libgcc/config/i386/crtfastmath.c

/*
 * Copyright (C) 2005-2014 Free Software Foundation, Inc.
 *
 * This file is free software; you can redistribute it and/or modify it
 * under the terms of the GNU General Public License as published by the
 * Free Software Foundation; either version 3, or (at your option) any
 * later version.
 * 
 * This file is distributed in the hope that it will be useful, but
 * WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 * General Public License for more details.
 * 
 * Under Section 7 of GPL version 3, you are granted additional
 * permissions described in the GCC Runtime Library Exception, version
 * 3.1, as published by the Free Software Foundation.
 *
 * You should have received a copy of the GNU General Public License and
 * a copy of the GCC Runtime Library Exception along with this program;
 * see the files COPYING3 and COPYING.RUNTIME respectively.  If not, see
 * <http://www.gnu.org/licenses/>.
 */

#ifndef _SOFT_FLOAT
#define MXCSR_DAZ (1 << 6)	/* Enable denormals are zero mode */
#define MXCSR_FTZ (1 << 15)	/* Enable flush to zero mode */

#ifndef __x86_64__
/* All 64-bit targets have SSE and DAZ;
   only check them explicitly for 32-bit ones. */
#include "cpuid.h"
#endif

#if !defined __x86_64__ && defined __sun__ && defined __svr4__
#include <signal.h>
#include <ucontext.h>

static volatile sig_atomic_t sigill_caught;

static void
sigill_hdlr (int sig __attribute((unused)),
	     siginfo_t *sip __attribute__((unused)),
	     ucontext_t *ucp)
{
  sigill_caught = 1;
  /* Set PC to the instruction after the faulting one to skip over it,
     otherwise we enter an infinite loop.  3 is the size of the movaps
     instruction.  */
  ucp->uc_mcontext.gregs[EIP] += 3;
  setcontext (ucp);
}
#endif

static void __attribute__((constructor))
#ifndef __x86_64__
/* The i386 ABI only requires 4-byte stack alignment, so this is necessary
   to make sure the fxsave struct gets correct alignment.
   See PR27537 and PR28621.  */
__attribute__ ((force_align_arg_pointer))
#endif
set_fast_math (void)
{
#ifndef __x86_64__
  unsigned int eax, ebx, ecx, edx;

  if (!__get_cpuid (1, &eax, &ebx, &ecx, &edx))
    return;

  if (edx & bit_SSE)
    {
      unsigned int mxcsr;
  
#if defined __sun__ && defined __svr4__
      /* Solaris 2 before Solaris 9 4/04 cannot execute SSE instructions even
	 if the CPU supports them.  Programs receive SIGILL instead, so check
	 for that at runtime.  */
      struct sigaction act, oact;

      act.sa_handler = sigill_hdlr;
      sigemptyset (&act.sa_mask);
      /* Need to set SA_SIGINFO so a ucontext_t * is passed to the handler.  */
      act.sa_flags = SA_SIGINFO;
      sigaction (SIGILL, &act, &oact);

      /* We need a single SSE instruction here so the handler can safely skip
	 over it.  */
      __asm__ volatile ("movaps %xmm0,%xmm0");

      sigaction (SIGILL, &oact, NULL);

      if (sigill_caught)
	return;
#endif /* __sun__ && __svr4__ */

      if (edx & bit_FXSAVE)
	{
	  /* Check if DAZ is available.  */
	  struct
	    {
	      unsigned short cwd;
	      unsigned short swd;
	      unsigned short twd;
	      unsigned short fop;
	      unsigned int fip;
	      unsigned int fcs;
	      unsigned int foo;
	      unsigned int fos;
	      unsigned int mxcsr;
	      unsigned int mxcsr_mask;
	      unsigned int st_space[32];
	      unsigned int xmm_space[32];
	      unsigned int padding[56];
	    } __attribute__ ((aligned (16))) fxsave;

	  /* This is necessary since some implementations of FXSAVE
	     do not modify reserved areas within the image.  */
	  fxsave.mxcsr_mask = 0;

	  __builtin_ia32_fxsave (&fxsave);

	  mxcsr = fxsave.mxcsr;

	  if (fxsave.mxcsr_mask & MXCSR_DAZ)
	    mxcsr |= MXCSR_DAZ;
	}
      else
	mxcsr = __builtin_ia32_stmxcsr ();

      mxcsr |= MXCSR_FTZ;
      __builtin_ia32_ldmxcsr (mxcsr);
    }
#else
  unsigned int mxcsr = __builtin_ia32_stmxcsr ();
  mxcsr |= MXCSR_DAZ | MXCSR_FTZ;
  __builtin_ia32_ldmxcsr (mxcsr);
#endif
}
#endif