aboutsummaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
authoraurel32 <aurel32@c046a42c-6fe2-441c-8c8c-71466251a162>2008-04-12 20:14:54 +0000
committeraurel32 <aurel32@c046a42c-6fe2-441c-8c8c-71466251a162>2008-04-12 20:14:54 +0000
commitf54b3f920f12fb4fb41e259f793a853860a7d2ec (patch)
treea1fdf84ed97f34461db71ad03cfbc0e146cf5e28
parent339dea2774fbfb7092cc35a061849c0cb448618d (diff)
HPPA (PA-RISC) host support
(Stuart Brady) git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@4199 c046a42c-6fe2-441c-8c8c-71466251a162
-rw-r--r--Makefile.target8
-rwxr-xr-xconfigure7
-rw-r--r--cpu-all.h11
-rw-r--r--cpu-exec.c29
-rw-r--r--dis-asm.h4
-rw-r--r--disas.c2
-rw-r--r--dyngen-exec.h7
-rw-r--r--dyngen.c129
-rw-r--r--hppa-dis.c2834
-rw-r--r--hppa.ld214
-rw-r--r--tcg/hppa/tcg-target.c980
-rw-r--r--tcg/hppa/tcg-target.h204
-rw-r--r--tcg/i386/tcg-target.c3
-rw-r--r--tcg/sparc/tcg-target.c3
-rw-r--r--tcg/tcg-dyngen.c50
-rw-r--r--tcg/tcg.c30
-rw-r--r--tcg/x86_64/tcg-target.c3
17 files changed, 4505 insertions, 13 deletions
diff --git a/Makefile.target b/Makefile.target
index 7bedbf763..9bfee4169 100644
--- a/Makefile.target
+++ b/Makefile.target
@@ -128,6 +128,11 @@ ifeq ($(ARCH),alpha)
CFLAGS+=-msmall-data
endif
+ifeq ($(ARCH),hppa)
+OP_CFLAGS=-O1 -fno-delayed-branch
+BASE_LDFLAGS+=-Wl,-T,$(SRC_PATH)/$(ARCH).ld
+endif
+
ifeq ($(ARCH),ia64)
CFLAGS+=-mno-sdata
OP_CFLAGS+=-mno-sdata
@@ -267,6 +272,9 @@ endif
ifeq ($(findstring sh4, $(TARGET_ARCH) $(ARCH)),sh4)
LIBOBJS+=sh4-dis.o
endif
+ifeq ($(findstring hppa, $(TARGET_BASE_ARCH) $(ARCH)),hppa)
+LIBOBJS+=hppa-dis.o
+endif
ifeq ($(findstring s390, $(TARGET_ARCH) $(ARCH)),s390)
LIBOBJS+=s390-dis.o
endif
diff --git a/configure b/configure
index eaf24e51a..a2f54b83c 100755
--- a/configure
+++ b/configure
@@ -50,6 +50,9 @@ case "$cpu" in
cris)
cpu="cris"
;;
+ parisc|parisc64)
+ cpu="hppa"
+ ;;
ia64)
cpu="ia64"
;;
@@ -576,6 +579,7 @@ else
# if cross compiling, cannot launch a program, so make a static guess
if test "$cpu" = "armv4b" \
+ -o "$cpu" = "hppa" \
-o "$cpu" = "m68k" \
-o "$cpu" = "mips" \
-o "$cpu" = "mips64" \
@@ -865,6 +869,9 @@ elif test "$cpu" = "armv4l" ; then
elif test "$cpu" = "cris" ; then
echo "ARCH=cris" >> $config_mak
echo "#define HOST_CRIS 1" >> $config_h
+elif test "$cpu" = "hppa" ; then
+ echo "ARCH=hppa" >> $config_mak
+ echo "#define HOST_HPPA 1" >> $config_h
elif test "$cpu" = "ia64" ; then
echo "ARCH=ia64" >> $config_mak
echo "#define HOST_IA64 1" >> $config_h
diff --git a/cpu-all.h b/cpu-all.h
index 2a2b1970e..21ed731a5 100644
--- a/cpu-all.h
+++ b/cpu-all.h
@@ -20,7 +20,7 @@
#ifndef CPU_ALL_H
#define CPU_ALL_H
-#if defined(__arm__) || defined(__sparc__) || defined(__mips__)
+#if defined(__arm__) || defined(__sparc__) || defined(__mips__) || defined(__hppa__)
#define WORDS_ALIGNED
#endif
@@ -952,6 +952,15 @@ static inline int64_t cpu_get_real_ticks(void)
return val;
}
+#elif defined(__hppa__)
+
+static inline int64_t cpu_get_real_ticks(void)
+{
+ int val;
+ asm volatile ("mfctl %%cr16, %0" : "=r"(val));
+ return val;
+}
+
#elif defined(__ia64)
static inline int64_t cpu_get_real_ticks(void)
diff --git a/cpu-exec.c b/cpu-exec.c
index a3e6e8e12..a17ea2f73 100644
--- a/cpu-exec.c
+++ b/cpu-exec.c
@@ -657,6 +657,17 @@ int cpu_exec(CPUState *env1)
"o0", "o1", "o2", "o3", "o4", "o5",
"l0", "l1", "l2", "l3", "l4", "l5",
"l6", "l7");
+#elif defined(__hppa__)
+ asm volatile ("ble 0(%%sr4,%1)\n"
+ "copy %%r31,%%r18\n"
+ "copy %%r28,%0\n"
+ : "=r" (T0)
+ : "r" (gen_func)
+ : "r1", "r2", "r3", "r4", "r5", "r6", "r7",
+ "r8", "r9", "r10", "r11", "r12", "r13",
+ "r18", "r19", "r20", "r21", "r22", "r23",
+ "r24", "r25", "r26", "r27", "r28", "r29",
+ "r30", "r31");
#elif defined(__arm__)
asm volatile ("mov pc, %0\n\t"
".global exec_loop\n\t"
@@ -1488,6 +1499,24 @@ int cpu_signal_handler(int host_signum, void *pinfo,
is_write, &uc->uc_sigmask, puc);
}
+#elif defined(__hppa__)
+
+int cpu_signal_handler(int host_signum, void *pinfo,
+ void *puc)
+{
+ struct siginfo *info = pinfo;
+ struct ucontext *uc = puc;
+ unsigned long pc;
+ int is_write;
+
+ pc = uc->uc_mcontext.sc_iaoq[0];
+ /* FIXME: compute is_write */
+ is_write = 0;
+ return handle_cpu_signal(pc, (unsigned long)info->si_addr,
+ is_write,
+ &uc->uc_sigmask, puc);
+}
+
#else
#error host CPU specific signal handler needed
diff --git a/dis-asm.h b/dis-asm.h
index dbe1465fd..de6e9f87e 100644
--- a/dis-asm.h
+++ b/dis-asm.h
@@ -157,6 +157,10 @@ enum bfd_architecture
#define bfd_mach_ppc_7400 7400
bfd_arch_rs6000, /* IBM RS/6000 */
bfd_arch_hppa, /* HP PA RISC */
+#define bfd_mach_hppa10 10
+#define bfd_mach_hppa11 11
+#define bfd_mach_hppa20 20
+#define bfd_mach_hppa20w 25
bfd_arch_d10v, /* Mitsubishi D10V */
bfd_arch_z8k, /* Zilog Z8000 */
#define bfd_mach_z8001 1
diff --git a/disas.c b/disas.c
index ecc15344f..b1192474c 100644
--- a/disas.c
+++ b/disas.c
@@ -279,6 +279,8 @@ void disas(FILE *out, void *code, unsigned long size)
print_insn = print_insn_m68k;
#elif defined(__s390__)
print_insn = print_insn_s390;
+#elif defined(__hppa__)
+ print_insn = print_insn_hppa;
#else
fprintf(out, "0x%lx: Asm output not supported on this arch\n",
(long) code);
diff --git a/dyngen-exec.h b/dyngen-exec.h
index 45a1ece40..1cccec809 100644
--- a/dyngen-exec.h
+++ b/dyngen-exec.h
@@ -124,6 +124,11 @@ extern int printf(const char *, ...);
#define AREG1 "r4"
#define AREG2 "r5"
#define AREG3 "r6"
+#elif defined(__hppa__)
+#define AREG0 "r17"
+#define AREG1 "r14"
+#define AREG2 "r15"
+#define AREG3 "r16"
#elif defined(__mips__)
#define AREG0 "fp"
#define AREG1 "s0"
@@ -279,6 +284,8 @@ extern int __op_jmp0, __op_jmp1, __op_jmp2, __op_jmp3;
#elif defined(__mips__)
#define EXIT_TB() asm volatile ("jr $ra")
#define GOTO_LABEL_PARAM(n) asm volatile (".set noat; la $1, " ASM_NAME(__op_gen_label) #n "; jr $1; .set at")
+#elif defined(__hppa__)
+#define GOTO_LABEL_PARAM(n) asm volatile ("b,n " ASM_NAME(__op_gen_label) #n)
#else
#error unsupported CPU
#endif
diff --git a/dyngen.c b/dyngen.c
index 3b073c18f..e4ea3b137 100644
--- a/dyngen.c
+++ b/dyngen.c
@@ -117,6 +117,13 @@
#define elf_check_arch(x) ((x) == EM_68K)
#define ELF_USES_RELOCA
+#elif defined(HOST_HPPA)
+
+#define ELF_CLASS ELFCLASS32
+#define ELF_ARCH EM_PARISC
+#define elf_check_arch(x) ((x) == EM_PARISC)
+#define ELF_USES_RELOCA
+
#elif defined(HOST_MIPS)
#define ELF_CLASS ELFCLASS32
@@ -1223,7 +1230,7 @@ int get_reloc_expr(char *name, int name_size, const char *sym_name)
snprintf(name, name_size, "param%s", p);
return 1;
} else {
-#ifdef HOST_SPARC
+#if defined(HOST_SPARC) || defined(HOST_HPPA)
if (sym_name[0] == '.')
snprintf(name, name_size,
"(long)(&__dot_%s)",
@@ -1661,6 +1668,43 @@ void gen_code(const char *name, host_ulong offset, host_ulong size,
error("rts expected at the end of %s", name);
copy_size = p - p_start;
}
+#elif defined(HOST_HPPA)
+ {
+ uint8_t *p;
+ p = p_start;
+ while (p < p_end) {
+ uint32_t insn = get32((uint32_t *)p);
+ if (insn == 0x6bc23fd9 || /* stw rp,-14(sp) */
+ insn == 0x08030241 || /* copy r3,r1 */
+ insn == 0x081e0243 || /* copy sp,r3 */
+ (insn & 0xffffc000) == 0x37de0000 || /* ldo x(sp),sp */
+ (insn & 0xffffc000) == 0x6fc10000) /* stwm r1,x(sp) */
+ p += 4;
+ else
+ break;
+ }
+ start_offset += p - p_start;
+ p_start = p;
+ p = p_end - 4;
+
+ while (p > p_start) {
+ uint32_t insn = get32((uint32_t *)p);
+ if ((insn & 0xffffc000) == 0x347e0000 || /* ldo x(r3),sp */
+ (insn & 0xffe0c000) == 0x4fc00000 || /* ldwm x(sp),rx */
+ (insn & 0xffffc000) == 0x37de0000 || /* ldo x(sp),sp */
+ insn == 0x48623fd9 || /* ldw -14(r3),rp */
+ insn == 0xe840c000 || /* bv r0(rp) */
+ insn == 0xe840c002) /* bv,n r0(rp) */
+ p -= 4;
+ else
+ break;
+ }
+ p += 4;
+ if (p <= p_start)
+ error("empty code for %s", name);
+
+ copy_size = p - p_start;
+ }
#elif defined(HOST_MIPS) || defined(HOST_MIPS64)
{
#define INSN_RETURN 0x03e00008
@@ -1746,7 +1790,7 @@ void gen_code(const char *name, host_ulong offset, host_ulong size,
!strstart(sym_name, "__op_param", NULL) &&
!strstart(sym_name, "__op_jmp", NULL) &&
!strstart(sym_name, "__op_gen_label", NULL)) {
-#if defined(HOST_SPARC)
+#if defined(HOST_SPARC) || defined(HOST_HPPA)
if (sym_name[0] == '.') {
fprintf(outfile,
"extern char __dot_%s __asm__(\"%s\");\n",
@@ -1774,8 +1818,13 @@ void gen_code(const char *name, host_ulong offset, host_ulong size,
}
}
+#ifdef __hppa__
+ fprintf(outfile, " memcpy(gen_code_ptr, (void *)((char *)__canonicalize_funcptr_for_compare(%s)+%d), %d);\n",
+ name, (int)(start_offset - offset), copy_size);
+#else
fprintf(outfile, " memcpy(gen_code_ptr, (void *)((char *)&%s+%d), %d);\n",
name, (int)(start_offset - offset), copy_size);
+#endif
/* emit code offset information */
{
@@ -2581,6 +2630,82 @@ void gen_code(const char *name, host_ulong offset, host_ulong size,
}
}
}
+#elif defined(HOST_HPPA)
+ {
+ char relname[256];
+ int type, is_label;
+ int addend;
+ int reloc_offset;
+ for (i = 0, rel = relocs; i < nb_relocs; i++, rel++) {
+ if (rel->r_offset >= start_offset &&
+ rel->r_offset < start_offset + copy_size) {
+ sym_name = get_rel_sym_name(rel);
+ sym_name = strtab + symtab[ELF32_R_SYM(rel->r_info)].st_name;
+ is_label = get_reloc_expr(relname, sizeof(relname), sym_name);
+ type = ELF32_R_TYPE(rel->r_info);
+ addend = rel->r_addend;
+ reloc_offset = rel->r_offset - start_offset;
+
+ if (is_label) {
+ switch (type) {
+ case R_PARISC_PCREL17F:
+ fprintf(outfile,
+" tcg_out_reloc(s, gen_code_ptr + %d, %d, %s, %d);\n",
+ reloc_offset, type, relname, addend);
+ break;
+ default:
+ error("unsupported hppa label relocation (%d)", type);
+ }
+ } else {
+ switch (type) {
+ case R_PARISC_DIR21L:
+ fprintf(outfile,
+" hppa_patch21l((uint32_t *)(gen_code_ptr + %d), %s, %d);\n",
+ reloc_offset, relname, addend);
+ break;
+ case R_PARISC_DIR14R:
+ fprintf(outfile,
+" hppa_patch14r((uint32_t *)(gen_code_ptr + %d), %s, %d);\n",
+ reloc_offset, relname, addend);
+ break;
+ case R_PARISC_PCREL17F:
+ if (strstart(sym_name, "__op_gen_label", NULL)) {
+ fprintf(outfile,
+" hppa_patch17f((uint32_t *)(gen_code_ptr + %d), %s, %d);\n",
+ reloc_offset, relname, addend);
+ } else {
+ fprintf(outfile,
+" HPPA_RECORD_BRANCH(hppa_stubs, (uint32_t *)(gen_code_ptr + %d), %s);\n",
+ reloc_offset, relname);
+ }
+ break;
+ case R_PARISC_DPREL21L:
+ if (strstart(sym_name, "__op_param", &p))
+ fprintf(outfile,
+" hppa_load_imm21l((uint32_t *)(gen_code_ptr + %d), param%s, %d);\n",
+ reloc_offset, p, addend);
+ else
+ fprintf(outfile,
+" hppa_patch21l_dprel((uint32_t *)(gen_code_ptr + %d), %s, %d);\n",
+ reloc_offset, relname, addend);
+ break;
+ case R_PARISC_DPREL14R:
+ if (strstart(sym_name, "__op_param", &p))
+ fprintf(outfile,
+" hppa_load_imm14r((uint32_t *)(gen_code_ptr + %d), param%s, %d);\n",
+ reloc_offset, p, addend);
+ else
+ fprintf(outfile,
+" hppa_patch14r_dprel((uint32_t *)(gen_code_ptr + %d), %s, %d);\n",
+ reloc_offset, relname, addend);
+ break;
+ default:
+ error("unsupported hppa relocation (%d)", type);
+ }
+ }
+ }
+ }
+ }
#elif defined(HOST_MIPS) || defined(HOST_MIPS64)
{
for (i = 0, rel = relocs; i < nb_relocs; i++, rel++) {
diff --git a/hppa-dis.c b/hppa-dis.c
new file mode 100644
index 000000000..e7e680375
--- /dev/null
+++ b/hppa-dis.c
@@ -0,0 +1,2834 @@
+/* Disassembler for the PA-RISC. Somewhat derived from sparc-pinsn.c.
+ Copyright 1989, 1990, 1992, 1993, 1994, 1995, 1998, 1999, 2000, 2001, 2003,
+ 2005 Free Software Foundation, Inc.
+
+ Contributed by the Center for Software Science at the
+ University of Utah (pa-gdb-bugs@cs.utah.edu).
+
+ This program 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 2 of the License, or
+ (at your option) any later version.
+
+ 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, Inc., 51 Franklin Street - Fifth Floor, Boston,
+ MA 02110-1301, USA. */
+
+#include "dis-asm.h"
+
+/* HP PA-RISC SOM object file format: definitions internal to BFD.
+ Copyright 1990, 1991, 1992, 1993, 1994, 1995, 1996, 1998, 1999, 2000,
+ 2003 Free Software Foundation, Inc.
+
+ Contributed by the Center for Software Science at the
+ University of Utah (pa-gdb-bugs@cs.utah.edu).
+
+ This file is part of BFD, the Binary File Descriptor library.
+
+ This program 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 2 of the License, or
+ (at your option) any later version.
+
+ 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, Inc., 51 Franklin Street - Fifth Floor, Boston, MA 02110-1301, USA. */
+
+#ifndef _LIBHPPA_H
+#define _LIBHPPA_H
+
+#define BYTES_IN_WORD 4
+#define PA_PAGESIZE 0x1000
+
+/* The PA instruction set variants. */
+enum pa_arch {pa10 = 10, pa11 = 11, pa20 = 20, pa20w = 25};
+
+/* HP PA-RISC relocation types */
+
+enum hppa_reloc_field_selector_type
+ {
+ R_HPPA_FSEL = 0x0,
+ R_HPPA_LSSEL = 0x1,
+ R_HPPA_RSSEL = 0x2,
+ R_HPPA_LSEL = 0x3,
+ R_HPPA_RSEL = 0x4,
+ R_HPPA_LDSEL = 0x5,
+ R_HPPA_RDSEL = 0x6,
+ R_HPPA_LRSEL = 0x7,
+ R_HPPA_RRSEL = 0x8,
+ R_HPPA_NSEL = 0x9,
+ R_HPPA_NLSEL = 0xa,
+ R_HPPA_NLRSEL = 0xb,
+ R_HPPA_PSEL = 0xc,
+ R_HPPA_LPSEL = 0xd,
+ R_HPPA_RPSEL = 0xe,
+ R_HPPA_TSEL = 0xf,
+ R_HPPA_LTSEL = 0x10,
+ R_HPPA_RTSEL = 0x11,
+ R_HPPA_LTPSEL = 0x12,
+ R_HPPA_RTPSEL = 0x13
+ };
+
+/* /usr/include/reloc.h defines these to constants. We want to use
+ them in enums, so #undef them before we start using them. We might
+ be able to fix this another way by simply managing not to include
+ /usr/include/reloc.h, but currently GDB picks up these defines
+ somewhere. */
+#undef e_fsel
+#undef e_lssel
+#undef e_rssel
+#undef e_lsel
+#undef e_rsel
+#undef e_ldsel
+#undef e_rdsel
+#undef e_lrsel
+#undef e_rrsel
+#undef e_nsel
+#undef e_nlsel
+#undef e_nlrsel
+#undef e_psel
+#undef e_lpsel
+#undef e_rpsel
+#undef e_tsel
+#undef e_ltsel
+#undef e_rtsel
+#undef e_one
+#undef e_two
+#undef e_pcrel
+#undef e_con
+#undef e_plabel
+#undef e_abs
+
+/* for compatibility */
+enum hppa_reloc_field_selector_type_alt
+ {
+ e_fsel = R_HPPA_FSEL,
+ e_lssel = R_HPPA_LSSEL,
+ e_rssel = R_HPPA_RSSEL,
+ e_lsel = R_HPPA_LSEL,
+ e_rsel = R_HPPA_RSEL,
+ e_ldsel = R_HPPA_LDSEL,
+ e_rdsel = R_HPPA_RDSEL,
+ e_lrsel = R_HPPA_LRSEL,
+ e_rrsel = R_HPPA_RRSEL,
+ e_nsel = R_HPPA_NSEL,
+ e_nlsel = R_HPPA_NLSEL,
+ e_nlrsel = R_HPPA_NLRSEL,
+ e_psel = R_HPPA_PSEL,
+ e_lpsel = R_HPPA_LPSEL,
+ e_rpsel = R_HPPA_RPSEL,
+ e_tsel = R_HPPA_TSEL,
+ e_ltsel = R_HPPA_LTSEL,
+ e_rtsel = R_HPPA_RTSEL,
+ e_ltpsel = R_HPPA_LTPSEL,
+ e_rtpsel = R_HPPA_RTPSEL
+ };
+
+enum hppa_reloc_expr_type
+ {
+ R_HPPA_E_ONE = 0,
+ R_HPPA_E_TWO = 1,
+ R_HPPA_E_PCREL = 2,
+ R_HPPA_E_CON = 3,
+ R_HPPA_E_PLABEL = 7,
+ R_HPPA_E_ABS = 18
+ };
+
+/* for compatibility */
+enum hppa_reloc_expr_type_alt
+ {
+ e_one = R_HPPA_E_ONE,
+ e_two = R_HPPA_E_TWO,
+ e_pcrel = R_HPPA_E_PCREL,
+ e_con = R_HPPA_E_CON,
+ e_plabel = R_HPPA_E_PLABEL,
+ e_abs = R_HPPA_E_ABS
+ };
+
+
+/* Relocations for function calls must be accompanied by parameter
+ relocation bits. These bits describe exactly where the caller has
+ placed the function's arguments and where it expects to find a return
+ value.
+
+ Both ELF and SOM encode this information within the addend field
+ of the call relocation. (Note this could break very badly if one
+ was to make a call like bl foo + 0x12345678).
+
+ The high order 10 bits contain parameter relocation information,
+ the low order 22 bits contain the constant offset. */
+
+#define HPPA_R_ARG_RELOC(a) \
+ (((a) >> 22) & 0x3ff)
+#define HPPA_R_CONSTANT(a) \
+ ((((bfd_signed_vma)(a)) << (BFD_ARCH_SIZE-22)) >> (BFD_ARCH_SIZE-22))
+#define HPPA_R_ADDEND(r, c) \
+ (((r) << 22) + ((c) & 0x3fffff))
+
+
+/* Some functions to manipulate PA instructions. */
+
+/* Declare the functions with the unused attribute to avoid warnings. */
+static inline int sign_extend (int, int) ATTRIBUTE_UNUSED;
+static inline int low_sign_extend (int, int) ATTRIBUTE_UNUSED;
+static inline int sign_unext (int, int) ATTRIBUTE_UNUSED;
+static inline int low_sign_unext (int, int) ATTRIBUTE_UNUSED;
+static inline int re_assemble_3 (int) ATTRIBUTE_UNUSED;
+static inline int re_assemble_12 (int) ATTRIBUTE_UNUSED;
+static inline int re_assemble_14 (int) ATTRIBUTE_UNUSED;
+static inline int re_assemble_16 (int) ATTRIBUTE_UNUSED;
+static inline int re_assemble_17 (int) ATTRIBUTE_UNUSED;
+static inline int re_assemble_21 (int) ATTRIBUTE_UNUSED;
+static inline int re_assemble_22 (int) ATTRIBUTE_UNUSED;
+static inline bfd_signed_vma hppa_field_adjust
+ (bfd_vma, bfd_signed_vma, enum hppa_reloc_field_selector_type_alt)
+ ATTRIBUTE_UNUSED;
+static inline int hppa_rebuild_insn (int, int, int) ATTRIBUTE_UNUSED;
+
+
+/* The *sign_extend functions are used to assemble various bitfields
+ taken from an instruction and return the resulting immediate
+ value. */
+
+static inline int
+sign_extend (int x, int len)
+{
+ int signbit = (1 << (len - 1));
+ int mask = (signbit << 1) - 1;
+ return ((x & mask) ^ signbit) - signbit;
+}
+
+static inline int
+low_sign_extend (int x, int len)
+{
+ return (x >> 1) - ((x & 1) << (len - 1));
+}
+
+
+/* The re_assemble_* functions prepare an immediate value for
+ insertion into an opcode. pa-risc uses all sorts of weird bitfields
+ in the instruction to hold the value. */
+
+static inline int
+sign_unext (int x, int len)
+{
+ int len_ones;
+
+ len_ones = (1 << len) - 1;
+
+ return x & len_ones;
+}
+
+static inline int
+low_sign_unext (int x, int len)
+{
+ int temp;
+ int sign;
+
+ sign = (x >> (len-1)) & 1;
+
+ temp = sign_unext (x, len-1);
+
+ return (temp << 1) | sign;
+}
+
+static inline int
+re_assemble_3 (int as3)
+{
+ return (( (as3 & 4) << (13-2))
+ | ((as3 & 3) << (13+1)));
+}
+
+static inline int
+re_assemble_12 (int as12)
+{
+ return (( (as12 & 0x800) >> 11)
+ | ((as12 & 0x400) >> (10 - 2))
+ | ((as12 & 0x3ff) << (1 + 2)));
+}
+
+static inline int
+re_assemble_14 (int as14)
+{
+ return (( (as14 & 0x1fff) << 1)
+ | ((as14 & 0x2000) >> 13));
+}
+
+static inline int
+re_assemble_16 (int as16)
+{
+ int s, t;
+
+ /* Unusual 16-bit encoding, for wide mode only. */
+ t = (as16 << 1) & 0xffff;
+ s = (as16 & 0x8000);
+ return (t ^ s ^ (s >> 1)) | (s >> 15);
+}
+
+static inline int
+re_assemble_17 (int as17)
+{
+ return (( (as17 & 0x10000) >> 16)
+ | ((as17 & 0x0f800) << (16 - 11))
+ | ((as17 & 0x00400) >> (10 - 2))
+ | ((as17 & 0x003ff) << (1 + 2)));
+}
+
+static inline int
+re_assemble_21 (int as21)
+{
+ return (( (as21 & 0x100000) >> 20)
+ | ((as21 & 0x0ffe00) >> 8)
+ | ((as21 & 0x000180) << 7)
+ | ((as21 & 0x00007c) << 14)
+ | ((as21 & 0x000003) << 12));
+}
+
+static inline int
+re_assemble_22 (int as22)
+{
+ return (( (as22 & 0x200000) >> 21)
+ | ((as22 & 0x1f0000) << (21 - 16))
+ | ((as22 & 0x00f800) << (16 - 11))
+ | ((as22 & 0x000400) >> (10 - 2))
+ | ((as22 & 0x0003ff) << (1 + 2)));
+}
+
+
+/* Handle field selectors for PA instructions.
+ The L and R (and LS, RS etc.) selectors are used in pairs to form a
+ full 32 bit address. eg.
+
+ LDIL L'start,%r1 ; put left part into r1
+ LDW R'start(%r1),%r2 ; add r1 and right part to form address
+
+ This function returns sign extended values in all cases.
+*/
+
+static inline bfd_signed_vma
+hppa_field_adjust (bfd_vma sym_val,
+ bfd_signed_vma addend,
+ enum hppa_reloc_field_selector_type_alt r_field)
+{
+ bfd_signed_vma value;
+
+ value = sym_val + addend;
+ switch (r_field)
+ {
+ case e_fsel:
+ /* F: No change. */
+ break;
+
+ case e_nsel:
+ /* N: null selector. I don't really understand what this is all
+ about, but HP's documentation says "this indicates that zero
+ bits are to be used for the displacement on the instruction.
+ This fixup is used to identify three-instruction sequences to
+ access data (for importing shared library data)." */
+ value = 0;
+ break;
+
+ case e_lsel:
+ case e_nlsel:
+ /* L: Select top 21 bits. */
+ value = value >> 11;
+ break;
+
+ case e_rsel:
+ /* R: Select bottom 11 bits. */
+ value = value & 0x7ff;
+ break;
+
+ case e_lssel:
+ /* LS: Round to nearest multiple of 2048 then select top 21 bits. */
+ value = value + 0x400;
+ value = value >> 11;
+ break;
+
+ case e_rssel:
+ /* RS: Select bottom 11 bits for LS.
+ We need to return a value such that 2048 * LS'x + RS'x == x.
+ ie. RS'x = x - ((x + 0x400) & -0x800)
+ this is just a sign extension from bit 21. */
+ value = ((value & 0x7ff) ^ 0x400) - 0x400;
+ break;
+
+ case e_ldsel:
+ /* LD: Round to next multiple of 2048 then select top 21 bits.
+ Yes, if we are already on a multiple of 2048, we go up to the
+ next one. RD in this case will be -2048. */
+ value = value + 0x800;
+ value = value >> 11;
+ break;
+
+ case e_rdsel:
+ /* RD: Set bits 0-20 to one. */
+ value = value | -0x800;
+ break;
+
+ case e_lrsel:
+ case e_nlrsel:
+ /* LR: L with rounding of the addend to nearest 8k. */
+ value = sym_val + ((addend + 0x1000) & -0x2000);
+ value = value >> 11;
+ break;
+
+ case e_rrsel:
+ /* RR: R with rounding of the addend to nearest 8k.
