@@ -121,6 +121,7 @@ extern void kvmppc_book3s_queue_irqprio(struct kvm_vcpu *vcpu, unsigned int vec)
extern void kvmppc_set_bat(struct kvm_vcpu *vcpu, struct kvmppc_bat *bat,
bool upper, u32 val);
extern void kvmppc_giveup_ext(struct kvm_vcpu *vcpu, ulong msr);
+extern int kvmppc_emulate_paired_single(struct kvm_run *run, struct kvm_vcpu *vcpu);
extern u32 kvmppc_trampoline_lowmem;
extern u32 kvmppc_trampoline_enter;
@@ -41,6 +41,7 @@ kvm-objs-$(CONFIG_KVM_E500) := $(kvm-e500-objs)
kvm-book3s_64-objs := \
$(common-objs-y) \
fpu.o \
+ book3s_paired_singles.o \
book3s.o \
book3s_64_emulate.o \
book3s_64_interrupts.o \
@@ -200,6 +200,9 @@ int kvmppc_core_emulate_op(struct kvm_run *run, struct kvm_vcpu *vcpu,
emulated = EMULATE_FAIL;
}
+ if (emulated == EMULATE_FAIL)
+ emulated = kvmppc_emulate_paired_single(run, vcpu);
+
return emulated;
}
new file mode 100644
@@ -0,0 +1,1356 @@
+/*
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, version 2, as
+ * published by the Free Software Foundation.
+ *
+ * This program 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.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
+ *
+ * Copyright Novell Inc 2010
+ *
+ * Authors: Alexander Graf <agraf@suse.de>
+ */
+
+#include <asm/kvm.h>
+#include <asm/kvm_ppc.h>
+#include <asm/disassemble.h>
+#include <asm/kvm_book3s.h>
+#include <asm/kvm_fpu.h>
+#include <asm/reg.h>
+#include <asm/cacheflush.h>
+#include <linux/vmalloc.h>
+
+/* #define DEBUG */
+
+#ifdef DEBUG
+#define dprintk printk
+#else
+#define dprintk(...) do { } while(0);
+#endif
+
+#define OP_LFS 48
+#define OP_LFSU 49
+#define OP_LFD 50
+#define OP_LFDU 51
+#define OP_STFS 52
+#define OP_STFSU 53
+#define OP_STFD 54
+#define OP_STFDU 55
+#define OP_PSQ_L 56
+#define OP_PSQ_LU 57
+#define OP_PSQ_ST 60
+#define OP_PSQ_STU 61
+
+#define OP_31_LFSX 535
+#define OP_31_LFSUX 567
+#define OP_31_LFDX 599
+#define OP_31_LFDUX 631
+#define OP_31_STFSX 663
+#define OP_31_STFSUX 695
+#define OP_31_STFX 727
+#define OP_31_STFUX 759
+#define OP_31_LWIZX 887
+#define OP_31_STFIWX 983
+
+#define OP_59_FADDS 21
+#define OP_59_FSUBS 20
+#define OP_59_FSQRTS 22
+#define OP_59_FDIVS 18
+#define OP_59_FRES 24
+#define OP_59_FMULS 25
+#define OP_59_FRSQRTES 26
+#define OP_59_FMSUBS 28
+#define OP_59_FMADDS 29
+#define OP_59_FNMSUBS 30
+#define OP_59_FNMADDS 31
+#define OP_59_FCFIDS 846
+#define OP_59_FCFIDUS 974
+
+#define OP_63_FCMPU 0
+#define OP_63_FCPSGN 8
+#define OP_63_FRSP 12
+#define OP_63_FCTIW 14
+#define OP_63_FCTIWZ 15
+#define OP_63_FDIV 18
+#define OP_63_FADD 21
+#define OP_63_FSQRT 22
+#define OP_63_FSEL 23
+#define OP_63_FRE 24
+#define OP_63_FMUL 25
+#define OP_63_FRSQRTE 26
+#define OP_63_FMSUB 28
+#define OP_63_FMADD 29
+#define OP_63_FNMSUB 30
+#define OP_63_FNMADD 31
+#define OP_63_FCMPO 32
+#define OP_63_MTFSB1 38
+#define OP_63_FSUB 20
+#define OP_63_FNEG 40
+#define OP_63_MCRFS 64
+#define OP_63_MTFSB0 70
+#define OP_63_FMR 72
+#define OP_63_MTFSFI 134
+#define OP_63_FCTIWU 142
+#define OP_63_FCTIWUZ 143
+#define OP_63_FNABS 136
+#define OP_63_FTDIV 128
+#define OP_63_FTSQRT 160
+#define OP_63_FABS 264
+#define OP_63_FRIN 392
+#define OP_63_FRIZ 424
+#define OP_63_FRIP 456
+#define OP_63_FRIM 488
+#define OP_63_MFFS 583
+#define OP_63_MTFSF 711
+#define OP_63_FCTID 814
+#define OP_63_FCTIDZ 815
+#define OP_63_FCFID 846
+#define OP_63_FCTIDU 942
+#define OP_63_FCTIDUZ 943
+#define OP_63_FCFIDU 974
+
+#define OP_4X_PS_CMPU0 0
+#define OP_4X_PSQ_LX 6
+#define OP_4XW_PSQ_STX 7
+#define OP_4A_PS_SUM0 10
+#define OP_4A_PS_SUM1 11
+#define OP_4A_PS_MULS0 12
+#define OP_4A_PS_MULS1 13
+#define OP_4A_PS_MADDS0 14
+#define OP_4A_PS_MADDS1 15
+#define OP_4A_PS_DIV 18
+#define OP_4A_PS_SUB 20
+#define OP_4A_PS_ADD 21
+#define OP_4A_PS_SEL 23
+#define OP_4A_PS_RES 24
+#define OP_4A_PS_MUL 25
+#define OP_4A_PS_RSQRTE 26
+#define OP_4A_PS_MSUB 28
+#define OP_4A_PS_MADD 29
+#define OP_4A_PS_NMSUB 30
+#define OP_4A_PS_NMADD 31
+#define OP_4X_PS_CMPO0 32
+#define OP_4X_PSQ_LUX 38
+#define OP_4XW_PSQ_STUX 39
+#define OP_4X_PS_NEG 40
+#define OP_4X_PS_CMPU1 64
+#define OP_4X_PS_MR 72
+#define OP_4X_PS_CMPO1 96
+#define OP_4X_PS_NABS 136
+#define OP_4X_PS_ABS 264
