[PATCH] D18190: [ELF] - Error out when R_X86_64_PC32/R_X86_64_32 are used against preemptible symbol when linking shared object.

[Sean Silva via llvm-commits]

On Wed, Mar 16, 2016 at 9:08 AM, George Rimar via llvm-commits <
llvm-commits at lists.llvm.org> wrote:

> > If you are not familiar with that test, then please take time to
> understand it. You were about to modify the input to that particular test.
> If you don't understand what you > are doing, you may be changing the
> test in such a way that it would make no longer sense.
>
>
> I modified the test in part of my needs. That change seemed and
> seems obvious for me. I agree with your position though, let return back to
> this talk tomorrow after I`ll investigate this testcase more deeply.
>

I wrote r259831 and it is simply an error in reducing the test case that I
ended up using -shared. It avoids needing another file. If you can
regenerate the .o file using the command line in the test using llvm33,
that will be sufficient (adding appropriate other flags for
-relocation-model). btw, the way I created this test was to have:

// RUN: /path/to/llvm33/bin/llvm-mc -filetype=obj
-triple=x86_64-unknown-linux %s -o %S/Inputs/llvm33-rela-outside-group.o

Only at the end when going to submit the test did I replace `RUN:
/path/to/llvm33/bin/llvm-mc` with `llvm33/llvm-mc`.

Sorry for creating all this complexity for you! If you want, I can
regenerate it when I'm back at my desk next week.

Another possibility is to change the test to not use -shared, but that
involves adding a new .s file for _start and IMO is less preferred since it
fails to emphasize the thing that is being fixed.

-- Sean Silva


>
> George.
>
> _______________________________________________
> llvm-commits mailing list
> llvm-commits at lists.llvm.org
> http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-commits
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20160316/29878efb/attachment.html>