2010-06-02 19:58:16 +02:00
|
|
|
/*
|
2012-03-02 00:26:31 +01:00
|
|
|
* Copyright (c) 2010-2012 ARM Limited
|
2010-06-02 19:58:16 +02:00
|
|
|
* All rights reserved
|
|
|
|
*
|
|
|
|
* The license below extends only to copyright in the software and shall
|
|
|
|
* not be construed as granting a license to any other intellectual
|
|
|
|
* property including but not limited to intellectual property relating
|
|
|
|
* to a hardware implementation of the functionality of the software
|
|
|
|
* licensed hereunder. You may use the software subject to the license
|
|
|
|
* terms below provided that you ensure that this notice is replicated
|
|
|
|
* unmodified and in its entirety in all distributions of the software,
|
|
|
|
* modified or unmodified, in source code or in binary form.
|
|
|
|
*
|
|
|
|
* Redistribution and use in source and binary forms, with or without
|
|
|
|
* modification, are permitted provided that the following conditions are
|
|
|
|
* met: redistributions of source code must retain the above copyright
|
|
|
|
* notice, this list of conditions and the following disclaimer;
|
|
|
|
* redistributions in binary form must reproduce the above copyright
|
|
|
|
* notice, this list of conditions and the following disclaimer in the
|
|
|
|
* documentation and/or other materials provided with the distribution;
|
|
|
|
* neither the name of the copyright holders nor the names of its
|
|
|
|
* contributors may be used to endorse or promote products derived from
|
|
|
|
* this software without specific prior written permission.
|
|
|
|
*
|
|
|
|
* THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
|
|
|
|
* "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
|
|
|
|
* LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
|
|
|
|
* A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
|
|
|
|
* OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
|
|
|
|
* SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
|
|
|
|
* LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
|
|
|
|
* DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
|
|
|
|
* THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
|
|
|
|
* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
|
|
|
|
* OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
|
|
|
|
*
|
|
|
|
* Authors: Gabe Black
|
|
|
|
* Ali Saidi
|
|
|
|
*/
|
|
|
|
|
|
|
|
#include "arch/arm/isa.hh"
|
2012-01-31 16:46:03 +01:00
|
|
|
#include "config/use_checker.hh"
|
2011-04-15 19:44:32 +02:00
|
|
|
#include "debug/Arm.hh"
|
|
|
|
#include "debug/MiscRegs.hh"
|
2010-09-14 04:26:03 +02:00
|
|
|
#include "sim/faults.hh"
|
2011-02-23 22:10:48 +01:00
|
|
|
#include "sim/stat_control.hh"
|
2011-05-05 03:38:28 +02:00
|
|
|
#include "sim/system.hh"
|
2010-06-02 19:58:16 +02:00
|
|
|
|
2012-01-31 16:46:03 +01:00
|
|
|
#if USE_CHECKER
|
|
|
|
#include "cpu/checker/cpu.hh"
|
|
|
|
#endif
|
|
|
|
|
2010-06-02 19:58:16 +02:00
|
|
|
namespace ArmISA
|
|
|
|
{
|
|
|
|
|
2010-06-02 19:58:17 +02:00
|
|
|
void
|
|
|
|
ISA::clear()
|
|
|
|
{
|
|
|
|
SCTLR sctlr_rst = miscRegs[MISCREG_SCTLR_RST];
|
2011-05-14 00:27:00 +02:00
|
|
|
uint32_t midr = miscRegs[MISCREG_MIDR];
|
2010-06-02 19:58:17 +02:00
|
|
|
memset(miscRegs, 0, sizeof(miscRegs));
|
|
|
|
CPSR cpsr = 0;
|
|
|
|
cpsr.mode = MODE_USER;
|
|
|
|
miscRegs[MISCREG_CPSR] = cpsr;
|
|
|
|
updateRegMap(cpsr);
|
|
|
|
|
|
|
|
SCTLR sctlr = 0;
|
2010-08-23 18:18:41 +02:00
|
|
|
sctlr.te = (bool)sctlr_rst.te;
|
2010-06-02 19:58:17 +02:00
|
|
|
sctlr.nmfi = (bool)sctlr_rst.nmfi;
|
|
|
|
sctlr.v = (bool)sctlr_rst.v;
|
|
|
|
sctlr.u = 1;
|
|
|
|
sctlr.xp = 1;
|
|
|
|
sctlr.rao2 = 1;
|
|
|
|
sctlr.rao3 = 1;
|
|
|
|
sctlr.rao4 = 1;
|
|
|
|
miscRegs[MISCREG_SCTLR] = sctlr;
|
|
|
|
miscRegs[MISCREG_SCTLR_RST] = sctlr_rst;
|
|
|
|
|
2011-05-14 00:27:00 +02:00
|
|
|
// Preserve MIDR accross reset
|
|
|
|
miscRegs[MISCREG_MIDR] = midr;
|
|
|
|
|
2010-06-02 19:58:17 +02:00
|
|
|
/* Start with an event in the mailbox */
|
|
|
|
miscRegs[MISCREG_SEV_MAILBOX] = 1;
|
|
|
|
|
|
|
|
// Separate Instruction and Data TLBs.