+#define OP_4X_PS_MERGE00 528
+#define OP_4X_PS_MERGE01 560
+#define OP_4X_PS_MERGE10 592
+#define OP_4X_PS_MERGE11 624
+
+#define SCALAR_NONE 0
+#define SCALAR_HIGH (1 << 0)
+#define SCALAR_LOW (1 << 1)
+#define SCALAR_NO_PS0 (1 << 2)
+#define SCALAR_NO_PS1 (1 << 3)
+
+#define GQR_ST_TYPE_MASK 0x00000007
+#define GQR_ST_TYPE_SHIFT 0
+#define GQR_ST_SCALE_MASK 0x00003f00
+#define GQR_ST_SCALE_SHIFT 8
+#define GQR_LD_TYPE_MASK 0x00070000
+#define GQR_LD_TYPE_SHIFT 16
+#define GQR_LD_SCALE_MASK 0x3f000000
+#define GQR_LD_SCALE_SHIFT 24
+
+#define GQR_QUANTIZE_FLOAT 0
+#define GQR_QUANTIZE_U8 4
+#define GQR_QUANTIZE_U16 5
+#define GQR_QUANTIZE_S8 6
+#define GQR_QUANTIZE_S16 7
+
+#define FPU_LS_SINGLE 0
+#define FPU_LS_DOUBLE 1
+#define FPU_LS_SINGLE_LOW 2
+
+static void call_fpu_inst(u32 inst, u64 *out, u64 *in1, u64 *in2, u64 *in3,
+ u32 *cr, u32 *fpscr)
+{
+ u32 cr_val = 0;
+ u32 *call_stack;
+ u64 inout[5] = { 0, 0, 0, 0, 0 };
+
+ if (fpscr)
+ inout[0] = *fpscr;
+ if (in1)
+ inout[1] = *in1;
+ if (in2)
+ inout[2] = *in2;
+ if (in3)
+ inout[3] = *in3;
+ if (cr)
+ cr_val = *cr;
+
+ dprintk(KERN_INFO "FPU Emulator 0x%x ( 0x%llx, 0x%llx, 0x%llx )", inst,
+ inout[1], inout[2], inout[3]);
+
+ call_stack = &kvmppc_call_stack[(smp_processor_id() * 2)];
+ call_stack[0] = inst;
+ /* call_stack[1] is INS_BLR */
+
+ flush_icache_range((ulong)call_stack, (ulong)&call_stack[1]);
+
+ dprintk(KERN_INFO "FPU call stack -> 0x%p (%x %x)\n",
+ call_stack, call_stack[0], call_stack[1]);
+
+ __asm__ volatile (
+ "lfd 0, (0*8)(%[inout]) ;"
+ "mtfsf 255, 0, 1, 0 ;"
+ "lfd 1, (1*8)(%[inout]) ;"
+ "lfd 2, (2*8)(%[inout]) ;"
+ "lfd 3, (3*8)(%[inout]) ;"
+ "mtctr %[call_addr] ;"
+ "mtcr %[cr_in] ;"
+ "bctrl ;"
+ "mffs 1 ;"
+ "stfd 1, (0*8)(%[inout]) ;"
+ "stfd 4, (4*8)(%[inout]) ;"
+ "mfcr %[cr_out] ;"
+ : [cr_out]"=r"(cr_val),
+ "=m"(inout[0]),
+ "=m"(inout[4])
+ : [cr_in]"r"(cr_val),
+ [inout]"b"(inout),
+ "m"(inout[0]),
+ "m"(inout[1]),
+ "m"(inout[2]),
+ "m"(inout[3]),
+ [call_addr]"r"(call_stack)
+ : "cc", "lr", "ctr");
+
+ dprintk(KERN_INFO "FPU Emulator result = %llx\n", inout[4]);
+
+ if (fpscr)
+ *fpscr = inout[0];
+ if (out)
+ *out = inout[4];
+ if (cr)
+ *cr = cr_val;
+}
+
+static void kvmppc_inject_pf(struct kvm_vcpu *vcpu, ulong eaddr, bool is_store)
+{
+ u64 dsisr;
+
+ vcpu->arch.msr = kvmppc_set_field(vcpu->arch.msr, 33, 36, 0);
+ vcpu->arch.msr = kvmppc_set_field(vcpu->arch.msr, 42, 47, 0);
+ vcpu->arch.dear = eaddr;
+ /* Page Fault */
+ dsisr = kvmppc_set_field(0, 33, 33, 1);
+ if (is_store)
+ to_book3s(vcpu)->dsisr = kvmppc_set_field(dsisr, 38, 38, 1);
+ kvmppc_book3s_queue_irqprio(vcpu, BOOK3S_INTERRUPT_DATA_STORAGE);
+}
+
+static int kvmppc_emulate_fpr_load(struct kvm_run *run, struct kvm_vcpu *vcpu,
+ int rs, ulong addr, int ls_type)
+{
+ int emulated = EMULATE_FAIL;
+ struct thread_struct t;
+ int r;
+ char tmp[8];
+ int len = sizeof(u32);
+
+ if (ls_type == FPU_LS_DOUBLE)
+ len = sizeof(u64);
+
+ t.fpscr.val = vcpu->arch.fpscr;
+
+ /* read from memory */
+ r = kvmppc_ld(vcpu, &addr, len, tmp, true);
+ vcpu->arch.paddr_accessed = addr;
+
+ if (r < 0) {
+ kvmppc_inject_pf(vcpu, addr, false);
+ goto done_load;
+ } else if (r == EMULATE_DO_MMIO) {
+ emulated = kvmppc_handle_load(run, vcpu, REG_FPR | rs, len, 1);
+ goto done_load;
+ }
+
+ emulated = EMULATE_DONE;
+
+ /* put in registers */
+ switch (ls_type) {
+ case FPU_LS_SINGLE:
+ cvt_fd((float*)tmp, (double*)&vcpu->arch.fpr[rs], &t);
+ vcpu->arch.qpr[rs] = *((u32*)tmp);
+ break;
+ case FPU_LS_DOUBLE:
+ vcpu->arch.fpr[rs] = *((u64*)tmp);
+ break;
+ }
+
+ dprintk(KERN_INFO "KVM: FPR_LD [0x%llx] at 0x%lx (%d)\n", *(u64*)tmp,
+ addr, len);
+
+done_load:
+ return emulated;
+}
+
+static int kvmppc_emulate_fpr_store(struct kvm_run *run, struct kvm_vcpu *vcpu,
+ int rs, ulong addr, int ls_type)
+{
+ int emulated = EMULATE_FAIL;
+ struct thread_struct t;
+ int r;
+ char tmp[8];
+ u64 val;
+ int len;
+
+ t.fpscr.val = vcpu->arch.fpscr;
+
+ switch (ls_type) {
+ case FPU_LS_SINGLE:
+ cvt_df((double*)&vcpu->arch.fpr[rs], (float*)tmp, &t);
+ val = *((u32*)tmp);
+ len = sizeof(u32);
+ break;
+ case FPU_LS_SINGLE_LOW:
+ *((u32*)tmp) = vcpu->arch.fpr[rs];