+ We need to return a value such that 2048 * LR'x + RR'x == x
+ ie. RR'x = s+a - (s + (((a + 0x1000) & -0x2000) & -0x800))
+ . = s+a - ((s & -0x800) + ((a + 0x1000) & -0x2000))
+ . = (s & 0x7ff) + a - ((a + 0x1000) & -0x2000) */
+ value = (sym_val & 0x7ff) + (((addend & 0x1fff) ^ 0x1000) - 0x1000);
+ break;
+
+ default:
+ abort ();
+ }
+ return value;
+}
+
+/* PA-RISC OPCODES */
+#define get_opcode(insn) (((insn) >> 26) & 0x3f)
+
+enum hppa_opcode_type
+{
+ /* None of the opcodes in the first group generate relocs, so we
+ aren't too concerned about them. */
+ OP_SYSOP = 0x00,
+ OP_MEMMNG = 0x01,
+ OP_ALU = 0x02,
+ OP_NDXMEM = 0x03,
+ OP_SPOP = 0x04,
+ OP_DIAG = 0x05,
+ OP_FMPYADD = 0x06,
+ OP_UNDEF07 = 0x07,
+ OP_COPRW = 0x09,
+ OP_COPRDW = 0x0b,
+ OP_COPR = 0x0c,
+ OP_FLOAT = 0x0e,
+ OP_PRDSPEC = 0x0f,
+ OP_UNDEF15 = 0x15,
+ OP_UNDEF1d = 0x1d,
+ OP_FMPYSUB = 0x26,
+ OP_FPFUSED = 0x2e,
+ OP_SHEXDP0 = 0x34,
+ OP_SHEXDP1 = 0x35,
+ OP_SHEXDP2 = 0x36,
+ OP_UNDEF37 = 0x37,
+ OP_SHEXDP3 = 0x3c,
+ OP_SHEXDP4 = 0x3d,
+ OP_MULTMED = 0x3e,
+ OP_UNDEF3f = 0x3f,
+
+ OP_LDIL = 0x08,
+ OP_ADDIL = 0x0a,
+
+ OP_LDO = 0x0d,
+ OP_LDB = 0x10,
+ OP_LDH = 0x11,
+ OP_LDW = 0x12,
+ OP_LDWM = 0x13,
+ OP_STB = 0x18,
+ OP_STH = 0x19,
+ OP_STW = 0x1a,
+ OP_STWM = 0x1b,
+
+ OP_LDD = 0x14,
+ OP_STD = 0x1c,
+
+ OP_FLDW = 0x16,
+ OP_LDWL = 0x17,
+ OP_FSTW = 0x1e,
+ OP_STWL = 0x1f,
+
+ OP_COMBT = 0x20,
+ OP_COMIBT = 0x21,
+ OP_COMBF = 0x22,
+ OP_COMIBF = 0x23,
+ OP_CMPBDT = 0x27,
+ OP_ADDBT = 0x28,
+ OP_ADDIBT = 0x29,
+ OP_ADDBF = 0x2a,
+ OP_ADDIBF = 0x2b,
+ OP_CMPBDF = 0x2f,
+ OP_BVB = 0x30,
+ OP_BB = 0x31,
+ OP_MOVB = 0x32,
+ OP_MOVIB = 0x33,
+ OP_CMPIBD = 0x3b,
+
+ OP_COMICLR = 0x24,
+ OP_SUBI = 0x25,
+ OP_ADDIT = 0x2c,
+ OP_ADDI = 0x2d,
+
+ OP_BE = 0x38,
+ OP_BLE = 0x39,
+ OP_BL = 0x3a
+};
+
+
+/* Insert VALUE into INSN using R_FORMAT to determine exactly what
+ bits to change. */
+
+static inline int
+hppa_rebuild_insn (int insn, int value, int r_format)
+{
+ switch (r_format)
+ {
+ case 11:
+ return (insn & ~ 0x7ff) | low_sign_unext (value, 11);
+
+ case 12:
+ return (insn & ~ 0x1ffd) | re_assemble_12 (value);
+
+
+ case 10:
+ return (insn & ~ 0x3ff1) | re_assemble_14 (value & -8);
+
+ case -11:
+ return (insn & ~ 0x3ff9) | re_assemble_14 (value & -4);
+
+ case 14:
+ return (insn & ~ 0x3fff) | re_assemble_14 (value);
+
+
+ case -10:
+ return (insn & ~ 0xfff1) | re_assemble_16 (value & -8);
+
+ case -16:
+ return (insn & ~ 0xfff9) | re_assemble_16 (value & -4);
+
+ case 16:
+ return (insn & ~ 0xffff) | re_assemble_16 (value);
+
+
+ case 17:
+ return (insn & ~ 0x1f1ffd) | re_assemble_17 (value);
+
+ case 21:
+ return (insn & ~ 0x1fffff) | re_assemble_21 (value);
+
+ case 22:
+ return (insn & ~ 0x3ff1ffd) | re_assemble_22 (value);
+
+ case 32:
+ return value;
+
+ default:
+ abort ();
+ }
+ return insn;
+}
+
+#endif /* _LIBHPPA_H */
+/* Table of opcodes for the PA-RISC.
+ Copyright 1990, 1991, 1992, 1993, 1994, 1995, 1996, 1998, 1999, 2000,
+ 2001, 2002, 2003, 2004, 2005
+ Free Software Foundation, Inc.
+
+ Contributed by the Center for Software Science at the
+ University of Utah (pa-gdb-bugs@cs.utah.edu).
+
+This file is part of GAS, the GNU Assembler, and GDB, the GNU disassembler.
+
+GAS/GDB 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 1, or (at your option)
+any later version.
+
+GAS/GDB 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 GAS or GDB; see the file COPYING. If not, write to
+the Free Software Foundation, 51 Franklin Street - Fifth Floor, Boston, MA 02110-1301, USA. */
+
+#if !defined(__STDC__) && !defined(const)
+#define const
+#endif
+
+/*
+ * Structure of an opcode table entry.
+ */
+
+/* There are two kinds of delay slot nullification: normal which is
+ * controled by the nullification bit, and conditional, which depends
+ * on the direction of the branch and its success or failure.
+ *
+ * NONE is unfortunately #defined in the hiux system include files.
+ * #undef it away.
+ */
+#undef NONE
+struct pa_opcode
+{
+ const char *name;
+ unsigned long int match; /* Bits that must be set... */
+ unsigned long int mask; /* ... in these bits. */
+ char *args;
+ enum pa_arch arch;
+ char flags;
+};
+
+/* Enables strict matching. Opcodes with match errors are skipped
+ when this bit is set. */
+#define FLAG_STRICT 0x1
+
+/*
+ All hppa opcodes are 32 bits.
+
+ The match component is a mask saying which bits must match a
+ particular opcode in order for an instruction to be an instance
+ of that opcode.
+
+ The args component is a string containing one character for each operand of
+ the instruction. Characters used as a prefix allow any second character to
+ be used without conflicting with the main operand characters.
+
+ Bit positions in this description follow HP usage of lsb = 31,
+ "at" is lsb of field.
+
+ In the args field, the following characters must match exactly:
+
+ '+,() '
+
+ In the args field, the following characters are unused:
+
+ ' " - / 34 6789:; '
+ '@ C M [\] '
+ '` e g } '
+
+ Here are all the characters:
+
+ ' !"#$%&'()*+-,./0123456789:;<=>?'
+ '@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_'
+ '`abcdefghijklmnopqrstuvwxyz{|}~ '
+
+Kinds of operands:
+ x integer register field at 15.
+ b integer register field at 10.
+ t integer register field at 31.
+ a integer register field at 10 and 15 (for PERMH)
+ 5 5 bit immediate at 15.
+ s 2 bit space specifier at 17.
+ S 3 bit space specifier at 18.
+ V 5 bit immediate value at 31
+ i 11 bit immediate value at 31
+ j 14 bit immediate value at 31
+ k 21 bit immediate value at 31
+ l 16 bit immediate value at 31 (wide mode only, unusual encoding).
+ n nullification for branch instructions
+ N nullification for spop and copr instructions
+ w 12 bit branch displacement
+ W 17 bit branch displacement (PC relative)
+ X 22 bit branch displacement (PC relative)
+ z 17 bit branch displacement (just a number, not an address)
+
+Also these:
+
+ . 2 bit shift amount at 25
+ * 4 bit shift amount at 25
+ p 5 bit shift count at 26 (to support the SHD instruction) encoded as
+ 31-p
+ ~ 6 bit shift count at 20,22:26 encoded as 63-~.
+ P 5 bit bit position at 26
+ q 6 bit bit position at 20,22:26
+ T 5 bit field length at 31 (encoded as 32-T)
+ % 6 bit field length at 23,27:31 (variable extract/deposit)
+ | 6 bit field length at 19,27:31 (fixed extract/deposit)
+ A 13 bit immediate at 18 (to support the BREAK instruction)
+ ^ like b, but describes a control register
+ ! sar (cr11) register
+ D 26 bit immediate at 31 (to support the DIAG instruction)
+ $ 9 bit immediate at 28 (to support POPBTS)
+
+ v 3 bit Special Function Unit identifier at 25
+ O 20 bit Special Function Unit operation split between 15 bits at 20
+ and 5 bits at 31
+ o 15 bit Special Function Unit operation at 20
+ 2 22 bit Special Function Unit operation split between 17 bits at 20
+ and 5 bits at 31
+ 1 15 bit Special Function Unit operation split between 10 bits at 20
+ and 5 bits at 31
+ 0 10 bit Special Function Unit operation split between 5 bits at 20
+ and 5 bits at 31
+ u 3 bit coprocessor unit identifier at 25
+ F Source Floating Point Operand Format Completer encoded 2 bits at 20
+ I Source Floating Point Operand Format Completer encoded 1 bits at 20
+ (for 0xe format FP instructions)
+ G Destination Floating Point Operand Format Completer encoded 2 bits at 18
+ H Floating Point Operand Format at 26 for 'fmpyadd' and 'fmpysub'
+ (very similar to 'F')
+
+ r 5 bit immediate value at 31 (for the break instruction)
+ (very similar to V above, except the value is unsigned instead of
+ low_sign_ext)
+ R 5 bit immediate value at 15 (for the ssm, rsm, probei instructions)
+ (same as r above, except the value is in a different location)
+ U 10 bit immediate value at 15 (for SSM, RSM on pa2.0)
+ Q 5 bit immediate value at 10 (a bit position specified in
+ the bb instruction. It's the same as r above, except the
+ value is in a different location)
+ B 5 bit immediate value at 10 (a bit position specified in
+ the bb instruction. Similar to Q, but 64 bit handling is
+ different.
+ Z %r1 -- implicit target of addil instruction.
+ L ,%r2 completer for new syntax branch
+ { Source format completer for fcnv
+ _ Destination format completer for fcnv
+ h cbit for fcmp
+ = gfx tests for ftest
+ d 14 bit offset for single precision FP long load/store.
+ # 14 bit offset for double precision FP load long/store.
+ J Yet another 14 bit offset for load/store with ma,mb completers.
+ K Yet another 14 bit offset for load/store with ma,mb completers.
+ y 16 bit offset for word aligned load/store (PA2.0 wide).
+ & 16 bit offset for dword aligned load/store (PA2.0 wide).
+ < 16 bit offset for load/store with ma,mb completers (PA2.0 wide).
+ > 16 bit offset for load/store with ma,mb completers (PA2.0 wide).
+ Y %sr0,%r31 -- implicit target of be,l instruction.
+ @ implicit immediate value of 0
+
+Completer operands all have 'c' as the prefix:
+
+ cx indexed load and store completer.
+ cX indexed load and store completer. Like cx, but emits a space
+ after in disassembler.
+ cm short load and store completer.
+ cM short load and store completer. Like cm, but emits a space
+ after in disassembler.
+ cq long load and store completer (like cm, but inserted into a
+ different location in the target instruction).
+ cs store bytes short completer.
+ cA store bytes short completer. Like cs, but emits a space
+ after in disassembler.
+ ce long load/store completer for LDW/STW with a different encoding
+ than the others
+ cc load cache control hint
+ cd load and clear cache control hint
+ cC store cache control hint
+ co ordered access
+
+ cp branch link and push completer
+ cP branch pop completer
+ cl branch link completer
+ cg branch gate completer
+
+ cw read/write completer for PROBE
+ cW wide completer for MFCTL
+ cL local processor completer for cache control
+ cZ System Control Completer (to support LPA, LHA, etc.)
+
+ ci correction completer for DCOR
+ ca add completer
+ cy 32 bit add carry completer
+ cY 64 bit add carry completer
+ cv signed overflow trap completer
+ ct trap on condition completer for ADDI, SUB
+ cT trap on condition completer for UADDCM
+ cb 32 bit borrow completer for SUB
+ cB 64 bit borrow completer for SUB
+
+ ch left/right half completer
+ cH signed/unsigned saturation completer
+ cS signed/unsigned completer at 21
+ cz zero/sign extension completer.
+ c* permutation completer
+
+Condition operands all have '?' as the prefix:
+
+ ?f Floating point compare conditions (encoded as 5 bits at 31)
+
+ ?a add conditions
+ ?A 64 bit add conditions
+ ?@ add branch conditions followed by nullify
+ ?d non-negated add branch conditions
+ ?D negated add branch conditions
+ ?w wide mode non-negated add branch conditions
+ ?W wide mode negated add branch conditions
+
+ ?s compare/subtract conditions
+ ?S 64 bit compare/subtract conditions
+ ?t non-negated compare and branch conditions
+ ?n 32 bit compare and branch conditions followed by nullify
+ ?N 64 bit compare and branch conditions followed by nullify
+ ?Q 64 bit compare and branch conditions for CMPIB instruction
+
+ ?l logical conditions
+ ?L 64 bit logical conditions
+
+ ?b branch on bit conditions
+ ?B 64 bit branch on bit conditions
+
+ ?x shift/extract/deposit conditions
+ ?X 64 bit shift/extract/deposit conditions
+ ?y shift/extract/deposit conditions followed by nullify for conditional
+ branches
+
+ ?u unit conditions
+ ?U 64 bit unit conditions
+
+Floating point registers all have 'f' as a prefix:
+
+ ft target register at 31
+ fT target register with L/R halves at 31
+ fa operand 1 register at 10
+ fA operand 1 register with L/R halves at 10
+ fX Same as fA, except prints a space before register during disasm
+ fb operand 2 register at 15
+ fB operand 2 register with L/R halves at 15
+ fC operand 3 register with L/R halves at 16:18,21:23
+ fe Like fT, but encoding is different.
+ fE Same as fe, except prints a space before register during disasm.
+ fx target register at 15 (only for PA 2.0 long format FLDD/FSTD).
+
+Float registers for fmpyadd and fmpysub:
+
+ fi mult operand 1 register at 10
+ fj mult operand 2 register at 15
+ fk mult target register at 20
+ fl add/sub operand register at 25
+ fm add/sub target register at 31
+
+*/
+
+
+#if 0
+/* List of characters not to put a space after. Note that
+ "," is included, as the "spopN" operations use literal
+ commas in their completer sections. */
+static const char *const completer_chars = ",CcY<>?!@+&U~FfGHINnOoZMadu|/=0123%e$m}";
+#endif
+
+/* The order of the opcodes in this table is significant:
+
+ * The assembler requires that all instances of the same mnemonic be
+ consecutive. If they aren't, the assembler will bomb at runtime.
+
+ * Immediate fields use pa_get_absolute_expression to parse the
+ string. It will generate a "bad expression" error if passed
+ a register name. Thus, register index variants of an opcode
+ need to precede immediate variants.
+
+ * The disassembler does not care about the order of the opcodes
+ except in cases where implicit addressing is used.
+
+ Here are the rules for ordering the opcodes of a mnemonic:
+
+ 1) Opcodes with FLAG_STRICT should precede opcodes without
+ FLAG_STRICT.
+
+ 2) Opcodes with FLAG_STRICT should be ordered as follows:
+ register index opcodes, short immediate opcodes, and finally
+ long immediate opcodes. When both pa10 and pa11 variants
+ of the same opcode are available, the pa10 opcode should
+ come first for correct architectural promotion.
+
+ 3) When implicit addressing is available for an opcode, the
+ implicit opcode should precede the explicit opcode.
+
+ 4) Opcodes without FLAG_STRICT should be ordered as follows:
+ register index opcodes, long immediate opcodes, and finally
+ short immediate opcodes. */
+
+static const struct pa_opcode pa_opcodes[] =
+{
+
+/* Pseudo-instructions. */
+
+{ "ldi", 0x34000000, 0xffe00000, "l,x", pa20w, 0},/* ldo val(r0),r */
+{ "ldi", 0x34000000, 0xffe0c000, "j,x", pa10, 0},/* ldo val(r0),r */
+
+{ "cmpib", 0xec000000, 0xfc000000, "?Qn5,b,w", pa20, FLAG_STRICT},
+{ "cmpib", 0x84000000, 0xf4000000, "?nn5,b,w", pa10, FLAG_STRICT},
+{ "comib", 0x84000000, 0xfc000000, "?nn5,b,w", pa10, 0}, /* comib{tf}*/
+/* This entry is for the disassembler only. It will never be used by
+ assembler. */
+{ "comib", 0x8c000000, 0xfc000000, "?nn5,b,w", pa10, 0}, /* comib{tf}*/
+{ "cmpb", 0x9c000000, 0xdc000000, "?Nnx,b,w", pa20, FLAG_STRICT},
+{ "cmpb", 0x80000000, 0xf4000000, "?nnx,b,w", pa10, FLAG_STRICT},
+{ "comb", 0x80000000, 0xfc000000, "?nnx,b,w", pa10, 0}, /* comb{tf} */
+/* This entry is for the disassembler only. It will never be used by
+ assembler. */
+{ "comb", 0x88000000, 0xfc000000, "?nnx,b,w", pa10, 0}, /* comb{tf} */
+{ "addb", 0xa0000000, 0xf4000000, "?Wnx,b,w", pa20w, FLAG_STRICT},
+{ "addb", 0xa0000000, 0xfc000000, "?@nx,b,w", pa10, 0}, /* addb{tf} */
+/* This entry is for the disassembler only. It will never be used by
+ assembler. */
+{ "addb", 0xa8000000, 0xfc000000, "?@nx,b,w", pa10, 0},
+{ "addib", 0xa4000000, 0xf4000000, "?Wn5,b,w", pa20w, FLAG_STRICT},
+{ "addib", 0xa4000000, 0xfc000000, "?@n5,b,w", pa10, 0}, /* addib{tf}*/
+/* This entry is for the disassembler only. It will never be used by
+ assembler. */
+{ "addib", 0xac000000, 0xfc000000, "?@n5,b,w", pa10, 0}, /* addib{tf}*/
+{ "nop", 0x08000240, 0xffffffff, "", pa10, 0}, /* or 0,0,0 */
+{ "copy", 0x08000240, 0xffe0ffe0, "x,t", pa10, 0}, /* or r,0,t */
+{ "mtsar", 0x01601840, 0xffe0ffff, "x", pa10, 0}, /* mtctl r,cr11 */
+
+/* Loads and Stores for integer registers. */
+
+{ "ldd", 0x0c0000c0, 0xfc00d3c0, "cxccx(b),t", pa20, FLAG_STRICT},
+{ "ldd", 0x0c0000c0, 0xfc0013c0, "cxccx(s,b),t", pa20, FLAG_STRICT},
+{ "ldd", 0x0c0010e0, 0xfc1ff3e0, "cocc@(b),t", pa20, FLAG_STRICT},
+{ "ldd", 0x0c0010e0, 0xfc1f33e0, "cocc@(s,b),t", pa20, FLAG_STRICT},
+{ "ldd", 0x0c0010c0, 0xfc00d3c0, "cmcc5(b),t", pa20, FLAG_STRICT},
+{ "ldd", 0x0c0010c0, 0xfc0013c0, "cmcc5(s,b),t", pa20, FLAG_STRICT},
+{ "ldd", 0x50000000, 0xfc000002, "cq&(b),x", pa20w, FLAG_STRICT},
+{ "ldd", 0x50000000, 0xfc00c002, "cq#(b),x", pa20, FLAG_STRICT},
+{ "ldd", 0x50000000, 0xfc000002, "cq#(s,b),x", pa20, FLAG_STRICT},
+{ "ldw", 0x0c000080, 0xfc00dfc0, "cXx(b),t", pa10, FLAG_STRICT},
+{ "ldw", 0x0c000080, 0xfc001fc0, "cXx(s,b),t", pa10, FLAG_STRICT},
+{ "ldw", 0x0c000080, 0xfc00d3c0, "cxccx(b),t", pa11, FLAG_STRICT},
+{ "ldw", 0x0c000080, 0xfc0013c0, "cxccx(s,b),t", pa11, FLAG_STRICT},
+{ "ldw", 0x0c0010a0, 0xfc1ff3e0, "cocc@(b),t", pa20, FLAG_STRICT},
+{ "ldw", 0x0c0010a0, 0xfc1f33e0, "cocc@(s,b),t", pa20, FLAG_STRICT},
+{ "ldw", 0x0c001080, 0xfc00dfc0, "cM5(b),t", pa10, FLAG_STRICT},
+{ "ldw", 0x0c001080, 0xfc001fc0, "cM5(s,b),t", pa10, FLAG_STRICT},
+{ "ldw", 0x0c001080, 0xfc00d3c0, "cmcc5(b),t", pa11, FLAG_STRICT},
+{ "ldw", 0x0c001080, 0xfc0013c0, "cmcc5(s,b),t", pa11, FLAG_STRICT},
+{ "ldw", 0x4c000000, 0xfc000000, "ce<(b),x", pa20w, FLAG_STRICT},
+{ "ldw", 0x5c000004, 0xfc000006, "ce>(b),x", pa20w, FLAG_STRICT},
+{ "ldw", 0x48000000, 0xfc000000, "l(b),x", pa20w, FLAG_STRICT},
+{ "ldw", 0x5c000004, 0xfc00c006, "ceK(b),x", pa20, FLAG_STRICT},
+{ "ldw", 0x5c000004, 0xfc000006, "ceK(s,b),x", pa20, FLAG_STRICT},
+{ "ldw", 0x4c000000, 0xfc00c000, "ceJ(b),x", pa10, FLAG_STRICT},
+{ "ldw", 0x4c000000, 0xfc000000, "ceJ(s,b),x", pa10, FLAG_STRICT},
+{ "ldw", 0x48000000, 0xfc00c000, "j(b),x", pa10, 0},
+{ "ldw", 0x48000000, 0xfc000000, "j(s,b),x", pa10, 0},
+{ "ldh", 0x0c000040, 0xfc00dfc0, "cXx(b),t", pa10, FLAG_STRICT},
+{ "ldh", 0x0c000040, 0xfc001fc0, "cXx(s,b),t", pa10, FLAG_STRICT},
+{ "ldh", 0x0c000040, 0xfc00d3c0, "cxccx(b),t", pa11, FLAG_STRICT},
+{ "ldh", 0x0c000040, 0xfc0013c0, "cxccx(s,b),t", pa11, FLAG_STRICT},
+{ "ldh", 0x0c001060, 0xfc1ff3e0, "cocc@(b),t", pa20, FLAG_STRICT},
+{ "ldh", 0x0c001060, 0xfc1f33e0, "cocc@(s,b),t", pa20, FLAG_STRICT},
+{ "ldh", 0x0c001040, 0xfc00dfc0, "cM5(b),t", pa10, FLAG_STRICT},
+{ "ldh", 0x0c001040, 0xfc001fc0, "cM5(s,b),t", pa10, FLAG_STRICT},
+{ "ldh", 0x0c001040, 0xfc00d3c0, "cmcc5(b),t", pa11, FLAG_STRICT},
+{ "ldh", 0x0c001040, 0xfc0013c0, "cmcc5(s,b),t", pa11, FLAG_STRICT},
+{ "ldh", 0x44000000, 0xfc000000, "l(b),x", pa20w, FLAG_STRICT},
+{ "ldh", 0x44000000, 0xfc00c000, "j(b),x", pa10, 0},
+{ "ldh", 0x44000000, 0xfc000000, "j(s,b),x", pa10, 0},
+{ "ldb", 0x0c000000, 0xfc00dfc0, "cXx(b),t", pa10, FLAG_STRICT},
+{ "ldb", 0x0c000000, 0xfc001fc0, "cXx(s,b),t", pa10, FLAG_STRICT},
+{ "ldb", 0x0c000000, 0xfc00d3c0, "cxccx(b),t", pa11, FLAG_STRICT},
+{ "ldb", 0x0c000000, 0xfc0013c0, "cxccx(s,b),t", pa11, FLAG_STRICT},
+{ "ldb", 0x0c001020, 0xfc1ff3e0, "cocc@(b),t", pa20, FLAG_STRICT},
+{ "ldb", 0x0c001020, 0xfc1f33e0, "cocc@(s,b),t", pa20, FLAG_STRICT},
+{ "ldb", 0x0c001000, 0xfc00dfc0, "cM5(b),t", pa10, FLAG_STRICT},
+{ "ldb", 0x0c001000, 0xfc001fc0, "cM5(s,b),t", pa10, FLAG_STRICT},
+{ "ldb", 0x0c001000, 0xfc00d3c0, "cmcc5(b),t", pa11, FLAG_STRICT},
+{ "ldb", 0x0c001000, 0xfc0013c0, "cmcc5(s,b),t", pa11, FLAG_STRICT},
+{ "ldb", 0x40000000, 0xfc000000, "l(b),x", pa20w, FLAG_STRICT},
+{ "ldb", 0x40000000, 0xfc00c000, "j(b),x", pa10, 0},
+{ "ldb", 0x40000000, 0xfc000000, "j(s,b),x", pa10, 0},
+{ "std", 0x0c0012e0, 0xfc00f3ff, "cocCx,@(b)", pa20, FLAG_STRICT},
+{ "std", 0x0c0012e0, 0xfc0033ff, "cocCx,@(s,b)", pa20, FLAG_STRICT},
+{ "std", 0x0c0012c0, 0xfc00d3c0, "cmcCx,V(b)", pa20, FLAG_STRICT},
+{ "std", 0x0c0012c0, 0xfc0013c0, "cmcCx,V(s,b)", pa20, FLAG_STRICT},