|
|
|
|
miscRegs[MISCREG_TLBTR] = 1;
|
|
|
|
|
|
|
|
MVFR0 mvfr0 = 0;
|
|
|
|
mvfr0.advSimdRegisters = 2;
|
|
|
|
mvfr0.singlePrecision = 2;
|
|
|
|
mvfr0.doublePrecision = 2;
|
|
|
|
mvfr0.vfpExceptionTrapping = 0;
|
|
|
|
mvfr0.divide = 1;
|
|
|
|
mvfr0.squareRoot = 1;
|
|
|
|
mvfr0.shortVectors = 1;
|
|
|
|
mvfr0.roundingModes = 1;
|
|
|
|
miscRegs[MISCREG_MVFR0] = mvfr0;
|
|
|
|
|
|
|
|
MVFR1 mvfr1 = 0;
|
|
|
|
mvfr1.flushToZero = 1;
|
|
|
|
mvfr1.defaultNaN = 1;
|
|
|
|
mvfr1.advSimdLoadStore = 1;
|
|
|
|
mvfr1.advSimdInteger = 1;
|
|
|
|
mvfr1.advSimdSinglePrecision = 1;
|
|
|
|
mvfr1.advSimdHalfPrecision = 1;
|
|
|
|
mvfr1.vfpHalfPrecision = 1;
|
|
|
|
miscRegs[MISCREG_MVFR1] = mvfr1;
|
|
|
|
|
|
|
|
miscRegs[MISCREG_MPIDR] = 0;
|
|
|
|
|
2010-06-02 19:58:18 +02:00
|
|
|
// Reset values of PRRR and NMRR are implementation dependent
|
|
|
|
|
|
|
|
miscRegs[MISCREG_PRRR] =
|
|
|
|
(1 << 19) | // 19
|
|
|
|
(0 << 18) | // 18
|
|
|
|
(0 << 17) | // 17
|
|
|
|
(1 << 16) | // 16
|
|
|
|
(2 << 14) | // 15:14
|
|
|
|
(0 << 12) | // 13:12
|
|
|
|
(2 << 10) | // 11:10
|
|
|
|
(2 << 8) | // 9:8
|
|
|
|
(2 << 6) | // 7:6
|
|
|
|
(2 << 4) | // 5:4
|
|
|
|
(1 << 2) | // 3:2
|
|
|
|
0; // 1:0
|
|
|
|
miscRegs[MISCREG_NMRR] =
|
|
|
|
(1 << 30) | // 31:30
|
|
|
|
(0 << 26) | // 27:26
|
|
|
|
(0 << 24) | // 25:24
|
|
|
|
(3 << 22) | // 23:22
|
|
|
|
(2 << 20) | // 21:20
|
|
|
|
(0 << 18) | // 19:18
|
|
|
|
(0 << 16) | // 17:16
|
|
|
|
(1 << 14) | // 15:14
|
|
|
|
(0 << 12) | // 13:12
|
|
|
|
(2 << 10) | // 11:10
|
|
|
|
(0 << 8) | // 9:8
|
|
|
|
(3 << 6) | // 7:6
|
|
|
|
(2 << 4) | // 5:4
|
|
|
|
(0 << 2) | // 3:2
|
|
|
|
0; // 1:0
|
|
|
|
|
2010-08-26 02:10:42 +02:00
|
|
|
miscRegs[MISCREG_CPACR] = 0;
|
|
|
|
miscRegs[MISCREG_FPSID] = 0x410430A0;
|
2011-03-18 01:20:20 +01:00
|
|
|
|
|
|
|
// See section B4.1.84 of ARM ARM
|
|
|
|
// All values are latest for ARMv7-A profile
|
2011-08-19 22:08:07 +02:00
|
|
|
miscRegs[MISCREG_ID_ISAR0] = 0x02101111;
|
2011-03-18 01:20:20 +01:00
|
|
|
miscRegs[MISCREG_ID_ISAR1] = 0x02112111;
|
|
|
|
miscRegs[MISCREG_ID_ISAR2] = 0x21232141;
|
|
|
|
miscRegs[MISCREG_ID_ISAR3] = 0x01112131;
|
|
|
|
miscRegs[MISCREG_ID_ISAR4] = 0x10010142;
|
|
|
|
miscRegs[MISCREG_ID_ISAR5] = 0x00000000;
|
|
|
|
|
2010-06-02 19:58:17 +02:00
|
|
|
//XXX We need to initialize the rest of the state.
|
|
|
|
}
|
|
|
|
|
2010-06-02 19:58:16 +02:00
|
|
|
MiscReg
|
|
|
|
ISA::readMiscRegNoEffect(int misc_reg)
|
|
|
|
{
|
|
|
|
assert(misc_reg < NumMiscRegs);
|
2010-08-23 18:18:41 +02:00
|
|
|
|
|
|
|
int flat_idx;
|
|
|
|
if (misc_reg == MISCREG_SPSR)
|
|
|
|
flat_idx = flattenMiscIndex(misc_reg);
|
|
|
|
else
|
|
|
|
flat_idx = misc_reg;
|
|
|
|
MiscReg val = miscRegs[flat_idx];
|
|
|
|
|
|
|
|
DPRINTF(MiscRegs, "Reading From misc reg %d (%d) : %#x\n",
|
|
|
|
misc_reg, flat_idx, val);
|
|
|
|
return val;
|
2010-06-02 19:58:16 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
MiscReg
|
|
|
|
ISA::readMiscReg(int misc_reg, ThreadContext *tc)
|
|
|
|
{
|
|
|
|
if (misc_reg == MISCREG_CPSR) {
|
|
|
|
CPSR cpsr = miscRegs[misc_reg];
|
ISA,CPU,etc: Create an ISA defined PC type that abstracts out ISA behaviors.
