an unrelated symbol on x86_64 too, even if using a symbol named `_DllMain` there also would be kinda odd.) IIRC you can do e.g. `mangle("DllMain")` to get the right symbol name. And the same goes for the `__imp_` form too, on i386 it would be `__imp__DllMain`, so `Twine("__imp_") + mangle("DllMain")` or something along those lines.

https://github.com/llvm/llvm-project/pull/147152