+ val = vcpu->arch.fpr[rs] & 0xffffffff;
+ len = sizeof(u32);
+ break;
+ case FPU_LS_DOUBLE:
+ *((u64*)tmp) = vcpu->arch.fpr[rs];
+ val = vcpu->arch.fpr[rs];
+ len = sizeof(u64);
+ break;
+ default:
+ val = 0;
+ len = 0;
+ }
+
+ r = kvmppc_st(vcpu, &addr, len, tmp, true);
+ vcpu->arch.paddr_accessed = addr;
+ if (r < 0) {
+ kvmppc_inject_pf(vcpu, addr, true);
+ } else if (r == EMULATE_DO_MMIO) {
+ emulated = kvmppc_handle_store(run, vcpu, val, len, 1);
+ } else {
+ emulated = EMULATE_DONE;
+ }
+
+ dprintk(KERN_INFO "KVM: FPR_ST [0x%llx] at 0x%lx (%d)\n",
+ val, addr, len);
+
+ return emulated;
+}
+
+static int kvmppc_emulate_psq_load(struct kvm_run *run, struct kvm_vcpu *vcpu,
+ int rs, ulong addr, bool w, int i)
+{
+ int emulated = EMULATE_FAIL;
+ struct thread_struct t;
+ int r;
+ float one = 1.0;
+ u32 tmp[2];
+
+ t.fpscr.val = vcpu->arch.fpscr;
+
+ /* read from memory */
+ if (w) {
+ r = kvmppc_ld(vcpu, &addr, sizeof(u32), tmp, true);
+ memcpy(&tmp[1], &one, sizeof(u32));
+ } else {
+ r = kvmppc_ld(vcpu, &addr, sizeof(u32) * 2, tmp, true);
+ }
+ vcpu->arch.paddr_accessed = addr;
+ if (r < 0) {
+ kvmppc_inject_pf(vcpu, addr, false);
+ goto done_load;
+ } else if ((r == EMULATE_DO_MMIO) && w) {
+ emulated = kvmppc_handle_load(run, vcpu, REG_FPR | rs, 4, 1);
+ vcpu->arch.qpr[rs] = tmp[1];
+ goto done_load;
+ } else if (r == EMULATE_DO_MMIO) {
+ emulated = kvmppc_handle_load(run, vcpu, REG_FQPR | rs, 8, 1);
+ goto done_load;
+ }
+
+ emulated = EMULATE_DONE;
+
+ /* put in registers */
+ cvt_fd((float*)&tmp[0], (double*)&vcpu->arch.fpr[rs], &t);
+ vcpu->arch.qpr[rs] = tmp[1];
+
+ dprintk(KERN_INFO "KVM: PSQ_LD [0x%x, 0x%x] at 0x%lx (%d)\n", tmp[0],
+ tmp[1], addr, w ? 4 : 8);
+
+done_load:
+ return emulated;
+}
+
+static int kvmppc_emulate_psq_store(struct kvm_run *run, struct kvm_vcpu *vcpu,
+ int rs, ulong addr, bool w, int i)
+{
+ int emulated = EMULATE_FAIL;
+ struct thread_struct t;
+ int r;
+ u32 tmp[2];
+ int len = w ? sizeof(u32) : sizeof(u64);
+
+ t.fpscr.val = vcpu->arch.fpscr;
+
+ cvt_df((double*)&vcpu->arch.fpr[rs], (float*)&tmp[0], &t);
+ tmp[1] = vcpu->arch.qpr[rs];
+
+ r = kvmppc_st(vcpu, &addr, len, tmp, true);
+ vcpu->arch.paddr_accessed = addr;
+ if (r < 0) {
+ kvmppc_inject_pf(vcpu, addr, true);
+ } else if ((r == EMULATE_DO_MMIO) && w) {
+ emulated = kvmppc_handle_store(run, vcpu, tmp[0], 4, 1);
+ } else if (r == EMULATE_DO_MMIO) {
+ u64 val = ((u64)tmp[0] << 32) | tmp[1];
+ emulated = kvmppc_handle_store(run, vcpu, val, 8, 1);
+ } else {
+ emulated = EMULATE_DONE;
+ }
+
+ dprintk(KERN_INFO "KVM: PSQ_ST [0x%x, 0x%x] at 0x%lx (%d)\n",
+ tmp[0], tmp[1], addr, len);
+
+ return emulated;
+}
+
+/*
+ * Cuts out inst bits with ordering according to spec.
+ * That means the leftmost bit is zero. All given bits are included.
+ */
+static inline u32 inst_get_field(u32 inst, int msb, int lsb)
+{
+ return kvmppc_get_field(inst, msb + 32, lsb + 32);
+}
+
+/*
+ * Replaces inst bits with ordering according to spec.
+ */
+static inline u32 inst_set_field(u32 inst, int msb, int lsb, int value)
+{
+ return kvmppc_set_field(inst, msb + 32, lsb + 32, value);
+}
+
+#define FPU_HAS_FRT (1 << 0)
+#define FPU_HAS_FRA (1 << 1)
+#define FPU_HAS_FRB (1 << 2)
+#define FPU_HAS_FRC (1 << 3)
+#define FPU_SYNC_QPR (1 << 4)
+
+static void kvmppc_do_fpu_inst(struct kvm_vcpu *vcpu, u32 inst, int flags)
+{
+ int frt = inst_get_field(inst, 6, 10);
+
+ u64 *fpr_a = NULL;
+ u64 *fpr_b = NULL;
+ u64 *fpr_c = NULL;
+ u64 *fpr_t = NULL;
+
+ if (flags & FPU_HAS_FRA) {
+ int fra = inst_get_field(inst, 11, 15);
+ fpr_a = &vcpu->arch.fpr[fra];
+ inst = inst_set_field(inst, 11, 15, 1);
+ }
+ if (flags & FPU_HAS_FRB) {
+ int frb = inst_get_field(inst, 16, 20);
+ fpr_b = &vcpu->arch.fpr[frb];
+ inst = inst_set_field(inst, 16, 20, 2);
+ }
+ if (flags & FPU_HAS_FRC) {
+ int frc = inst_get_field(inst, 21, 25);
+ fpr_c = &vcpu->arch.fpr[frc];
+ inst = inst_set_field(inst, 21, 25, 3);
+ }
+ if (flags & FPU_HAS_FRT) {
+ fpr_t = &vcpu->arch.fpr[frt];
+ inst = inst_set_field(inst, 6, 10, 4);
+ }
+
+ call_fpu_inst(inst, fpr_t, fpr_a, fpr_b, fpr_c,
+ &get_paca()->shadow_vcpu.cr,
+ &vcpu->arch.fpscr);
+
+ barrier();
+
+ if (flags & FPU_SYNC_QPR) {
+ struct thread_struct t;
