[llvm-bugs] [Bug 35635] New: Should we implement --no-allow-shlib-undefined?

via llvm-bugs llvm-bugs at lists.llvm.org
Mon Dec 11 16:35:07 PST 2017


            Bug ID: 35635
           Summary: Should we implement --no-allow-shlib-undefined?
           Product: lld
           Version: unspecified
          Hardware: All
                OS: All
            Status: NEW
          Severity: enhancement
          Priority: P
         Component: ELF
          Assignee: unassignedbugs at nondot.org
          Reporter: ruiu at google.com
                CC: llvm-bugs at lists.llvm.org

Currently lld behaves as if --allow-shlib-undefined is given, and there's no
way to check whether undefined symbols in DSOs can be resolved or not.

The lack of the option could result in a bad user experience. For example,
imagine that a user forgot to implement some function in a DSO. The DSO links
fine. And all executables that uses the DSO link fine with lld. But the
executables would fail at run-time because the symbol cannot be resolved by the
dynamic linker. Even worse, they don't always fail at start-up time because of
PLT entries are resolved lazily.

Maybe we should evaluate the impact of implementing --no-allow-shlib-undefined.
If it can be implemented easily, we probably should do that.

You are receiving this mail because:
You are on the CC list for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-bugs/attachments/20171212/41279ed7/attachment.html>

More information about the llvm-bugs mailing list