This change is a low level and pervasive reorganization of how PCs are managed
in M5. Back when Alpha was the only ISA, there were only 2 PCs to worry about,
the PC and the NPC, and the lsb of the PC signaled whether or not you were in
PAL mode. As other ISAs were added, we had to add an NNPC, micro PC and next
micropc, x86 and ARM introduced variable length instruction sets, and ARM
started to keep track of mode bits in the PC. Each CPU model handled PCs in
its own custom way that needed to be updated individually to handle the new
dimensions of variability, or, in the case of ARMs mode-bit-in-the-pc hack,
the complexity could be hidden in the ISA at the ISA implementation's expense.
Areas like the branch predictor hadn't been updated to handle branch delay
slots or micropcs, and it turns out that had introduced a significant (10s of
percent) performance bug in SPARC and to a lesser extend MIPS. Rather than
perpetuate the problem by reworking O3 again to handle the PC features needed
by x86, this change was introduced to rework PC handling in a more modular,
transparent, and hopefully efficient way.
PC type:
Rather than having the superset of all possible elements of PC state declared
in each of the CPU models, each ISA defines its own PCState type which has
exactly the elements it needs. A cross product of canned PCState classes are
defined in the new "generic" ISA directory for ISAs with/without delay slots
and microcode. These are either typedef-ed or subclassed by each ISA. To read
or write this structure through a *Context, you use the new pcState() accessor
which reads or writes depending on whether it has an argument. If you just
want the address of the current or next instruction or the current micro PC,
you can get those through read-only accessors on either the PCState type or
the *Contexts. These are instAddr(), nextInstAddr(), and microPC(). Note the
move away from readPC. That name is ambiguous since it's not clear whether or
not it should be the actual address to fetch from, or if it should have extra
bits in it like the PAL mode bit. Each class is free to define its own
functions to get at whatever values it needs however it needs to to be used in
ISA specific code. Eventually Alpha's PAL mode bit could be moved out of the
PC and into a separate field like ARM.
These types can be reset to a particular pc (where npc = pc +
sizeof(MachInst), nnpc = npc + sizeof(MachInst), upc = 0, nupc = 1 as
appropriate), printed, serialized, and compared. There is a branching()
function which encapsulates code in the CPU models that checked if an
instruction branched or not. Exactly what that means in the context of branch
delay slots which can skip an instruction when not taken is ambiguous, and
ideally this function and its uses can be eliminated. PCStates also generally
know how to advance themselves in various ways depending on if they point at
an instruction, a microop, or the last microop of a macroop. More on that
later.
Ideally, accessing all the PCs at once when setting them will improve
performance of M5 even though more data needs to be moved around. This is
because often all the PCs need to be manipulated together, and by getting them
all at once you avoid multiple function calls. Also, the PCs of a particular
thread will have spatial locality in the cache. Previously they were grouped
by element in arrays which spread out accesses.
Advancing the PC:
The PCs were previously managed entirely by the CPU which had to know about PC
semantics, try to figure out which dimension to increment the PC in, what to
set NPC/NNPC, etc. These decisions are best left to the ISA in conjunction
with the PC type itself. Because most of the information about how to
increment the PC (mainly what type of instruction it refers to) is contained
in the instruction object, a new advancePC virtual function was added to the
StaticInst class. Subclasses provide an implementation that moves around the
right element of the PC with a minimal amount of decision making. In ISAs like
Alpha, the instructions always simply assign NPC to PC without having to worry
about micropcs, nnpcs, etc. The added cost of a virtual function call should
be outweighed by not having to figure out as much about what to do with the
PCs and mucking around with the extra elements.
One drawback of making the StaticInsts advance the PC is that you have to
actually have one to advance the PC. This would, superficially, seem to
require decoding an instruction before fetch could advance. This is, as far as
I can tell, realistic. fetch would advance through memory addresses, not PCs,
perhaps predicting new memory addresses using existing ones. More
sophisticated decisions about control flow would be made later on, after the
instruction was decoded, and handed back to fetch. If branching needs to
happen, some amount of decoding needs to happen to see that it's a branch,
what the target is, etc. This could get a little more complicated if that gets
done by the predecoder, but I'm choosing to ignore that for now.
Variable length instructions:
To handle variable length instructions in x86 and ARM, the predecoder now
takes in the current PC by reference to the getExtMachInst function. It can
modify the PC however it needs to (by setting NPC to be the PC + instruction
length, for instance). This could be improved since the CPU doesn't know if
the PC was modified and always has to write it back.
ISA parser:
To support the new API, all PC related operand types were removed from the
parser and replaced with a PCState type. There are two warts on this
implementation. First, as with all the other operand types, the PCState still
has to have a valid operand type even though it doesn't use it. Second, using
syntax like PCS.npc(target) doesn't work for two reasons, this looks like the
syntax for operand type overriding, and the parser can't figure out if you're
reading or writing. Instructions that use the PCS operand (which I've
consistently called it) need to first read it into a local variable,
manipulate it, and then write it back out.