+
+ t.fpscr.val = vcpu->arch.fpscr;
+ cvt_df((double*)fpr_t, (float*)&vcpu->arch.qpr[frt], &t);
+ }
+}
+
+bool kvmppc_inst_is_paired_single(struct kvm_vcpu *vcpu, u32 inst)
+{
+ if (!(vcpu->arch.hflags & BOOK3S_HFLAG_PAIRED_SINGLE))
+ return false;
+
+ switch (get_op(inst)) {
+ case OP_PSQ_L:
+ case OP_PSQ_LU:
+ case OP_PSQ_ST:
+ case OP_PSQ_STU:
+ case OP_LFS:
+ case OP_LFSU:
+ case OP_LFD:
+ case OP_LFDU:
+ case OP_STFS:
+ case OP_STFSU:
+ case OP_STFD:
+ case OP_STFDU:
+ return true;
+ case 4:
+ /* X form */
+ switch (inst_get_field(inst, 21, 30)) {
+ case OP_4X_PS_CMPU0:
+ case OP_4X_PSQ_LX:
+ case OP_4X_PS_CMPO0:
+ case OP_4X_PSQ_LUX:
+ case OP_4X_PS_NEG:
+ case OP_4X_PS_CMPU1:
+ case OP_4X_PS_MR:
+ case OP_4X_PS_CMPO1:
+ case OP_4X_PS_NABS:
+ case OP_4X_PS_ABS:
+ case OP_4X_PS_MERGE00:
+ case OP_4X_PS_MERGE01:
+ case OP_4X_PS_MERGE10:
+ case OP_4X_PS_MERGE11:
+ return true;
+ }
+ /* XW form */
+ switch (inst_get_field(inst, 25, 30)) {
+ case OP_4XW_PSQ_STX:
+ case OP_4XW_PSQ_STUX:
+ return true;
+ }
+ /* A form */
+ switch (inst_get_field(inst, 26, 30)) {
+ case OP_4A_PS_SUM1:
+ case OP_4A_PS_SUM0:
+ case OP_4A_PS_MULS0:
+ case OP_4A_PS_MULS1:
+ case OP_4A_PS_MADDS0:
+ case OP_4A_PS_MADDS1:
+ case OP_4A_PS_DIV:
+ case OP_4A_PS_SUB:
+ case OP_4A_PS_ADD:
+ case OP_4A_PS_SEL:
+ case OP_4A_PS_RES:
+ case OP_4A_PS_MUL:
+ case OP_4A_PS_RSQRTE:
+ case OP_4A_PS_MSUB:
+ case OP_4A_PS_MADD:
+ case OP_4A_PS_NMSUB:
+ case OP_4A_PS_NMADD:
+ return true;
+ }
+ break;
+ case 59:
+ switch (inst_get_field(inst, 21, 30)) {
+ case OP_59_FADDS:
+ case OP_59_FSUBS:
+ case OP_59_FDIVS:
+ case OP_59_FRES:
+ case OP_59_FRSQRTES:
+ return true;
+ }
+ switch (inst_get_field(inst, 26, 30)) {
+ case OP_59_FMULS:
+ case OP_59_FMSUBS:
+ case OP_59_FMADDS:
+ case OP_59_FNMSUBS:
+ case OP_59_FNMADDS:
+ return true;
+ }
+ break;
+ case 63:
+ switch (inst_get_field(inst, 21, 30)) {
+ case OP_63_MTFSB0:
+ case OP_63_MTFSB1:
+ case OP_63_MTFSF:
+ case OP_63_MTFSFI:
+ case OP_63_MCRFS:
+ case OP_63_MFFS:
+ case OP_63_FCMPU:
+ case OP_63_FCMPO:
+ case OP_63_FNEG:
+ case OP_63_FMR:
+ case OP_63_FABS:
+ case OP_63_FRSP:
+ case OP_63_FDIV:
+ case OP_63_FADD:
+ case OP_63_FSUB:
+ case OP_63_FCTIW:
+ case OP_63_FCTIWZ:
+ case OP_63_FRSQRTE:
+ case OP_63_FCPSGN:
+ return true;
+ }
+ switch (inst_get_field(inst, 26, 30)) {
+ case OP_63_FMUL:
+ case OP_63_FSEL:
+ case OP_63_FMSUB:
+ case OP_63_FMADD:
+ case OP_63_FNMSUB:
+ case OP_63_FNMADD:
+ return true;
+ }
+ break;
+ case 31:
+ switch (inst_get_field(inst, 21, 30)) {
+ case OP_31_LFSX:
+ case OP_31_LFSUX:
+ case OP_31_LFDX:
+ case OP_31_LFDUX:
+ case OP_31_STFSX:
+ case OP_31_STFSUX:
+ case OP_31_STFX:
+ case OP_31_STFUX:
+ case OP_31_STFIWX:
+ return true;
+ }
+ break;
+ }
+
+ return false;
+}
+
+static int get_d_signext(u32 inst)
+{
+ int d = inst & 0x8ff;
+
+ if (d & 0x800)
+ return -(d & 0x7ff);
+
+ return (d & 0x7ff);
+}
+
+static int kvmppc_ps_three_in(struct kvm_vcpu *vcpu, bool rc,
+ int reg_out, int reg_in1, int reg_in2,
+ int reg_in3, int scalar,
+ void (*func)(struct thread_struct *t,
+ u32 *dst, u32 *src1,
+ u32 *src2, u32 *src3))
+{
+ u32 *qpr = vcpu->arch.qpr;
+ u64 *fpr = vcpu->arch.fpr;
+ u32 ps0_out;
+ u32 ps0_in1, ps0_in2, ps0_in3;
+ u32 ps1_in1, ps1_in2, ps1_in3;
+ struct thread_struct t;
+ t.fpscr.val = vcpu->arch.fpscr;
+
+ /* RC */
+ WARN_ON(rc);
+
+ /* PS0 */
+ cvt_df((double*)&fpr[reg_in1], (float*)&ps0_in1, &t);
+ cvt_df((double*)&fpr[reg_in2], (float*)&ps0_in2, &t);
+ cvt_df((double*)&fpr[reg_in3], (float*)&ps0_in3, &t);
+
+ if (scalar & SCALAR_LOW)
+ ps0_in2 = qpr[reg_in2];
+
+ func(&t, &ps0_out, &ps0_in1, &ps0_in2, &ps0_in3);
+
+ dprintk(KERN_INFO "PS3 ps0 -> f(0x%x, 0x%x, 0x%x) = 0x%x\n",
+ ps0_in1, ps0_in2, ps0_in3, ps0_out);
+
+ if (!(scalar & SCALAR_NO_PS0))
+ cvt_fd((float*)&ps0_out, (double*)&fpr[reg_out], &t);
+
+ /* PS1 */
+ ps1_in1 = qpr[reg_in1];
+ ps1_in2 = qpr[reg_in2];
+ ps1_in3 = qpr[reg_in3];
+
+ if (scalar & SCALAR_HIGH)
+ ps1_in2 = ps0_in2;
+
+ if (!(scalar & SCALAR_NO_PS1))
+ func(&t, &qpr[reg_out], &ps1_in1, &ps1_in2, &ps1_in3);
+
+ dprintk(KERN_INFO "PS3 ps1 -> f(0x%x, 0x%x, 0x%x) = 0x%x\n",
+ ps1_in1, ps1_in2, ps1_in3, qpr[reg_out]);