+{ "std", 0x70000000, 0xfc000002, "cqx,&(b)", pa20w, FLAG_STRICT},
+{ "std", 0x70000000, 0xfc00c002, "cqx,#(b)", pa20, FLAG_STRICT},
+{ "std", 0x70000000, 0xfc000002, "cqx,#(s,b)", pa20, FLAG_STRICT},
+{ "stw", 0x0c0012a0, 0xfc00f3ff, "cocCx,@(b)", pa20, FLAG_STRICT},
+{ "stw", 0x0c0012a0, 0xfc0033ff, "cocCx,@(s,b)", pa20, FLAG_STRICT},
+{ "stw", 0x0c001280, 0xfc00dfc0, "cMx,V(b)", pa10, FLAG_STRICT},
+{ "stw", 0x0c001280, 0xfc001fc0, "cMx,V(s,b)", pa10, FLAG_STRICT},
+{ "stw", 0x0c001280, 0xfc00d3c0, "cmcCx,V(b)", pa11, FLAG_STRICT},
+{ "stw", 0x0c001280, 0xfc0013c0, "cmcCx,V(s,b)", pa11, FLAG_STRICT},
+{ "stw", 0x6c000000, 0xfc000000, "cex,<(b)", pa20w, FLAG_STRICT},
+{ "stw", 0x7c000004, 0xfc000006, "cex,>(b)", pa20w, FLAG_STRICT},
+{ "stw", 0x68000000, 0xfc000000, "x,l(b)", pa20w, FLAG_STRICT},
+{ "stw", 0x7c000004, 0xfc00c006, "cex,K(b)", pa20, FLAG_STRICT},
+{ "stw", 0x7c000004, 0xfc000006, "cex,K(s,b)", pa20, FLAG_STRICT},
+{ "stw", 0x6c000000, 0xfc00c000, "cex,J(b)", pa10, FLAG_STRICT},
+{ "stw", 0x6c000000, 0xfc000000, "cex,J(s,b)", pa10, FLAG_STRICT},
+{ "stw", 0x68000000, 0xfc00c000, "x,j(b)", pa10, 0},
+{ "stw", 0x68000000, 0xfc000000, "x,j(s,b)", pa10, 0},
+{ "sth", 0x0c001260, 0xfc00f3ff, "cocCx,@(b)", pa20, FLAG_STRICT},
+{ "sth", 0x0c001260, 0xfc0033ff, "cocCx,@(s,b)", pa20, FLAG_STRICT},
+{ "sth", 0x0c001240, 0xfc00dfc0, "cMx,V(b)", pa10, FLAG_STRICT},
+{ "sth", 0x0c001240, 0xfc001fc0, "cMx,V(s,b)", pa10, FLAG_STRICT},
+{ "sth", 0x0c001240, 0xfc00d3c0, "cmcCx,V(b)", pa11, FLAG_STRICT},
+{ "sth", 0x0c001240, 0xfc0013c0, "cmcCx,V(s,b)", pa11, FLAG_STRICT},
+{ "sth", 0x64000000, 0xfc000000, "x,l(b)", pa20w, FLAG_STRICT},
+{ "sth", 0x64000000, 0xfc00c000, "x,j(b)", pa10, 0},
+{ "sth", 0x64000000, 0xfc000000, "x,j(s,b)", pa10, 0},
+{ "stb", 0x0c001220, 0xfc00f3ff, "cocCx,@(b)", pa20, FLAG_STRICT},
+{ "stb", 0x0c001220, 0xfc0033ff, "cocCx,@(s,b)", pa20, FLAG_STRICT},
+{ "stb", 0x0c001200, 0xfc00dfc0, "cMx,V(b)", pa10, FLAG_STRICT},
+{ "stb", 0x0c001200, 0xfc001fc0, "cMx,V(s,b)", pa10, FLAG_STRICT},
+{ "stb", 0x0c001200, 0xfc00d3c0, "cmcCx,V(b)", pa11, FLAG_STRICT},
+{ "stb", 0x0c001200, 0xfc0013c0, "cmcCx,V(s,b)", pa11, FLAG_STRICT},
+{ "stb", 0x60000000, 0xfc000000, "x,l(b)", pa20w, FLAG_STRICT},
+{ "stb", 0x60000000, 0xfc00c000, "x,j(b)", pa10, 0},
+{ "stb", 0x60000000, 0xfc000000, "x,j(s,b)", pa10, 0},
+{ "ldwm", 0x4c000000, 0xfc00c000, "j(b),x", pa10, 0},
+{ "ldwm", 0x4c000000, 0xfc000000, "j(s,b),x", pa10, 0},
+{ "stwm", 0x6c000000, 0xfc00c000, "x,j(b)", pa10, 0},
+{ "stwm", 0x6c000000, 0xfc000000, "x,j(s,b)", pa10, 0},
+{ "ldwx", 0x0c000080, 0xfc00dfc0, "cXx(b),t", pa10, FLAG_STRICT},
+{ "ldwx", 0x0c000080, 0xfc001fc0, "cXx(s,b),t", pa10, FLAG_STRICT},
+{ "ldwx", 0x0c000080, 0xfc00d3c0, "cxccx(b),t", pa11, FLAG_STRICT},
+{ "ldwx", 0x0c000080, 0xfc0013c0, "cxccx(s,b),t", pa11, FLAG_STRICT},
+{ "ldwx", 0x0c000080, 0xfc00dfc0, "cXx(b),t", pa10, 0},
+{ "ldwx", 0x0c000080, 0xfc001fc0, "cXx(s,b),t", pa10, 0},
+{ "ldhx", 0x0c000040, 0xfc00dfc0, "cXx(b),t", pa10, FLAG_STRICT},
+{ "ldhx", 0x0c000040, 0xfc001fc0, "cXx(s,b),t", pa10, FLAG_STRICT},
+{ "ldhx", 0x0c000040, 0xfc00d3c0, "cxccx(b),t", pa11, FLAG_STRICT},
+{ "ldhx", 0x0c000040, 0xfc0013c0, "cxccx(s,b),t", pa11, FLAG_STRICT},
+{ "ldhx", 0x0c000040, 0xfc00dfc0, "cXx(b),t", pa10, 0},
+{ "ldhx", 0x0c000040, 0xfc001fc0, "cXx(s,b),t", pa10, 0},
+{ "ldbx", 0x0c000000, 0xfc00dfc0, "cXx(b),t", pa10, FLAG_STRICT},
+{ "ldbx", 0x0c000000, 0xfc001fc0, "cXx(s,b),t", pa10, FLAG_STRICT},
+{ "ldbx", 0x0c000000, 0xfc00d3c0, "cxccx(b),t", pa11, FLAG_STRICT},
+{ "ldbx", 0x0c000000, 0xfc0013c0, "cxccx(s,b),t", pa11, FLAG_STRICT},
+{ "ldbx", 0x0c000000, 0xfc00dfc0, "cXx(b),t", pa10, 0},
+{ "ldbx", 0x0c000000, 0xfc001fc0, "cXx(s,b),t", pa10, 0},
+{ "ldwa", 0x0c000180, 0xfc00dfc0, "cXx(b),t", pa10, FLAG_STRICT},
+{ "ldwa", 0x0c000180, 0xfc00d3c0, "cxccx(b),t", pa11, FLAG_STRICT},
+{ "ldwa", 0x0c0011a0, 0xfc1ff3e0, "cocc@(b),t", pa20, FLAG_STRICT},
+{ "ldwa", 0x0c001180, 0xfc00dfc0, "cM5(b),t", pa10, FLAG_STRICT},
+{ "ldwa", 0x0c001180, 0xfc00d3c0, "cmcc5(b),t", pa11, FLAG_STRICT},
+{ "ldcw", 0x0c0001c0, 0xfc00dfc0, "cXx(b),t", pa10, FLAG_STRICT},
+{ "ldcw", 0x0c0001c0, 0xfc001fc0, "cXx(s,b),t", pa10, FLAG_STRICT},
+{ "ldcw", 0x0c0001c0, 0xfc00d3c0, "cxcdx(b),t", pa11, FLAG_STRICT},
+{ "ldcw", 0x0c0001c0, 0xfc0013c0, "cxcdx(s,b),t", pa11, FLAG_STRICT},
+{ "ldcw", 0x0c0011c0, 0xfc00dfc0, "cM5(b),t", pa10, FLAG_STRICT},
+{ "ldcw", 0x0c0011c0, 0xfc001fc0, "cM5(s,b),t", pa10, FLAG_STRICT},
+{ "ldcw", 0x0c0011c0, 0xfc00d3c0, "cmcd5(b),t", pa11, FLAG_STRICT},
+{ "ldcw", 0x0c0011c0, 0xfc0013c0, "cmcd5(s,b),t", pa11, FLAG_STRICT},
+{ "stwa", 0x0c0013a0, 0xfc00d3ff, "cocCx,@(b)", pa20, FLAG_STRICT},
+{ "stwa", 0x0c001380, 0xfc00dfc0, "cMx,V(b)", pa10, FLAG_STRICT},
+{ "stwa", 0x0c001380, 0xfc00d3c0, "cmcCx,V(b)", pa11, FLAG_STRICT},
+{ "stby", 0x0c001300, 0xfc00dfc0, "cAx,V(b)", pa10, FLAG_STRICT},
+{ "stby", 0x0c001300, 0xfc001fc0, "cAx,V(s,b)", pa10, FLAG_STRICT},
+{ "stby", 0x0c001300, 0xfc00d3c0, "cscCx,V(b)", pa11, FLAG_STRICT},
+{ "stby", 0x0c001300, 0xfc0013c0, "cscCx,V(s,b)", pa11, FLAG_STRICT},
+{ "ldda", 0x0c000100, 0xfc00d3c0, "cxccx(b),t", pa20, FLAG_STRICT},
+{ "ldda", 0x0c001120, 0xfc1ff3e0, "cocc@(b),t", pa20, FLAG_STRICT},
+{ "ldda", 0x0c001100, 0xfc00d3c0, "cmcc5(b),t", pa20, FLAG_STRICT},
+{ "ldcd", 0x0c000140, 0xfc00d3c0, "cxcdx(b),t", pa20, FLAG_STRICT},
+{ "ldcd", 0x0c000140, 0xfc0013c0, "cxcdx(s,b),t", pa20, FLAG_STRICT},
+{ "ldcd", 0x0c001140, 0xfc00d3c0, "cmcd5(b),t", pa20, FLAG_STRICT},
+{ "ldcd", 0x0c001140, 0xfc0013c0, "cmcd5(s,b),t", pa20, FLAG_STRICT},
+{ "stda", 0x0c0013e0, 0xfc00f3ff, "cocCx,@(b)", pa20, FLAG_STRICT},
+{ "stda", 0x0c0013c0, 0xfc00d3c0, "cmcCx,V(b)", pa20, FLAG_STRICT},
+{ "ldwax", 0x0c000180, 0xfc00dfc0, "cXx(b),t", pa10, FLAG_STRICT},
+{ "ldwax", 0x0c000180, 0xfc00d3c0, "cxccx(b),t", pa11, FLAG_STRICT},
+{ "ldwax", 0x0c000180, 0xfc00dfc0, "cXx(b),t", pa10, 0},
+{ "ldcwx", 0x0c0001c0, 0xfc00dfc0, "cXx(b),t", pa10, FLAG_STRICT},
+{ "ldcwx", 0x0c0001c0, 0xfc001fc0, "cXx(s,b),t", pa10, FLAG_STRICT},
+{ "ldcwx", 0x0c0001c0, 0xfc00d3c0, "cxcdx(b),t", pa11, FLAG_STRICT},
+{ "ldcwx", 0x0c0001c0, 0xfc0013c0, "cxcdx(s,b),t", pa11, FLAG_STRICT},
+{ "ldcwx", 0x0c0001c0, 0xfc00dfc0, "cXx(b),t", pa10, 0},
+{ "ldcwx", 0x0c0001c0, 0xfc001fc0, "cXx(s,b),t", pa10, 0},
+{ "ldws", 0x0c001080, 0xfc00dfc0, "cM5(b),t", pa10, FLAG_STRICT},
+{ "ldws", 0x0c001080, 0xfc001fc0, "cM5(s,b),t", pa10, FLAG_STRICT},
+{ "ldws", 0x0c001080, 0xfc00d3c0, "cmcc5(b),t", pa11, FLAG_STRICT},
+{ "ldws", 0x0c001080, 0xfc0013c0, "cmcc5(s,b),t", pa11, FLAG_STRICT},
+{ "ldws", 0x0c001080, 0xfc00dfc0, "cM5(b),t", pa10, 0},
+{ "ldws", 0x0c001080, 0xfc001fc0, "cM5(s,b),t", pa10, 0},
+{ "ldhs", 0x0c001040, 0xfc00dfc0, "cM5(b),t", pa10, FLAG_STRICT},
+{ "ldhs", 0x0c001040, 0xfc001fc0, "cM5(s,b),t", pa10, FLAG_STRICT},
+{ "ldhs", 0x0c001040, 0xfc00d3c0, "cmcc5(b),t", pa11, FLAG_STRICT},
+{ "ldhs", 0x0c001040, 0xfc0013c0, "cmcc5(s,b),t", pa11, FLAG_STRICT},
+{ "ldhs", 0x0c001040, 0xfc00dfc0, "cM5(b),t", pa10, 0},
+{ "ldhs", 0x0c001040, 0xfc001fc0, "cM5(s,b),t", pa10, 0},
+{ "ldbs", 0x0c001000, 0xfc00dfc0, "cM5(b),t", pa10, FLAG_STRICT},
+{ "ldbs", 0x0c001000, 0xfc001fc0, "cM5(s,b),t", pa10, FLAG_STRICT},
+{ "ldbs", 0x0c001000, 0xfc00d3c0, "cmcc5(b),t", pa11, FLAG_STRICT},
+{ "ldbs", 0x0c001000, 0xfc0013c0, "cmcc5(s,b),t", pa11, FLAG_STRICT},
+{ "ldbs", 0x0c001000, 0xfc00dfc0, "cM5(b),t", pa10, 0},
+{ "ldbs", 0x0c001000, 0xfc001fc0, "cM5(s,b),t", pa10, 0},
+{ "ldwas", 0x0c001180, 0xfc00dfc0, "cM5(b),t", pa10, FLAG_STRICT},
+{ "ldwas", 0x0c001180, 0xfc00d3c0, "cmcc5(b),t", pa11, FLAG_STRICT},
+{ "ldwas", 0x0c001180, 0xfc00dfc0, "cM5(b),t", pa10, 0},
+{ "ldcws", 0x0c0011c0, 0xfc00dfc0, "cM5(b),t", pa10, FLAG_STRICT},
+{ "ldcws", 0x0c0011c0, 0xfc001fc0, "cM5(s,b),t", pa10, FLAG_STRICT},
+{ "ldcws", 0x0c0011c0, 0xfc00d3c0, "cmcd5(b),t", pa11, FLAG_STRICT},
+{ "ldcws", 0x0c0011c0, 0xfc0013c0, "cmcd5(s,b),t", pa11, FLAG_STRICT},
+{ "ldcws", 0x0c0011c0, 0xfc00dfc0, "cM5(b),t", pa10, 0},
+{ "ldcws", 0x0c0011c0, 0xfc001fc0, "cM5(s,b),t", pa10, 0},
+{ "stws", 0x0c001280, 0xfc00dfc0, "cMx,V(b)", pa10, FLAG_STRICT},
+{ "stws", 0x0c001280, 0xfc001fc0, "cMx,V(s,b)", pa10, FLAG_STRICT},
+{ "stws", 0x0c001280, 0xfc00d3c0, "cmcCx,V(b)", pa11, FLAG_STRICT},
+{ "stws", 0x0c001280, 0xfc0013c0, "cmcCx,V(s,b)", pa11, FLAG_STRICT},
+{ "stws", 0x0c001280, 0xfc00dfc0, "cMx,V(b)", pa10, 0},
+{ "stws", 0x0c001280, 0xfc001fc0, "cMx,V(s,b)", pa10, 0},
+{ "sths", 0x0c001240, 0xfc00dfc0, "cMx,V(b)", pa10, FLAG_STRICT},
+{ "sths", 0x0c001240, 0xfc001fc0, "cMx,V(s,b)", pa10, FLAG_STRICT},
+{ "sths", 0x0c001240, 0xfc00d3c0, "cmcCx,V(b)", pa11, FLAG_STRICT},
+{ "sths", 0x0c001240, 0xfc0013c0, "cmcCx,V(s,b)", pa11, FLAG_STRICT},
+{ "sths", 0x0c001240, 0xfc00dfc0, "cMx,V(b)", pa10, 0},
+{ "sths", 0x0c001240, 0xfc001fc0, "cMx,V(s,b)", pa10, 0},
+{ "stbs", 0x0c001200, 0xfc00dfc0, "cMx,V(b)", pa10, FLAG_STRICT},
+{ "stbs", 0x0c001200, 0xfc001fc0, "cMx,V(s,b)", pa10, FLAG_STRICT},
+{ "stbs", 0x0c001200, 0xfc00d3c0, "cmcCx,V(b)", pa11, FLAG_STRICT},
+{ "stbs", 0x0c001200, 0xfc0013c0, "cmcCx,V(s,b)", pa11, FLAG_STRICT},
+{ "stbs", 0x0c001200, 0xfc00dfc0, "cMx,V(b)", pa10, 0},
+{ "stbs", 0x0c001200, 0xfc001fc0, "cMx,V(s,b)", pa10, 0},
+{ "stwas", 0x0c001380, 0xfc00dfc0, "cMx,V(b)", pa10, FLAG_STRICT},
+{ "stwas", 0x0c001380, 0xfc00d3c0, "cmcCx,V(b)", pa11, FLAG_STRICT},
+{ "stwas", 0x0c001380, 0xfc00dfc0, "cMx,V(b)", pa10, 0},
+{ "stdby", 0x0c001340, 0xfc00d3c0, "cscCx,V(b)", pa20, FLAG_STRICT},
+{ "stdby", 0x0c001340, 0xfc0013c0, "cscCx,V(s,b)", pa20, FLAG_STRICT},
+{ "stbys", 0x0c001300, 0xfc00dfc0, "cAx,V(b)", pa10, FLAG_STRICT},
+{ "stbys", 0x0c001300, 0xfc001fc0, "cAx,V(s,b)", pa10, FLAG_STRICT},
+{ "stbys", 0x0c001300, 0xfc00d3c0, "cscCx,V(b)", pa11, FLAG_STRICT},
+{ "stbys", 0x0c001300, 0xfc0013c0, "cscCx,V(s,b)", pa11, FLAG_STRICT},
+{ "stbys", 0x0c001300, 0xfc00dfc0, "cAx,V(b)", pa10, 0},
+{ "stbys", 0x0c001300, 0xfc001fc0, "cAx,V(s,b)", pa10, 0},
+
+/* Immediate instructions. */
+{ "ldo", 0x34000000, 0xfc000000, "l(b),x", pa20w, 0},
+{ "ldo", 0x34000000, 0xfc00c000, "j(b),x", pa10, 0},
+{ "ldil", 0x20000000, 0xfc000000, "k,b", pa10, 0},
+{ "addil", 0x28000000, 0xfc000000, "k,b,Z", pa10, 0},
+{ "addil", 0x28000000, 0xfc000000, "k,b", pa10, 0},
+
+/* Branching instructions. */
+{ "b", 0xe8008000, 0xfc00e000, "cpnXL", pa20, FLAG_STRICT},
+{ "b", 0xe800a000, 0xfc00e000, "clnXL", pa20, FLAG_STRICT},
+{ "b", 0xe8000000, 0xfc00e000, "clnW,b", pa10, FLAG_STRICT},
+{ "b", 0xe8002000, 0xfc00e000, "cgnW,b", pa10, FLAG_STRICT},
+{ "b", 0xe8000000, 0xffe0e000, "nW", pa10, 0}, /* b,l foo,r0 */
+{ "bl", 0xe8000000, 0xfc00e000, "nW,b", pa10, 0},
+{ "gate", 0xe8002000, 0xfc00e000, "nW,b", pa10, 0},
+{ "blr", 0xe8004000, 0xfc00e001, "nx,b", pa10, 0},
+{ "bv", 0xe800c000, 0xfc00fffd, "nx(b)", pa10, 0},
+{ "bv", 0xe800c000, 0xfc00fffd, "n(b)", pa10, 0},
+{ "bve", 0xe800f001, 0xfc1ffffd, "cpn(b)L", pa20, FLAG_STRICT},
+{ "bve", 0xe800f000, 0xfc1ffffd, "cln(b)L", pa20, FLAG_STRICT},
+{ "bve", 0xe800d001, 0xfc1ffffd, "cPn(b)", pa20, FLAG_STRICT},
+{ "bve", 0xe800d000, 0xfc1ffffd, "n(b)", pa20, FLAG_STRICT},
+{ "be", 0xe4000000, 0xfc000000, "clnz(S,b),Y", pa10, FLAG_STRICT},
+{ "be", 0xe4000000, 0xfc000000, "clnz(b),Y", pa10, FLAG_STRICT},
+{ "be", 0xe0000000, 0xfc000000, "nz(S,b)", pa10, 0},
+{ "be", 0xe0000000, 0xfc000000, "nz(b)", pa10, 0},
+{ "ble", 0xe4000000, 0xfc000000, "nz(S,b)", pa10, 0},
+{ "movb", 0xc8000000, 0xfc000000, "?ynx,b,w", pa10, 0},
+{ "movib", 0xcc000000, 0xfc000000, "?yn5,b,w", pa10, 0},
+{ "combt", 0x80000000, 0xfc000000, "?tnx,b,w", pa10, 0},
+{ "combf", 0x88000000, 0xfc000000, "?tnx,b,w", pa10, 0},
+{ "comibt", 0x84000000, 0xfc000000, "?tn5,b,w", pa10, 0},
+{ "comibf", 0x8c000000, 0xfc000000, "?tn5,b,w", pa10, 0},
+{ "addbt", 0xa0000000, 0xfc000000, "?dnx,b,w", pa10, 0},
+{ "addbf", 0xa8000000, 0xfc000000, "?dnx,b,w", pa10, 0},
+{ "addibt", 0xa4000000, 0xfc000000, "?dn5,b,w", pa10, 0},
+{ "addibf", 0xac000000, 0xfc000000, "?dn5,b,w", pa10, 0},
+{ "bb", 0xc0004000, 0xffe06000, "?bnx,!,w", pa10, FLAG_STRICT},
+{ "bb", 0xc0006000, 0xffe06000, "?Bnx,!,w", pa20, FLAG_STRICT},
+{ "bb", 0xc4004000, 0xfc006000, "?bnx,Q,w", pa10, FLAG_STRICT},
+{ "bb", 0xc4004000, 0xfc004000, "?Bnx,B,w", pa20, FLAG_STRICT},
+{ "bvb", 0xc0004000, 0xffe04000, "?bnx,w", pa10, 0},
+{ "clrbts", 0xe8004005, 0xffffffff, "", pa20, FLAG_STRICT},
+{ "popbts", 0xe8004005, 0xfffff007, "$", pa20, FLAG_STRICT},
+{ "pushnom", 0xe8004001, 0xffffffff, "", pa20, FLAG_STRICT},
+{ "pushbts", 0xe8004001, 0xffe0ffff, "x", pa20, FLAG_STRICT},
+
+/* Computation Instructions. */
+
+{ "cmpclr", 0x080008a0, 0xfc000fe0, "?Sx,b,t", pa20, FLAG_STRICT},
+{ "cmpclr", 0x08000880, 0xfc000fe0, "?sx,b,t", pa10, FLAG_STRICT},
+{ "comclr", 0x08000880, 0xfc000fe0, "?sx,b,t", pa10, 0},
+{ "or", 0x08000260, 0xfc000fe0, "?Lx,b,t", pa20, FLAG_STRICT},
+{ "or", 0x08000240, 0xfc000fe0, "?lx,b,t", pa10, 0},
+{ "xor", 0x080002a0, 0xfc000fe0, "?Lx,b,t", pa20, FLAG_STRICT},
+{ "xor", 0x08000280, 0xfc000fe0, "?lx,b,t", pa10, 0},
+{ "and", 0x08000220, 0xfc000fe0, "?Lx,b,t", pa20, FLAG_STRICT},
+{ "and", 0x08000200, 0xfc000fe0, "?lx,b,t", pa10, 0},
+{ "andcm", 0x08000020, 0xfc000fe0, "?Lx,b,t", pa20, FLAG_STRICT},
+{ "andcm", 0x08000000, 0xfc000fe0, "?lx,b,t", pa10, 0},
+{ "uxor", 0x080003a0, 0xfc000fe0, "?Ux,b,t", pa20, FLAG_STRICT},
+{ "uxor", 0x08000380, 0xfc000fe0, "?ux,b,t", pa10, 0},
+{ "uaddcm", 0x080009a0, 0xfc000fa0, "cT?Ux,b,t", pa20, FLAG_STRICT},
+{ "uaddcm", 0x08000980, 0xfc000fa0, "cT?ux,b,t", pa10, FLAG_STRICT},
+{ "uaddcm", 0x08000980, 0xfc000fe0, "?ux,b,t", pa10, 0},
+{ "uaddcmt", 0x080009c0, 0xfc000fe0, "?ux,b,t", pa10, 0},
+{ "dcor", 0x08000ba0, 0xfc1f0fa0, "ci?Ub,t", pa20, FLAG_STRICT},
+{ "dcor", 0x08000b80, 0xfc1f0fa0, "ci?ub,t", pa10, FLAG_STRICT},
+{ "dcor", 0x08000b80, 0xfc1f0fe0, "?ub,t", pa10, 0},
+{ "idcor", 0x08000bc0, 0xfc1f0fe0, "?ub,t", pa10, 0},
+{ "addi", 0xb0000000, 0xfc000000, "ct?ai,b,x", pa10, FLAG_STRICT},
+{ "addi", 0xb4000000, 0xfc000000, "cv?ai,b,x", pa10, FLAG_STRICT},
+{ "addi", 0xb4000000, 0xfc000800, "?ai,b,x", pa10, 0},
+{ "addio", 0xb4000800, 0xfc000800, "?ai,b,x", pa10, 0},
+{ "addit", 0xb0000000, 0xfc000800, "?ai,b,x", pa10, 0},
+{ "addito", 0xb0000800, 0xfc000800, "?ai,b,x", pa10, 0},
+{ "add", 0x08000720, 0xfc0007e0, "cY?Ax,b,t", pa20, FLAG_STRICT},
+{ "add", 0x08000700, 0xfc0007e0, "cy?ax,b,t", pa10, FLAG_STRICT},
+{ "add", 0x08000220, 0xfc0003e0, "ca?Ax,b,t", pa20, FLAG_STRICT},
+{ "add", 0x08000200, 0xfc0003e0, "ca?ax,b,t", pa10, FLAG_STRICT},
+{ "add", 0x08000600, 0xfc000fe0, "?ax,b,t", pa10, 0},
+{ "addl", 0x08000a00, 0xfc000fe0, "?ax,b,t", pa10, 0},
+{ "addo", 0x08000e00, 0xfc000fe0, "?ax,b,t", pa10, 0},
+{ "addc", 0x08000700, 0xfc000fe0, "?ax,b,t", pa10, 0},
+{ "addco", 0x08000f00, 0xfc000fe0, "?ax,b,t", pa10, 0},
+{ "sub", 0x080004e0, 0xfc0007e0, "ct?Sx,b,t", pa20, FLAG_STRICT},
+{ "sub", 0x080004c0, 0xfc0007e0, "ct?sx,b,t", pa10, FLAG_STRICT},
+{ "sub", 0x08000520, 0xfc0007e0, "cB?Sx,b,t", pa20, FLAG_STRICT},
+{ "sub", 0x08000500, 0xfc0007e0, "cb?sx,b,t", pa10, FLAG_STRICT},
+{ "sub", 0x08000420, 0xfc0007e0, "cv?Sx,b,t", pa20, FLAG_STRICT},
+{ "sub", 0x08000400, 0xfc0007e0, "cv?sx,b,t", pa10, FLAG_STRICT},
+{ "sub", 0x08000400, 0xfc000fe0, "?sx,b,t", pa10, 0},
+{ "subo", 0x08000c00, 0xfc000fe0, "?sx,b,t", pa10, 0},
+{ "subb", 0x08000500, 0xfc000fe0, "?sx,b,t", pa10, 0},
+{ "subbo", 0x08000d00, 0xfc000fe0, "?sx,b,t", pa10, 0},
+{ "subt", 0x080004c0, 0xfc000fe0, "?sx,b,t", pa10, 0},
+{ "subto", 0x08000cc0, 0xfc000fe0, "?sx,b,t", pa10, 0},
+{ "ds", 0x08000440, 0xfc000fe0, "?sx,b,t", pa10, 0},
+{ "subi", 0x94000000, 0xfc000000, "cv?si,b,x", pa10, FLAG_STRICT},
+{ "subi", 0x94000000, 0xfc000800, "?si,b,x", pa10, 0},
+{ "subio", 0x94000800, 0xfc000800, "?si,b,x", pa10, 0},
+{ "cmpiclr", 0x90000800, 0xfc000800, "?Si,b,x", pa20, FLAG_STRICT},
+{ "cmpiclr", 0x90000000, 0xfc000800, "?si,b,x", pa10, FLAG_STRICT},
+{ "comiclr", 0x90000000, 0xfc000800, "?si,b,x", pa10, 0},
+{ "shladd", 0x08000220, 0xfc000320, "ca?Ax,.,b,t", pa20, FLAG_STRICT},
+{ "shladd", 0x08000200, 0xfc000320, "ca?ax,.,b,t", pa10, FLAG_STRICT},
+{ "sh1add", 0x08000640, 0xfc000fe0, "?ax,b,t", pa10, 0},
+{ "sh1addl", 0x08000a40, 0xfc000fe0, "?ax,b,t", pa10, 0},
+{ "sh1addo", 0x08000e40, 0xfc000fe0, "?ax,b,t", pa10, 0},
+{ "sh2add", 0x08000680, 0xfc000fe0, "?ax,b,t", pa10, 0},
+{ "sh2addl", 0x08000a80, 0xfc000fe0, "?ax,b,t", pa10, 0},
+{ "sh2addo", 0x08000e80, 0xfc000fe0, "?ax,b,t", pa10, 0},
+{ "sh3add", 0x080006c0, 0xfc000fe0, "?ax,b,t", pa10, 0},
+{ "sh3addl", 0x08000ac0, 0xfc000fe0, "?ax,b,t", pa10, 0},
+{ "sh3addo", 0x08000ec0, 0xfc000fe0, "?ax,b,t", pa10, 0},
+
+/* Subword Operation Instructions. */
+
+{ "hadd", 0x08000300, 0xfc00ff20, "cHx,b,t", pa20, FLAG_STRICT},
+{ "havg", 0x080002c0, 0xfc00ffe0, "x,b,t", pa20, FLAG_STRICT},
+{ "hshl", 0xf8008800, 0xffe0fc20, "x,*,t", pa20, FLAG_STRICT},
+{ "hshladd", 0x08000700, 0xfc00ff20, "x,.,b,t", pa20, FLAG_STRICT},
+{ "hshr", 0xf800c800, 0xfc1ff820, "cSb,*,t", pa20, FLAG_STRICT},
+{ "hshradd", 0x08000500, 0xfc00ff20, "x,.,b,t", pa20, FLAG_STRICT},
+{ "hsub", 0x08000100, 0xfc00ff20, "cHx,b,t", pa20, FLAG_STRICT},
+{ "mixh", 0xf8008400, 0xfc009fe0, "chx,b,t", pa20, FLAG_STRICT},
+{ "mixw", 0xf8008000, 0xfc009fe0, "chx,b,t", pa20, FLAG_STRICT},
+{ "permh", 0xf8000000, 0xfc009020, "c*a,t", pa20, FLAG_STRICT},
+
+
+/* Extract and Deposit Instructions. */
+
+{ "shrpd", 0xd0000200, 0xfc001fe0, "?Xx,b,!,t", pa20, FLAG_STRICT},
+{ "shrpd", 0xd0000400, 0xfc001400, "?Xx,b,~,t", pa20, FLAG_STRICT},
+{ "shrpw", 0xd0000000, 0xfc001fe0, "?xx,b,!,t", pa10, FLAG_STRICT},
+{ "shrpw", 0xd0000800, 0xfc001c00, "?xx,b,p,t", pa10, FLAG_STRICT},
+{ "vshd", 0xd0000000, 0xfc001fe0, "?xx,b,t", pa10, 0},
+{ "shd", 0xd0000800, 0xfc001c00, "?xx,b,p,t", pa10, 0},
+{ "extrd", 0xd0001200, 0xfc001ae0, "cS?Xb,!,%,x", pa20, FLAG_STRICT},
+{ "extrd", 0xd8000000, 0xfc000000, "cS?Xb,q,|,x", pa20, FLAG_STRICT},
+{ "extrw", 0xd0001000, 0xfc001be0, "cS?xb,!,T,x", pa10, FLAG_STRICT},
+{ "extrw", 0xd0001800, 0xfc001800, "cS?xb,P,T,x", pa10, FLAG_STRICT},
+{ "vextru", 0xd0001000, 0xfc001fe0, "?xb,T,x", pa10, 0},
+{ "vextrs", 0xd0001400, 0xfc001fe0, "?xb,T,x", pa10, 0},
+{ "extru", 0xd0001800, 0xfc001c00, "?xb,P,T,x", pa10, 0},
+{ "extrs", 0xd0001c00, 0xfc001c00, "?xb,P,T,x", pa10, 0},
+{ "depd", 0xd4000200, 0xfc001ae0, "cz?Xx,!,%,b", pa20, FLAG_STRICT},
+{ "depd", 0xf0000000, 0xfc000000, "cz?Xx,~,|,b", pa20, FLAG_STRICT},
+{ "depdi", 0xd4001200, 0xfc001ae0, "cz?X5,!,%,b", pa20, FLAG_STRICT},
+{ "depdi", 0xf4000000, 0xfc000000, "cz?X5,~,|,b", pa20, FLAG_STRICT},
+{ "depw", 0xd4000000, 0xfc001be0, "cz?xx,!,T,b", pa10, FLAG_STRICT},
+{ "depw", 0xd4000800, 0xfc001800, "cz?xx,p,T,b", pa10, FLAG_STRICT},
+{ "depwi", 0xd4001000, 0xfc001be0, "cz?x5,!,T,b", pa10, FLAG_STRICT},
+{ "depwi", 0xd4001800, 0xfc001800, "cz?x5,p,T,b", pa10, FLAG_STRICT},
+{ "zvdep", 0xd4000000, 0xfc001fe0, "?xx,T,b", pa10, 0},
+{ "vdep", 0xd4000400, 0xfc001fe0, "?xx,T,b", pa10, 0},
+{ "zdep", 0xd4000800, 0xfc001c00, "?xx,p,T,b", pa10, 0},
+{ "dep", 0xd4000c00, 0xfc001c00, "?xx,p,T,b", pa10, 0},
+{ "zvdepi", 0xd4001000, 0xfc001fe0, "?x5,T,b", pa10, 0},
+{ "vdepi", 0xd4001400, 0xfc001fe0, "?x5,T,b", pa10, 0},
+{ "zdepi", 0xd4001800, 0xfc001c00, "?x5,p,T,b", pa10, 0},
+{ "depi", 0xd4001c00, 0xfc001c00, "?x5,p,T,b", pa10, 0},
+
+/* System Control Instructions. */