Return address stack:
The return address stack needed a little extra help because, in the presence
of branch delay slots, it has to merge together elements of the return PC and
the call PC. To handle that, a buildRetPC utility function was added. There
are basically only two versions in all the ISAs, but it didn't seem short
enough to put into the generic ISA directory. Also, the branch predictor code
in O3 and InOrder were adjusted so that they always store the PC of the actual
call instruction in the RAS, not the next PC. If the call instruction is a
microop, the next PC refers to the next microop in the same macroop which is
probably not desirable. The buildRetPC function advances the PC intelligently
to the next macroop (in an ISA specific way) so that that case works.
Change in stats:
There were no change in stats except in MIPS and SPARC in the O3 model. MIPS
runs in about 9% fewer ticks. SPARC runs with 30%-50% fewer ticks, which could
likely be improved further by setting call/return instruction flags and taking
advantage of the RAS.
TODO:
Add != operators to the PCState classes, defined trivially to be !(a==b).
Smooth out places where PCs are split apart, passed around, and put back
together later. I think this might happen in SPARC's fault code. Add ISA
specific constructors that allow setting PC elements without calling a bunch
of accessors. Try to eliminate the need for the branching() function. Factor
out Alpha's PAL mode pc bit into a separate flag field, and eliminate places
where it's blindly masked out or tested in the PC.
2010-10-31 08:07:20 +01:00
|
|
|
PCState pc = tc->pcState();
|
|
|
|
cpsr.j = pc.jazelle() ? 1 : 0;
|
|
|
|
cpsr.t = pc.thumb() ? 1 : 0;
|
2010-06-02 19:58:16 +02:00
|
|
|
return cpsr;
|
|
|
|
}
|
2010-11-15 21:04:04 +01:00
|
|
|
if (misc_reg >= MISCREG_CP15_UNIMP_START)
|
2010-06-02 19:58:16 +02:00
|
|
|
panic("Unimplemented CP15 register %s read.\n",
|
|
|
|
miscRegName[misc_reg]);
|
2010-11-15 21:04:04 +01:00
|
|
|
|
2010-06-02 19:58:16 +02:00
|
|
|
switch (misc_reg) {
|
2011-05-05 03:38:28 +02:00
|
|
|
case MISCREG_MPIDR:
|
|
|
|
return tc->cpuId();
|
|
|
|
break;
|
2011-07-15 18:53:34 +02:00
|
|
|
case MISCREG_ID_MMFR0:
|
|
|
|
return 0x03; // VMSAv7 support
|
|
|
|
case MISCREG_ID_MMFR2:
|
|
|
|
return 0x01230000; // no HW access | WFI stalling | ISB and DSB
|
|
|
|
// | all TLB maintenance | no Harvard
|
2011-05-05 03:38:28 +02:00
|
|
|
case MISCREG_ID_MMFR3:
|
|
|
|
return 0xF0102211; // SuperSec | Coherent TLB | Bcast Maint |
|
|
|
|
// BP Maint | Cache Maint Set/way | Cache Maint MVA
|
2010-06-02 19:58:16 +02:00
|
|
|
case MISCREG_CLIDR:
|
2010-11-08 20:58:24 +01:00
|
|
|
warn_once("The clidr register always reports 0 caches.\n");
|
2011-07-15 18:53:34 +02:00
|
|
|
warn_once("clidr LoUIS field of 0b001 to match current "
|
|
|
|
"ARM implementations.\n");
|
|
|
|
return 0x00200000;
|
2010-06-02 19:58:16 +02:00
|
|
|
case MISCREG_CCSIDR:
|
2010-11-08 20:58:24 +01:00
|
|
|
warn_once("The ccsidr register isn't implemented and "
|
2010-06-02 19:58:16 +02:00
|
|
|
"always reads as 0.\n");
|
|
|
|
break;
|
|
|
|
case MISCREG_ID_PFR0:
|
2010-08-23 18:18:40 +02:00
|
|
|
warn("Returning thumbEE disabled for now since we don't support CP14"
|
|
|
|
"config registers and jumping to ThumbEE vectors\n");
|
|
|
|
return 0x0031; // !ThumbEE | !Jazelle | Thumb | ARM
|
2011-05-14 00:27:00 +02:00
|
|
|
case MISCREG_ID_PFR1:
|
|
|
|
warn("reading unimplmented register ID_PFR1");
|
|
|
|
return 0;
|
2010-08-23 18:18:40 +02:00
|
|
|
case MISCREG_CTR:
|
|
|
|
return 0x86468006; // V7, 64 byte cache line, load/exclusive is exact
|
|
|
|
case MISCREG_ACTLR:
|
|
|
|
warn("Not doing anything for miscreg ACTLR\n");
|
|
|
|
break;
|
|
|
|
case MISCREG_PMCR:
|
|
|
|
case MISCREG_PMCCNTR:
|
|
|
|
case MISCREG_PMSELR:
|
2011-05-14 00:27:00 +02:00
|
|
|
warn("Not doing anything for read to miscreg %s\n",
|
2010-08-23 18:18:40 +02:00
|
|
|
miscRegName[misc_reg]);
|
|
|
|
break;
|
2011-05-14 00:27:01 +02:00
|
|
|
case MISCREG_CPSR_Q:
|
|
|
|
panic("shouldn't be reading this register seperately\n");
|
2010-12-08 01:19:57 +01:00
|
|
|
case MISCREG_FPSCR_QC:
|
|
|
|
return readMiscRegNoEffect(MISCREG_FPSCR) & ~FpscrQcMask;
|
|
|
|
case MISCREG_FPSCR_EXC:
|
|
|
|
return readMiscRegNoEffect(MISCREG_FPSCR) & ~FpscrExcMask;
|
2011-09-13 19:06:13 +02:00
|
|
|
case MISCREG_L2CTLR:
|
2012-03-02 00:26:31 +01:00
|
|
|
{
|
|
|
|
// mostly unimplemented, just set NumCPUs field from sim and return
|
|
|
|
L2CTLR l2ctlr = 0;
|
|
|
|
// b00:1CPU to b11:4CPUs
|
|
|
|
l2ctlr.numCPUs = tc->getSystemPtr()->numContexts() - 1;
|
|
|
|
return l2ctlr;
|
|
|
|
}
|
|
|
|
case MISCREG_DBGDIDR:
|
|
|
|
/* For now just implement the version number.
