/*      $NetBSD: spe.c,v 1.11 2020/07/06 09:34:16 rin Exp $     */

/*-
* Copyright (c) 2011 The NetBSD Foundation, Inc.
* All rights reserved.
*
* This code is derived from software contributed to The NetBSD Foundation
* by Matt Thomas of 3am Software Foundry.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
*    notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
*    notice, this list of conditions and the following disclaimer in the
*    documentation and/or other materials provided with the distribution.
*
* THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS
* ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED
* TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
* PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS
* BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
* CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
* SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
* INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
* CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
* ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
* POSSIBILITY OF SUCH DAMAGE.
*/

#include <sys/cdefs.h>
__KERNEL_RCSID(0, "$NetBSD: spe.c,v 1.11 2020/07/06 09:34:16 rin Exp $");

#ifdef _KERNEL_OPT
#include "opt_altivec.h"
#endif

#ifdef PPC_HAVE_SPE

#include <sys/param.h>
#include <sys/proc.h>
#include <sys/systm.h>
#include <sys/atomic.h>
#include <sys/siginfo.h>
#include <sys/pcu.h>

#include <powerpc/altivec.h>
#include <powerpc/spr.h>
#include <powerpc/booke/spr.h>
#include <powerpc/psl.h>
#include <powerpc/pcb.h>

static void vec_state_load(lwp_t *, u_int);
static void vec_state_save(lwp_t *);
static void vec_state_release(lwp_t *);

const pcu_ops_t vec_ops = {
       .pcu_id = PCU_VEC,
       .pcu_state_load = vec_state_load,
       .pcu_state_save = vec_state_save,
       .pcu_state_release = vec_state_release,
};

bool
vec_used_p(lwp_t *l)
{
       return pcu_valid_p(&vec_ops, l);
}

void
vec_mark_used(lwp_t *l)
{
       pcu_discard(&vec_ops, l, true);
}

void
vec_state_load(lwp_t *l, u_int flags)
{
       struct pcb * const pcb = lwp_getpcb(l);

       if ((flags & PCU_VALID) == 0) {
               memset(&pcb->pcb_vr, 0, sizeof(pcb->pcb_vr));
               vec_mark_used(l);
       }

       /*
        * Enable SPE temporarily (and disable interrupts).
        */
       const register_t msr = mfmsr();
       mtmsr((msr & ~PSL_EE) | PSL_SPV);
       __asm volatile ("isync");

       /*
        * Call an assembly routine to do load everything.
        */
       vec_load_from_vreg(&pcb->pcb_vr);
       __asm volatile ("sync");


       /*
        * Restore MSR (turn off SPE)
        */
       mtmsr(msr);
       __asm volatile ("isync");

       /*
        * Set PSL_SPV so vectors will be enabled on return to user.
        */
       l->l_md.md_utf->tf_srr1 |= PSL_SPV;
}

void
vec_state_save(lwp_t *l)
{
       struct pcb * const pcb = lwp_getpcb(l);

       /*
        * Turn on SPE, turn off interrupts.
        */
       const register_t msr = mfmsr();
       mtmsr((msr & ~PSL_EE) | PSL_SPV);
       __asm volatile ("isync");

       /*
        * Save the vector state which is best done in assembly.
        */
       vec_unload_to_vreg(&pcb->pcb_vr);
       __asm volatile ("sync");

       /*
        * Restore MSR (turn off SPE)
        */
       mtmsr(msr);
       __asm volatile ("isync");
}

void
vec_state_release(lwp_t *l)
{
       /*
        * Turn off SPV so the next SPE instruction will cause a
        * SPE unavailable exception
        */
       l->l_md.md_utf->tf_srr1 &= ~PSL_SPV;
}

void
vec_restore_from_mcontext(lwp_t *l, const mcontext_t *mcp)
{
       struct pcb * const pcb = lwp_getpcb(l);
       const union __vr *vr = mcp->__vrf.__vrs;

       vec_save(l);