+
+{ "break", 0x00000000, 0xfc001fe0, "r,A", pa10, 0},
+{ "rfi", 0x00000c00, 0xffffff1f, "cr", pa10, FLAG_STRICT},
+{ "rfi", 0x00000c00, 0xffffffff, "", pa10, 0},
+{ "rfir", 0x00000ca0, 0xffffffff, "", pa11, 0},
+{ "ssm", 0x00000d60, 0xfc00ffe0, "U,t", pa20, FLAG_STRICT},
+{ "ssm", 0x00000d60, 0xffe0ffe0, "R,t", pa10, 0},
+{ "rsm", 0x00000e60, 0xfc00ffe0, "U,t", pa20, FLAG_STRICT},
+{ "rsm", 0x00000e60, 0xffe0ffe0, "R,t", pa10, 0},
+{ "mtsm", 0x00001860, 0xffe0ffff, "x", pa10, 0},
+{ "ldsid", 0x000010a0, 0xfc1fffe0, "(b),t", pa10, 0},
+{ "ldsid", 0x000010a0, 0xfc1f3fe0, "(s,b),t", pa10, 0},
+{ "mtsp", 0x00001820, 0xffe01fff, "x,S", pa10, 0},
+{ "mtctl", 0x00001840, 0xfc00ffff, "x,^", pa10, 0},
+{ "mtsarcm", 0x016018C0, 0xffe0ffff, "x", pa20, FLAG_STRICT},
+{ "mfia", 0x000014A0, 0xffffffe0, "t", pa20, FLAG_STRICT},
+{ "mfsp", 0x000004a0, 0xffff1fe0, "S,t", pa10, 0},
+{ "mfctl", 0x016048a0, 0xffffffe0, "cW!,t", pa20, FLAG_STRICT},
+{ "mfctl", 0x000008a0, 0xfc1fffe0, "^,t", pa10, 0},
+{ "sync", 0x00000400, 0xffffffff, "", pa10, 0},
+{ "syncdma", 0x00100400, 0xffffffff, "", pa10, 0},
+{ "probe", 0x04001180, 0xfc00ffa0, "cw(b),x,t", pa10, FLAG_STRICT},
+{ "probe", 0x04001180, 0xfc003fa0, "cw(s,b),x,t", pa10, FLAG_STRICT},
+{ "probei", 0x04003180, 0xfc00ffa0, "cw(b),R,t", pa10, FLAG_STRICT},
+{ "probei", 0x04003180, 0xfc003fa0, "cw(s,b),R,t", pa10, FLAG_STRICT},
+{ "prober", 0x04001180, 0xfc00ffe0, "(b),x,t", pa10, 0},
+{ "prober", 0x04001180, 0xfc003fe0, "(s,b),x,t", pa10, 0},
+{ "proberi", 0x04003180, 0xfc00ffe0, "(b),R,t", pa10, 0},
+{ "proberi", 0x04003180, 0xfc003fe0, "(s,b),R,t", pa10, 0},
+{ "probew", 0x040011c0, 0xfc00ffe0, "(b),x,t", pa10, 0},
+{ "probew", 0x040011c0, 0xfc003fe0, "(s,b),x,t", pa10, 0},
+{ "probewi", 0x040031c0, 0xfc00ffe0, "(b),R,t", pa10, 0},
+{ "probewi", 0x040031c0, 0xfc003fe0, "(s,b),R,t", pa10, 0},
+{ "lpa", 0x04001340, 0xfc00ffc0, "cZx(b),t", pa10, 0},
+{ "lpa", 0x04001340, 0xfc003fc0, "cZx(s,b),t", pa10, 0},
+{ "lci", 0x04001300, 0xfc00ffe0, "x(b),t", pa11, 0},
+{ "lci", 0x04001300, 0xfc003fe0, "x(s,b),t", pa11, 0},
+{ "pdtlb", 0x04001600, 0xfc00ffdf, "cLcZx(b)", pa20, FLAG_STRICT},
+{ "pdtlb", 0x04001600, 0xfc003fdf, "cLcZx(s,b)", pa20, FLAG_STRICT},
+{ "pdtlb", 0x04001600, 0xfc1fffdf, "cLcZ@(b)", pa20, FLAG_STRICT},
+{ "pdtlb", 0x04001600, 0xfc1f3fdf, "cLcZ@(s,b)", pa20, FLAG_STRICT},
+{ "pdtlb", 0x04001200, 0xfc00ffdf, "cZx(b)", pa10, 0},
+{ "pdtlb", 0x04001200, 0xfc003fdf, "cZx(s,b)", pa10, 0},
+{ "pitlb", 0x04000600, 0xfc001fdf, "cLcZx(S,b)", pa20, FLAG_STRICT},
+{ "pitlb", 0x04000600, 0xfc1f1fdf, "cLcZ@(S,b)", pa20, FLAG_STRICT},
+{ "pitlb", 0x04000200, 0xfc001fdf, "cZx(S,b)", pa10, 0},
+{ "pdtlbe", 0x04001240, 0xfc00ffdf, "cZx(b)", pa10, 0},
+{ "pdtlbe", 0x04001240, 0xfc003fdf, "cZx(s,b)", pa10, 0},
+{ "pitlbe", 0x04000240, 0xfc001fdf, "cZx(S,b)", pa10, 0},
+{ "idtlba", 0x04001040, 0xfc00ffff, "x,(b)", pa10, 0},
+{ "idtlba", 0x04001040, 0xfc003fff, "x,(s,b)", pa10, 0},
+{ "iitlba", 0x04000040, 0xfc001fff, "x,(S,b)", pa10, 0},
+{ "idtlbp", 0x04001000, 0xfc00ffff, "x,(b)", pa10, 0},
+{ "idtlbp", 0x04001000, 0xfc003fff, "x,(s,b)", pa10, 0},
+{ "iitlbp", 0x04000000, 0xfc001fff, "x,(S,b)", pa10, 0},
+{ "pdc", 0x04001380, 0xfc00ffdf, "cZx(b)", pa10, 0},
+{ "pdc", 0x04001380, 0xfc003fdf, "cZx(s,b)", pa10, 0},
+{ "fdc", 0x04001280, 0xfc00ffdf, "cZx(b)", pa10, FLAG_STRICT},
+{ "fdc", 0x04001280, 0xfc003fdf, "cZx(s,b)", pa10, FLAG_STRICT},
+{ "fdc", 0x04003280, 0xfc00ffff, "5(b)", pa20, FLAG_STRICT},
+{ "fdc", 0x04003280, 0xfc003fff, "5(s,b)", pa20, FLAG_STRICT},
+{ "fdc", 0x04001280, 0xfc00ffdf, "cZx(b)", pa10, 0},
+{ "fdc", 0x04001280, 0xfc003fdf, "cZx(s,b)", pa10, 0},
+{ "fic", 0x040013c0, 0xfc00dfdf, "cZx(b)", pa20, FLAG_STRICT},
+{ "fic", 0x04000280, 0xfc001fdf, "cZx(S,b)", pa10, 0},
+{ "fdce", 0x040012c0, 0xfc00ffdf, "cZx(b)", pa10, 0},
+{ "fdce", 0x040012c0, 0xfc003fdf, "cZx(s,b)", pa10, 0},
+{ "fice", 0x040002c0, 0xfc001fdf, "cZx(S,b)", pa10, 0},
+{ "diag", 0x14000000, 0xfc000000, "D", pa10, 0},
+{ "idtlbt", 0x04001800, 0xfc00ffff, "x,b", pa20, FLAG_STRICT},
+{ "iitlbt", 0x04000800, 0xfc00ffff, "x,b", pa20, FLAG_STRICT},
+
+/* These may be specific to certain versions of the PA. Joel claimed
+ they were 72000 (7200?) specific. However, I'm almost certain the
+ mtcpu/mfcpu were undocumented, but available in the older 700 machines. */
+{ "mtcpu", 0x14001600, 0xfc00ffff, "x,^", pa10, 0},
+{ "mfcpu", 0x14001A00, 0xfc00ffff, "^,x", pa10, 0},
+{ "tocen", 0x14403600, 0xffffffff, "", pa10, 0},
+{ "tocdis", 0x14401620, 0xffffffff, "", pa10, 0},
+{ "shdwgr", 0x14402600, 0xffffffff, "", pa10, 0},
+{ "grshdw", 0x14400620, 0xffffffff, "", pa10, 0},
+
+/* gfw and gfr are not in the HP PA 1.1 manual, but they are in either
+ the Timex FPU or the Mustang ERS (not sure which) manual. */
+{ "gfw", 0x04001680, 0xfc00ffdf, "cZx(b)", pa11, 0},
+{ "gfw", 0x04001680, 0xfc003fdf, "cZx(s,b)", pa11, 0},
+{ "gfr", 0x04001a80, 0xfc00ffdf, "cZx(b)", pa11, 0},
+{ "gfr", 0x04001a80, 0xfc003fdf, "cZx(s,b)", pa11, 0},
+
+/* Floating Point Coprocessor Instructions. */
+
+{ "fldw", 0x24000000, 0xfc00df80, "cXx(b),fT", pa10, FLAG_STRICT},
+{ "fldw", 0x24000000, 0xfc001f80, "cXx(s,b),fT", pa10, FLAG_STRICT},
+{ "fldw", 0x24000000, 0xfc00d380, "cxccx(b),fT", pa11, FLAG_STRICT},
+{ "fldw", 0x24000000, 0xfc001380, "cxccx(s,b),fT", pa11, FLAG_STRICT},
+{ "fldw", 0x24001020, 0xfc1ff3a0, "cocc@(b),fT", pa20, FLAG_STRICT},
+{ "fldw", 0x24001020, 0xfc1f33a0, "cocc@(s,b),fT", pa20, FLAG_STRICT},
+{ "fldw", 0x24001000, 0xfc00df80, "cM5(b),fT", pa10, FLAG_STRICT},
+{ "fldw", 0x24001000, 0xfc001f80, "cM5(s,b),fT", pa10, FLAG_STRICT},
+{ "fldw", 0x24001000, 0xfc00d380, "cmcc5(b),fT", pa11, FLAG_STRICT},
+{ "fldw", 0x24001000, 0xfc001380, "cmcc5(s,b),fT", pa11, FLAG_STRICT},
+{ "fldw", 0x5c000000, 0xfc000004, "y(b),fe", pa20w, FLAG_STRICT},
+{ "fldw", 0x58000000, 0xfc000000, "cJy(b),fe", pa20w, FLAG_STRICT},
+{ "fldw", 0x5c000000, 0xfc00c004, "d(b),fe", pa20, FLAG_STRICT},
+{ "fldw", 0x5c000000, 0xfc000004, "d(s,b),fe", pa20, FLAG_STRICT},
+{ "fldw", 0x58000000, 0xfc00c000, "cJd(b),fe", pa20, FLAG_STRICT},
+{ "fldw", 0x58000000, 0xfc000000, "cJd(s,b),fe", pa20, FLAG_STRICT},
+{ "fldd", 0x2c000000, 0xfc00dfc0, "cXx(b),ft", pa10, FLAG_STRICT},
+{ "fldd", 0x2c000000, 0xfc001fc0, "cXx(s,b),ft", pa10, FLAG_STRICT},
+{ "fldd", 0x2c000000, 0xfc00d3c0, "cxccx(b),ft", pa11, FLAG_STRICT},
+{ "fldd", 0x2c000000, 0xfc0013c0, "cxccx(s,b),ft", pa11, FLAG_STRICT},
+{ "fldd", 0x2c001020, 0xfc1ff3e0, "cocc@(b),ft", pa20, FLAG_STRICT},
+{ "fldd", 0x2c001020, 0xfc1f33e0, "cocc@(s,b),ft", pa20, FLAG_STRICT},
+{ "fldd", 0x2c001000, 0xfc00dfc0, "cM5(b),ft", pa10, FLAG_STRICT},
+{ "fldd", 0x2c001000, 0xfc001fc0, "cM5(s,b),ft", pa10, FLAG_STRICT},
+{ "fldd", 0x2c001000, 0xfc00d3c0, "cmcc5(b),ft", pa11, FLAG_STRICT},
+{ "fldd", 0x2c001000, 0xfc0013c0, "cmcc5(s,b),ft", pa11, FLAG_STRICT},
+{ "fldd", 0x50000002, 0xfc000002, "cq&(b),fx", pa20w, FLAG_STRICT},
+{ "fldd", 0x50000002, 0xfc00c002, "cq#(b),fx", pa20, FLAG_STRICT},
+{ "fldd", 0x50000002, 0xfc000002, "cq#(s,b),fx", pa20, FLAG_STRICT},
+{ "fstw", 0x24000200, 0xfc00df80, "cXfT,x(b)", pa10, FLAG_STRICT},
+{ "fstw", 0x24000200, 0xfc001f80, "cXfT,x(s,b)", pa10, FLAG_STRICT},
+{ "fstw", 0x24000200, 0xfc00d380, "cxcCfT,x(b)", pa11, FLAG_STRICT},
+{ "fstw", 0x24000200, 0xfc001380, "cxcCfT,x(s,b)", pa11, FLAG_STRICT},
+{ "fstw", 0x24001220, 0xfc1ff3a0, "cocCfT,@(b)", pa20, FLAG_STRICT},
+{ "fstw", 0x24001220, 0xfc1f33a0, "cocCfT,@(s,b)", pa20, FLAG_STRICT},
+{ "fstw", 0x24001200, 0xfc00df80, "cMfT,5(b)", pa10, FLAG_STRICT},
+{ "fstw", 0x24001200, 0xfc001f80, "cMfT,5(s,b)", pa10, FLAG_STRICT},
+{ "fstw", 0x24001200, 0xfc00df80, "cMfT,5(b)", pa10, FLAG_STRICT},
+{ "fstw", 0x24001200, 0xfc001f80, "cMfT,5(s,b)", pa10, FLAG_STRICT},
+{ "fstw", 0x7c000000, 0xfc000004, "fE,y(b)", pa20w, FLAG_STRICT},
+{ "fstw", 0x78000000, 0xfc000000, "cJfE,y(b)", pa20w, FLAG_STRICT},
+{ "fstw", 0x7c000000, 0xfc00c004, "fE,d(b)", pa20, FLAG_STRICT},
+{ "fstw", 0x7c000000, 0xfc000004, "fE,d(s,b)", pa20, FLAG_STRICT},
+{ "fstw", 0x78000000, 0xfc00c000, "cJfE,d(b)", pa20, FLAG_STRICT},
+{ "fstw", 0x78000000, 0xfc000000, "cJfE,d(s,b)", pa20, FLAG_STRICT},
+{ "fstd", 0x2c000200, 0xfc00dfc0, "cXft,x(b)", pa10, FLAG_STRICT},
+{ "fstd", 0x2c000200, 0xfc001fc0, "cXft,x(s,b)", pa10, FLAG_STRICT},
+{ "fstd", 0x2c000200, 0xfc00d3c0, "cxcCft,x(b)", pa11, FLAG_STRICT},
+{ "fstd", 0x2c000200, 0xfc0013c0, "cxcCft,x(s,b)", pa11, FLAG_STRICT},
+{ "fstd", 0x2c001220, 0xfc1ff3e0, "cocCft,@(b)", pa20, FLAG_STRICT},
+{ "fstd", 0x2c001220, 0xfc1f33e0, "cocCft,@(s,b)", pa20, FLAG_STRICT},
+{ "fstd", 0x2c001200, 0xfc00dfc0, "cMft,5(b)", pa10, FLAG_STRICT},
+{ "fstd", 0x2c001200, 0xfc001fc0, "cMft,5(s,b)", pa10, FLAG_STRICT},
+{ "fstd", 0x2c001200, 0xfc00d3c0, "cmcCft,5(b)", pa11, FLAG_STRICT},
+{ "fstd", 0x2c001200, 0xfc0013c0, "cmcCft,5(s,b)", pa11, FLAG_STRICT},
+{ "fstd", 0x70000002, 0xfc000002, "cqfx,&(b)", pa20w, FLAG_STRICT},
+{ "fstd", 0x70000002, 0xfc00c002, "cqfx,#(b)", pa20, FLAG_STRICT},
+{ "fstd", 0x70000002, 0xfc000002, "cqfx,#(s,b)", pa20, FLAG_STRICT},
+{ "fldwx", 0x24000000, 0xfc00df80, "cXx(b),fT", pa10, FLAG_STRICT},
+{ "fldwx", 0x24000000, 0xfc001f80, "cXx(s,b),fT", pa10, FLAG_STRICT},
+{ "fldwx", 0x24000000, 0xfc00d380, "cxccx(b),fT", pa11, FLAG_STRICT},
+{ "fldwx", 0x24000000, 0xfc001380, "cxccx(s,b),fT", pa11, FLAG_STRICT},
+{ "fldwx", 0x24000000, 0xfc00df80, "cXx(b),fT", pa10, 0},
+{ "fldwx", 0x24000000, 0xfc001f80, "cXx(s,b),fT", pa10, 0},
+{ "flddx", 0x2c000000, 0xfc00dfc0, "cXx(b),ft", pa10, FLAG_STRICT},
+{ "flddx", 0x2c000000, 0xfc001fc0, "cXx(s,b),ft", pa10, FLAG_STRICT},
+{ "flddx", 0x2c000000, 0xfc00d3c0, "cxccx(b),ft", pa11, FLAG_STRICT},
+{ "flddx", 0x2c000000, 0xfc0013c0, "cxccx(s,b),ft", pa11, FLAG_STRICT},
+{ "flddx", 0x2c000000, 0xfc00dfc0, "cXx(b),ft", pa10, 0},
+{ "flddx", 0x2c000000, 0xfc001fc0, "cXx(s,b),ft", pa10, 0},
+{ "fstwx", 0x24000200, 0xfc00df80, "cxfT,x(b)", pa10, FLAG_STRICT},
+{ "fstwx", 0x24000200, 0xfc001f80, "cxfT,x(s,b)", pa10, FLAG_STRICT},
+{ "fstwx", 0x24000200, 0xfc00d380, "cxcCfT,x(b)", pa11, FLAG_STRICT},
+{ "fstwx", 0x24000200, 0xfc001380, "cxcCfT,x(s,b)", pa11, FLAG_STRICT},
+{ "fstwx", 0x24000200, 0xfc00df80, "cxfT,x(b)", pa10, 0},
+{ "fstwx", 0x24000200, 0xfc001f80, "cxfT,x(s,b)", pa10, 0},
+{ "fstdx", 0x2c000200, 0xfc00dfc0, "cxft,x(b)", pa10, FLAG_STRICT},
+{ "fstdx", 0x2c000200, 0xfc001fc0, "cxft,x(s,b)", pa10, FLAG_STRICT},
+{ "fstdx", 0x2c000200, 0xfc00d3c0, "cxcCft,x(b)", pa11, FLAG_STRICT},
+{ "fstdx", 0x2c000200, 0xfc0013c0, "cxcCft,x(s,b)", pa11, FLAG_STRICT},
+{ "fstdx", 0x2c000200, 0xfc00dfc0, "cxft,x(b)", pa10, 0},
+{ "fstdx", 0x2c000200, 0xfc001fc0, "cxft,x(s,b)", pa10, 0},
+{ "fstqx", 0x3c000200, 0xfc00dfc0, "cxft,x(b)", pa10, 0},
+{ "fstqx", 0x3c000200, 0xfc001fc0, "cxft,x(s,b)", pa10, 0},
+{ "fldws", 0x24001000, 0xfc00df80, "cm5(b),fT", pa10, FLAG_STRICT},
+{ "fldws", 0x24001000, 0xfc001f80, "cm5(s,b),fT", pa10, FLAG_STRICT},
+{ "fldws", 0x24001000, 0xfc00d380, "cmcc5(b),fT", pa11, FLAG_STRICT},
+{ "fldws", 0x24001000, 0xfc001380, "cmcc5(s,b),fT", pa11, FLAG_STRICT},
+{ "fldws", 0x24001000, 0xfc00df80, "cm5(b),fT", pa10, 0},
+{ "fldws", 0x24001000, 0xfc001f80, "cm5(s,b),fT", pa10, 0},
+{ "fldds", 0x2c001000, 0xfc00dfc0, "cm5(b),ft", pa10, FLAG_STRICT},
+{ "fldds", 0x2c001000, 0xfc001fc0, "cm5(s,b),ft", pa10, FLAG_STRICT},
+{ "fldds", 0x2c001000, 0xfc00d3c0, "cmcc5(b),ft", pa11, FLAG_STRICT},
+{ "fldds", 0x2c001000, 0xfc0013c0, "cmcc5(s,b),ft", pa11, FLAG_STRICT},
+{ "fldds", 0x2c001000, 0xfc00dfc0, "cm5(b),ft", pa10, 0},
+{ "fldds", 0x2c001000, 0xfc001fc0, "cm5(s,b),ft", pa10, 0},
+{ "fstws", 0x24001200, 0xfc00df80, "cmfT,5(b)", pa10, FLAG_STRICT},
+{ "fstws", 0x24001200, 0xfc001f80, "cmfT,5(s,b)", pa10, FLAG_STRICT},
+{ "fstws", 0x24001200, 0xfc00d380, "cmcCfT,5(b)", pa11, FLAG_STRICT},
+{ "fstws", 0x24001200, 0xfc001380, "cmcCfT,5(s,b)", pa11, FLAG_STRICT},
+{ "fstws", 0x24001200, 0xfc00df80, "cmfT,5(b)", pa10, 0},
+{ "fstws", 0x24001200, 0xfc001f80, "cmfT,5(s,b)", pa10, 0},
+{ "fstds", 0x2c001200, 0xfc00dfc0, "cmft,5(b)", pa10, FLAG_STRICT},
+{ "fstds", 0x2c001200, 0xfc001fc0, "cmft,5(s,b)", pa10, FLAG_STRICT},
+{ "fstds", 0x2c001200, 0xfc00d3c0, "cmcCft,5(b)", pa11, FLAG_STRICT},
+{ "fstds", 0x2c001200, 0xfc0013c0, "cmcCft,5(s,b)", pa11, FLAG_STRICT},
+{ "fstds", 0x2c001200, 0xfc00dfc0, "cmft,5(b)", pa10, 0},
+{ "fstds", 0x2c001200, 0xfc001fc0, "cmft,5(s,b)", pa10, 0},
+{ "fstqs", 0x3c001200, 0xfc00dfc0, "cmft,5(b)", pa10, 0},
+{ "fstqs", 0x3c001200, 0xfc001fc0, "cmft,5(s,b)", pa10, 0},
+{ "fadd", 0x30000600, 0xfc00e7e0, "Ffa,fb,fT", pa10, 0},
+{ "fadd", 0x38000600, 0xfc00e720, "IfA,fB,fT", pa10, 0},
+{ "fsub", 0x30002600, 0xfc00e7e0, "Ffa,fb,fT", pa10, 0},
+{ "fsub", 0x38002600, 0xfc00e720, "IfA,fB,fT", pa10, 0},
+{ "fmpy", 0x30004600, 0xfc00e7e0, "Ffa,fb,fT", pa10, 0},
+{ "fmpy", 0x38004600, 0xfc00e720, "IfA,fB,fT", pa10, 0},
+{ "fdiv", 0x30006600, 0xfc00e7e0, "Ffa,fb,fT", pa10, 0},
+{ "fdiv", 0x38006600, 0xfc00e720, "IfA,fB,fT", pa10, 0},
+{ "fsqrt", 0x30008000, 0xfc1fe7e0, "Ffa,fT", pa10, 0},
+{ "fsqrt", 0x38008000, 0xfc1fe720, "FfA,fT", pa10, 0},
+{ "fabs", 0x30006000, 0xfc1fe7e0, "Ffa,fT", pa10, 0},
+{ "fabs", 0x38006000, 0xfc1fe720, "FfA,fT", pa10, 0},
+{ "frem", 0x30008600, 0xfc00e7e0, "Ffa,fb,fT", pa10, 0},
+{ "frem", 0x38008600, 0xfc00e720, "FfA,fB,fT", pa10, 0},
+{ "frnd", 0x3000a000, 0xfc1fe7e0, "Ffa,fT", pa10, 0},
+{ "frnd", 0x3800a000, 0xfc1fe720, "FfA,fT", pa10, 0},
+{ "fcpy", 0x30004000, 0xfc1fe7e0, "Ffa,fT", pa10, 0},
+{ "fcpy", 0x38004000, 0xfc1fe720, "FfA,fT", pa10, 0},
+{ "fcnvff", 0x30000200, 0xfc1f87e0, "FGfa,fT", pa10, 0},
+{ "fcnvff", 0x38000200, 0xfc1f8720, "FGfA,fT", pa10, 0},
+{ "fcnvxf", 0x30008200, 0xfc1f87e0, "FGfa,fT", pa10, 0},
+{ "fcnvxf", 0x38008200, 0xfc1f8720, "FGfA,fT", pa10, 0},
+{ "fcnvfx", 0x30010200, 0xfc1f87e0, "FGfa,fT", pa10, 0},
+{ "fcnvfx", 0x38010200, 0xfc1f8720, "FGfA,fT", pa10, 0},
+{ "fcnvfxt", 0x30018200, 0xfc1f87e0, "FGfa,fT", pa10, 0},
+{ "fcnvfxt", 0x38018200, 0xfc1f8720, "FGfA,fT", pa10, 0},
+{ "fmpyfadd", 0xb8000000, 0xfc000020, "IfA,fB,fC,fT", pa20, FLAG_STRICT},
+{ "fmpynfadd", 0xb8000020, 0xfc000020, "IfA,fB,fC,fT", pa20, FLAG_STRICT},
+{ "fneg", 0x3000c000, 0xfc1fe7e0, "Ffa,fT", pa20, FLAG_STRICT},
+{ "fneg", 0x3800c000, 0xfc1fe720, "IfA,fT", pa20, FLAG_STRICT},
+{ "fnegabs", 0x3000e000, 0xfc1fe7e0, "Ffa,fT", pa20, FLAG_STRICT},
+{ "fnegabs", 0x3800e000, 0xfc1fe720, "IfA,fT", pa20, FLAG_STRICT},
+{ "fcnv", 0x30000200, 0xfc1c0720, "{_fa,fT", pa20, FLAG_STRICT},
+{ "fcnv", 0x38000200, 0xfc1c0720, "FGfA,fT", pa20, FLAG_STRICT},
+{ "fcmp", 0x30000400, 0xfc00e7e0, "F?ffa,fb", pa10, FLAG_STRICT},
+{ "fcmp", 0x38000400, 0xfc00e720, "I?ffA,fB", pa10, FLAG_STRICT},
+{ "fcmp", 0x30000400, 0xfc0007e0, "F?ffa,fb,h", pa20, FLAG_STRICT},
+{ "fcmp", 0x38000400, 0xfc000720, "I?ffA,fB,h", pa20, FLAG_STRICT},
+{ "fcmp", 0x30000400, 0xfc00e7e0, "F?ffa,fb", pa10, 0},
+{ "fcmp", 0x38000400, 0xfc00e720, "I?ffA,fB", pa10, 0},
+{ "xmpyu", 0x38004700, 0xfc00e720, "fX,fB,fT", pa11, 0},
+{ "fmpyadd", 0x18000000, 0xfc000000, "Hfi,fj,fk,fl,fm", pa11, 0},
+{ "fmpysub", 0x98000000, 0xfc000000, "Hfi,fj,fk,fl,fm", pa11, 0},
+{ "ftest", 0x30002420, 0xffffffff, "", pa10, FLAG_STRICT},
+{ "ftest", 0x30002420, 0xffffffe0, ",=", pa20, FLAG_STRICT},
+{ "ftest", 0x30000420, 0xffff1fff, "m", pa20, FLAG_STRICT},
+{ "fid", 0x30000000, 0xffffffff, "", pa11, 0},
+
+/* Performance Monitor Instructions. */
+
+{ "pmdis", 0x30000280, 0xffffffdf, "N", pa20, FLAG_STRICT},
+{ "pmenb", 0x30000680, 0xffffffff, "", pa20, FLAG_STRICT},
+
+/* Assist Instructions. */
+
+{ "spop0", 0x10000000, 0xfc000600, "v,ON", pa10, 0},
+{ "spop1", 0x10000200, 0xfc000600, "v,oNt", pa10, 0},
+{ "spop2", 0x10000400, 0xfc000600, "v,1Nb", pa10, 0},
+{ "spop3", 0x10000600, 0xfc000600, "v,0Nx,b", pa10, 0},
+{ "copr", 0x30000000, 0xfc000000, "u,2N", pa10, 0},
+{ "cldw", 0x24000000, 0xfc00de00, "ucXx(b),t", pa10, FLAG_STRICT},
+{ "cldw", 0x24000000, 0xfc001e00, "ucXx(s,b),t", pa10, FLAG_STRICT},
+{ "cldw", 0x24000000, 0xfc00d200, "ucxccx(b),t", pa11, FLAG_STRICT},
+{ "cldw", 0x24000000, 0xfc001200, "ucxccx(s,b),t", pa11, FLAG_STRICT},
+{ "cldw", 0x24001000, 0xfc00d200, "ucocc@(b),t", pa20, FLAG_STRICT},
+{ "cldw", 0x24001000, 0xfc001200, "ucocc@(s,b),t", pa20, FLAG_STRICT},
+{ "cldw", 0x24001000, 0xfc00de00, "ucM5(b),t", pa10, FLAG_STRICT},
+{ "cldw", 0x24001000, 0xfc001e00, "ucM5(s,b),t", pa10, FLAG_STRICT},
+{ "cldw", 0x24001000, 0xfc00d200, "ucmcc5(b),t", pa11, FLAG_STRICT},
+{ "cldw", 0x24001000, 0xfc001200, "ucmcc5(s,b),t", pa11, FLAG_STRICT},
+{ "cldd", 0x2c000000, 0xfc00de00, "ucXx(b),t", pa10, FLAG_STRICT},
+{ "cldd", 0x2c000000, 0xfc001e00, "ucXx(s,b),t", pa10, FLAG_STRICT},
+{ "cldd", 0x2c000000, 0xfc00d200, "ucxccx(b),t", pa11, FLAG_STRICT},
+{ "cldd", 0x2c000000, 0xfc001200, "ucxccx(s,b),t", pa11, FLAG_STRICT},
+{ "cldd", 0x2c001000, 0xfc00d200, "ucocc@(b),t", pa20, FLAG_STRICT},
+{ "cldd", 0x2c001000, 0xfc001200, "ucocc@(s,b),t", pa20, FLAG_STRICT},
+{ "cldd", 0x2c001000, 0xfc00de00, "ucM5(b),t", pa10, FLAG_STRICT},
+{ "cldd", 0x2c001000, 0xfc001e00, "ucM5(s,b),t", pa10, FLAG_STRICT},
+{ "cldd", 0x2c001000, 0xfc00d200, "ucmcc5(b),t", pa11, FLAG_STRICT},
+{ "cldd", 0x2c001000, 0xfc001200, "ucmcc5(s,b),t", pa11, FLAG_STRICT},
+{ "cstw", 0x24000200, 0xfc00de00, "ucXt,x(b)", pa10, FLAG_STRICT},
+{ "cstw", 0x24000200, 0xfc001e00, "ucXt,x(s,b)", pa10, FLAG_STRICT},
+{ "cstw", 0x24000200, 0xfc00d200, "ucxcCt,x(b)", pa11, FLAG_STRICT},
+{ "cstw", 0x24000200, 0xfc001200, "ucxcCt,x(s,b)", pa11, FLAG_STRICT},
+{ "cstw", 0x24001200, 0xfc00d200, "ucocCt,@(b)", pa20, FLAG_STRICT},
+{ "cstw", 0x24001200, 0xfc001200, "ucocCt,@(s,b)", pa20, FLAG_STRICT},
+{ "cstw", 0x24001200, 0xfc00de00, "ucMt,5(b)", pa10, FLAG_STRICT},
+{ "cstw", 0x24001200, 0xfc001e00, "ucMt,5(s,b)", pa10, FLAG_STRICT},
+{ "cstw", 0x24001200, 0xfc00d200, "ucmcCt,5(b)", pa11, FLAG_STRICT},
+{ "cstw", 0x24001200, 0xfc001200, "ucmcCt,5(s,b)", pa11, FLAG_STRICT},
+{ "cstd", 0x2c000200, 0xfc00de00, "ucXt,x(b)", pa10, FLAG_STRICT},
+{ "cstd", 0x2c000200, 0xfc001e00, "ucXt,x(s,b)", pa10, FLAG_STRICT},
+{ "cstd", 0x2c000200, 0xfc00d200, "ucxcCt,x(b)", pa11, FLAG_STRICT},
+{ "cstd", 0x2c000200, 0xfc001200, "ucxcCt,x(s,b)", pa11, FLAG_STRICT},
+{ "cstd", 0x2c001200, 0xfc00d200, "ucocCt,@(b)", pa20, FLAG_STRICT},
+{ "cstd", 0x2c001200, 0xfc001200, "ucocCt,@(s,b)", pa20, FLAG_STRICT},
+{ "cstd", 0x2c001200, 0xfc00de00, "ucMt,5(b)", pa10, FLAG_STRICT},
+{ "cstd", 0x2c001200, 0xfc001e00, "ucMt,5(s,b)", pa10, FLAG_STRICT},
+{ "cstd", 0x2c001200, 0xfc00d200, "ucmcCt,5(b)", pa11, FLAG_STRICT},
+{ "cstd", 0x2c001200, 0xfc001200, "ucmcCt,5(s,b)", pa11, FLAG_STRICT},
+{ "cldwx", 0x24000000, 0xfc00de00, "ucXx(b),t", pa10, FLAG_STRICT},
+{ "cldwx", 0x24000000, 0xfc001e00, "ucXx(s,b),t", pa10, FLAG_STRICT},
+{ "cldwx", 0x24000000, 0xfc00d200, "ucxccx(b),t", pa11, FLAG_STRICT},
+{ "cldwx", 0x24000000, 0xfc001200, "ucxccx(s,b),t", pa11, FLAG_STRICT},
+{ "cldwx", 0x24000000, 0xfc00de00, "ucXx(b),t", pa10, 0},
+{ "cldwx", 0x24000000, 0xfc001e00, "ucXx(s,b),t", pa10, 0},
+{ "clddx", 0x2c000000, 0xfc00de00, "ucXx(b),t", pa10, FLAG_STRICT},
+{ "clddx", 0x2c000000, 0xfc001e00, "ucXx(s,b),t", pa10, FLAG_STRICT},
+{ "clddx", 0x2c000000, 0xfc00d200, "ucxccx(b),t", pa11, FLAG_STRICT},
+{ "clddx", 0x2c000000, 0xfc001200, "ucxccx(s,b),t", pa11, FLAG_STRICT},
+{ "clddx", 0x2c000000, 0xfc00de00, "ucXx(b),t", pa10, 0},