|
|
|
|
* Return 0 as we don't support debug architecture yet.
|
|
|
|
*/
|
|
|
|
return 0;
|
|
|
|
case MISCREG_DBGDSCR_INT:
|
|
|
|
return 0;
|
2010-06-02 19:58:16 +02:00
|
|
|
}
|
|
|
|
return readMiscRegNoEffect(misc_reg);
|
|
|
|
}
|
|
|
|
|
|
|
|
void
|
|
|
|
ISA::setMiscRegNoEffect(int misc_reg, const MiscReg &val)
|
|
|
|
{
|
|
|
|
assert(misc_reg < NumMiscRegs);
|
2010-08-23 18:18:41 +02:00
|
|
|
|
|
|
|
int flat_idx;
|
|
|
|
if (misc_reg == MISCREG_SPSR)
|
|
|
|
flat_idx = flattenMiscIndex(misc_reg);
|
|
|
|
else
|
|
|
|
flat_idx = misc_reg;
|
|
|
|
miscRegs[flat_idx] = val;
|
|
|
|
|
|
|
|
DPRINTF(MiscRegs, "Writing to misc reg %d (%d) : %#x\n", misc_reg,
|
|
|
|
flat_idx, val);
|
2010-06-02 19:58:16 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
void
|
|
|
|
ISA::setMiscReg(int misc_reg, const MiscReg &val, ThreadContext *tc)
|
|
|
|
{
|
2010-11-15 21:04:03 +01:00
|
|
|
|
2010-06-02 19:58:16 +02:00
|
|
|
MiscReg newVal = val;
|
2011-05-05 03:38:28 +02:00
|
|
|
int x;
|
|
|
|
System *sys;
|
|
|
|
ThreadContext *oc;
|
|
|
|
|
2010-06-02 19:58:16 +02:00
|
|
|
if (misc_reg == MISCREG_CPSR) {
|
|
|
|
updateRegMap(val);
|
2010-11-15 21:04:03 +01:00
|
|
|
|
|
|
|
|
|
|
|
CPSR old_cpsr = miscRegs[MISCREG_CPSR];
|
|
|
|
int old_mode = old_cpsr.mode;
|
2010-06-02 19:58:16 +02:00
|
|
|
CPSR cpsr = val;
|
2010-11-15 21:04:03 +01:00
|
|
|
if (old_mode != cpsr.mode) {
|
|
|
|
tc->getITBPtr()->invalidateMiscReg();
|
|
|
|
tc->getDTBPtr()->invalidateMiscReg();
|
|
|
|
}
|
|
|
|
|
2010-08-23 18:18:41 +02:00
|
|
|
DPRINTF(Arm, "Updating CPSR from %#x to %#x f:%d i:%d a:%d mode:%#x\n",
|
|
|
|
miscRegs[misc_reg], cpsr, cpsr.f, cpsr.i, cpsr.a, cpsr.mode);
|
ISA,CPU,etc: Create an ISA defined PC type that abstracts out ISA behaviors.
This change is a low level and pervasive reorganization of how PCs are managed
in M5. Back when Alpha was the only ISA, there were only 2 PCs to worry about,
the PC and the NPC, and the lsb of the PC signaled whether or not you were in
PAL mode. As other ISAs were added, we had to add an NNPC, micro PC and next
micropc, x86 and ARM introduced variable length instruction sets, and ARM
started to keep track of mode bits in the PC. Each CPU model handled PCs in
its own custom way that needed to be updated individually to handle the new
dimensions of variability, or, in the case of ARMs mode-bit-in-the-pc hack,
the complexity could be hidden in the ISA at the ISA implementation's expense.
Areas like the branch predictor hadn't been updated to handle branch delay
slots or micropcs, and it turns out that had introduced a significant (10s of
percent) performance bug in SPARC and to a lesser extend MIPS. Rather than
perpetuate the problem by reworking O3 again to handle the PC features needed
by x86, this change was introduced to rework PC handling in a more modular,
transparent, and hopefully efficient way.
