Ian Lance Taylor - Re: SH: PR target/24445 (original) (raw)

This is the mail archive of the gcc@gcc.gnu.orgmailing list for the GCC project.

• From: Ian Lance Taylor
• To: Kaz Kojima
• Cc: joern dot rennecke at st dot com, gcc at gcc dot gnu dot org, aoliva at redhat dot com
• Date: 02 Nov 2005 20:20:13 -0800
• Subject: Re: SH: PR target/24445
• References: <43694C6F.4060609@st.com><20051103.093951.28780963.kkojima@rr.iij4u.or.jp><m3pspiijl0.fsf@gossamer.airs.com><20051103.111005.102582223.kkojima@rr.iij4u.or.jp>

Kaz Kojima kkojima@rr.iij4u.or.jp writes:

Like this?

@@ -2861,6 +2861,12 @@ expand_call (tree exp, rtx target, int i

      If they refer to the same register, this move will be a no-op,
      except when function inlining is being done.  */

•  /* We have to copy a return value in a CLASS_LIKELY_SPILLED hard

•     reg to a plain register.  */

•  if (REG_P (valreg)

•      && HARD_REGISTER_P (valreg)

•      && CLASS_LIKELY_SPILLED_P (REGNO_REG_CLASS (REGNO (valreg))))

•    valreg = copy_to_reg (valreg);
    emit_move_insn (target, valreg);
  
    /* If we are setting a MEM, this code must be executed.  Since it is

In general, yes. But looking at it we should probably only call copy_to_reg if TARGET is not itself a pseudo-register.

And I think I would put the new code before the "TARGET and VALREG cannot be equal" comment.

That patch is OK if it works and passes testing, and if you add a test case and a ChangeLog entry.

Thanks.

Ian

• Follow-Ups:
  • Re: SH: PR target/24445
    * From: Kaz Kojima
• References:
  • Re: SH: PR target/24445
    * From: Joern RENNECKE
  • Re: SH: PR target/24445
    * From: Kaz Kojima
  • Re: SH: PR target/24445
    * From: Ian Lance Taylor
  • Re: SH: PR target/24445
    * From: Kaz Kojima