+{ "clddx", 0x2c000000, 0xfc001e00, "ucXx(s,b),t", pa10, 0},
+{ "cstwx", 0x24000200, 0xfc00de00, "ucXt,x(b)", pa10, FLAG_STRICT},
+{ "cstwx", 0x24000200, 0xfc001e00, "ucXt,x(s,b)", pa10, FLAG_STRICT},
+{ "cstwx", 0x24000200, 0xfc00d200, "ucxcCt,x(b)", pa11, FLAG_STRICT},
+{ "cstwx", 0x24000200, 0xfc001200, "ucxcCt,x(s,b)", pa11, FLAG_STRICT},
+{ "cstwx", 0x24000200, 0xfc00de00, "ucXt,x(b)", pa10, 0},
+{ "cstwx", 0x24000200, 0xfc001e00, "ucXt,x(s,b)", pa10, 0},
+{ "cstdx", 0x2c000200, 0xfc00de00, "ucXt,x(b)", pa10, FLAG_STRICT},
+{ "cstdx", 0x2c000200, 0xfc001e00, "ucXt,x(s,b)", pa10, FLAG_STRICT},
+{ "cstdx", 0x2c000200, 0xfc00d200, "ucxcCt,x(b)", pa11, FLAG_STRICT},
+{ "cstdx", 0x2c000200, 0xfc001200, "ucxcCt,x(s,b)", pa11, FLAG_STRICT},
+{ "cstdx", 0x2c000200, 0xfc00de00, "ucXt,x(b)", pa10, 0},
+{ "cstdx", 0x2c000200, 0xfc001e00, "ucXt,x(s,b)", pa10, 0},
+{ "cldws", 0x24001000, 0xfc00de00, "ucM5(b),t", pa10, FLAG_STRICT},
+{ "cldws", 0x24001000, 0xfc001e00, "ucM5(s,b),t", pa10, FLAG_STRICT},
+{ "cldws", 0x24001000, 0xfc00d200, "ucmcc5(b),t", pa11, FLAG_STRICT},
+{ "cldws", 0x24001000, 0xfc001200, "ucmcc5(s,b),t", pa11, FLAG_STRICT},
+{ "cldws", 0x24001000, 0xfc00de00, "ucM5(b),t", pa10, 0},
+{ "cldws", 0x24001000, 0xfc001e00, "ucM5(s,b),t", pa10, 0},
+{ "cldds", 0x2c001000, 0xfc00de00, "ucM5(b),t", pa10, FLAG_STRICT},
+{ "cldds", 0x2c001000, 0xfc001e00, "ucM5(s,b),t", pa10, FLAG_STRICT},
+{ "cldds", 0x2c001000, 0xfc00d200, "ucmcc5(b),t", pa11, FLAG_STRICT},
+{ "cldds", 0x2c001000, 0xfc001200, "ucmcc5(s,b),t", pa11, FLAG_STRICT},
+{ "cldds", 0x2c001000, 0xfc00de00, "ucM5(b),t", pa10, 0},
+{ "cldds", 0x2c001000, 0xfc001e00, "ucM5(s,b),t", pa10, 0},
+{ "cstws", 0x24001200, 0xfc00de00, "ucMt,5(b)", pa10, FLAG_STRICT},
+{ "cstws", 0x24001200, 0xfc001e00, "ucMt,5(s,b)", pa10, FLAG_STRICT},
+{ "cstws", 0x24001200, 0xfc00d200, "ucmcCt,5(b)", pa11, FLAG_STRICT},
+{ "cstws", 0x24001200, 0xfc001200, "ucmcCt,5(s,b)", pa11, FLAG_STRICT},
+{ "cstws", 0x24001200, 0xfc00de00, "ucMt,5(b)", pa10, 0},
+{ "cstws", 0x24001200, 0xfc001e00, "ucMt,5(s,b)", pa10, 0},
+{ "cstds", 0x2c001200, 0xfc00de00, "ucMt,5(b)", pa10, FLAG_STRICT},
+{ "cstds", 0x2c001200, 0xfc001e00, "ucMt,5(s,b)", pa10, FLAG_STRICT},
+{ "cstds", 0x2c001200, 0xfc00d200, "ucmcCt,5(b)", pa11, FLAG_STRICT},
+{ "cstds", 0x2c001200, 0xfc001200, "ucmcCt,5(s,b)", pa11, FLAG_STRICT},
+{ "cstds", 0x2c001200, 0xfc00de00, "ucMt,5(b)", pa10, 0},
+{ "cstds", 0x2c001200, 0xfc001e00, "ucMt,5(s,b)", pa10, 0},
+
+/* More pseudo instructions which must follow the main table. */
+{ "call", 0xe800f000, 0xfc1ffffd, "n(b)", pa20, FLAG_STRICT},
+{ "call", 0xe800a000, 0xffe0e000, "nW", pa10, FLAG_STRICT},
+{ "ret", 0xe840d000, 0xfffffffd, "n", pa20, FLAG_STRICT},
+
+};
+
+#define NUMOPCODES ((sizeof pa_opcodes)/(sizeof pa_opcodes[0]))
+
+/* SKV 12/18/92. Added some denotations for various operands. */
+
+#define PA_IMM11_AT_31 'i'
+#define PA_IMM14_AT_31 'j'
+#define PA_IMM21_AT_31 'k'
+#define PA_DISP12 'w'
+#define PA_DISP17 'W'
+
+#define N_HPPA_OPERAND_FORMATS 5
+
+/* Integer register names, indexed by the numbers which appear in the
+ opcodes. */
+static const char *const reg_names[] =
+{
+ "flags", "r1", "rp", "r3", "r4", "r5", "r6", "r7", "r8", "r9",
+ "r10", "r11", "r12", "r13", "r14", "r15", "r16", "r17", "r18", "r19",
+ "r20", "r21", "r22", "r23", "r24", "r25", "r26", "dp", "ret0", "ret1",
+ "sp", "r31"
+};
+
+/* Floating point register names, indexed by the numbers which appear in the
+ opcodes. */
+static const char *const fp_reg_names[] =
+{
+ "fpsr", "fpe2", "fpe4", "fpe6",
+ "fr4", "fr5", "fr6", "fr7", "fr8",
+ "fr9", "fr10", "fr11", "fr12", "fr13", "fr14", "fr15",
+ "fr16", "fr17", "fr18", "fr19", "fr20", "fr21", "fr22", "fr23",
+ "fr24", "fr25", "fr26", "fr27", "fr28", "fr29", "fr30", "fr31"
+};
+
+typedef unsigned int CORE_ADDR;
+
+/* Get at various relevent fields of an instruction word. */
+
+#define MASK_5 0x1f
+#define MASK_10 0x3ff
+#define MASK_11 0x7ff
+#define MASK_14 0x3fff
+#define MASK_16 0xffff
+#define MASK_21 0x1fffff
+
+/* These macros get bit fields using HP's numbering (MSB = 0). */
+
+#define GET_FIELD(X, FROM, TO) \
+ ((X) >> (31 - (TO)) & ((1 << ((TO) - (FROM) + 1)) - 1))
+
+#define GET_BIT(X, WHICH) \
+ GET_FIELD (X, WHICH, WHICH)
+
+/* Some of these have been converted to 2-d arrays because they
+ consume less storage this way. If the maintenance becomes a
+ problem, convert them back to const 1-d pointer arrays. */
+static const char *const control_reg[] =
+{
+ "rctr", "cr1", "cr2", "cr3", "cr4", "cr5", "cr6", "cr7",
+ "pidr1", "pidr2", "ccr", "sar", "pidr3", "pidr4",
+ "iva", "eiem", "itmr", "pcsq", "pcoq", "iir", "isr",
+ "ior", "ipsw", "eirr", "tr0", "tr1", "tr2", "tr3",
+ "tr4", "tr5", "tr6", "tr7"
+};
+
+static const char *const compare_cond_names[] =
+{
+ "", ",=", ",<", ",<=", ",<<", ",<<=", ",sv", ",od",
+ ",tr", ",<>", ",>=", ",>", ",>>=", ",>>", ",nsv", ",ev"
+};
+static const char *const compare_cond_64_names[] =
+{
+ "", ",*=", ",*<", ",*<=", ",*<<", ",*<<=", ",*sv", ",*od",
+ ",*tr", ",*<>", ",*>=", ",*>", ",*>>=", ",*>>", ",*nsv", ",*ev"
+};
+static const char *const cmpib_cond_64_names[] =
+{
+ ",*<<", ",*=", ",*<", ",*<=", ",*>>=", ",*<>", ",*>=", ",*>"
+};
+static const char *const add_cond_names[] =
+{
+ "", ",=", ",<", ",<=", ",nuv", ",znv", ",sv", ",od",
+ ",tr", ",<>", ",>=", ",>", ",uv", ",vnz", ",nsv", ",ev"
+};
+static const char *const add_cond_64_names[] =
+{
+ "", ",*=", ",*<", ",*<=", ",*nuv", ",*znv", ",*sv", ",*od",
+ ",*tr", ",*<>", ",*>=", ",*>", ",*uv", ",*vnz", ",*nsv", ",*ev"
+};
+static const char *const wide_add_cond_names[] =
+{
+ "", ",=", ",<", ",<=", ",nuv", ",*=", ",*<", ",*<=",
+ ",tr", ",<>", ",>=", ",>", ",uv", ",*<>", ",*>=", ",*>"
+};
+static const char *const logical_cond_names[] =
+{
+ "", ",=", ",<", ",<=", 0, 0, 0, ",od",
+ ",tr", ",<>", ",>=", ",>", 0, 0, 0, ",ev"};
+static const char *const logical_cond_64_names[] =
+{
+ "", ",*=", ",*<", ",*<=", 0, 0, 0, ",*od",
+ ",*tr", ",*<>", ",*>=", ",*>", 0, 0, 0, ",*ev"};
+static const char *const unit_cond_names[] =
+{
+ "", ",swz", ",sbz", ",shz", ",sdc", ",swc", ",sbc", ",shc",
+ ",tr", ",nwz", ",nbz", ",nhz", ",ndc", ",nwc", ",nbc", ",nhc"
+};
+static const char *const unit_cond_64_names[] =
+{
+ "", ",*swz", ",*sbz", ",*shz", ",*sdc", ",*swc", ",*sbc", ",*shc",
+ ",*tr", ",*nwz", ",*nbz", ",*nhz", ",*ndc", ",*nwc", ",*nbc", ",*nhc"
+};
+static const char *const shift_cond_names[] =
+{
+ "", ",=", ",<", ",od", ",tr", ",<>", ",>=", ",ev"
+};
+static const char *const shift_cond_64_names[] =
+{
+ "", ",*=", ",*<", ",*od", ",*tr", ",*<>", ",*>=", ",*ev"
+};
+static const char *const bb_cond_64_names[] =
+{
+ ",*<", ",*>="
+};
+static const char *const index_compl_names[] = {"", ",m", ",s", ",sm"};
+static const char *const short_ldst_compl_names[] = {"", ",ma", "", ",mb"};
+static const char *const short_bytes_compl_names[] =
+{
+ "", ",b,m", ",e", ",e,m"
+};
+static const char *const float_format_names[] = {",sgl", ",dbl", "", ",quad"};
+static const char *const fcnv_fixed_names[] = {",w", ",dw", "", ",qw"};
+static const char *const fcnv_ufixed_names[] = {",uw", ",udw", "", ",uqw"};
+static const char *const float_comp_names[] =
+{
+ ",false?", ",false", ",?", ",!<=>", ",=", ",=t", ",?=", ",!<>",
+ ",!?>=", ",<", ",?<", ",!>=", ",!?>", ",<=", ",?<=", ",!>",
+ ",!?<=", ",>", ",?>", ",!<=", ",!?<", ",>=", ",?>=", ",!<",
+ ",!?=", ",<>", ",!=", ",!=t", ",!?", ",<=>", ",true?", ",true"
+};
+static const char *const signed_unsigned_names[] = {",u", ",s"};
+static const char *const mix_half_names[] = {",l", ",r"};
+static const char *const saturation_names[] = {",us", ",ss", 0, ""};
+static const char *const read_write_names[] = {",r", ",w"};
+static const char *const add_compl_names[] = { 0, "", ",l", ",tsv" };
+
+/* For a bunch of different instructions form an index into a
+ completer name table. */
+#define GET_COMPL(insn) (GET_FIELD (insn, 26, 26) | \
+ GET_FIELD (insn, 18, 18) << 1)
+
+#define GET_COND(insn) (GET_FIELD ((insn), 16, 18) + \
+ (GET_FIELD ((insn), 19, 19) ? 8 : 0))
+
+/* Utility function to print registers. Put these first, so gcc's function
+ inlining can do its stuff. */
+
+#define fputs_filtered(STR,F) (*info->fprintf_func) (info->stream, "%s", STR)
+
+static void
+fput_reg (unsigned reg, disassemble_info *info)
+{
+ (*info->fprintf_func) (info->stream, reg ? reg_names[reg] : "r0");
+}
+
+static void
+fput_fp_reg (unsigned reg, disassemble_info *info)
+{
+ (*info->fprintf_func) (info->stream, reg ? fp_reg_names[reg] : "fr0");
+}
+
+static void
+fput_fp_reg_r (unsigned reg, disassemble_info *info)
+{
+ /* Special case floating point exception registers. */
+ if (reg < 4)
+ (*info->fprintf_func) (info->stream, "fpe%d", reg * 2 + 1);
+ else
+ (*info->fprintf_func) (info->stream, "%sR",
+ reg ? fp_reg_names[reg] : "fr0");
+}
+
+static void
+fput_creg (unsigned reg, disassemble_info *info)
+{
+ (*info->fprintf_func) (info->stream, control_reg[reg]);
+}
+
+/* Print constants with sign. */
+
+static void
+fput_const (unsigned num, disassemble_info *info)
+{
+ if ((int) num < 0)
+ (*info->fprintf_func) (info->stream, "-%x", - (int) num);
+ else
+ (*info->fprintf_func) (info->stream, "%x", num);
+}
+
+/* Routines to extract various sized constants out of hppa
+ instructions. */
+
+/* Extract a 3-bit space register number from a be, ble, mtsp or mfsp. */
+static int
+extract_3 (unsigned word)
+{
+ return GET_FIELD (word, 18, 18) << 2 | GET_FIELD (word, 16, 17);
+}
+
+static int
+extract_5_load (unsigned word)
+{
+ return low_sign_extend (word >> 16 & MASK_5, 5);
+}
+
+/* Extract the immediate field from a st{bhw}s instruction. */
+
+static int
+extract_5_store (unsigned word)
+{
+ return low_sign_extend (word & MASK_5, 5);
+}
+
+/* Extract the immediate field from a break instruction. */
+
+static unsigned
+extract_5r_store (unsigned word)
+{
+ return (word & MASK_5);
+}
+
+/* Extract the immediate field from a {sr}sm instruction. */
+
+static unsigned
+extract_5R_store (unsigned word)
+{
+ return (word >> 16 & MASK_5);
+}
+
+/* Extract the 10 bit immediate field from a {sr}sm instruction. */
+
+static unsigned
+extract_10U_store (unsigned word)
+{
+ return (word >> 16 & MASK_10);
+}
+
+/* Extract the immediate field from a bb instruction. */
+
+static unsigned
+extract_5Q_store (unsigned word)
+{
+ return (word >> 21 & MASK_5);
+}
+
+/* Extract an 11 bit immediate field. */
+
+static int
+extract_11 (unsigned word)
+{
+ return low_sign_extend (word & MASK_11, 11);
+}
+
+/* Extract a 14 bit immediate field. */
+
+static int
+extract_14 (unsigned word)
+{
+ return low_sign_extend (word & MASK_14, 14);
+}
+
+/* Extract a 16 bit immediate field (PA2.0 wide only). */
+
+static int
+extract_16 (unsigned word)
+{
+ int m15, m0, m1;
+
+ m0 = GET_BIT (word, 16);
+ m1 = GET_BIT (word, 17);
+ m15 = GET_BIT (word, 31);
+ word = (word >> 1) & 0x1fff;
+ word = word | (m15 << 15) | ((m15 ^ m0) << 14) | ((m15 ^ m1) << 13);
+ return sign_extend (word, 16);
+}
+
+/* Extract a 21 bit constant. */
+
+static int
+extract_21 (unsigned word)
+{
+ int val;
+
+ word &= MASK_21;
+ word <<= 11;
+ val = GET_FIELD (word, 20, 20);
+ val <<= 11;
+ val |= GET_FIELD (word, 9, 19);
+ val <<= 2;
+ val |= GET_FIELD (word, 5, 6);
+ val <<= 5;
+ val |= GET_FIELD (word, 0, 4);
+ val <<= 2;
+ val |= GET_FIELD (word, 7, 8);
+ return sign_extend (val, 21) << 11;
+}
+
+/* Extract a 12 bit constant from branch instructions. */
+
+static int
+extract_12 (unsigned word)
+{
+ return sign_extend (GET_FIELD (word, 19, 28)
+ | GET_FIELD (word, 29, 29) << 10
+ | (word & 0x1) << 11, 12) << 2;
+}
+
+/* Extract a 17 bit constant from branch instructions, returning the
+ 19 bit signed value. */
+
+static int
+extract_17 (unsigned word)
+{
+ return sign_extend (GET_FIELD (word, 19, 28)
+ | GET_FIELD (word, 29, 29) << 10
+ | GET_FIELD (word, 11, 15) << 11
+ | (word & 0x1) << 16, 17) << 2;
+}
+
+static int
+extract_22 (unsigned word)
+{
+ return sign_extend (GET_FIELD (word, 19, 28)
+ | GET_FIELD (word, 29, 29) << 10
+ | GET_FIELD (word, 11, 15) << 11
+ | GET_FIELD (word, 6, 10) << 16
+ | (word & 0x1) << 21, 22) << 2;
+}
+
+/* Print one instruction. */
+
+int
+print_insn_hppa (bfd_vma memaddr, disassemble_info *info)
+{
+ bfd_byte buffer[4];
+ unsigned int insn, i;
+
+ {
+ int status =
+ (*info->read_memory_func) (memaddr, buffer, sizeof (buffer), info);
+ if (status != 0)
+ {
+ (*info->memory_error_func) (status, memaddr, info);
+ return -1;
+ }
+ }
+
+ insn = bfd_getb32 (buffer);
+
+ for (i = 0; i < NUMOPCODES; ++i)
+ {
+ const struct pa_opcode *opcode = &pa_opcodes[i];
+
+ if ((insn & opcode->mask) == opcode->match)
+ {
+ const char *s;
+#ifndef BFD64
+ if (opcode->arch == pa20w)
+ continue;
+#endif
+ (*info->fprintf_func) (info->stream, "%s", opcode->name);
+
+ if (!strchr ("cfCY?-+nHNZFIuv{", opcode->args[0]))
+ (*info->fprintf_func) (info->stream, " ");
+ for (s = opcode->args; *s != '\0'; ++s)
+ {
+ switch (*s)
+ {
+ case 'x':
+ fput_reg (GET_FIELD (insn, 11, 15), info);
+ break;
+ case 'a':
+ case 'b':
+ fput_reg (GET_FIELD (insn, 6, 10), info);
+ break;
+ case '^':
+ fput_creg (GET_FIELD (insn, 6, 10), info);
+ break;
+ case 't':
+ fput_reg (GET_FIELD (insn, 27, 31), info);
+ break;
+
+ /* Handle floating point registers. */
+ case 'f':
+ switch (*++s)
+ {
+ case 't':
+ fput_fp_reg (GET_FIELD (insn, 27, 31), info);
+ break;
+ case 'T':
+ if (GET_FIELD (insn, 25, 25))
+ fput_fp_reg_r (GET_FIELD (insn, 27, 31), info);
+ else
+ fput_fp_reg (GET_FIELD (insn, 27, 31), info);
+ break;
+ case 'a':
+ if (GET_FIELD (insn, 25, 25))
+ fput_fp_reg_r (GET_FIELD (insn, 6, 10), info);
+ else
+ fput_fp_reg (GET_FIELD (insn, 6, 10), info);
+ break;
+
+ /* 'fA' will not generate a space before the regsiter
+ name. Normally that is fine. Except that it
+ causes problems with xmpyu which has no FP format
+ completer. */
+ case 'X':
+ fputs_filtered (" ", info);
+ /* FALLTHRU */
+
+ case 'A':
+ if (GET_FIELD (insn, 24, 24))
+ fput_fp_reg_r (GET_FIELD (insn, 6, 10), info);
+ else
+ fput_fp_reg (GET_FIELD (insn, 6, 10), info);
+ break;
+ case 'b':
+ if (GET_FIELD (insn, 25, 25))
+ fput_fp_reg_r (GET_FIELD (insn, 11, 15), info);
+ else
+ fput_fp_reg (GET_FIELD (insn, 11, 15), info);
+ break;
+ case 'B':
+ if (GET_FIELD (insn, 19, 19))
+ fput_fp_reg_r (GET_FIELD (insn, 11, 15), info);
+ else
+ fput_fp_reg (GET_FIELD (insn, 11, 15), info);
+ break;
+ case 'C':
+ {
+ int reg = GET_FIELD (insn, 21, 22);
+ reg |= GET_FIELD (insn, 16, 18) << 2;
+ if (GET_FIELD (insn, 23, 23) != 0)
+ fput_fp_reg_r (reg, info);
+ else
+ fput_fp_reg (reg, info);
+ break;
+ }
+ case 'i':
+ {
+ int reg = GET_FIELD (insn, 6, 10);
+
+ reg |= (GET_FIELD (insn, 26, 26) << 4);
+ fput_fp_reg (reg, info);
+ break;
+ }
+ case 'j':
+ {
+ int reg = GET_FIELD (insn, 11, 15);
+
+ reg |= (GET_FIELD (insn, 26, 26) << 4);
+ fput_fp_reg (reg, info);
+ break;
+ }
+ case 'k':
+ {
+ int reg = GET_FIELD (insn, 27, 31);
+
+ reg |= (GET_FIELD (insn, 26, 26) << 4);
+ fput_fp_reg (reg, info);
+ break;
+ }
+ case 'l':
+ {
+ int reg = GET_FIELD (insn, 21, 25);
+
+ reg |= (GET_FIELD (insn, 26, 26) << 4);
+ fput_fp_reg (reg, info);
+ break;
+ }
+ case 'm':
+ {
+ int reg = GET_FIELD (insn, 16, 20);
+
+ reg |= (GET_FIELD (insn, 26, 26) << 4);
+ fput_fp_reg (reg, info);
+ break;
+ }
+
+ /* 'fe' will not generate a space before the register
+ name. Normally that is fine. Except that it
+ causes problems with fstw fe,y(b) which has no FP
+ format completer. */
+ case 'E':
+ fputs_filtered (" ", info);
+ /* FALLTHRU */
+
+ case 'e':
+ if (GET_FIELD (insn, 30, 30))
+ fput_fp_reg_r (GET_FIELD (insn, 11, 15), info);
+ else
+ fput_fp_reg (GET_FIELD (insn, 11, 15), info);
+ break;
+ case 'x':
+ fput_fp_reg (GET_FIELD (insn, 11, 15), info);
+ break;
+ }
+ break;
+
+ case '5':
+ fput_const (extract_5_load (insn), info);
+ break;
+ case 's':
+ {
+ int space = GET_FIELD (insn, 16, 17);
+ /* Zero means implicit addressing, not use of sr0. */
+ if (space != 0)
+ (*info->fprintf_func) (info->stream, "sr%d", space);
+ }
+ break;
+
+ case 'S':
+ (*info->fprintf_func) (info->stream, "sr%d",
+ extract_3 (insn));
+ break;
+
+ /* Handle completers. */
+ case 'c':
+ switch (*++s)
+ {
+ case 'x':
+ (*info->fprintf_func)
+ (info->stream, "%s",
+ index_compl_names[GET_COMPL (insn)]);
+ break;
+ case 'X':
+ (*info->fprintf_func)
+ (info->stream, "%s ",
+ index_compl_names[GET_COMPL (insn)]);
+ break;
+ case 'm':
+ (*info->fprintf_func)
+ (info->stream, "%s",
+ short_ldst_compl_names[GET_COMPL (insn)]);
+ break;
+ case 'M':
+ (*info->fprintf_func)
+ (info->stream, "%s ",
+ short_ldst_compl_names[GET_COMPL (insn)]);
+ break;
+ case 'A':
+ (*info->fprintf_func)
+ (info->stream, "%s ",
+ short_bytes_compl_names[GET_COMPL (insn)]);
+ break;
+ case 's':
+ (*info->fprintf_func)
+ (info->stream, "%s",
+ short_bytes_compl_names[GET_COMPL (insn)]);
+ break;
+ case 'c':
+ case 'C':
+ switch (GET_FIELD (insn, 20, 21))
+ {
+ case 1:
+ (*info->fprintf_func) (info->stream, ",bc ");
+ break;
+ case 2:
+ (*info->fprintf_func) (info->stream, ",sl ");
+ break;
+ default:
+ (*info->fprintf_func) (info->stream, " ");
+ }
+ break;
+ case 'd':
+ switch (GET_FIELD (insn, 20, 21))
+ {
+ case 1:
+ (*info->fprintf_func) (info->stream, ",co ");
+ break;
+ default:
+ (*info->fprintf_func) (info->stream, " ");
+ }
+ break;
+ case 'o':
+ (*info->fprintf_func) (info->stream, ",o");
+ break;
+ case 'g':
+ (*info->fprintf_func) (info->stream, ",gate");
+ break;
+ case 'p':
+ (*info->fprintf_func) (info->stream, ",l,push");
+ break;
+ case 'P':
+ (*info->fprintf_func) (info->stream, ",pop");
+ break;
+ case 'l':
+ case 'L':
+ (*info->fprintf_func) (info->stream, ",l");
+ break;
+ case 'w':
+ (*info->fprintf_func)
+ (info->stream, "%s ",
+ read_write_names[GET_FIELD (insn, 25, 25)]);
+ break;
+ case 'W':
+ (*info->fprintf_func) (info->stream, ",w ");
+ break;
+ case 'r':
+ if (GET_FIELD (insn, 23, 26) == 5)
+ (*info->fprintf_func) (info->stream, ",r");
+ break;
+ case 'Z':
+ if (GET_FIELD (insn, 26, 26))
+ (*info->fprintf_func) (info->stream, ",m ");
+ else
+ (*info->fprintf_func) (info->stream, " ");
+ break;
+ case 'i':
+ if (GET_FIELD (insn, 25, 25))
+ (*info->fprintf_func) (info->stream, ",i");
+ break;
+ case 'z':
+ if (!GET_FIELD (insn, 21, 21))
+ (*info->fprintf_func) (info->stream, ",z");
+ break;
+ case 'a':
+ (*info->fprintf_func)
+ (info->stream, "%s",
+ add_compl_names[GET_FIELD (insn, 20, 21)]);
+ break;
+ case 'Y':
+ (*info->fprintf_func)
+ (info->stream, ",dc%s",
+ add_compl_names[GET_FIELD (insn, 20, 21)]);
+ break;
+ case 'y':
+ (*info->fprintf_func)
+ (info->stream, ",c%s",
+ add_compl_names[GET_FIELD (insn, 20, 21)]);
+ break;
+ case 'v':
+ if (GET_FIELD (insn, 20, 20))
+ (*info->fprintf_func) (info->stream, ",tsv");
+ break;
+ case 't':
+ (*info->fprintf_func) (info->stream, ",tc");
+ if (GET_FIELD (insn, 20, 20))
+ (*info->fprintf_func) (info->stream, ",tsv");
+ break;
+ case 'B':
+ (*info->fprintf_func) (info->stream, ",db");
+ if (GET_FIELD (insn, 20, 20))
+ (*info->fprintf_func) (info->stream, ",tsv");
+ break;
+ case 'b':
+ (*info->fprintf_func) (info->stream, ",b");
+ if (GET_FIELD (insn, 20, 20))
+ (*info->fprintf_func) (info->stream, ",tsv");
+ break;
+ case 'T':
+ if (GET_FIELD (insn, 25, 25))
+ (*info->fprintf_func) (info->stream, ",tc");
+ break;
+ case 'S':
+ /* EXTRD/W has a following condition. */
+ if (*(s + 1) == '?')