PC type:
Rather than having the superset of all possible elements of PC state declared
in each of the CPU models, each ISA defines its own PCState type which has
exactly the elements it needs. A cross product of canned PCState classes are
defined in the new "generic" ISA directory for ISAs with/without delay slots
and microcode. These are either typedef-ed or subclassed by each ISA. To read
or write this structure through a *Context, you use the new pcState() accessor
which reads or writes depending on whether it has an argument. If you just
want the address of the current or next instruction or the current micro PC,
you can get those through read-only accessors on either the PCState type or
the *Contexts. These are instAddr(), nextInstAddr(), and microPC(). Note the
move away from readPC. That name is ambiguous since it's not clear whether or
not it should be the actual address to fetch from, or if it should have extra
bits in it like the PAL mode bit. Each class is free to define its own
functions to get at whatever values it needs however it needs to to be used in
ISA specific code. Eventually Alpha's PAL mode bit could be moved out of the
PC and into a separate field like ARM.
These types can be reset to a particular pc (where npc = pc +
sizeof(MachInst), nnpc = npc + sizeof(MachInst), upc = 0, nupc = 1 as
appropriate), printed, serialized, and compared. There is a branching()
function which encapsulates code in the CPU models that checked if an
instruction branched or not. Exactly what that means in the context of branch
delay slots which can skip an instruction when not taken is ambiguous, and
ideally this function and its uses can be eliminated. PCStates also generally
know how to advance themselves in various ways depending on if they point at
an instruction, a microop, or the last microop of a macroop. More on that
later.
Ideally, accessing all the PCs at once when setting them will improve
performance of M5 even though more data needs to be moved around. This is
because often all the PCs need to be manipulated together, and by getting them
all at once you avoid multiple function calls. Also, the PCs of a particular
thread will have spatial locality in the cache. Previously they were grouped
by element in arrays which spread out accesses.
Advancing the PC:
The PCs were previously managed entirely by the CPU which had to know about PC
semantics, try to figure out which dimension to increment the PC in, what to
set NPC/NNPC, etc. These decisions are best left to the ISA in conjunction
with the PC type itself. Because most of the information about how to
increment the PC (mainly what type of instruction it refers to) is contained
in the instruction object, a new advancePC virtual function was added to the
StaticInst class. Subclasses provide an implementation that moves around the
right element of the PC with a minimal amount of decision making. In ISAs like
Alpha, the instructions always simply assign NPC to PC without having to worry
about micropcs, nnpcs, etc. The added cost of a virtual function call should
be outweighed by not having to figure out as much about what to do with the
PCs and mucking around with the extra elements.
One drawback of making the StaticInsts advance the PC is that you have to
actually have one to advance the PC. This would, superficially, seem to
require decoding an instruction before fetch could advance. This is, as far as
I can tell, realistic. fetch would advance through memory addresses, not PCs,
perhaps predicting new memory addresses using existing ones. More
sophisticated decisions about control flow would be made later on, after the
instruction was decoded, and handed back to fetch. If branching needs to
happen, some amount of decoding needs to happen to see that it's a branch,
what the target is, etc. This could get a little more complicated if that gets
done by the predecoder, but I'm choosing to ignore that for now.
Variable length instructions:
To handle variable length instructions in x86 and ARM, the predecoder now
takes in the current PC by reference to the getExtMachInst function. It can
modify the PC however it needs to (by setting NPC to be the PC + instruction
length, for instance). This could be improved since the CPU doesn't know if
the PC was modified and always has to write it back.
ISA parser:
To support the new API, all PC related operand types were removed from the
parser and replaced with a PCState type. There are two warts on this
implementation. First, as with all the other operand types, the PCState still
has to have a valid operand type even though it doesn't use it. Second, using
syntax like PCS.npc(target) doesn't work for two reasons, this looks like the
syntax for operand type overriding, and the parser can't figure out if you're
reading or writing. Instructions that use the PCS operand (which I've
consistently called it) need to first read it into a local variable,
manipulate it, and then write it back out.
Return address stack:
The return address stack needed a little extra help because, in the presence
of branch delay slots, it has to merge together elements of the return PC and
the call PC. To handle that, a buildRetPC utility function was added. There
are basically only two versions in all the ISAs, but it didn't seem short
enough to put into the generic ISA directory. Also, the branch predictor code
in O3 and InOrder were adjusted so that they always store the PC of the actual
call instruction in the RAS, not the next PC. If the call instruction is a
microop, the next PC refers to the next microop in the same macroop which is
probably not desirable. The buildRetPC function advances the PC intelligently
to the next macroop (in an ISA specific way) so that that case works.
Change in stats:
There were no change in stats except in MIPS and SPARC in the O3 model. MIPS
runs in about 9% fewer ticks. SPARC runs with 30%-50% fewer ticks, which could
likely be improved further by setting call/return instruction flags and taking
advantage of the RAS.
TODO:
Add != operators to the PCState classes, defined trivially to be !(a==b).
Smooth out places where PCs are split apart, passed around, and put back
together later. I think this might happen in SPARC's fault code. Add ISA
specific constructors that allow setting PC elements without calling a bunch
of accessors. Try to eliminate the need for the branching() function. Factor
out Alpha's PAL mode pc bit into a separate flag field, and eliminate places
where it's blindly masked out or tested in the PC.
2010-10-31 08:07:20 +01:00
|
|
|
PCState pc = tc->pcState();
|
|
|
|
pc.nextThumb(cpsr.t);
|
|
|
|
pc.nextJazelle(cpsr.j);
|
2012-01-31 16:46:03 +01:00
|
|
|
#if USE_CHECKER
|
|
|
|
tc->pcStateNoRecord(pc);
|
|
|
|
#else
|
ISA,CPU,etc: Create an ISA defined PC type that abstracts out ISA behaviors.