+
+ return EMULATE_DONE;
+}
+
+static int kvmppc_ps_two_in(struct kvm_vcpu *vcpu, bool rc,
+ int reg_out, int reg_in1, int reg_in2,
+ int scalar,
+ void (*func)(struct thread_struct *t,
+ u32 *dst, u32 *src1,
+ u32 *src2))
+{
+ u32 *qpr = vcpu->arch.qpr;
+ u64 *fpr = vcpu->arch.fpr;
+ u32 ps0_out;
+ u32 ps0_in1, ps0_in2;
+ u32 ps1_out;
+ u32 ps1_in1, ps1_in2;
+ struct thread_struct t;
+ t.fpscr.val = vcpu->arch.fpscr;
+
+ /* RC */
+ WARN_ON(rc);
+
+ /* PS0 */
+ cvt_df((double*)&fpr[reg_in1], (float*)&ps0_in1, &t);
+
+ if (scalar & SCALAR_LOW)
+ ps0_in2 = qpr[reg_in2];
+ else
+ cvt_df((double*)&fpr[reg_in2], (float*)&ps0_in2, &t);
+
+ func(&t, &ps0_out, &ps0_in1, &ps0_in2);
+
+ if (!(scalar & SCALAR_NO_PS0)) {
+ dprintk(KERN_INFO "PS2 ps0 -> f(0x%x, 0x%x) = 0x%x\n",
+ ps0_in1, ps0_in2, ps0_out);
+
+ cvt_fd((float*)&ps0_out, (double*)&fpr[reg_out], &t);
+ }
+
+ /* PS1 */
+ ps1_in1 = qpr[reg_in1];
+ ps1_in2 = qpr[reg_in2];
+
+ if (scalar & SCALAR_HIGH)
+ ps1_in2 = ps0_in2;
+
+ func(&t, &ps1_out, &ps1_in1, &ps1_in2);
+
+ if (!(scalar & SCALAR_NO_PS1)) {
+ qpr[reg_out] = ps1_out;
+
+ dprintk(KERN_INFO "PS2 ps1 -> f(0x%x, 0x%x) = 0x%x\n",
+ ps1_in1, ps1_in2, qpr[reg_out]);
+ }
+
+ return EMULATE_DONE;
+}
+
+static int kvmppc_ps_one_in(struct kvm_vcpu *vcpu, bool rc,
+ int reg_out, int reg_in,
+ void (*func)(struct thread_struct *t,
+ u32 *dst, u32 *src1))
+{
+ u32 *qpr = vcpu->arch.qpr;
+ u64 *fpr = vcpu->arch.fpr;
+ u32 ps0_out, ps0_in;
+ u32 ps1_in;
+ struct thread_struct t;
+ t.fpscr.val = vcpu->arch.fpscr;
+
+ /* RC */
+ WARN_ON(rc);
+
+ /* PS0 */
+ cvt_df((double*)&fpr[reg_in], (float*)&ps0_in, &t);
+ func(&t, &ps0_out, &ps0_in);
+
+ dprintk(KERN_INFO "PS1 ps0 -> f(0x%x) = 0x%x\n",
+ ps0_in, ps0_out);
+
+ cvt_fd((float*)&ps0_out, (double*)&fpr[reg_out], &t);
+
+ /* PS1 */
+ ps1_in = qpr[reg_in];
+ func(&t, &qpr[reg_out], &ps1_in);
+
+ dprintk(KERN_INFO "PS1 ps1 -> f(0x%x) = 0x%x\n",
+ ps1_in, qpr[reg_out]);
+
+ return EMULATE_DONE;
+}
+
+int kvmppc_emulate_paired_single(struct kvm_run *run, struct kvm_vcpu *vcpu)
+{
+ u32 inst = vcpu->arch.last_inst;
+ enum emulation_result emulated = EMULATE_DONE;
+
+ int ax_rd = inst_get_field(inst, 6, 10);
+ int ax_ra = inst_get_field(inst, 11, 15);
+ int ax_rb = inst_get_field(inst, 16, 20);
+ int ax_rc = inst_get_field(inst, 21, 25);
+ short full_d = inst_get_field(inst, 16, 31);
+
+ bool rcomp = (inst & 1) ? true : false;
+ struct thread_struct t;
+#ifdef DEBUG
+ int i;
+#endif
+
+ t.fpscr.val = vcpu->arch.fpscr;
+
+ if (!kvmppc_inst_is_paired_single(vcpu, inst))
+ return EMULATE_FAIL;
+
+ if (!(vcpu->arch.msr & MSR_FP)) {
+ kvmppc_book3s_queue_irqprio(vcpu, BOOK3S_INTERRUPT_FP_UNAVAIL);
+ return EMULATE_AGAIN;
+ }
+
+ kvmppc_giveup_ext(vcpu, MSR_FP);
+ preempt_disable();
+ enable_kernel_fp();
+ /* Do we need to clear FE0 / FE1 here? Don't think so. */
+
+#ifdef DEBUG
+ for (i = 0; i < ARRAY_SIZE(vcpu->arch.fpr); i++) {
+ u32 f;
+ cvt_df((double*)&vcpu->arch.fpr[i], (float*)&f, &t);
+ dprintk(KERN_INFO "FPR[%d] = 0x%x / 0x%llx QPR[%d] = 0x%x\n",
+ i, f, vcpu->arch.fpr[i], i, vcpu->arch.qpr[i]);
+ }
+#endif
+
+ switch (get_op(inst)) {
+ case OP_PSQ_L:
+ {
+ ulong addr = ax_ra ? kvmppc_get_gpr(vcpu, ax_ra) : 0;
+ bool w = inst_get_field(inst, 16, 16) ? true : false;
+ int i = inst_get_field(inst, 17, 19);
+
+ addr += get_d_signext(inst);
+ emulated = kvmppc_emulate_psq_load(run, vcpu, ax_rd, addr, w, i);
+ break;
+ }
+ case OP_PSQ_LU:
+ {
+ ulong addr = kvmppc_get_gpr(vcpu, ax_ra);
+ bool w = inst_get_field(inst, 16, 16) ? true : false;
+ int i = inst_get_field(inst, 17, 19);
+
+ addr += get_d_signext(inst);
+ emulated = kvmppc_emulate_psq_load(run, vcpu, ax_rd, addr, w, i);
+
+ if (emulated == EMULATE_DONE)
+ kvmppc_set_gpr(vcpu, ax_ra, addr);
+ break;
+ }
+ case OP_PSQ_ST:
+ {
+ ulong addr = ax_ra ? kvmppc_get_gpr(vcpu, ax_ra) : 0;
+ bool w = inst_get_field(inst, 16, 16) ? true : false;
+ int i = inst_get_field(inst, 17, 19);
+
+ addr += get_d_signext(inst);
+ emulated = kvmppc_emulate_psq_store(run, vcpu, ax_rd, addr, w, i);
+ break;
+ }
+ case OP_PSQ_STU:
+ {
+ ulong addr = kvmppc_get_gpr(vcpu, ax_ra);
+ bool w = inst_get_field(inst, 16, 16) ? true : false;