+ (*info->fprintf_func)
+ (info->stream, "%s",
+ signed_unsigned_names[GET_FIELD (insn, 21, 21)]);
+ else
+ (*info->fprintf_func)
+ (info->stream, "%s ",
+ signed_unsigned_names[GET_FIELD (insn, 21, 21)]);
+ break;
+ case 'h':
+ (*info->fprintf_func)
+ (info->stream, "%s",
+ mix_half_names[GET_FIELD (insn, 17, 17)]);
+ break;
+ case 'H':
+ (*info->fprintf_func)
+ (info->stream, "%s ",
+ saturation_names[GET_FIELD (insn, 24, 25)]);
+ break;
+ case '*':
+ (*info->fprintf_func)
+ (info->stream, ",%d%d%d%d ",
+ GET_FIELD (insn, 17, 18), GET_FIELD (insn, 20, 21),
+ GET_FIELD (insn, 22, 23), GET_FIELD (insn, 24, 25));
+ break;
+
+ case 'q':
+ {
+ int m, a;
+
+ m = GET_FIELD (insn, 28, 28);
+ a = GET_FIELD (insn, 29, 29);
+
+ if (m && !a)
+ fputs_filtered (",ma ", info);
+ else if (m && a)
+ fputs_filtered (",mb ", info);
+ else
+ fputs_filtered (" ", info);
+ break;
+ }
+
+ case 'J':
+ {
+ int opc = GET_FIELD (insn, 0, 5);
+
+ if (opc == 0x16 || opc == 0x1e)
+ {
+ if (GET_FIELD (insn, 29, 29) == 0)
+ fputs_filtered (",ma ", info);
+ else
+ fputs_filtered (",mb ", info);
+ }
+ else
+ fputs_filtered (" ", info);
+ break;
+ }
+
+ case 'e':
+ {
+ int opc = GET_FIELD (insn, 0, 5);
+
+ if (opc == 0x13 || opc == 0x1b)
+ {
+ if (GET_FIELD (insn, 18, 18) == 1)
+ fputs_filtered (",mb ", info);
+ else
+ fputs_filtered (",ma ", info);
+ }
+ else if (opc == 0x17 || opc == 0x1f)
+ {
+ if (GET_FIELD (insn, 31, 31) == 1)
+ fputs_filtered (",ma ", info);
+ else
+ fputs_filtered (",mb ", info);
+ }
+ else
+ fputs_filtered (" ", info);
+
+ break;
+ }
+ }
+ break;
+
+ /* Handle conditions. */
+ case '?':
+ {
+ s++;
+ switch (*s)
+ {
+ case 'f':
+ (*info->fprintf_func)
+ (info->stream, "%s ",
+ float_comp_names[GET_FIELD (insn, 27, 31)]);
+ break;
+
+ /* These four conditions are for the set of instructions
+ which distinguish true/false conditions by opcode
+ rather than by the 'f' bit (sigh): comb, comib,
+ addb, addib. */
+ case 't':
+ fputs_filtered
+ (compare_cond_names[GET_FIELD (insn, 16, 18)], info);
+ break;
+ case 'n':
+ fputs_filtered
+ (compare_cond_names[GET_FIELD (insn, 16, 18)
+ + GET_FIELD (insn, 4, 4) * 8],
+ info);
+ break;
+ case 'N':
+ fputs_filtered
+ (compare_cond_64_names[GET_FIELD (insn, 16, 18)
+ + GET_FIELD (insn, 2, 2) * 8],
+ info);
+ break;
+ case 'Q':
+ fputs_filtered
+ (cmpib_cond_64_names[GET_FIELD (insn, 16, 18)],
+ info);
+ break;
+ case '@':
+ fputs_filtered
+ (add_cond_names[GET_FIELD (insn, 16, 18)
+ + GET_FIELD (insn, 4, 4) * 8],
+ info);
+ break;
+ case 's':
+ (*info->fprintf_func)
+ (info->stream, "%s ",
+ compare_cond_names[GET_COND (insn)]);
+ break;
+ case 'S':
+ (*info->fprintf_func)
+ (info->stream, "%s ",
+ compare_cond_64_names[GET_COND (insn)]);
+ break;
+ case 'a':
+ (*info->fprintf_func)
+ (info->stream, "%s ",
+ add_cond_names[GET_COND (insn)]);
+ break;
+ case 'A':
+ (*info->fprintf_func)
+ (info->stream, "%s ",
+ add_cond_64_names[GET_COND (insn)]);
+ break;
+ case 'd':
+ (*info->fprintf_func)
+ (info->stream, "%s",
+ add_cond_names[GET_FIELD (insn, 16, 18)]);
+ break;
+
+ case 'W':
+ (*info->fprintf_func)
+ (info->stream, "%s",
+ wide_add_cond_names[GET_FIELD (insn, 16, 18) +
+ GET_FIELD (insn, 4, 4) * 8]);
+ break;
+
+ case 'l':
+ (*info->fprintf_func)
+ (info->stream, "%s ",
+ logical_cond_names[GET_COND (insn)]);
+ break;
+ case 'L':
+ (*info->fprintf_func)
+ (info->stream, "%s ",
+ logical_cond_64_names[GET_COND (insn)]);
+ break;
+ case 'u':
+ (*info->fprintf_func)
+ (info->stream, "%s ",
+ unit_cond_names[GET_COND (insn)]);
+ break;
+ case 'U':
+ (*info->fprintf_func)
+ (info->stream, "%s ",
+ unit_cond_64_names[GET_COND (insn)]);
+ break;
+ case 'y':
+ case 'x':
+ case 'b':
+ (*info->fprintf_func)
+ (info->stream, "%s",
+ shift_cond_names[GET_FIELD (insn, 16, 18)]);
+
+ /* If the next character in args is 'n', it will handle
+ putting out the space. */
+ if (s[1] != 'n')
+ (*info->fprintf_func) (info->stream, " ");
+ break;
+ case 'X':
+ (*info->fprintf_func)
+ (info->stream, "%s ",
+ shift_cond_64_names[GET_FIELD (insn, 16, 18)]);
+ break;
+ case 'B':
+ (*info->fprintf_func)
+ (info->stream, "%s",
+ bb_cond_64_names[GET_FIELD (insn, 16, 16)]);
+
+ /* If the next character in args is 'n', it will handle
+ putting out the space. */
+ if (s[1] != 'n')
+ (*info->fprintf_func) (info->stream, " ");
+ break;
+ }
+ break;
+ }
+
+ case 'V':
+ fput_const (extract_5_store (insn), info);
+ break;
+ case 'r':
+ fput_const (extract_5r_store (insn), info);
+ break;
+ case 'R':
+ fput_const (extract_5R_store (insn), info);
+ break;
+ case 'U':
+ fput_const (extract_10U_store (insn), info);
+ break;
+ case 'B':
+ case 'Q':
+ fput_const (extract_5Q_store (insn), info);
+ break;
+ case 'i':
+ fput_const (extract_11 (insn), info);
+ break;
+ case 'j':
+ fput_const (extract_14 (insn), info);
+ break;
+ case 'k':
+ fputs_filtered ("L%", info);
+ fput_const (extract_21 (insn), info);
+ break;
+ case '<':
+ case 'l':
+ /* 16-bit long disp., PA2.0 wide only. */
+ fput_const (extract_16 (insn), info);
+ break;
+ case 'n':
+ if (insn & 0x2)
+ (*info->fprintf_func) (info->stream, ",n ");
+ else
+ (*info->fprintf_func) (info->stream, " ");
+ break;
+ case 'N':
+ if ((insn & 0x20) && s[1])
+ (*info->fprintf_func) (info->stream, ",n ");
+ else if (insn & 0x20)
+ (*info->fprintf_func) (info->stream, ",n");
+ else if (s[1])
+ (*info->fprintf_func) (info->stream, " ");
+ break;
+ case 'w':
+ (*info->print_address_func)
+ (memaddr + 8 + extract_12 (insn), info);
+ break;
+ case 'W':
+ /* 17 bit PC-relative branch. */
+ (*info->print_address_func)
+ ((memaddr + 8 + extract_17 (insn)), info);
+ break;
+ case 'z':
+ /* 17 bit displacement. This is an offset from a register
+ so it gets disasssembled as just a number, not any sort
+ of address. */
+ fput_const (extract_17 (insn), info);
+ break;
+
+ case 'Z':
+ /* addil %r1 implicit output. */
+ fputs_filtered ("r1", info);
+ break;
+
+ case 'Y':
+ /* be,l %sr0,%r31 implicit output. */
+ fputs_filtered ("sr0,r31", info);
+ break;
+
+ case '@':
+ (*info->fprintf_func) (info->stream, "0");
+ break;
+
+ case '.':
+ (*info->fprintf_func) (info->stream, "%d",
+ GET_FIELD (insn, 24, 25));
+ break;
+ case '*':
+ (*info->fprintf_func) (info->stream, "%d",
+ GET_FIELD (insn, 22, 25));
+ break;
+ case '!':
+ fputs_filtered ("sar", info);
+ break;
+ case 'p':
+ (*info->fprintf_func) (info->stream, "%d",
+ 31 - GET_FIELD (insn, 22, 26));
+ break;
+ case '~':
+ {
+ int num;
+ num = GET_FIELD (insn, 20, 20) << 5;
+ num |= GET_FIELD (insn, 22, 26);
+ (*info->fprintf_func) (info->stream, "%d", 63 - num);
+ break;
+ }
+ case 'P':
+ (*info->fprintf_func) (info->stream, "%d",
+ GET_FIELD (insn, 22, 26));
+ break;
+ case 'q':
+ {
+ int num;
+ num = GET_FIELD (insn, 20, 20) << 5;
+ num |= GET_FIELD (insn, 22, 26);
+ (*info->fprintf_func) (info->stream, "%d", num);
+ break;
+ }
+ case 'T':
+ (*info->fprintf_func) (info->stream, "%d",
+ 32 - GET_FIELD (insn, 27, 31));
+ break;
+ case '%':
+ {
+ int num;
+ num = (GET_FIELD (insn, 23, 23) + 1) * 32;
+ num -= GET_FIELD (insn, 27, 31);
+ (*info->fprintf_func) (info->stream, "%d", num);
+ break;
+ }
+ case '|':
+ {
+ int num;
+ num = (GET_FIELD (insn, 19, 19) + 1) * 32;
+ num -= GET_FIELD (insn, 27, 31);
+ (*info->fprintf_func) (info->stream, "%d", num);
+ break;
+ }
+ case '$':
+ fput_const (GET_FIELD (insn, 20, 28), info);
+ break;
+ case 'A':
+ fput_const (GET_FIELD (insn, 6, 18), info);
+ break;
+ case 'D':
+ fput_const (GET_FIELD (insn, 6, 31), info);
+ break;
+ case 'v':
+ (*info->fprintf_func) (info->stream, ",%d",
+ GET_FIELD (insn, 23, 25));
+ break;
+ case 'O':
+ fput_const ((GET_FIELD (insn, 6,20) << 5 |
+ GET_FIELD (insn, 27, 31)), info);
+ break;
+ case 'o':
+ fput_const (GET_FIELD (insn, 6, 20), info);
+ break;
+ case '2':
+ fput_const ((GET_FIELD (insn, 6, 22) << 5 |
+ GET_FIELD (insn, 27, 31)), info);
+ break;
+ case '1':
+ fput_const ((GET_FIELD (insn, 11, 20) << 5 |
+ GET_FIELD (insn, 27, 31)), info);
+ break;
+ case '0':
+ fput_const ((GET_FIELD (insn, 16, 20) << 5 |
+ GET_FIELD (insn, 27, 31)), info);
+ break;
+ case 'u':
+ (*info->fprintf_func) (info->stream, ",%d",
+ GET_FIELD (insn, 23, 25));
+ break;
+ case 'F':
+ /* If no destination completer and not before a completer
+ for fcmp, need a space here. */
+ if (s[1] == 'G' || s[1] == '?')
+ fputs_filtered
+ (float_format_names[GET_FIELD (insn, 19, 20)], info);
+ else
+ (*info->fprintf_func)
+ (info->stream, "%s ",
+ float_format_names[GET_FIELD (insn, 19, 20)]);
+ break;
+ case 'G':
+ (*info->fprintf_func)
+ (info->stream, "%s ",
+ float_format_names[GET_FIELD (insn, 17, 18)]);
+ break;
+ case 'H':
+ if (GET_FIELD (insn, 26, 26) == 1)
+ (*info->fprintf_func) (info->stream, "%s ",
+ float_format_names[0]);
+ else
+ (*info->fprintf_func) (info->stream, "%s ",
+ float_format_names[1]);
+ break;
+ case 'I':
+ /* If no destination completer and not before a completer
+ for fcmp, need a space here. */
+ if (s[1] == '?')
+ fputs_filtered
+ (float_format_names[GET_FIELD (insn, 20, 20)], info);
+ else
+ (*info->fprintf_func)
+ (info->stream, "%s ",
+ float_format_names[GET_FIELD (insn, 20, 20)]);
+ break;
+
+ case 'J':
+ fput_const (extract_14 (insn), info);
+ break;
+
+ case '#':
+ {
+ int sign = GET_FIELD (insn, 31, 31);
+ int imm10 = GET_FIELD (insn, 18, 27);
+ int disp;
+
+ if (sign)
+ disp = (-1 << 10) | imm10;
+ else
+ disp = imm10;
+
+ disp <<= 3;
+ fput_const (disp, info);
+ break;
+ }
+ case 'K':
+ case 'd':
+ {
+ int sign = GET_FIELD (insn, 31, 31);
+ int imm11 = GET_FIELD (insn, 18, 28);
+ int disp;
+
+ if (sign)
+ disp = (-1 << 11) | imm11;
+ else
+ disp = imm11;
+
+ disp <<= 2;
+ fput_const (disp, info);
+ break;
+ }
+
+ case '>':
+ case 'y':
+ {
+ /* 16-bit long disp., PA2.0 wide only. */
+ int disp = extract_16 (insn);
+ disp &= ~3;
+ fput_const (disp, info);
+ break;
+ }
+
+ case '&':
+ {
+ /* 16-bit long disp., PA2.0 wide only. */
+ int disp = extract_16 (insn);
+ disp &= ~7;
+ fput_const (disp, info);
+ break;
+ }
+
+ case '_':
+ break; /* Dealt with by '{' */
+
+ case '{':
+ {
+ int sub = GET_FIELD (insn, 14, 16);
+ int df = GET_FIELD (insn, 17, 18);
+ int sf = GET_FIELD (insn, 19, 20);
+ const char * const * source = float_format_names;
+ const char * const * dest = float_format_names;
+ char *t = "";
+
+ if (sub == 4)
+ {
+ fputs_filtered (",UND ", info);
+ break;
+ }
+ if ((sub & 3) == 3)
+ t = ",t";
+ if ((sub & 3) == 1)
+ source = sub & 4 ? fcnv_ufixed_names : fcnv_fixed_names;
+ if (sub & 2)
+ dest = sub & 4 ? fcnv_ufixed_names : fcnv_fixed_names;
+
+ (*info->fprintf_func) (info->stream, "%s%s%s ",
+ t, source[sf], dest[df]);
+ break;
+ }
+
+ case 'm':
+ {
+ int y = GET_FIELD (insn, 16, 18);
+
+ if (y != 1)
+ fput_const ((y ^ 1) - 1, info);
+ }
+ break;
+
+ case 'h':
+ {
+ int cbit;
+
+ cbit = GET_FIELD (insn, 16, 18);
+
+ if (cbit > 0)
+ (*info->fprintf_func) (info->stream, ",%d", cbit - 1);
+ break;
+ }
+
+ case '=':
+ {
+ int cond = GET_FIELD (insn, 27, 31);
+
+ switch (cond)
+ {
+ case 0: fputs_filtered (" ", info); break;
+ case 1: fputs_filtered ("acc ", info); break;
+ case 2: fputs_filtered ("rej ", info); break;
+ case 5: fputs_filtered ("acc8 ", info); break;
+ case 6: fputs_filtered ("rej8 ", info); break;
+ case 9: fputs_filtered ("acc6 ", info); break;
+ case 13: fputs_filtered ("acc4 ", info); break;
+ case 17: fputs_filtered ("acc2 ", info); break;
+ default: break;
+ }
+ break;
+ }
+
+ case 'X':
+ (*info->print_address_func)
+ (memaddr + 8 + extract_22 (insn), info);
+ break;
+ case 'L':
+ fputs_filtered (",rp", info);
+ break;
+ default:
+ (*info->fprintf_func) (info->stream, "%c", *s);
+ break;
+ }
+ }
+ return sizeof (insn);
+ }
+ }
+ (*info->fprintf_func) (info->stream, "#%8x", insn);
+ return sizeof (insn);
+}
diff --git a/hppa.ld b/hppa.ld
new file mode 100644
index 000000000..8ef7fa572
--- /dev/null
+++ b/hppa.ld
@@ -0,0 +1,214 @@
+/* Default linker script, for normal executables */
+OUTPUT_FORMAT("elf32-hppa-linux", "elf32-hppa-linux",
+ "elf32-hppa-linux")
+OUTPUT_ARCH(hppa:hppa1.1)
+ENTRY(_start)
+SEARCH_DIR("/usr/hppa-linux-gnu/lib"); SEARCH_DIR("/usr/local/lib"); SEARCH_DIR("/lib"); SEARCH_DIR("/usr/lib");
+SECTIONS
+{
+ /* Read-only sections, merged into text segment: */
+ PROVIDE (__executable_start = 0x60000000); . = 0x60000000 + SIZEOF_HEADERS;
+ .interp : { *(.interp) }
+ .hash : { *(.hash) }
+ .dynsym : { *(.dynsym) }
+ .dynstr : { *(.dynstr) }
+ .gnu.version : { *(.gnu.version) }
+ .gnu.version_d : { *(.gnu.version_d) }
+ .gnu.version_r : { *(.gnu.version_r) }
+ .rel.init : { *(.rel.init) }
+ .rela.init : { *(.rela.init) }
+ .rel.text : { *(.rel.text .rel.text.* .rel.gnu.linkonce.t.*) }
+ .rela.text : { *(.rela.text .rela.text.* .rela.gnu.linkonce.t.*) }
+ .rel.fini : { *(.rel.fini) }
+ .rela.fini : { *(.rela.fini) }
+ .rel.rodata : { *(.rel.rodata .rel.rodata.* .rel.gnu.linkonce.r.*) }
+ .rela.rodata : { *(.rela.rodata .rela.rodata.* .rela.gnu.linkonce.r.*) }
+ .rel.data.rel.ro : { *(.rel.data.rel.ro* .rel.gnu.linkonce.d.rel.ro.*) }
+ .rela.data.rel.ro : { *(.rela.data.rel.ro* .rela.gnu.linkonce.d.rel.ro.*) }
+ .rel.data : { *(.rel.data .rel.data.* .rel.gnu.linkonce.d.*) }
+ .rela.data : { *(.rela.data .rela.data.* .rela.gnu.linkonce.d.*) }
+ .rel.tdata : { *(.rel.tdata .rel.tdata.* .rel.gnu.linkonce.td.*) }
+ .rela.tdata : { *(.rela.tdata .rela.tdata.* .rela.gnu.linkonce.td.*) }
+ .rel.tbss : { *(.rel.tbss .rel.tbss.* .rel.gnu.linkonce.tb.*) }
+ .rela.tbss : { *(.rela.tbss .rela.tbss.* .rela.gnu.linkonce.tb.*) }
+ .rel.ctors : { *(.rel.ctors) }
+ .rela.ctors : { *(.rela.ctors) }
+ .rel.dtors : { *(.rel.dtors) }
+ .rela.dtors : { *(.rela.dtors) }
+ .rel.got : { *(.rel.got) }
+ .rela.got : { *(.rela.got) }
+ .rel.sdata : { *(.rel.sdata .rel.sdata.* .rel.gnu.linkonce.s.*) }
+ .rela.sdata : { *(.rela.sdata .rela.sdata.* .rela.gnu.linkonce.s.*) }
+ .rel.sbss : { *(.rel.sbss .rel.sbss.* .rel.gnu.linkonce.sb.*) }
+ .rela.sbss : { *(.rela.sbss .rela.sbss.* .rela.gnu.linkonce.sb.*) }
+ .rel.sdata2 : { *(.rel.sdata2 .rel.sdata2.* .rel.gnu.linkonce.s2.*) }
+ .rela.sdata2 : { *(.rela.sdata2 .rela.sdata2.* .rela.gnu.linkonce.s2.*) }
+ .rel.sbss2 : { *(.rel.sbss2 .rel.sbss2.* .rel.gnu.linkonce.sb2.*) }
+ .rela.sbss2 : { *(.rela.sbss2 .rela.sbss2.* .rela.gnu.linkonce.sb2.*) }
+ .rel.bss : { *(.rel.bss .rel.bss.* .rel.gnu.linkonce.b.*) }
+ .rela.bss : { *(.rela.bss .rela.bss.* .rela.gnu.linkonce.b.*) }
+ .rel.plt : { *(.rel.plt) }
+ .rela.plt : { *(.rela.plt) }
+ .init :
+ {
+ KEEP (*(.init))
+ } =0x08000240
+ .text :
+ {
+ *(.text .stub .text.* .gnu.linkonce.t.*)
+ KEEP (*(.text.*personality*))
+ /* .gnu.warning sections are handled specially by elf32.em. */
+ *(.gnu.warning)
+ } =0x08000240
+ .fini :
+ {
+ KEEP (*(.fini))
+ } =0x08000240
+ PROVIDE (__etext = .);
+ PROVIDE (_etext = .);
+ PROVIDE (etext = .);
+ .rodata : { *(.rodata .rodata.* .gnu.linkonce.r.*) }
+ .rodata1 : { *(.rodata1) }
+ .sdata2 :
+ {
+ *(.sdata2 .sdata2.* .gnu.linkonce.s2.*)
+ }
+ .sbss2 : { *(.sbss2 .sbss2.* .gnu.linkonce.sb2.*) }
+ .PARISC.unwind : { *(.PARISC.unwind) }
+ .eh_frame_hdr : { *(.eh_frame_hdr) }
+ .eh_frame : ONLY_IF_RO { KEEP (*(.eh_frame)) }
+ .gcc_except_table : ONLY_IF_RO { *(.gcc_except_table .gcc_except_table.*) }
+ /* Adjust the address for the data segment. We want to adjust up to
+ the same address within the page on the next page up. */
+ . = ALIGN(0x10000) + (. & (0x10000 - 1));
+ /* Exception handling */
+ .eh_frame : ONLY_IF_RW { KEEP (*(.eh_frame)) }
+ .gcc_except_table : ONLY_IF_RW { *(.gcc_except_table .gcc_except_table.*) }
+ /* Thread Local Storage sections */
+ .tdata : { *(.tdata .tdata.* .gnu.linkonce.td.*) }
+ .tbss : { *(.tbss .tbss.* .gnu.linkonce.tb.*) *(.tcommon) }
+ .preinit_array :
+ {
+ PROVIDE_HIDDEN (__preinit_array_start = .);
+ KEEP (*(.preinit_array))
+ PROVIDE_HIDDEN (__preinit_array_end = .);
+ }
+ .init_array :
+ {
+ PROVIDE_HIDDEN (__init_array_start = .);
+ KEEP (*(SORT(.init_array.*)))
+ KEEP (*(.init_array))
+ PROVIDE_HIDDEN (__init_array_end = .);
+ }
+ .fini_array :
+ {
+ PROVIDE_HIDDEN (__fini_array_start = .);
+ KEEP (*(.fini_array))
+ KEEP (*(SORT(.fini_array.*)))
+ PROVIDE_HIDDEN (__fini_array_end = .);
+ }
+ .ctors :
+ {
+ /* gcc uses crtbegin.o to find the start of
+ the constructors, so we make sure it is
+ first. Because this is a wildcard, it
+ doesn't matter if the user does not
+ actually link against crtbegin.o; the
+ linker won't look for a file to match a
+ wildcard. The wildcard also means that it
+ doesn't matter which directory crtbegin.o
+ is in. */
+ KEEP (*crtbegin*.o(.ctors))
+ /* We don't want to include the .ctor section from
+ the crtend.o file until after the sorted ctors.
+ The .ctor section from the crtend file contains the
+ end of ctors marker and it must be last */
+ KEEP (*(EXCLUDE_FILE (*crtend*.o ) .ctors))
+ KEEP (*(SORT(.ctors.*)))
+ KEEP (*(.ctors))
+ }
+ .dtors :
+ {
+ KEEP (*crtbegin*.o(.dtors))
+ KEEP (*(EXCLUDE_FILE (*crtend*.o ) .dtors))
+ KEEP (*(SORT(.dtors.*)))
+ KEEP (*(.dtors))
+ }
+ .jcr : { KEEP (*(.jcr)) }
+ .data.rel.ro : { *(.data.rel.ro.local* .gnu.linkonce.d.rel.ro.local.*) *(.data.rel.ro* .gnu.linkonce.d.rel.ro.*) }
+ .dynamic : { *(.dynamic) }
+ .data :
+ {
+ PROVIDE ($global$ = .);
+ *(.data .data.* .gnu.linkonce.d.*)
+ KEEP (*(.gnu.linkonce.d.*personality*))
+ SORT(CONSTRUCTORS)
+ }
+ .data1 : { *(.data1) }
+ .plt : { *(.plt) }
+ .got : { *(.got.plt) *(.got) }
+ /* We want the small data sections together, so single-instruction offsets
+ can access them all, and initialized data all before uninitialized, so
+ we can shorten the on-disk segment size. */
+ .sdata :
+ {
+ *(.sdata .sdata.* .gnu.linkonce.s.*)
+ }
+ _edata = .; PROVIDE (edata = .);
+ __bss_start = .;
+ .sbss :
+ {
+ *(.dynsbss)
+ *(.sbss .sbss.* .gnu.linkonce.sb.*)
+ *(.scommon)
+ }
+ .bss :
+ {
+ *(.dynbss)
+ *(.bss .bss.* .gnu.linkonce.b.*)
+ *(COMMON)
+ /* Align here to ensure that the .bss section occupies space up to
+ _end. Align after .bss to ensure correct alignment even if the
+ .bss section disappears because there are no input sections.
