Re: bin/47301: Miscompilation by bundled GCC

[Torbjorn Granlund <tg%gmplib.org@localhost>]

christos%zoulas.com@localhost (Christos Zoulas) writes:

  The following reply was made to PR bin/47301; it has been noted by GNATS.
  
  From: christos%zoulas.com@localhost (Christos Zoulas)
  To: gnats-bugs%NetBSD.org@localhost, toolchain-manager%netbsd.org@localhost, 
        gnats-admin%netbsd.org@localhost, netbsd-bugs%netbsd.org@localhost, 
tg%gmplib.org@localhost
  Cc: 
  Subject: Re: bin/47301: Miscompilation by bundled GCC
  Date: Mon, 10 Dec 2012 15:37:59 -0500
  
   On Dec 10,  4:40pm, Paul_Koning%Dell.com@localhost 
(<Paul_Koning%Dell.com@localhost>) wrote:
   -- Subject: Re: bin/47301: Miscompilation by bundled GCC
   
   |  > i.e., reload puts both pseudo 261 and pseudo 262 into hard reg 1 in
   |  > the same live range.  That is bad.
   |  >
   |  > Very few current backends support mem-to-mem copying in a single
   |  > instruction.  It might be something that no longer works in reload.
   |  >
   |  > Assuming my assumption is correct, a possible fix would be to disallow
   |  > mem-to-mem insns, at least until reload.
   
   Yes, but that dissallowing should happen in the MI code not in the backends,
   right? If so, do you know how to do it/where to put it?
   
No, in vax.md.  

-- 
Torbjörn