+ int i = inst_get_field(inst, 17, 19);
+
+ addr += get_d_signext(inst);
+ emulated = kvmppc_emulate_psq_store(run, vcpu, ax_rd, addr, w, i);
+
+ if (emulated == EMULATE_DONE)
+ kvmppc_set_gpr(vcpu, ax_ra, addr);
+ break;
+ }
+ case 4:
+ /* X form */
+ switch (inst_get_field(inst, 21, 30)) {
+ case OP_4X_PS_CMPU0:
+ /* XXX */
+ emulated = EMULATE_FAIL;
+ break;
+ case OP_4X_PSQ_LX:
+ {
+ ulong addr = ax_ra ? kvmppc_get_gpr(vcpu, ax_ra) : 0;
+ bool w = inst_get_field(inst, 21, 21) ? true : false;
+ int i = inst_get_field(inst, 22, 24);
+
+ addr += kvmppc_get_gpr(vcpu, ax_rb);
+ emulated = kvmppc_emulate_psq_load(run, vcpu, ax_rd, addr, w, i);
+ break;
+ }
+ case OP_4X_PS_CMPO0:
+ /* XXX */
+ emulated = EMULATE_FAIL;
+ break;
+ case OP_4X_PSQ_LUX:
+ {
+ ulong addr = kvmppc_get_gpr(vcpu, ax_ra);
+ bool w = inst_get_field(inst, 21, 21) ? true : false;
+ int i = inst_get_field(inst, 22, 24);
+
+ addr += kvmppc_get_gpr(vcpu, ax_rb);
+ emulated = kvmppc_emulate_psq_load(run, vcpu, ax_rd, addr, w, i);
+
+ if (emulated == EMULATE_DONE)
+ kvmppc_set_gpr(vcpu, ax_ra, addr);
+ break;
+ }
+ case OP_4X_PS_NEG:
+ vcpu->arch.fpr[ax_rd] = vcpu->arch.fpr[ax_rb];
+ vcpu->arch.fpr[ax_rd] ^= 0x8000000000000000ULL;
+ vcpu->arch.qpr[ax_rd] = vcpu->arch.qpr[ax_rb];
+ vcpu->arch.qpr[ax_rd] ^= 0x80000000;
+ break;
+ case OP_4X_PS_CMPU1:
+ /* XXX */
+ emulated = EMULATE_FAIL;
+ break;
+ case OP_4X_PS_MR:
+ WARN_ON(rcomp);
+ vcpu->arch.fpr[ax_rd] = vcpu->arch.fpr[ax_rb];
+ vcpu->arch.qpr[ax_rd] = vcpu->arch.qpr[ax_rb];
+ break;
+ case OP_4X_PS_CMPO1:
+ /* XXX */
+ emulated = EMULATE_FAIL;
+ break;
+ case OP_4X_PS_NABS:
+ WARN_ON(rcomp);
+ vcpu->arch.fpr[ax_rd] = vcpu->arch.fpr[ax_rb];
+ vcpu->arch.fpr[ax_rd] |= 0x8000000000000000ULL;
+ vcpu->arch.qpr[ax_rd] = vcpu->arch.qpr[ax_rb];
+ vcpu->arch.qpr[ax_rd] |= 0x80000000;
+ break;
+ case OP_4X_PS_ABS:
+ WARN_ON(rcomp);
+ vcpu->arch.fpr[ax_rd] = vcpu->arch.fpr[ax_rb];
+ vcpu->arch.fpr[ax_rd] &= ~0x8000000000000000ULL;
+ vcpu->arch.qpr[ax_rd] = vcpu->arch.qpr[ax_rb];
+ vcpu->arch.qpr[ax_rd] &= ~0x80000000;
+ break;
+ case OP_4X_PS_MERGE00:
+ WARN_ON(rcomp);
+ vcpu->arch.fpr[ax_rd] = vcpu->arch.fpr[ax_ra];
+ /* vcpu->arch.qpr[ax_rd] = vcpu->arch.fpr[ax_rb]; */
+ cvt_df((double*)&vcpu->arch.fpr[ax_rb],
+ (float*)&vcpu->arch.qpr[ax_rd], &t);
+ break;
+ case OP_4X_PS_MERGE01:
+ WARN_ON(rcomp);
+ vcpu->arch.fpr[ax_rd] = vcpu->arch.fpr[ax_ra];
+ vcpu->arch.qpr[ax_rd] = vcpu->arch.qpr[ax_rb];
+ break;
+ case OP_4X_PS_MERGE10:
+ WARN_ON(rcomp);
+ /* vcpu->arch.fpr[ax_rd] = vcpu->arch.qpr[ax_ra]; */
+ cvt_fd((float*)&vcpu->arch.qpr[ax_ra],
+ (double*)&vcpu->arch.fpr[ax_rd], &t);
+ /* vcpu->arch.qpr[ax_rd] = vcpu->arch.fpr[ax_rb]; */
+ cvt_df((double*)&vcpu->arch.fpr[ax_rb],
+ (float*)&vcpu->arch.qpr[ax_rd], &t);
+ break;
+ case OP_4X_PS_MERGE11:
+ WARN_ON(rcomp);
+ /* vcpu->arch.fpr[ax_rd] = vcpu->arch.qpr[ax_ra]; */
+ cvt_fd((float*)&vcpu->arch.qpr[ax_ra],
+ (double*)&vcpu->arch.fpr[ax_rd], &t);
+ vcpu->arch.qpr[ax_rd] = vcpu->arch.qpr[ax_rb];
+ break;
+ }
+ /* XW form */
+ switch (inst_get_field(inst, 25, 30)) {
+ case OP_4XW_PSQ_STX:
+ {
+ ulong addr = ax_ra ? kvmppc_get_gpr(vcpu, ax_ra) : 0;
+ bool w = inst_get_field(inst, 21, 21) ? true : false;
+ int i = inst_get_field(inst, 22, 24);
+
+ addr += kvmppc_get_gpr(vcpu, ax_rb);
+ emulated = kvmppc_emulate_psq_store(run, vcpu, ax_rd, addr, w, i);
+ break;
+ }
+ case OP_4XW_PSQ_STUX:
+ {
+ ulong addr = kvmppc_get_gpr(vcpu, ax_ra);
+ bool w = inst_get_field(inst, 21, 21) ? true : false;
+ int i = inst_get_field(inst, 22, 24);
+
+ addr += kvmppc_get_gpr(vcpu, ax_rb);
+ emulated = kvmppc_emulate_psq_store(run, vcpu, ax_rd, addr, w, i);
+
+ if (emulated == EMULATE_DONE)
+ kvmppc_set_gpr(vcpu, ax_ra, addr);
+ break;
+ }
+ }
+ /* A form */
+ switch (inst_get_field(inst, 26, 30)) {
+ case OP_4A_PS_SUM1:
+ emulated = kvmppc_ps_two_in(vcpu, rcomp, ax_rd,
+ ax_rb, ax_ra, SCALAR_NO_PS0 | SCALAR_HIGH, fp_fadds);
+ vcpu->arch.fpr[ax_rd] = vcpu->arch.fpr[ax_rc];
+ break;
+ case OP_4A_PS_SUM0:
+ emulated = kvmppc_ps_two_in(vcpu, rcomp, ax_rd,
+ ax_ra, ax_rb, SCALAR_NO_PS1 | SCALAR_LOW, fp_fadds);
+ vcpu->arch.qpr[ax_rd] = vcpu->arch.qpr[ax_rc];
