Re: Test case for bug that occurred in 2.27 on mips

["Maciej W. Rozycki" <macro at mips dot com>]

Hi,

 Apologies for the delay in response, I've had a hectic time recently.

> I encountered an issue linking the attached file on binutils 2.27 when
> targeting mips. I suspect the issue is already fixed, as I see another
> resolved bug relating to similar functionality in a newer version.
> I don't have the latest binutils to confirm it right now.
> 
> I assume there might be some value in it as a test case that has
> recently failed, please consider it.

 A test case is always valuable, although one that only relies on binutils 
components, such as GAS and LD, has a higher chance for reproducibility.  
In this case however I was able to trigger the same assertion failure, as 
well as another one, with a MIPS/Linux compiler and binutils master.

 That is actually unsurprising given that your program defines an ifunc, 
and we still have no support for ifuncs in the MIPS port of binutils.  An 
implementation has been proposed and the review has started, but has not 
completed yet.  The most recent version submitted has been archived here: 
<https://sourceware.org/ml/binutils/2017-01/msg00412.html>.

 I hope this helps.  Let me know if you have any further questions or 
comments.  Thank you for your interest in GNU binutils.

  Maciej