Invalid read in mpz_powm_ui

Torbjorn Granlund tg at
Thu Dec 6 15:03:52 CET 2012

Zimmermann Paul <Paul.Zimmermann at> writes:

  thanks. The MPFR and MPC builds with gmp-devel work again.
Thanks for feedback!

  >   PS: aren't the nightly builds of GMP tested against valgrind?
  > Nope.  Perhaps we should.  I ran it now, I and found 4 errors (including
  > yours).  Two of these have not yet been analysed, but one of these
  > unanalysed problems is related to code not yet used by any documented
  > interface.  The 4th error was harmless and due to how we do mpn_incr_u.
  > I implemented a simple workaround to silence valgrind.
  I fully agree you should!
I gave it a shot, but neither on a AMD FX nor on an ARM v7 does it
complete GMP tests in a default build.  It hits unsupported instructions
like vmovsd (for AMD64) and umaal (for ARM).

The vmovsd was generated by the compiler or part of some library
(perhaps libc).  The umaal is used from GMP's assembly code.

I understand that it is hard to keep up with the relentless addition of
instructions in the x86 world.  The umaal is not new, however.

I'll wait for valgrind 3.8 to arrive to FreeBSD's ports or GNU/Linux's
various package systems.  (I am unable to report these problems, since
they request me to "create an account", something I will not bother to


More information about the gmp-bugs mailing list