This change is a low level and pervasive reorganization of how PCs are managed
in M5. Back when Alpha was the only ISA, there were only 2 PCs to worry about,
the PC and the NPC, and the lsb of the PC signaled whether or not you were in
PAL mode. As other ISAs were added, we had to add an NNPC, micro PC and next
micropc, x86 and ARM introduced variable length instruction sets, and ARM
started to keep track of mode bits in the PC. Each CPU model handled PCs in
its own custom way that needed to be updated individually to handle the new
dimensions of variability, or, in the case of ARMs mode-bit-in-the-pc hack,
the complexity could be hidden in the ISA at the ISA implementation's expense.
Areas like the branch predictor hadn't been updated to handle branch delay
slots or micropcs, and it turns out that had introduced a significant (10s of
percent) performance bug in SPARC and to a lesser extend MIPS. Rather than
perpetuate the problem by reworking O3 again to handle the PC features needed
by x86, this change was introduced to rework PC handling in a more modular,
transparent, and hopefully efficient way.
PC type:
Rather than having the superset of all possible elements of PC state declared
in each of the CPU models, each ISA defines its own PCState type which has
exactly the elements it needs. A cross product of canned PCState classes are
defined in the new "generic" ISA directory for ISAs with/without delay slots
and microcode. These are either typedef-ed or subclassed by each ISA. To read
or write this structure through a *Context, you use the new pcState() accessor
which reads or writes depending on whether it has an argument. If you just
want the address of the current or next instruction or the current micro PC,
you can get those through read-only accessors on either the PCState type or
the *Contexts. These are instAddr(), nextInstAddr(), and microPC(). Note the
move away from readPC. That name is ambiguous since it's not clear whether or
not it should be the actual address to fetch from, or if it should have extra
bits in it like the PAL mode bit. Each class is free to define its own
functions to get at whatever values it needs however it needs to to be used in
ISA specific code. Eventually Alpha's PAL mode bit could be moved out of the
PC and into a separate field like ARM.
These types can be reset to a particular pc (where npc = pc +
sizeof(MachInst), nnpc = npc + sizeof(MachInst), upc = 0, nupc = 1 as
appropriate), printed, serialized, and compared. There is a branching()
function which encapsulates code in the CPU models that checked if an
instruction branched or not. Exactly what that means in the context of branch
delay slots which can skip an instruction when not taken is ambiguous, and
ideally this function and its uses can be eliminated. PCStates also generally
know how to advance themselves in various ways depending on if they point at
an instruction, a microop, or the last microop of a macroop. More on that
later.
Ideally, accessing all the PCs at once when setting them will improve
performance of M5 even though more data needs to be moved around. This is
because often all the PCs need to be manipulated together, and by getting them
all at once you avoid multiple function calls. Also, the PCs of a particular
thread will have spatial locality in the cache. Previously they were grouped
by element in arrays which spread out accesses.
Advancing the PC:
The PCs were previously managed entirely by the CPU which had to know about PC
semantics, try to figure out which dimension to increment the PC in, what to
set NPC/NNPC, etc. These decisions are best left to the ISA in conjunction
with the PC type itself. Because most of the information about how to
increment the PC (mainly what type of instruction it refers to) is contained
in the instruction object, a new advancePC virtual function was added to the
StaticInst class. Subclasses provide an implementation that moves around the
right element of the PC with a minimal amount of decision making. In ISAs like
Alpha, the instructions always simply assign NPC to PC without having to worry
about micropcs, nnpcs, etc. The added cost of a virtual function call should
be outweighed by not having to figure out as much about what to do with the
PCs and mucking around with the extra elements.
One drawback of making the StaticInsts advance the PC is that you have to
actually have one to advance the PC. This would, superficially, seem to
require decoding an instruction before fetch could advance. This is, as far as
I can tell, realistic. fetch would advance through memory addresses, not PCs,
perhaps predicting new memory addresses using existing ones. More
sophisticated decisions about control flow would be made later on, after the
instruction was decoded, and handed back to fetch. If branching needs to
happen, some amount of decoding needs to happen to see that it's a branch,
what the target is, etc. This could get a little more complicated if that gets
done by the predecoder, but I'm choosing to ignore that for now.
Variable length instructions:
To handle variable length instructions in x86 and ARM, the predecoder now
takes in the current PC by reference to the getExtMachInst function. It can
modify the PC however it needs to (by setting NPC to be the PC + instruction
length, for instance). This could be improved since the CPU doesn't know if
the PC was modified and always has to write it back.
ISA parser:
To support the new API, all PC related operand types were removed from the
parser and replaced with a PCState type. There are two warts on this
implementation. First, as with all the other operand types, the PCState still
has to have a valid operand type even though it doesn't use it. Second, using
syntax like PCS.npc(target) doesn't work for two reasons, this looks like the
syntax for operand type overriding, and the parser can't figure out if you're
reading or writing. Instructions that use the PCS operand (which I've
consistently called it) need to first read it into a local variable,
manipulate it, and then write it back out.