+ break;
+ case OP_4A_PS_MULS0:
+ emulated = kvmppc_ps_two_in(vcpu, rcomp, ax_rd,
+ ax_ra, ax_rc, SCALAR_HIGH, fp_fmuls);
+ break;
+ case OP_4A_PS_MULS1:
+ emulated = kvmppc_ps_two_in(vcpu, rcomp, ax_rd,
+ ax_ra, ax_rc, SCALAR_LOW, fp_fmuls);
+ break;
+ case OP_4A_PS_MADDS0:
+ emulated = kvmppc_ps_three_in(vcpu, rcomp, ax_rd,
+ ax_ra, ax_rc, ax_rb, SCALAR_HIGH, fp_fmadds);
+ break;
+ case OP_4A_PS_MADDS1:
+ emulated = kvmppc_ps_three_in(vcpu, rcomp, ax_rd,
+ ax_ra, ax_rc, ax_rb, SCALAR_LOW, fp_fmadds);
+ break;
+ case OP_4A_PS_DIV:
+ emulated = kvmppc_ps_two_in(vcpu, rcomp, ax_rd,
+ ax_ra, ax_rb, SCALAR_NONE, fp_fdivs);
+ break;
+ case OP_4A_PS_SUB:
+ emulated = kvmppc_ps_two_in(vcpu, rcomp, ax_rd,
+ ax_ra, ax_rb, SCALAR_NONE, fp_fsubs);
+ break;
+ case OP_4A_PS_ADD:
+ emulated = kvmppc_ps_two_in(vcpu, rcomp, ax_rd,
+ ax_ra, ax_rb, SCALAR_NONE, fp_fadds);
+ break;
+ case OP_4A_PS_SEL:
+ emulated = kvmppc_ps_three_in(vcpu, rcomp, ax_rd,
+ ax_ra, ax_rc, ax_rb, SCALAR_NONE, fp_fsel);
+ break;
+ case OP_4A_PS_RES:
+ emulated = kvmppc_ps_one_in(vcpu, rcomp, ax_rd,
+ ax_rb, fp_fres);
+ break;
+ case OP_4A_PS_MUL:
+ emulated = kvmppc_ps_two_in(vcpu, rcomp, ax_rd,
+ ax_ra, ax_rc, SCALAR_NONE, fp_fmuls);
+ break;
+ case OP_4A_PS_RSQRTE:
+ emulated = kvmppc_ps_one_in(vcpu, rcomp, ax_rd,
+ ax_rb, fp_frsqrte);
+ break;
+ case OP_4A_PS_MSUB:
+ emulated = kvmppc_ps_three_in(vcpu, rcomp, ax_rd,
+ ax_ra, ax_rc, ax_rb, SCALAR_NONE, fp_fmsubs);
+ break;
+ case OP_4A_PS_MADD:
+ emulated = kvmppc_ps_three_in(vcpu, rcomp, ax_rd,
+ ax_ra, ax_rc, ax_rb, SCALAR_NONE, fp_fmadds);
+ break;
+ case OP_4A_PS_NMSUB:
+ emulated = kvmppc_ps_three_in(vcpu, rcomp, ax_rd,
+ ax_ra, ax_rc, ax_rb, SCALAR_NONE, fp_fnmsubs);
+ break;
+ case OP_4A_PS_NMADD:
+ emulated = kvmppc_ps_three_in(vcpu, rcomp, ax_rd,
+ ax_ra, ax_rc, ax_rb, SCALAR_NONE, fp_fnmadds);
+ break;
+ }
+ break;
+
+ /* Real FPU operations */
+
+ case OP_LFS:
+ {
+ ulong addr = (ax_ra ? kvmppc_get_gpr(vcpu, ax_ra) : 0) + full_d;
+
+ emulated = kvmppc_emulate_fpr_load(run, vcpu, ax_rd, addr,
+ FPU_LS_SINGLE);
+ break;
+ }
+ case OP_LFSU:
+ {
+ ulong addr = kvmppc_get_gpr(vcpu, ax_ra) + full_d;
+
+ emulated = kvmppc_emulate_fpr_load(run, vcpu, ax_rd, addr,
+ FPU_LS_SINGLE);
+
+ if (emulated == EMULATE_DONE)
+ kvmppc_set_gpr(vcpu, ax_ra, addr);
+ break;
+ }
+ case OP_LFD:
+ {
+ ulong addr = (ax_ra ? kvmppc_get_gpr(vcpu, ax_ra) : 0) + full_d;
+
+ emulated = kvmppc_emulate_fpr_load(run, vcpu, ax_rd, addr,
+ FPU_LS_DOUBLE);
+ break;
+ }
+ case OP_LFDU:
+ {
+ ulong addr = kvmppc_get_gpr(vcpu, ax_ra) + full_d;
+
+ emulated = kvmppc_emulate_fpr_load(run, vcpu, ax_rd, addr,
+ FPU_LS_DOUBLE);
+
+ if (emulated == EMULATE_DONE)
+ kvmppc_set_gpr(vcpu, ax_ra, addr);
+ break;
+ }
+ case OP_STFS:
+ {
+ ulong addr = (ax_ra ? kvmppc_get_gpr(vcpu, ax_ra) : 0) + full_d;
+
+ emulated = kvmppc_emulate_fpr_store(run, vcpu, ax_rd, addr,
+ FPU_LS_SINGLE);
+ break;
+ }
+ case OP_STFSU:
+ {
+ ulong addr = kvmppc_get_gpr(vcpu, ax_ra) + full_d;
+
+ emulated = kvmppc_emulate_fpr_store(run, vcpu, ax_rd, addr,
+ FPU_LS_SINGLE);
+
+ if (emulated == EMULATE_DONE)
+ kvmppc_set_gpr(vcpu, ax_ra, addr);
+ break;
+ }
+ case OP_STFD:
+ {
+ ulong addr = (ax_ra ? kvmppc_get_gpr(vcpu, ax_ra) : 0) + full_d;
+
+ emulated = kvmppc_emulate_fpr_store(run, vcpu, ax_rd, addr,
+ FPU_LS_DOUBLE);
+ break;
+ }
+ case OP_STFDU:
+ {
+ ulong addr = kvmppc_get_gpr(vcpu, ax_ra) + full_d;
+
+ emulated = kvmppc_emulate_fpr_store(run, vcpu, ax_rd, addr,
+ FPU_LS_DOUBLE);
+
+ if (emulated == EMULATE_DONE)
+ kvmppc_set_gpr(vcpu, ax_ra, addr);
+ break;
+ }
+ case 31:
+ switch (inst_get_field(inst, 21, 30)) {
+ case OP_31_LFSX:
+ {
+ ulong addr = ax_ra ? kvmppc_get_gpr(vcpu, ax_ra) : 0;
+
+ addr += kvmppc_get_gpr(vcpu, ax_rb);
+ emulated = kvmppc_emulate_fpr_load(run, vcpu, ax_rd,
+ addr, FPU_LS_SINGLE);
+ break;
+ }
+ case OP_31_LFSUX:
+ {
+ ulong addr = kvmppc_get_gpr(vcpu, ax_ra) +
+ kvmppc_get_gpr(vcpu, ax_rb);
+
+ emulated = kvmppc_emulate_fpr_load(run, vcpu, ax_rd,
+ addr, FPU_LS_SINGLE);
+
+ if (emulated == EMULATE_DONE)
+ kvmppc_set_gpr(vcpu, ax_ra, addr);
+ break;
+ }
+ case OP_31_LFDX:
+ {
+ ulong addr = (ax_ra ? kvmppc_get_gpr(vcpu, ax_ra) : 0) +
