+Thu Feb 22 22:48:57 1996 Andrew Cagney <cagney@highland.com.au>
+
+ * README, RUN, INSTALL: Update to reflect announcement
+ * psim: PSIM 1.0.1 released
+
Thu Feb 22 14:01:56 1996 Michael Meissner <meissner@tiktok.cygnus.com>
* emul_bugapi.c (emul_bugapi_do_read): New function to handle
Correct implementation of _OUTLN. Add support for _OUTSTR and
_PCRLR system calls.
- * psim.c (psim_options): Call device_add_string_property, not
- device_tree_add_parsed, since we want to add the model as a
- string, and 603/604 look like integers.
+Wed Feb 21 17:07:27 1996 Andrew Cagney <cagney@highland.com.au>
+
+ * NOTES: New file. Ramblings on why things were done they way
+ they were.
+
+ * psim.c (psim_options): Didn't enter the model value into the
+ device tree as a string.
+
+ * cpu.c (cpu_synchronize_context): Wrong test for conditional
+ flush of cache.
+
+ * emul_generic.c (emul_add_tree_hardware): reg value didn't match
+ bus address.
+
+ * ppc-opcode-flat: new file. Generate an instruction decode
+ function like ppc-opcode-complex but use switch statements.
+
+ * INSTALL: document new opcode file, add example configurations.
+
+Tue Feb 20 18:42:31 1996 Andrew Cagney <cagney@highland.com.au>
+
+ * main.c (main): rename psim instance (system) to simulation and
+ make global.
+ * main.c (error): print out performance even when an error occures.
+
+ * emul_bugapi.c (emul_bugapi_create): Fix argument passing.
- * emul_bugapi.c (emul_bugapi_create): Make formats type
- compatible.
+ * emul_generic.c (emul_add_tree_hardware): Move hardware devices
+ to 0x80000000 from 0x400000.
Mon Feb 19 22:54:40 1996 Andrew Cagney <cagney@highland.com.au>
--- /dev/null
+#
+# This file is part of the program psim.
+#
+# Copyright (C) 1994-1995, Andrew Cagney <cagney@highland.com.au>
+#
+# 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., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
+#
+# Instruction decode:
+#
+# The table that follows is used by gen to construct a decision tree
+# that can identify each possible instruction. Gen then outputs this
+# decision tree as (according to config) a table or switch statement
+# as the function idecode.
+#
+# In parallel to this, as mentioned above, WITH_EXPANDED_SEMANTICS
+# determines of the semantic functions themselves should be expanded
+# in a similar way.
+#
+# The table contains the following entries:
+#
+# <valid>
+#
+# Must be 1 for the entry to be considered. The last entry must be
+# zero.
+#
+# <first>
+# <last>
+#
+# Range of bits (within the instruction) that should be searched for
+# an instruction field. Within such ranges, gen looks for opcodes
+# (constants), registers (strings) and reserved bits (slash) and
+# according to the rules that follows includes or excludes them from
+# a possible instruction field.
+#
+# <force_first>
+# <force_last>
+#
+# If an instructioin field was found, enlarge the field size so that
+# it is forced to at least include bits starting from <force_first>
+# (<force_last>). To stop this occuring, use <force_first> = <last>
+# + 1 and <force_last> = <first> - 1.
+#
+# <force_slash>
+#
+# Treat `/' fields as a constant instead of variable when looking for
+# an instruction field.
+#
+# <force_expansion>
+#
+# Treat any contained register (string) fields as constant when
+# determining the instruction field. For the instruction decode (and
+# controled by IDECODE_EXPAND_SEMANTICS) this forces the expansion of
+# what would otherwize be non constant bits of an instruction.
+#
+# <use_switch>
+#
+# Should this table be expanded using a switch statement (val 1) and
+# if so, should it be padded with entries so as to force the compiler
+# to generate a jump table (val 2).
+#
+# <special_mask>
+# <special_value>
+# <special_rule>
+#
+# Special rule to fine tune how specific (or groups) of instructions
+# are expanded. The applicability of the rule is determined by
+#
+# <special_mask> != 0 && (instruction> & <special_mask>) == <special_value>
+#
+# Where <instruction> is obtained by looking only at constant fields
+# with in an instructions spec. When determining an expansion, the
+# rule is only considered when a node contains a single instruction.
+# <special_rule> can be any of:
+#
+# 0: for this instruction, expand by earlier rules
+# 1: expand bits <force_low> .. <force_hi> only
+# 2: boolean expansion of only zero/non-zero cases
+#
+ 0: 5: 0: 5:0:: 2:0x00000000:0x00000000:0
+21:31:32:-1:0:OE,LR,AA,Rc,LK:1:0x00000000:0x00000000:0
+ 6: 9: 6: 9:0:BO: 1:0xfc000000:0x40000000:1
+11:15:11:15:0:RA: 1:0xfc000000:0x38000000:2
+11:15:11:15:0:RA: 1:0xfc000000:0x3c000000:2