+ FIXME: Why do we need it? When there is no .bss section, we don't
+ pad the .data section. */
+ . = ALIGN(. != 0 ? 32 / 8 : 1);
+ }
+ . = ALIGN(32 / 8);
+ . = ALIGN(32 / 8);
+ _end = .; PROVIDE (end = .);
+ /* Stabs debugging sections. */
+ .stab 0 : { *(.stab) }
+ .stabstr 0 : { *(.stabstr) }
+ .stab.excl 0 : { *(.stab.excl) }
+ .stab.exclstr 0 : { *(.stab.exclstr) }
+ .stab.index 0 : { *(.stab.index) }
+ .stab.indexstr 0 : { *(.stab.indexstr) }
+ .comment 0 : { *(.comment) }
+ /* DWARF debug sections.
+ Symbols in the DWARF debugging sections are relative to the beginning
+ of the section so we begin them at 0. */
+ /* DWARF 1 */
+ .debug 0 : { *(.debug) }
+ .line 0 : { *(.line) }
+ /* GNU DWARF 1 extensions */
+ .debug_srcinfo 0 : { *(.debug_srcinfo) }
+ .debug_sfnames 0 : { *(.debug_sfnames) }
+ /* DWARF 1.1 and DWARF 2 */
+ .debug_aranges 0 : { *(.debug_aranges) }
+ .debug_pubnames 0 : { *(.debug_pubnames) }
+ /* DWARF 2 */
+ .debug_info 0 : { *(.debug_info .gnu.linkonce.wi.*) }
+ .debug_abbrev 0 : { *(.debug_abbrev) }
+ .debug_line 0 : { *(.debug_line) }
+ .debug_frame 0 : { *(.debug_frame) }
+ .debug_str 0 : { *(.debug_str) }
+ .debug_loc 0 : { *(.debug_loc) }
+ .debug_macinfo 0 : { *(.debug_macinfo) }
+ /* SGI/MIPS DWARF 2 extensions */
+ .debug_weaknames 0 : { *(.debug_weaknames) }
+ .debug_funcnames 0 : { *(.debug_funcnames) }
+ .debug_typenames 0 : { *(.debug_typenames) }
+ .debug_varnames 0 : { *(.debug_varnames) }
+ /DISCARD/ : { *(.note.GNU-stack) }
+}
diff --git a/tcg/hppa/tcg-target.c b/tcg/hppa/tcg-target.c
new file mode 100644
index 000000000..2d17c8190
--- /dev/null
+++ b/tcg/hppa/tcg-target.c
@@ -0,0 +1,980 @@
+/*
+ * Tiny Code Generator for QEMU
+ *
+ * Copyright (c) 2008 Fabrice Bellard
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+ * THE SOFTWARE.
+ */
+
+static const char * const tcg_target_reg_names[TCG_TARGET_NB_REGS] = {
+ "%r0",
+ "%r1",
+ "%rp",
+ "%r3",
+ "%r4",
+ "%r5",
+ "%r6",
+ "%r7",
+ "%r8",
+ "%r9",
+ "%r10",
+ "%r11",
+ "%r12",
+ "%r13",
+ "%r14",
+ "%r15",
+ "%r16",
+ "%r17",
+ "%r18",
+ "%r19",
+ "%r20",
+ "%r21",
+ "%r22",
+ "%r23",
+ "%r24",
+ "%r25",
+ "%r26",
+ "%dp",
+ "%ret0",
+ "%ret1",
+ "%sp",
+ "%r31",
+};
+
+static const int tcg_target_reg_alloc_order[] = {
+ TCG_REG_R4,
+ TCG_REG_R5,
+ TCG_REG_R6,
+ TCG_REG_R7,
+ TCG_REG_R8,
+ TCG_REG_R9,
+ TCG_REG_R10,
+ TCG_REG_R11,
+ TCG_REG_R12,
+ TCG_REG_R13,
+
+ TCG_REG_R17,
+ TCG_REG_R14,
+ TCG_REG_R15,
+ TCG_REG_R16,
+};
+
+static const int tcg_target_call_iarg_regs[4] = {
+ TCG_REG_R26,
+ TCG_REG_R25,
+ TCG_REG_R24,
+ TCG_REG_R23,
+};
+
+static const int tcg_target_call_oarg_regs[2] = {
+ TCG_REG_RET0,
+ TCG_REG_RET1,
+};
+
+static void patch_reloc(uint8_t *code_ptr, int type,
+ tcg_target_long value, tcg_target_long addend)
+{
+ switch (type) {
+ case R_PARISC_PCREL17F:
+ hppa_patch17f((uint32_t *)code_ptr, value, addend);
+ break;
+ default:
+ tcg_abort();
+ }
+}
+
+/* maximum number of register used for input function arguments */
+static inline int tcg_target_get_call_iarg_regs_count(int flags)
+{
+ return 4;
+}
+
+/* parse target specific constraints */
+int target_parse_constraint(TCGArgConstraint *ct, const char **pct_str)
+{
+ const char *ct_str;
+
+ ct_str = *pct_str;
+ switch (ct_str[0]) {
+ case 'r':
+ ct->ct |= TCG_CT_REG;
+ tcg_regset_set32(ct->u.regs, 0, 0xffffffff);
+ break;
+ case 'L': /* qemu_ld/st constraint */
+ ct->ct |= TCG_CT_REG;
+ tcg_regset_set32(ct->u.regs, 0, 0xffffffff);
+ tcg_regset_reset_reg(ct->u.regs, TCG_REG_R26);
+ tcg_regset_reset_reg(ct->u.regs, TCG_REG_R25);
+ tcg_regset_reset_reg(ct->u.regs, TCG_REG_R24);
+ tcg_regset_reset_reg(ct->u.regs, TCG_REG_R23);
+ break;
+ default:
+ return -1;
+ }
+ ct_str++;
+ *pct_str = ct_str;
+ return 0;
+}
+
+/* test if a constant matches the constraint */
+static inline int tcg_target_const_match(tcg_target_long val,
+ const TCGArgConstraint *arg_ct)
+{
+ int ct;
+
+ ct = arg_ct->ct;
+
+ /* TODO */
+
+ return 0;
+}
+
+#define INSN_OP(x) ((x) << 26)
+#define INSN_EXT3BR(x) ((x) << 13)
+#define INSN_EXT3SH(x) ((x) << 10)
+#define INSN_EXT4(x) ((x) << 6)
+#define INSN_EXT5(x) (x)
+#define INSN_EXT6(x) ((x) << 6)
+#define INSN_EXT7(x) ((x) << 6)
+#define INSN_EXT8A(x) ((x) << 6)
+#define INSN_EXT8B(x) ((x) << 5)
+#define INSN_T(x) (x)
+#define INSN_R1(x) ((x) << 16)
+#define INSN_R2(x) ((x) << 21)
+#define INSN_DEP_LEN(x) (32 - (x))
+#define INSN_SHDEP_CP(x) ((31 - (x)) << 5)
+#define INSN_SHDEP_P(x) ((x) << 5)
+#define INSN_COND(x) ((x) << 13)
+
+#define COND_NEVER 0
+#define COND_EQUAL 1
+#define COND_LT 2
+#define COND_LTEQ 3
+#define COND_LTU 4
+#define COND_LTUEQ 5
+#define COND_SV 6
+#define COND_OD 7
+
+
+/* Logical ADD */
+#define ARITH_ADD (INSN_OP(0x02) | INSN_EXT6(0x28))
+#define ARITH_AND (INSN_OP(0x02) | INSN_EXT6(0x08))
+#define ARITH_OR (INSN_OP(0x02) | INSN_EXT6(0x09))
+#define ARITH_XOR (INSN_OP(0x02) | INSN_EXT6(0x0a))
+#define ARITH_SUB (INSN_OP(0x02) | INSN_EXT6(0x10))
+
+#define SHD (INSN_OP(0x34) | INSN_EXT3SH(2))
+#define VSHD (INSN_OP(0x34) | INSN_EXT3SH(0))
+#define DEP (INSN_OP(0x35) | INSN_EXT3SH(3))
+#define ZDEP (INSN_OP(0x35) | INSN_EXT3SH(2))
+#define ZVDEP (INSN_OP(0x35) | INSN_EXT3SH(0))
+#define EXTRU (INSN_OP(0x34) | INSN_EXT3SH(6))
+#define EXTRS (INSN_OP(0x34) | INSN_EXT3SH(7))
+#define VEXTRS (INSN_OP(0x34) | INSN_EXT3SH(5))
+
+#define SUBI (INSN_OP(0x25))
+#define MTCTL (INSN_OP(0x00) | INSN_EXT8B(0xc2))
+
+#define BL (INSN_OP(0x3a) | INSN_EXT3BR(0))
+#define BLE_SR4 (INSN_OP(0x39) | (1 << 13))
+#define BV (INSN_OP(0x3a) | INSN_EXT3BR(6))
+#define BV_N (INSN_OP(0x3a) | INSN_EXT3BR(6) | 2)
+#define LDIL (INSN_OP(0x08))
+#define LDO (INSN_OP(0x0d))
+
+#define LDB (INSN_OP(0x10))
+#define LDH (INSN_OP(0x11))
+#define LDW (INSN_OP(0x12))
+#define LDWM (INSN_OP(0x13))
+
+#define STB (INSN_OP(0x18))
+#define STH (INSN_OP(0x19))
+#define STW (INSN_OP(0x1a))
+#define STWM (INSN_OP(0x1b))
+
+#define COMBT (INSN_OP(0x20))
+#define COMBF (INSN_OP(0x22))
+
+static int lowsignext(uint32_t val, int start, int length)
+{
+ return (((val << 1) & ~(~0 << length)) |
+ ((val >> (length - 1)) & 1)) << start;
+}
+
+static inline void tcg_out_mov(TCGContext *s, int ret, int arg)
+{
+ /* PA1.1 defines COPY as OR r,0,t */
+ tcg_out32(s, ARITH_OR | INSN_T(ret) | INSN_R1(arg) | INSN_R2(TCG_REG_R0));
+
+ /* PA2.0 defines COPY as LDO 0(r),t
+ * but hppa-dis.c is unaware of this definition */
+ /* tcg_out32(s, LDO | INSN_R1(ret) | INSN_R2(arg) | reassemble_14(0)); */
+}
+
+static inline void tcg_out_movi(TCGContext *s, TCGType type,
+ int ret, tcg_target_long arg)
+{
+ if (arg == (arg & 0x1fff)) {
+ tcg_out32(s, LDO | INSN_R1(ret) | INSN_R2(TCG_REG_R0) |
+ reassemble_14(arg));
+ } else {
+ tcg_out32(s, LDIL | INSN_R2(ret) |
+ reassemble_21(lrsel((uint32_t)arg, 0)));
+ if (arg & 0x7ff)
+ tcg_out32(s, LDO | INSN_R1(ret) | INSN_R2(ret) |
+ reassemble_14(rrsel((uint32_t)arg, 0)));
+ }
+}
+
+static inline void tcg_out_ld_raw(TCGContext *s, int ret,
+ tcg_target_long arg)
+{
+ tcg_out32(s, LDIL | INSN_R2(ret) |
+ reassemble_21(lrsel((uint32_t)arg, 0)));
+ tcg_out32(s, LDW | INSN_R1(ret) | INSN_R2(ret) |
+ reassemble_14(rrsel((uint32_t)arg, 0)));
+}
+
+static inline void tcg_out_ld_ptr(TCGContext *s, int ret,
+ tcg_target_long arg)
+{
+ tcg_out_ld_raw(s, ret, arg);
+}
+
+static inline void tcg_out_ldst(TCGContext *s, int ret, int addr, int offset,
+ int op)
+{
+ if (offset == (offset & 0xfff))
+ tcg_out32(s, op | INSN_R1(ret) | INSN_R2(addr) |
+ reassemble_14(offset));
+ else {
+ fprintf(stderr, "unimplemented %s with offset %d\n", __func__, offset);
+ tcg_abort();
+ }
+}
+
+static inline void tcg_out_ld(TCGContext *s, TCGType type, int ret,
+ int arg1, tcg_target_long arg2)
+{
+ fprintf(stderr, "unimplemented %s\n", __func__);
+ tcg_abort();
+}
+
+static inline void tcg_out_st(TCGContext *s, TCGType type, int ret,
+ int arg1, tcg_target_long arg2)
+{
+ fprintf(stderr, "unimplemented %s\n", __func__);
+ tcg_abort();
+}
+
+static inline void tcg_out_arith(TCGContext *s, int t, int r1, int r2, int op)
+{
+ tcg_out32(s, op | INSN_T(t) | INSN_R1(r1) | INSN_R2(r2));
+}
+
+static inline void tcg_out_arithi(TCGContext *s, int t, int r1,
+ tcg_target_long val, int op)
+{
+ tcg_out_movi(s, TCG_TYPE_PTR, TCG_REG_R20, val);
+ tcg_out_arith(s, t, r1, TCG_REG_R20, op);
+}
+
+static inline void tcg_out_addi(TCGContext *s, int reg, tcg_target_long val)
+{
+ tcg_out_arithi(s, reg, reg, val, ARITH_ADD);
+}
+
+static inline void tcg_out_nop(TCGContext *s)
+{
+ tcg_out32(s, ARITH_OR | INSN_T(TCG_REG_R0) | INSN_R1(TCG_REG_R0) |
+ INSN_R2(TCG_REG_R0));
+}
+
+static inline void tcg_out_ext8s(TCGContext *s, int ret, int arg) {
+ tcg_out32(s, EXTRS | INSN_R1(ret) | INSN_R2(arg) |
+ INSN_SHDEP_P(31) | INSN_DEP_LEN(8));
+}
+
+static inline void tcg_out_ext16s(TCGContext *s, int ret, int arg) {
+ tcg_out32(s, EXTRS | INSN_R1(ret) | INSN_R2(arg) |
+ INSN_SHDEP_P(31) | INSN_DEP_LEN(16));
+}
+
+static inline void tcg_out_bswap16(TCGContext *s, int ret, int arg) {
+ if(ret != arg)
+ tcg_out_mov(s, ret, arg);
+ tcg_out32(s, DEP | INSN_R2(ret) | INSN_R1(ret) |
+ INSN_SHDEP_CP(15) | INSN_DEP_LEN(8));
+ tcg_out32(s, SHD | INSN_T(ret) | INSN_R1(TCG_REG_R0) |
+ INSN_R2(ret) | INSN_SHDEP_CP(8));
+}
+
+static inline void tcg_out_bswap32(TCGContext *s, int ret, int arg, int temp) {
+ tcg_out32(s, SHD | INSN_T(temp) | INSN_R1(arg) |
+ INSN_R2(arg) | INSN_SHDEP_CP(16));
+ tcg_out32(s, DEP | INSN_R2(temp) | INSN_R1(temp) |
+ INSN_SHDEP_CP(15) | INSN_DEP_LEN(8));
+ tcg_out32(s, SHD | INSN_T(ret) | INSN_R1(arg) |
+ INSN_R2(temp) | INSN_SHDEP_CP(8));
+}
+
+static inline void tcg_out_call(TCGContext *s, void *func)
+{
+ uint32_t val = (uint32_t)__canonicalize_funcptr_for_compare(func);
+ tcg_out32(s, LDIL | INSN_R2(TCG_REG_R20) |
+ reassemble_21(lrsel(val, 0)));
+ tcg_out32(s, BLE_SR4 | INSN_R2(TCG_REG_R20) |
+ reassemble_17(rrsel(val, 0) >> 2));
+ tcg_out_mov(s, TCG_REG_RP, TCG_REG_R31);
+}
+
+#if defined(CONFIG_SOFTMMU)
+extern void __ldb_mmu(void);
+extern void __ldw_mmu(void);
+extern void __ldl_mmu(void);
+extern void __ldq_mmu(void);
+
+extern void __stb_mmu(void);
+extern void __stw_mmu(void);
+extern void __stl_mmu(void);
+extern void __stq_mmu(void);
+
+static void *qemu_ld_helpers[4] = {
+ __ldb_mmu,
+ __ldw_mmu,
+ __ldl_mmu,
+ __ldq_mmu,
+};
+
+static void *qemu_st_helpers[4] = {
+ __stb_mmu,
+ __stw_mmu,
+ __stl_mmu,
+ __stq_mmu,
+};
+#endif
+
+static void tcg_out_qemu_ld(TCGContext *s, const TCGArg *args, int opc)
+{
+ int addr_reg, data_reg, data_reg2, r0, r1, mem_index, s_bits, bswap;
+#if defined(CONFIG_SOFTMMU)
+ uint32_t *label1_ptr, *label2_ptr;
+#endif
+#if TARGET_LONG_BITS == 64
+#if defined(CONFIG_SOFTMMU)
+ uint32_t *label3_ptr;
+#endif
+ int addr_reg2;
+#endif
+
+ data_reg = *args++;
+ if (opc == 3)
+ data_reg2 = *args++;
+ else
+ data_reg2 = 0; /* surpress warning */
+ addr_reg = *args++;
+#if TARGET_LONG_BITS == 64
+ addr_reg2 = *args++;
+#endif
+ mem_index = *args;
+ s_bits = opc & 3;
+
+ r0 = TCG_REG_R26;
+ r1 = TCG_REG_R25;
+
+#if defined(CONFIG_SOFTMMU)
+ tcg_out_mov(s, r1, addr_reg);
+
+ tcg_out_mov(s, r0, addr_reg);
+
+ tcg_out32(s, SHD | INSN_T(r1) | INSN_R1(TCG_REG_R0) | INSN_R2(r1) |
+ INSN_SHDEP_CP(TARGET_PAGE_BITS - CPU_TLB_ENTRY_BITS));
+
+ tcg_out_arithi(s, r0, r0, TARGET_PAGE_MASK | ((1 << s_bits) - 1),
+ ARITH_AND);
+
+ tcg_out_arithi(s, r1, r1, (CPU_TLB_SIZE - 1) << CPU_TLB_ENTRY_BITS,
+ ARITH_AND);
+
+ tcg_out_arith(s, r1, r1, TCG_AREG0, ARITH_ADD);
+ tcg_out_arithi(s, r1, r1,
+ offsetof(CPUState, tlb_table[mem_index][0].addr_read),
+ ARITH_ADD);
+
+ tcg_out_ldst(s, TCG_REG_R20, r1, 0, LDW);
+
+#if TARGET_LONG_BITS == 32
+ /* if equal, jump to label1 */
+ label1_ptr = (uint32_t *)s->code_ptr;
+ tcg_out32(s, COMBT | INSN_R1(TCG_REG_R20) | INSN_R2(r0) |
+ INSN_COND(COND_EQUAL));
+ tcg_out_mov(s, r0, addr_reg); /* delay slot */
+#else
+ /* if not equal, jump to label3 */
+ label3_ptr = (uint32_t *)s->code_ptr;
+ tcg_out32(s, COMBF | INSN_R1(TCG_REG_R20) | INSN_R2(r0) |
+ INSN_COND(COND_EQUAL));
+ tcg_out_mov(s, r0, addr_reg); /* delay slot */
+
+ tcg_out_ldst(s, TCG_REG_R20, r1, 4, LDW);
+
+ /* if equal, jump to label1 */
+ label1_ptr = (uint32_t *)s->code_ptr;
+ tcg_out32(s, COMBT | INSN_R1(TCG_REG_R20) | INSN_R2(addr_reg2) |
+ INSN_COND(COND_EQUAL));
+ tcg_out_nop(s); /* delay slot */
+
+ /* label3: */
+ *label3_ptr |= reassemble_12((uint32_t *)s->code_ptr - label3_ptr - 2);
+#endif
+
+#if TARGET_LONG_BITS == 32
+ tcg_out_mov(s, TCG_REG_R26, addr_reg);
+ tcg_out_movi(s, TCG_TYPE_I32, TCG_REG_R25, mem_index);
+#else
+ tcg_out_mov(s, TCG_REG_R26, addr_reg);
+ tcg_out_mov(s, TCG_REG_R25, addr_reg2);
+ tcg_out_movi(s, TCG_TYPE_I32, TCG_REG_R24, mem_index);
+#endif
+
+ tcg_out_call(s, qemu_ld_helpers[s_bits]);
+
+ switch(opc) {
+ case 0 | 4:
+ tcg_out_ext8s(s, data_reg, TCG_REG_RET0);
+ break;
+ case 1 | 4:
+ tcg_out_ext16s(s, data_reg, TCG_REG_RET0);
+ break;
+ case 0:
+ case 1:
+ case 2:
+ default:
+ tcg_out_mov(s, data_reg, TCG_REG_RET0);
+ break;
+ case 3:
+ tcg_abort();
+ tcg_out_mov(s, data_reg, TCG_REG_RET0);
+ tcg_out_mov(s, data_reg2, TCG_REG_RET1);
+ break;
+ }
+
+ /* jump to label2 */
+ label2_ptr = (uint32_t *)s->code_ptr;
+ tcg_out32(s, BL | INSN_R2(TCG_REG_R0) | 2);
+
+ /* label1: */
+ *label1_ptr |= reassemble_12((uint32_t *)s->code_ptr - label1_ptr - 2);
+
+ tcg_out_arithi(s, TCG_REG_R20, r1,
+ offsetof(CPUTLBEntry, addend) - offsetof(CPUTLBEntry, addr_read),
+ ARITH_ADD);
+ tcg_out_ldst(s, TCG_REG_R20, TCG_REG_R20, 0, LDW);
+ tcg_out_arith(s, r0, r0, TCG_REG_R20, ARITH_ADD);
+#else
+ r0 = addr_reg;
+#endif
+
+#ifdef TARGET_WORDS_BIGENDIAN
+ bswap = 0;
+#else
+ bswap = 1;
+#endif
+ switch (opc) {
+ case 0:
+ tcg_out_ldst(s, data_reg, r0, 0, LDB);
+ break;
+ case 0 | 4:
+ tcg_out_ldst(s, data_reg, r0, 0, LDB);
+ tcg_out_ext8s(s, data_reg, data_reg);
+ break;
+ case 1:
+ tcg_out_ldst(s, data_reg, r0, 0, LDH);
+ if (bswap)
+ tcg_out_bswap16(s, data_reg, data_reg);
+ break;
+ case 1 | 4:
+ tcg_out_ldst(s, data_reg, r0, 0, LDH);
+ if (bswap)
+ tcg_out_bswap16(s, data_reg, data_reg);
+ tcg_out_ext16s(s, data_reg, data_reg);
+ break;
+ case 2:
+ tcg_out_ldst(s, data_reg, r0, 0, LDW);
+ if (bswap)
+ tcg_out_bswap32(s, data_reg, data_reg, TCG_REG_R20);
+ break;
+ case 3:
+ tcg_abort();
+ if (!bswap) {
+ tcg_out_ldst(s, data_reg, r0, 0, LDW);
+ tcg_out_ldst(s, data_reg2, r0, 4, LDW);
+ } else {
+ tcg_out_ldst(s, data_reg, r0, 4, LDW);
+ tcg_out_bswap32(s, data_reg, data_reg, TCG_REG_R20);
+ tcg_out_ldst(s, data_reg2, r0, 0, LDW);
+ tcg_out_bswap32(s, data_reg2, data_reg2, TCG_REG_R20);
+ }
+ break;
+ default:
+ tcg_abort();
+ }
+
+#if defined(CONFIG_SOFTMMU)
+ /* label2: */
+ *label2_ptr |= reassemble_17((uint32_t *)s->code_ptr - label2_ptr - 2);
+#endif
+}
+
+static void tcg_out_qemu_st(TCGContext *s, const TCGArg *args, int opc)
+{
+ int addr_reg, data_reg, data_reg2, r0, r1, mem_index, s_bits, bswap;
+#if defined(CONFIG_SOFTMMU)
+ uint32_t *label1_ptr, *label2_ptr;
+#endif
+#if TARGET_LONG_BITS == 64
+#if defined(CONFIG_SOFTMMU)
+ uint32_t *label3_ptr;
+#endif
+ int addr_reg2;
+#endif
+
+ data_reg = *args++;
+ if (opc == 3)
+ data_reg2 = *args++;
+ else
+ data_reg2 = 0; /* surpress warning */
+ addr_reg = *args++;
+#if TARGET_LONG_BITS == 64
+ addr_reg2 = *args++;
+#endif
+ mem_index = *args;
+
+ s_bits = opc;
+
+ r0 = TCG_REG_R26;
+ r1 = TCG_REG_R25;
+
+#if defined(CONFIG_SOFTMMU)
+ tcg_out_mov(s, r1, addr_reg);
+
+ tcg_out_mov(s, r0, addr_reg);
+
+ tcg_out32(s, SHD | INSN_T(r1) | INSN_R1(TCG_REG_R0) | INSN_R2(r1) |
+ INSN_SHDEP_CP(TARGET_PAGE_BITS - CPU_TLB_ENTRY_BITS));
+
+ tcg_out_arithi(s, r0, r0, TARGET_PAGE_MASK | ((1 << s_bits) - 1),
+ ARITH_AND);
+
+ tcg_out_arithi(s, r1, r1, (CPU_TLB_SIZE - 1) << CPU_TLB_ENTRY_BITS,
+ ARITH_AND);
+
+ tcg_out_arith(s, r1, r1, TCG_AREG0, ARITH_ADD);
+ tcg_out_arithi(s, r1, r1,
+ offsetof(CPUState, tlb_table[mem_index][0].addr_write),
+ ARITH_ADD);
+
+ tcg_out_ldst(s, TCG_REG_R20, r1, 0, LDW);
+
+#if TARGET_LONG_BITS == 32
+ /* if equal, jump to label1 */
+ label1_ptr = (uint32_t *)s->code_ptr;
+ tcg_out32(s, COMBT | INSN_R1(TCG_REG_R20) | INSN_R2(r0) |
+ INSN_COND(COND_EQUAL));
+ tcg_out_mov(s, r0, addr_reg); /* delay slot */
+#else
+ /* if not equal, jump to label3 */
+ label3_ptr = (uint32_t *)s->code_ptr;
+ tcg_out32(s, COMBF | INSN_R1(TCG_REG_R20) | INSN_R2(r0) |
+ INSN_COND(COND_EQUAL));
+ tcg_out_mov(s, r0, addr_reg); /* delay slot */
+
+ tcg_out_ldst(s, TCG_REG_R20, r1, 4, LDW);
+
+ /* if equal, jump to label1 */
+ label1_ptr = (uint32_t *)s->code_ptr;
+ tcg_out32(s, COMBT | INSN_R1(TCG_REG_R20) | INSN_R2(addr_reg2) |
+ INSN_COND(COND_EQUAL));
+ tcg_out_nop(s); /* delay slot */
+
+ /* label3: */
+ *label3_ptr |= reassemble_12((uint32_t *)s->code_ptr - label3_ptr - 2);
+#endif
+
+ tcg_out_mov(s, TCG_REG_R26, addr_reg);
+#if TARGET_LONG_BITS == 64
+ tcg_out_mov(s, TCG_REG_R25, addr_reg2);
+ if (opc == 3) {
+ tcg_abort();
+ tcg_out_mov(s, TCG_REG_R24, data_reg);
+ tcg_out_mov(s, TCG_REG_R23, data_reg2);
+ /* TODO: push mem_index */
+ tcg_abort();
+ } else {
+ switch(opc) {
+ case 0:
+ tcg_out32(s, EXTRU | INSN_R1(TCG_REG_R24) | INSN_R2(data_reg) |
+ INSN_SHDEP_P(31) | INSN_DEP_LEN(8));
+ break;
+ case 1:
+ tcg_out32(s, EXTRU | INSN_R1(TCG_REG_R24) | INSN_R2(data_reg) |
+ INSN_SHDEP_P(31) | INSN_DEP_LEN(16));
+ break;
+ case 2:
+ tcg_out_mov(s, TCG_REG_R24, data_reg);
+ break;
+ }
+ tcg_out_movi(s, TCG_TYPE_I32, TCG_REG_R23, mem_index);
+ }
+#else
+ if (opc == 3) {
+ tcg_abort();
+ tcg_out_mov(s, TCG_REG_R25, data_reg);
+ tcg_out_mov(s, TCG_REG_R24, data_reg2);
+ tcg_out_movi(s, TCG_TYPE_I32, TCG_REG_R23, mem_index);
+ } else {
+ switch(opc) {
+ case 0:
+ tcg_out32(s, EXTRU | INSN_R1(TCG_REG_R25) | INSN_R2(data_reg) |
+ INSN_SHDEP_P(31) | INSN_DEP_LEN(8));
+ break;
+ case 1:
+ tcg_out32(s, EXTRU | INSN_R1(TCG_REG_R25) | INSN_R2(data_reg) |
+ INSN_SHDEP_P(31) | INSN_DEP_LEN(16));
+ break;
+ case 2:
+ tcg_out_mov(s, TCG_REG_R25, data_reg);
+ break;
+ }
+ tcg_out_movi(s, TCG_TYPE_I32, TCG_REG_R24, mem_index);
+ }
+#endif
+ tcg_out_call(s, qemu_st_helpers[s_bits]);
+
+ /* jump to label2 */
+ label2_ptr = (uint32_t *)s->code_ptr;
+ tcg_out32(s, BL | INSN_R2(TCG_REG_R0) | 2);
+
+ /* label1: */
+ *label1_ptr |= reassemble_12((uint32_t *)s->code_ptr - label1_ptr - 2);
+
+ tcg_out_arithi(s, TCG_REG_R20, r1,
+ offsetof(CPUTLBEntry, addend) - offsetof(CPUTLBEntry, addr_write),
+ ARITH_ADD);
+ tcg_out_ldst(s, TCG_REG_R20, TCG_REG_R20, 0, LDW);
+ tcg_out_arith(s, r0, r0, TCG_REG_R20, ARITH_ADD);
+#else
+ r0 = addr_reg;
+#endif
+
+#ifdef TARGET_WORDS_BIGENDIAN
+ bswap = 0;
+#else
+ bswap = 1;
+#endif
+ switch (opc) {
+ case 0:
+ tcg_out_ldst(s, data_reg, r0, 0, STB);
+ break;
+ case 1:
+ if (bswap) {
+ tcg_out_bswap16(s, TCG_REG_R20, data_reg);
+ data_reg = TCG_REG_R20;
+ }
+ tcg_out_ldst(s, data_reg, r0, 0, STH);
+ break;
+ case 2:
+ if (bswap) {
+ tcg_out_bswap32(s, TCG_REG_R20, data_reg, TCG_REG_R20);
+ data_reg = TCG_REG_R20;
+ }
+ tcg_out_ldst(s, data_reg, r0, 0, STW);
+ break;
+ case 3:
+ tcg_abort();
+ if (!bswap) {