       /* grab the accumulator */
       pcb->pcb_vr.vreg[8][0] = vr->__vr32[2];
       pcb->pcb_vr.vreg[8][1] = vr->__vr32[3];

       /*
        * We store the high parts of each register in the first 8 vectors.
        */
       for (u_int i = 0; i < 8; i++, vr += 4) {
               pcb->pcb_vr.vreg[i][0] = vr[0].__vr32[0];
               pcb->pcb_vr.vreg[i][1] = vr[1].__vr32[0];
               pcb->pcb_vr.vreg[i][2] = vr[2].__vr32[0];
               pcb->pcb_vr.vreg[i][3] = vr[3].__vr32[0];
       }
       l->l_md.md_utf->tf_spefscr = pcb->pcb_vr.vscr = mcp->__vrf.__vscr;
       pcb->pcb_vr.vrsave = mcp->__vrf.__vrsave;
}

bool
vec_save_to_mcontext(lwp_t *l, mcontext_t *mcp, unsigned int *flagp)
{
       struct pcb * const pcb = lwp_getpcb(l);

       if (!vec_used_p(l))
               return false;

       vec_save(l);

       mcp->__gregs[_REG_MSR] |= PSL_SPV;

       union __vr *vr = mcp->__vrf.__vrs;
       const register_t *fixreg = l->l_md.md_utf->tf_fixreg;
       for (u_int i = 0; i < 32; i++, vr += 4, fixreg += 4) {
               vr[0].__vr32[0] = pcb->pcb_vr.vreg[i][0];
               vr[0].__vr32[1] = fixreg[0];
               vr[0].__vr32[2] = 0;
               vr[0].__vr32[3] = 0;
               vr[1].__vr32[0] = pcb->pcb_vr.vreg[i][1];
               vr[1].__vr32[1] = fixreg[1];
               vr[1].__vr32[2] = 0;
               vr[1].__vr32[3] = 0;
               vr[2].__vr32[0] = pcb->pcb_vr.vreg[i][2];
               vr[2].__vr32[1] = fixreg[2];
               vr[2].__vr32[2] = 0;
               vr[2].__vr32[3] = 0;
               vr[3].__vr32[0] = pcb->pcb_vr.vreg[i][3];
               vr[3].__vr32[1] = fixreg[3];
               vr[3].__vr32[2] = 0;
               vr[3].__vr32[3] = 0;
       }

       mcp->__vrf.__vrs[0].__vr32[2] = pcb->pcb_vr.vreg[8][0];
       mcp->__vrf.__vrs[0].__vr32[3] = pcb->pcb_vr.vreg[8][1];

       mcp->__vrf.__vrsave = pcb->pcb_vr.vrsave;
       mcp->__vrf.__vscr = l->l_md.md_utf->tf_spefscr;

       *flagp |= _UC_POWERPC_SPE;

       return true;
}

static const struct {
       uint32_t mask;
       int code;
} spefscr_siginfo_map[] = {
       { SPEFSCR_FINV|SPEFSCR_FINVH, FPE_FLTINV },
       { SPEFSCR_FOVF|SPEFSCR_FOVFH, FPE_FLTOVF },
       { SPEFSCR_FUNF|SPEFSCR_FUNFH, FPE_FLTUND },
       { SPEFSCR_FX  |SPEFSCR_FXH,   FPE_FLTRES },
       { SPEFSCR_FDBZ|SPEFSCR_FDBZH, FPE_FLTDIV },
       { SPEFSCR_OV  |SPEFSCR_OVH,   FPE_INTOVF },
};

int
vec_siginfo_code(const struct trapframe *tf)
{
       for (u_int i = 0; i < __arraycount(spefscr_siginfo_map); i++) {
               if (tf->tf_spefscr & spefscr_siginfo_map[i].mask)
                       return spefscr_siginfo_map[i].code;
       }
       return 0;
}

#endif /* PPC_HAVE_SPE */

You are viewing proxied material from ftp.icm.edu.pl. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.