[Bug ld/23839] ld generates a corrupted relocation table for .ARM.exidx sections

[nickc at redhat dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=23839

--- Comment #3 from Nick Clifton <nickc at redhat dot com> ---
(In reply to Philippe Daouadi from comment #2)

Hi Philippe,

> If you have a C++ project at hand that cross-compiles to ARM, maybe you can

Sorry - I do not have one of those. :-(  I did consider adding in the patch
anyway, although converting the printf() to an abort(), but there are cases
where it is valid for a relocation to have an offset that is larger than the
section size, so that would not work.

Unfortunately it seems that in order to fix this problem we are really
going to need a reproducer.

By the way, what configuration are you using ?

Cheers
  Nick

PS.  It may turn out that these large C++ programs really are too large
and that they cannot be linked successfully.  Obviously the linker should
not silently be producing broken executables, but in the end it may turn
out that there is no real solution to the underlying problem...

-- 
You are receiving this mail because:
You are on the CC list for the bug.