+ kvmppc_get_gpr(vcpu, ax_rb);
+
+ emulated = kvmppc_emulate_fpr_load(run, vcpu, ax_rd,
+ addr, FPU_LS_DOUBLE);
+ break;
+ }
+ case OP_31_LFDUX:
+ {
+ ulong addr = kvmppc_get_gpr(vcpu, ax_ra) +
+ kvmppc_get_gpr(vcpu, ax_rb);
+
+ emulated = kvmppc_emulate_fpr_load(run, vcpu, ax_rd,
+ addr, FPU_LS_DOUBLE);
+
+ if (emulated == EMULATE_DONE)
+ kvmppc_set_gpr(vcpu, ax_ra, addr);
+ break;
+ }
+ case OP_31_STFSX:
+ {
+ ulong addr = (ax_ra ? kvmppc_get_gpr(vcpu, ax_ra) : 0) +
+ kvmppc_get_gpr(vcpu, ax_rb);
+
+ emulated = kvmppc_emulate_fpr_store(run, vcpu, ax_rd,
+ addr, FPU_LS_SINGLE);
+ break;
+ }
+ case OP_31_STFSUX:
+ {
+ ulong addr = kvmppc_get_gpr(vcpu, ax_ra) +
+ kvmppc_get_gpr(vcpu, ax_rb);
+
+ emulated = kvmppc_emulate_fpr_store(run, vcpu, ax_rd,
+ addr, FPU_LS_SINGLE);
+
+ if (emulated == EMULATE_DONE)
+ kvmppc_set_gpr(vcpu, ax_ra, addr);
+ break;
+ }
+ case OP_31_STFX:
+ {
+ ulong addr = (ax_ra ? kvmppc_get_gpr(vcpu, ax_ra) : 0) +
+ kvmppc_get_gpr(vcpu, ax_rb);
+
+ emulated = kvmppc_emulate_fpr_store(run, vcpu, ax_rd,
+ addr, FPU_LS_DOUBLE);
+ break;
+ }
+ case OP_31_STFUX:
+ {
+ ulong addr = kvmppc_get_gpr(vcpu, ax_ra) +
+ kvmppc_get_gpr(vcpu, ax_rb);
+
+ emulated = kvmppc_emulate_fpr_store(run, vcpu, ax_rd,
+ addr, FPU_LS_DOUBLE);
+
+ if (emulated == EMULATE_DONE)
+ kvmppc_set_gpr(vcpu, ax_ra, addr);
+ break;
+ }
+ case OP_31_STFIWX:
+ {
+ ulong addr = (ax_ra ? kvmppc_get_gpr(vcpu, ax_ra) : 0) +
+ kvmppc_get_gpr(vcpu, ax_rb);
+
+ emulated = kvmppc_emulate_fpr_store(run, vcpu, ax_rd,
+ addr,
+ FPU_LS_SINGLE_LOW);
+ break;
+ }
+ break;
+ }
+ break;
+ case 59:
+ switch (inst_get_field(inst, 21, 30)) {
+ case OP_59_FADDS:
+ case OP_59_FSUBS:
+ case OP_59_FDIVS:
+ kvmppc_do_fpu_inst(vcpu, inst, FPU_HAS_FRA |
+ FPU_HAS_FRB | FPU_HAS_FRT |
+ FPU_SYNC_QPR);
+ break;
+ case OP_59_FRES:
+ case OP_59_FRSQRTES:
+ kvmppc_do_fpu_inst(vcpu, inst, FPU_HAS_FRT |
+ FPU_HAS_FRB | FPU_SYNC_QPR);
+ break;
+ }
+ switch (inst_get_field(inst, 26, 30)) {
+ case OP_59_FMULS:
+ kvmppc_do_fpu_inst(vcpu, inst, FPU_HAS_FRA |
+ FPU_HAS_FRC | FPU_HAS_FRT |
+ FPU_SYNC_QPR);
+ break;
+ case OP_59_FMSUBS:
+ case OP_59_FMADDS:
+ case OP_59_FNMSUBS:
+ case OP_59_FNMADDS:
+ kvmppc_do_fpu_inst(vcpu, inst, FPU_HAS_FRA |
+ FPU_HAS_FRB | FPU_HAS_FRC |
+ FPU_HAS_FRT | FPU_SYNC_QPR);
+ break;
+ }
+ break;
+ case 63:
+ switch (inst_get_field(inst, 21, 30)) {
+ case OP_63_MTFSB0:
+ case OP_63_MTFSB1:
+ case OP_63_MCRFS:
+ case OP_63_MTFSFI:
+ kvmppc_do_fpu_inst(vcpu, inst, 0);
+ break;
+ case OP_63_MFFS:
+ kvmppc_do_fpu_inst(vcpu, inst, FPU_HAS_FRT);
+ break;
+ case OP_63_MTFSF:
+ kvmppc_do_fpu_inst(vcpu, inst, FPU_HAS_FRB);
+ break;
+ case OP_63_FCMPU:
+ case OP_63_FCMPO:
+ kvmppc_do_fpu_inst(vcpu, inst, FPU_HAS_FRA |
+ FPU_HAS_FRB);
+ break;
+ case OP_63_FNEG:
+ case OP_63_FMR:
+ case OP_63_FABS:
+ kvmppc_do_fpu_inst(vcpu, inst, FPU_HAS_FRT |
+ FPU_HAS_FRB);
+ break;
+ case OP_63_FCPSGN:
+ case OP_63_FDIV:
+ case OP_63_FADD:
+ case OP_63_FSUB:
+ kvmppc_do_fpu_inst(vcpu, inst, FPU_HAS_FRA |
+ FPU_HAS_FRB | FPU_HAS_FRT);
+ break;
+ case OP_63_FCTIW:
+ case OP_63_FCTIWZ:
+ kvmppc_do_fpu_inst(vcpu, inst, FPU_HAS_FRB |
+ FPU_HAS_FRT);
+ break;
+ case OP_63_FRSP:
+ kvmppc_do_fpu_inst(vcpu, inst, FPU_HAS_FRB |
+ FPU_HAS_FRT | FPU_SYNC_QPR);
+ break;
+ case OP_63_FRSQRTE:
+ {
+ float f, one = 1.0f;
+
+ cvt_df((double*)&vcpu->arch.fpr[ax_rb], &f, &t);
+ /* f = sqrt(f) */
+ fp_fsqrts(&t, (u32*)&f, (u32*)&f);
+ /* f = 1.0f / f */
+ fp_fdivs(&t, (u32*)&f, (u32*)&one, (u32*)&f);
+ cvt_fd(&f, (double*)&vcpu->arch.fpr[ax_rd], &t);
+ break;
+ }
+ }
+ switch (inst_get_field(inst, 26, 30)) {
+ case OP_63_FMUL:
+ kvmppc_do_fpu_inst(vcpu, inst, FPU_HAS_FRA |
+ FPU_HAS_FRC | FPU_HAS_FRT);
+ break;
+ case OP_63_FSEL:
+ case OP_63_FMSUB:
+ case OP_63_FMADD:
+ case OP_63_FNMSUB:
+ case OP_63_FNMADD:
+ kvmppc_do_fpu_inst(vcpu, inst, FPU_HAS_FRA |
+ FPU_HAS_FRB | FPU_HAS_FRC |
+ FPU_HAS_FRT);
+ break;
+ }
+ break;
+ }
+
+#ifdef DEBUG
+ for (i = 0; i < ARRAY_SIZE(vcpu->arch.fpr); i++) {
+ u32 f;
+ cvt_df((double*)&vcpu->arch.fpr[i], (float*)&f, &t);
+ dprintk(KERN_INFO "FPR[%d] = 0x%x\n", i, f);
+ }
+#endif
+
+ preempt_enable();
+
+ return emulated;
+}