+ tcg_out_ldst(s, data_reg, r0, 0, STW);
+ tcg_out_ldst(s, data_reg2, r0, 4, STW);
+ } else {
+ tcg_out_bswap32(s, TCG_REG_R20, data_reg, TCG_REG_R20);
+ tcg_out_ldst(s, TCG_REG_R20, r0, 4, STW);
+ tcg_out_bswap32(s, TCG_REG_R20, data_reg2, TCG_REG_R20);
+ tcg_out_ldst(s, TCG_REG_R20, r0, 0, STW);
+ }
+ break;
+ default:
+ tcg_abort();
+ }
+
+#if defined(CONFIG_SOFTMMU)
+ /* label2: */
+ *label2_ptr |= reassemble_17((uint32_t *)s->code_ptr - label2_ptr - 2);
+#endif
+}
+
+static inline void tcg_out_op(TCGContext *s, int opc, const TCGArg *args,
+ const int *const_args)
+{
+ int c;
+
+ switch (opc) {
+ case INDEX_op_exit_tb:
+ tcg_out_movi(s, TCG_TYPE_PTR, TCG_REG_RET0, args[0]);
+ tcg_out32(s, BV_N | INSN_R2(TCG_REG_R18));
+ break;
+ case INDEX_op_goto_tb:
+ if (s->tb_jmp_offset) {
+ /* direct jump method */
+ fprintf(stderr, "goto_tb direct\n");
+ tcg_abort();
+ tcg_out_movi(s, TCG_TYPE_PTR, TCG_REG_R20, args[0]);
+ tcg_out32(s, BV_N | INSN_R2(TCG_REG_R20));
+ s->tb_jmp_offset[args[0]] = s->code_ptr - s->code_buf;
+ } else {
+ /* indirect jump method */
+ tcg_out_ld_ptr(s, TCG_REG_R20,
+ (tcg_target_long)(s->tb_next + args[0]));
+ tcg_out32(s, BV_N | INSN_R2(TCG_REG_R20));
+ }
+ s->tb_next_offset[args[0]] = s->code_ptr - s->code_buf;
+ break;
+ case INDEX_op_call:
+ tcg_out32(s, BLE_SR4 | INSN_R2(args[0]));
+ tcg_out_mov(s, TCG_REG_RP, TCG_REG_R31);
+ break;
+ case INDEX_op_jmp:
+ fprintf(stderr, "unimplemented jmp\n");
+ tcg_abort();
+ break;
+ case INDEX_op_br:
+ fprintf(stderr, "unimplemented br\n");
+ tcg_abort();
+ break;
+ case INDEX_op_movi_i32:
+ tcg_out_movi(s, TCG_TYPE_I32, args[0], (uint32_t)args[1]);
+ break;
+
+ case INDEX_op_ld8u_i32:
+ tcg_out_ldst(s, args[0], args[1], args[2], LDB);
+ break;
+ case INDEX_op_ld8s_i32:
+ tcg_out_ldst(s, args[0], args[1], args[2], LDB);
+ tcg_out_ext8s(s, args[0], args[0]);
+ break;
+ case INDEX_op_ld16u_i32:
+ tcg_out_ldst(s, args[0], args[1], args[2], LDH);
+ break;
+ case INDEX_op_ld16s_i32:
+ tcg_out_ldst(s, args[0], args[1], args[2], LDH);
+ tcg_out_ext16s(s, args[0], args[0]);
+ break;
+ case INDEX_op_ld_i32:
+ tcg_out_ldst(s, args[0], args[1], args[2], LDW);
+ break;
+
+ case INDEX_op_st8_i32:
+ tcg_out_ldst(s, args[0], args[1], args[2], STB);
+ break;
+ case INDEX_op_st16_i32:
+ tcg_out_ldst(s, args[0], args[1], args[2], STH);
+ break;
+ case INDEX_op_st_i32:
+ tcg_out_ldst(s, args[0], args[1], args[2], STW);
+ break;
+
+ case INDEX_op_sub_i32:
+ c = ARITH_SUB;
+ goto gen_arith;
+ case INDEX_op_and_i32:
+ c = ARITH_AND;
+ goto gen_arith;
+ case INDEX_op_or_i32:
+ c = ARITH_OR;
+ goto gen_arith;
+ case INDEX_op_xor_i32:
+ c = ARITH_XOR;
+ goto gen_arith;
+ case INDEX_op_add_i32:
+ c = ARITH_ADD;
+ goto gen_arith;
+
+ case INDEX_op_shl_i32:
+ tcg_out32(s, SUBI | INSN_R1(TCG_REG_R20) | INSN_R2(args[2]) |
+ lowsignext(0x1f, 0, 11));
+ tcg_out32(s, MTCTL | INSN_R2(11) | INSN_R1(TCG_REG_R20));
+ tcg_out32(s, ZVDEP | INSN_R2(args[0]) | INSN_R1(args[1]) |
+ INSN_DEP_LEN(32));
+ break;
+ case INDEX_op_shr_i32:
+ tcg_out32(s, MTCTL | INSN_R2(11) | INSN_R1(args[2]));
+ tcg_out32(s, VSHD | INSN_T(args[0]) | INSN_R1(TCG_REG_R0) |
+ INSN_R2(args[1]));
+ break;
+ case INDEX_op_sar_i32:
+ tcg_out32(s, SUBI | INSN_R1(TCG_REG_R20) | INSN_R2(args[2]) |
+ lowsignext(0x1f, 0, 11));
+ tcg_out32(s, MTCTL | INSN_R2(11) | INSN_R1(TCG_REG_R20));
+ tcg_out32(s, VEXTRS | INSN_R1(args[0]) | INSN_R2(args[1]) |
+ INSN_DEP_LEN(32));
+ break;
+
+ case INDEX_op_mul_i32:
+ fprintf(stderr, "unimplemented mul\n");
+ tcg_abort();
+ break;
+ case INDEX_op_mulu2_i32:
+ fprintf(stderr, "unimplemented mulu2\n");
+ tcg_abort();
+ break;
+ case INDEX_op_div2_i32:
+ fprintf(stderr, "unimplemented div2\n");
+ tcg_abort();
+ break;
+ case INDEX_op_divu2_i32:
+ fprintf(stderr, "unimplemented divu2\n");
+ tcg_abort();
+ break;
+
+ case INDEX_op_brcond_i32:
+ fprintf(stderr, "unimplemented brcond\n");
+ tcg_abort();
+ break;
+
+ case INDEX_op_qemu_ld8u:
+ tcg_out_qemu_ld(s, args, 0);
+ break;
+ case INDEX_op_qemu_ld8s:
+ tcg_out_qemu_ld(s, args, 0 | 4);
+ break;
+ case INDEX_op_qemu_ld16u:
+ tcg_out_qemu_ld(s, args, 1);
+ break;
+ case INDEX_op_qemu_ld16s:
+ tcg_out_qemu_ld(s, args, 1 | 4);
+ break;
+ case INDEX_op_qemu_ld32u:
+ tcg_out_qemu_ld(s, args, 2);
+ break;
+
+ case INDEX_op_qemu_st8:
+ tcg_out_qemu_st(s, args, 0);
+ break;
+ case INDEX_op_qemu_st16:
+ tcg_out_qemu_st(s, args, 1);
+ break;
+ case INDEX_op_qemu_st32:
+ tcg_out_qemu_st(s, args, 2);
+ break;
+
+ default:
+ fprintf(stderr, "unknown opcode 0x%x\n", opc);
+ tcg_abort();
+ }
+ return;
+
+gen_arith:
+ tcg_out_arith(s, args[0], args[1], args[2], c);
+}
+
+static const TCGTargetOpDef hppa_op_defs[] = {
+ { INDEX_op_exit_tb, { } },
+ { INDEX_op_goto_tb, { } },
+
+ { INDEX_op_call, { "r" } },
+ { INDEX_op_jmp, { "r" } },
+ { INDEX_op_br, { } },
+
+ { INDEX_op_mov_i32, { "r", "r" } },
+ { INDEX_op_movi_i32, { "r" } },
+ { INDEX_op_ld8u_i32, { "r", "r" } },
+ { INDEX_op_ld8s_i32, { "r", "r" } },
+ { INDEX_op_ld16u_i32, { "r", "r" } },
+ { INDEX_op_ld16s_i32, { "r", "r" } },
+ { INDEX_op_ld_i32, { "r", "r" } },
+ { INDEX_op_st8_i32, { "r", "r" } },
+ { INDEX_op_st16_i32, { "r", "r" } },
+ { INDEX_op_st_i32, { "r", "r" } },
+
+ { INDEX_op_add_i32, { "r", "r", "r" } },
+ { INDEX_op_sub_i32, { "r", "r", "r" } },
+ { INDEX_op_and_i32, { "r", "r", "r" } },
+ { INDEX_op_or_i32, { "r", "r", "r" } },
+ { INDEX_op_xor_i32, { "r", "r", "r" } },
+
+ { INDEX_op_shl_i32, { "r", "r", "r" } },
+ { INDEX_op_shr_i32, { "r", "r", "r" } },
+ { INDEX_op_sar_i32, { "r", "r", "r" } },
+
+ { INDEX_op_brcond_i32, { "r", "r" } },
+
+#if TARGET_LONG_BITS == 32
+ { INDEX_op_qemu_ld8u, { "r", "L" } },
+ { INDEX_op_qemu_ld8s, { "r", "L" } },
+ { INDEX_op_qemu_ld16u, { "r", "L" } },
+ { INDEX_op_qemu_ld16s, { "r", "L" } },
+ { INDEX_op_qemu_ld32u, { "r", "L" } },
+ { INDEX_op_qemu_ld64, { "r", "r", "L" } },
+
+ { INDEX_op_qemu_st8, { "L", "L" } },
+ { INDEX_op_qemu_st16, { "L", "L" } },
+ { INDEX_op_qemu_st32, { "L", "L" } },
+ { INDEX_op_qemu_st64, { "L", "L", "L" } },
+#else
+ { INDEX_op_qemu_ld8u, { "r", "L", "L" } },
+ { INDEX_op_qemu_ld8s, { "r", "L", "L" } },
+ { INDEX_op_qemu_ld16u, { "r", "L", "L" } },
+ { INDEX_op_qemu_ld16s, { "r", "L", "L" } },
+ { INDEX_op_qemu_ld32u, { "r", "L", "L" } },
+ { INDEX_op_qemu_ld32s, { "r", "L", "L" } },
+ { INDEX_op_qemu_ld64, { "r", "r", "L", "L" } },
+
+ { INDEX_op_qemu_st8, { "L", "L", "L" } },
+ { INDEX_op_qemu_st16, { "L", "L", "L" } },
+ { INDEX_op_qemu_st32, { "L", "L", "L" } },
+ { INDEX_op_qemu_st64, { "L", "L", "L", "L" } },
+#endif
+ { -1 },
+};
+
+void tcg_target_init(TCGContext *s)
+{
+ tcg_regset_set32(tcg_target_available_regs[TCG_TYPE_I32], 0, 0xffffffff);
+ tcg_regset_set32(tcg_target_call_clobber_regs, 0,
+ (1 << TCG_REG_R20) |
+ (1 << TCG_REG_R21) |
+ (1 << TCG_REG_R22) |
+ (1 << TCG_REG_R23) |
+ (1 << TCG_REG_R24) |
+ (1 << TCG_REG_R25) |
+ (1 << TCG_REG_R26));
+
+ tcg_regset_clear(s->reserved_regs);
+ tcg_regset_set_reg(s->reserved_regs, TCG_REG_R0); /* hardwired to zero */
+ tcg_regset_set_reg(s->reserved_regs, TCG_REG_R1); /* addil target */
+ tcg_regset_set_reg(s->reserved_regs, TCG_REG_RP); /* link register */
+ tcg_regset_set_reg(s->reserved_regs, TCG_REG_R3); /* frame pointer */
+ tcg_regset_set_reg(s->reserved_regs, TCG_REG_R18); /* return pointer */
+ tcg_regset_set_reg(s->reserved_regs, TCG_REG_R19); /* clobbered w/o pic */
+ tcg_regset_set_reg(s->reserved_regs, TCG_REG_R20); /* reserved */
+ tcg_regset_set_reg(s->reserved_regs, TCG_REG_DP); /* data pointer */
+ tcg_regset_set_reg(s->reserved_regs, TCG_REG_SP); /* stack pointer */
+ tcg_regset_set_reg(s->reserved_regs, TCG_REG_R31); /* ble link reg */
+
+ tcg_add_target_add_op_defs(hppa_op_defs);
+}
diff --git a/tcg/hppa/tcg-target.h b/tcg/hppa/tcg-target.h
new file mode 100644
index 000000000..8e2693d0e
--- /dev/null
+++ b/tcg/hppa/tcg-target.h
@@ -0,0 +1,204 @@
+/*
+ * Tiny Code Generator for QEMU
+ *
+ * Copyright (c) 2008 Fabrice Bellard
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to deal
+ * in the Software without restriction, including without limitation the rights
+ * to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+ * copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
+ * THE SOFTWARE.
+ */
+
+#define TCG_TARGET_HPPA 1
+
+#if defined(_PA_RISC1_1)
+#define TCG_TARGET_REG_BITS 32
+#else
+#error unsupported
+#endif
+
+#define TCG_TARGET_WORDS_BIGENDIAN
+
+#define TCG_TARGET_NB_REGS 32
+
+enum {
+ TCG_REG_R0 = 0,
+ TCG_REG_R1,
+ TCG_REG_RP,
+ TCG_REG_R3,
+ TCG_REG_R4,
+ TCG_REG_R5,
+ TCG_REG_R6,
+ TCG_REG_R7,
+ TCG_REG_R8,
+ TCG_REG_R9,
+ TCG_REG_R10,
+ TCG_REG_R11,
+ TCG_REG_R12,
+ TCG_REG_R13,
+ TCG_REG_R14,
+ TCG_REG_R15,
+ TCG_REG_R16,
+ TCG_REG_R17,
+ TCG_REG_R18,
+ TCG_REG_R19,
+ TCG_REG_R20,
+ TCG_REG_R21,
+ TCG_REG_R22,
+ TCG_REG_R23,
+ TCG_REG_R24,
+ TCG_REG_R25,
+ TCG_REG_R26,
+ TCG_REG_DP,
+ TCG_REG_RET0,
+ TCG_REG_RET1,
+ TCG_REG_SP,
+ TCG_REG_R31,
+};
+
+/* used for function call generation */
+#define TCG_REG_CALL_STACK TCG_REG_SP
+#define TCG_TARGET_STACK_ALIGN 16
+#define TCG_TARGET_STACK_GROWSUP
+
+/* optional instructions */
+//#define TCG_TARGET_HAS_ext8s_i32
+//#define TCG_TARGET_HAS_ext16s_i32
+//#define TCG_TARGET_HAS_bswap16_i32
+//#define TCG_TARGET_HAS_bswap_i32
+
+/* Note: must be synced with dyngen-exec.h */
+#define TCG_AREG0 TCG_REG_R17
+#define TCG_AREG1 TCG_REG_R14
+#define TCG_AREG2 TCG_REG_R15
+#define TCG_AREG3 TCG_REG_R16
+
+static inline void flush_icache_range(unsigned long start, unsigned long stop)
+{
+ start &= ~31;
+ while (start <= stop)
+ {
+ asm volatile ("fdc 0(%0)\n"
+ "sync\n"
+ "fic 0(%%sr4, %0)\n"
+ "sync\n"
+ : : "r"(start) : "memory");
+ start += 32;
+ }
+}
+
+/* supplied by libgcc */
+extern void *__canonicalize_funcptr_for_compare(void *);
+
+/* Field selection types defined by hppa */
+#define rnd(x) (((x)+0x1000)&~0x1fff)
+/* lsel: select left 21 bits */
+#define lsel(v,a) (((v)+(a))>>11)
+/* rsel: select right 11 bits */
+#define rsel(v,a) (((v)+(a))&0x7ff)
+/* lrsel with rounding of addend to nearest 8k */
+#define lrsel(v,a) (((v)+rnd(a))>>11)
+/* rrsel with rounding of addend to nearest 8k */
+#define rrsel(v,a) ((((v)+rnd(a))&0x7ff)+((a)-rnd(a)))
+
+#define mask(x,sz) ((x) & ~((1<<(sz))-1))
+
+static inline int reassemble_12(int as12)
+{
+ return (((as12 & 0x800) >> 11) |
+ ((as12 & 0x400) >> 8) |
+ ((as12 & 0x3ff) << 3));
+}
+
+static inline int reassemble_14(int as14)
+{
+ return (((as14 & 0x1fff) << 1) |
+ ((as14 & 0x2000) >> 13));
+}
+
+static inline int reassemble_17(int as17)
+{
+ return (((as17 & 0x10000) >> 16) |
+ ((as17 & 0x0f800) << 5) |
+ ((as17 & 0x00400) >> 8) |
+ ((as17 & 0x003ff) << 3));
+}
+
+static inline int reassemble_21(int as21)
+{
+ return (((as21 & 0x100000) >> 20) |
+ ((as21 & 0x0ffe00) >> 8) |
+ ((as21 & 0x000180) << 7) |
+ ((as21 & 0x00007c) << 14) |
+ ((as21 & 0x000003) << 12));
+}
+
+static inline void hppa_patch21l(uint32_t *insn, int val, int addend)
+{
+ val = lrsel(val, addend);
+ *insn = mask(*insn, 21) | reassemble_21(val);
+}
+
+static inline void hppa_patch14r(uint32_t *insn, int val, int addend)
+{
+ val = rrsel(val, addend);
+ *insn = mask(*insn, 14) | reassemble_14(val);
+}
+
+static inline void hppa_patch17r(uint32_t *insn, int val, int addend)
+{
+ val = rrsel(val, addend);
+ *insn = (*insn & ~0x1f1ffd) | reassemble_17(val);
+}
+
+
+static inline void hppa_patch21l_dprel(uint32_t *insn, int val, int addend)
+{
+ register unsigned int dp asm("r27");
+ hppa_patch21l(insn, val - dp, addend);
+}
+
+static inline void hppa_patch14r_dprel(uint32_t *insn, int val, int addend)
+{
+ register unsigned int dp asm("r27");
+ hppa_patch14r(insn, val - dp, addend);
+}
+
+static inline void hppa_patch17f(uint32_t *insn, int val, int addend)
+{
+ int dot = (int)insn & ~0x3;
+ int v = ((val + addend) - dot - 8) / 4;
+ if (v > (1 << 16) || v < -(1 << 16)) {
+ printf("cannot fit branch to offset %d [%08x->%08x]\n", v, dot, val);
+ abort();
+ }
+ *insn = (*insn & ~0x1f1ffd) | reassemble_17(v);
+}
+
+static inline void hppa_load_imm21l(uint32_t *insn, int val, int addend)
+{
+ /* Transform addil L'sym(%dp) to ldil L'val, %r1 */
+ *insn = 0x20200000 | reassemble_21(lrsel(val, 0));
+}
+
+static inline void hppa_load_imm14r(uint32_t *insn, int val, int addend)
+{
+ /* Transform ldw R'sym(%r1), %rN to ldo R'sym(%r1), %rN */
+ hppa_patch14r(insn, val, addend);
+ /* HACK */
+ if (addend == 0)
+ *insn = (*insn & ~0xfc000000) | (0x0d << 26);
+}
diff --git a/tcg/i386/tcg-target.c b/tcg/i386/tcg-target.c
index 54df7dd2e..51eb6dd3e 100644
--- a/tcg/i386/tcg-target.c
+++ b/tcg/i386/tcg-target.c
@@ -47,8 +47,9 @@ const int tcg_target_call_iarg_regs[3] = { TCG_REG_EAX, TCG_REG_EDX, TCG_REG_ECX
const int tcg_target_call_oarg_regs[2] = { TCG_REG_EAX, TCG_REG_EDX };
static void patch_reloc(uint8_t *code_ptr, int type,
- tcg_target_long value)
+ tcg_target_long value, tcg_target_long addend)
{
+ value += addend;
switch(type) {
case R_386_32:
*(uint32_t *)code_ptr = value;
diff --git a/tcg/sparc/tcg-target.c b/tcg/sparc/tcg-target.c
index 340ef94a1..fc58d5056 100644
--- a/tcg/sparc/tcg-target.c
+++ b/tcg/sparc/tcg-target.c
@@ -88,8 +88,9 @@ static const int tcg_target_call_oarg_regs[2] = {
};
static void patch_reloc(uint8_t *code_ptr, int type,
- tcg_target_long value)
+ tcg_target_long value, tcg_target_long addend)
{
+ value += addend;
switch (type) {
case R_SPARC_32:
if (value != (uint32_t)value)
diff --git a/tcg/tcg-dyngen.c b/tcg/tcg-dyngen.c
index 578037212..77502af04 100644
--- a/tcg/tcg-dyngen.c
+++ b/tcg/tcg-dyngen.c
@@ -465,10 +465,55 @@ static inline void ia64_apply_fixes (uint8_t **gen_code_pp,
#endif
#ifndef CONFIG_NO_DYNGEN_OP
+
+#if defined __hppa__
+struct hppa_branch_stub {
+ uint32_t *location;
+ long target;
+ struct hppa_branch_stub *next;
+};
+
+#define HPPA_RECORD_BRANCH(LIST, LOC, TARGET) \
+do { \
+ struct hppa_branch_stub *stub = alloca(sizeof(struct hppa_branch_stub)); \
+ stub->location = LOC; \
+ stub->target = TARGET; \
+ stub->next = LIST; \
+ LIST = stub; \
+} while (0)
+
+static inline void hppa_process_stubs(struct hppa_branch_stub *stub,
+ uint8_t **gen_code_pp)
+{
+ uint32_t *s = (uint32_t *)*gen_code_pp;
+ uint32_t *p = s + 1;
+
+ if (!stub) return;
+
+ for (; stub != NULL; stub = stub->next) {
+ unsigned long l = (unsigned long)p;
+ /* stub:
+ * ldil L'target, %r1
+ * be,n R'target(%sr4,%r1)
+ */
+ *p++ = 0x20200000 | reassemble_21(lrsel(stub->target, 0));
+ *p++ = 0xe0202002 | (reassemble_17(rrsel(stub->target, 0) >> 2));
+ hppa_patch17f(stub->location, l, 0);
+ }
+ /* b,l,n stub,%r0 */
+ *s = 0xe8000002 | reassemble_17((p - s) - 2);
+ *gen_code_pp = (uint8_t *)p;
+}
+#endif /* __hppa__ */
+
const TCGArg *dyngen_op(TCGContext *s, int opc, const TCGArg *opparam_ptr)
{
uint8_t *gen_code_ptr;
+#ifdef __hppa__
+ struct hppa_branch_stub *hppa_stubs = NULL;
+#endif
+
gen_code_ptr = s->code_ptr;
switch(opc) {
@@ -478,6 +523,11 @@ const TCGArg *dyngen_op(TCGContext *s, int opc, const TCGArg *opparam_ptr)
default:
tcg_abort();
}
+
+#ifdef __hppa__
+ hppa_process_stubs(hppa_stubs, &gen_code_ptr);
+#endif
+
s->code_ptr = gen_code_ptr;
return opparam_ptr;
}
diff --git a/tcg/tcg.c b/tcg/tcg.c
index 26e0ff80a..c62d06a64 100644
--- a/tcg/tcg.c
+++ b/tcg/tcg.c
@@ -53,7 +53,7 @@
static void patch_reloc(uint8_t *code_ptr, int type,
- tcg_target_long value);
+ tcg_target_long value, tcg_target_long addend);
TCGOpDef tcg_op_defs[] = {
#define DEF(s, n, copy_size) { #s, 0, 0, n, n, 0, copy_size },
@@ -100,7 +100,7 @@ void tcg_out_reloc(TCGContext *s, uint8_t *code_ptr, int type,
/* FIXME: This may break relocations on RISC targets that
modify instruction fields in place. The caller may not have
written the initial value. */
- patch_reloc(code_ptr, type, l->u.value + addend);
+ patch_reloc(code_ptr, type, l->u.value, addend);
} else {
/* add a new relocation entry */
r = tcg_malloc(sizeof(TCGRelocation));
@@ -123,7 +123,7 @@ static void tcg_out_label(TCGContext *s, int label_index,
tcg_abort();
r = l->u.first_reloc;
while (r != NULL) {
- patch_reloc(r->ptr, r->type, value + r->addend);
+ patch_reloc(r->ptr, r->type, value, r->addend);
r = r->next;
}
l->has_value = 1;
@@ -1442,7 +1442,7 @@ static int tcg_reg_alloc_call(TCGContext *s, const TCGOpDef *def,
int nb_iargs, nb_oargs, flags, nb_regs, i, reg, nb_params;
TCGArg arg, func_arg;
TCGTemp *ts;
- tcg_target_long stack_offset, call_stack_size;
+ tcg_target_long stack_offset, call_stack_size, func_addr;
int const_func_arg;
TCGRegSet allocated_regs;
const TCGArgConstraint *arg_ct;
@@ -1464,7 +1464,11 @@ static int tcg_reg_alloc_call(TCGContext *s, const TCGOpDef *def,
call_stack_size = (nb_params - nb_regs) * sizeof(tcg_target_long);
call_stack_size = (call_stack_size + TCG_TARGET_STACK_ALIGN - 1) &
~(TCG_TARGET_STACK_ALIGN - 1);
+#ifdef TCG_TARGET_STACK_GROWSUP
+ tcg_out_addi(s, TCG_REG_CALL_STACK, call_stack_size);
+#else
tcg_out_addi(s, TCG_REG_CALL_STACK, -call_stack_size);
+#endif
stack_offset = 0;
for(i = nb_regs; i < nb_params; i++) {
@@ -1487,7 +1491,11 @@ static int tcg_reg_alloc_call(TCGContext *s, const TCGOpDef *def,
} else {
tcg_abort();
}
+#ifdef TCG_TARGET_STACK_GROWSUP
+ stack_offset -= sizeof(tcg_target_long);
+#else
stack_offset += sizeof(tcg_target_long);
+#endif
}
/* assign input registers */
@@ -1516,6 +1524,10 @@ static int tcg_reg_alloc_call(TCGContext *s, const TCGOpDef *def,
func_arg = args[nb_oargs + nb_iargs - 1];
arg_ct = &def->args_ct[0];
ts = &s->temps[func_arg];
+ func_addr = ts->val;
+#ifdef HOST_HPPA
+ func_addr = (tcg_target_long)__canonicalize_funcptr_for_compare((void *)func_addr);
+#endif
const_func_arg = 0;
if (ts->val_type == TEMP_VAL_MEM) {
reg = tcg_reg_alloc(s, arg_ct->u.regs, allocated_regs);
@@ -1529,12 +1541,12 @@ static int tcg_reg_alloc_call(TCGContext *s, const TCGOpDef *def,
}
func_arg = reg;
} else if (ts->val_type == TEMP_VAL_CONST) {
- if (tcg_target_const_match(ts->val, arg_ct)) {
+ if (tcg_target_const_match(func_addr, arg_ct)) {
const_func_arg = 1;
- func_arg = ts->val;
+ func_arg = func_addr;
} else {
reg = tcg_reg_alloc(s, arg_ct->u.regs, allocated_regs);
- tcg_out_movi(s, ts->type, reg, ts->val);
+ tcg_out_movi(s, ts->type, reg, func_addr);
func_arg = reg;
}
} else {
@@ -1574,7 +1586,11 @@ static int tcg_reg_alloc_call(TCGContext *s, const TCGOpDef *def,
tcg_out_op(s, opc, &func_arg, &const_func_arg);
+#ifdef TCG_TARGET_STACK_GROWSUP
+ tcg_out_addi(s, TCG_REG_CALL_STACK, -call_stack_size);
+#else
tcg_out_addi(s, TCG_REG_CALL_STACK, call_stack_size);
+#endif
/* assign output registers and emit moves if needed */
for(i = 0; i < nb_oargs; i++) {
diff --git a/tcg/x86_64/tcg-target.c b/tcg/x86_64/tcg-target.c
index 9cd89b277..480ffcbc2 100644
--- a/tcg/x86_64/tcg-target.c
+++ b/tcg/x86_64/tcg-target.c
@@ -74,8 +74,9 @@ const int tcg_target_call_oarg_regs[2] = {
};
static void patch_reloc(uint8_t *code_ptr, int type,
- tcg_target_long value)
+ tcg_target_long value, tcg_target_long addend)
{
+ value += addend;
switch(type) {
case R_X86_64_32:
if (value != (uint32_t)value)