Return address stack:
The return address stack needed a little extra help because, in the presence
of branch delay slots, it has to merge together elements of the return PC and
the call PC. To handle that, a buildRetPC utility function was added. There
are basically only two versions in all the ISAs, but it didn't seem short
enough to put into the generic ISA directory. Also, the branch predictor code
in O3 and InOrder were adjusted so that they always store the PC of the actual
call instruction in the RAS, not the next PC. If the call instruction is a
microop, the next PC refers to the next microop in the same macroop which is
probably not desirable. The buildRetPC function advances the PC intelligently
to the next macroop (in an ISA specific way) so that that case works.
Change in stats:
There were no change in stats except in MIPS and SPARC in the O3 model. MIPS
runs in about 9% fewer ticks. SPARC runs with 30%-50% fewer ticks, which could
likely be improved further by setting call/return instruction flags and taking
advantage of the RAS.
TODO:
Add != operators to the PCState classes, defined trivially to be !(a==b).
Smooth out places where PCs are split apart, passed around, and put back
together later. I think this might happen in SPARC's fault code. Add ISA
specific constructors that allow setting PC elements without calling a bunch
of accessors. Try to eliminate the need for the branching() function. Factor
out Alpha's PAL mode pc bit into a separate flag field, and eliminate places
where it's blindly masked out or tested in the PC.
2010-10-31 08:07:20 +01:00
|
|
|
tc->pcState(pc);
|
2012-01-31 16:46:03 +01:00
|
|
|
#endif //USE_CHECKER
|
2010-06-02 19:58:16 +02:00
|
|
|
} else if (misc_reg >= MISCREG_CP15_UNIMP_START &&
|
2010-06-02 19:58:16 +02:00
|
|
|
misc_reg < MISCREG_CP15_END) {
|
|
|
|
panic("Unimplemented CP15 register %s wrote with %#x.\n",
|
|
|
|
miscRegName[misc_reg], val);
|
2010-06-02 19:58:16 +02:00
|
|
|
} else {
|
|
|
|
switch (misc_reg) {
|
|
|
|
case MISCREG_CPACR:
|
|
|
|
{
|
2011-04-04 18:42:28 +02:00
|
|
|
|
|
|
|
const uint32_t ones = (uint32_t)(-1);
|
|
|
|
CPACR cpacrMask = 0;
|
|
|
|
// Only cp10, cp11, and ase are implemented, nothing else should
|
|
|
|
// be writable
|
|
|
|
cpacrMask.cp10 = ones;
|
|
|
|
cpacrMask.cp11 = ones;
|
|
|
|
cpacrMask.asedis = ones;
|
|
|
|
newVal &= cpacrMask;
|
|
|
|
DPRINTF(MiscRegs, "Writing misc reg %s: %#x\n",
|
|
|
|
miscRegName[misc_reg], newVal);
|
2010-06-02 19:58:16 +02:00
|
|
|
}
|
|
|
|
break;
|
|
|
|
case MISCREG_CSSELR:
|
2010-11-08 20:58:24 +01:00
|
|
|
warn_once("The csselr register isn't implemented.\n");
|
2011-04-04 18:42:28 +02:00
|
|
|
return;
|
2010-06-02 19:58:16 +02:00
|
|
|
case MISCREG_FPSCR:
|
|
|
|
{
|
|
|
|
const uint32_t ones = (uint32_t)(-1);
|
|
|
|
FPSCR fpscrMask = 0;
|
|
|
|
fpscrMask.ioc = ones;
|
|
|
|
fpscrMask.dzc = ones;
|
|
|
|
fpscrMask.ofc = ones;
|
|
|
|
fpscrMask.ufc = ones;
|
|
|
|
fpscrMask.ixc = ones;
|
|
|
|
fpscrMask.idc = ones;
|
|
|
|
fpscrMask.len = ones;
|
|
|
|
fpscrMask.stride = ones;
|
|
|
|
fpscrMask.rMode = ones;
|
|
|
|
fpscrMask.fz = ones;
|
|
|
|
fpscrMask.dn = ones;
|
|
|
|
fpscrMask.ahp = ones;
|
|
|
|
fpscrMask.qc = ones;
|
|
|
|
fpscrMask.v = ones;
|
|
|
|
fpscrMask.c = ones;
|
|
|
|
fpscrMask.z = ones;
|
|
|
|
fpscrMask.n = ones;
|
|
|
|
newVal = (newVal & (uint32_t)fpscrMask) |
|
|
|
|
(miscRegs[MISCREG_FPSCR] & ~(uint32_t)fpscrMask);
|
|
|
|
}
|
|
|
|
break;
|
2011-05-14 00:27:01 +02:00
|
|
|
case MISCREG_CPSR_Q:
|
|
|
|
{
|
|
|
|
assert(!(newVal & ~CpsrMaskQ));
|
|
|
|
newVal = miscRegs[MISCREG_CPSR] | newVal;
|
|
|
|
misc_reg = MISCREG_CPSR;
|
|
|
|
}
|
|
|
|
break;
|
2010-12-08 01:19:57 +01:00
|
|
|
case MISCREG_FPSCR_QC:
|
|
|
|
{
|
|
|
|
newVal = miscRegs[MISCREG_FPSCR] | (newVal & FpscrQcMask);
|
|
|
|
misc_reg = MISCREG_FPSCR;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
case MISCREG_FPSCR_EXC:
|
|
|
|
{
|
|
|
|
newVal = miscRegs[MISCREG_FPSCR] | (newVal & FpscrExcMask);
|
|
|
|
misc_reg = MISCREG_FPSCR;
|
|
|
|
}
|
|
|
|
break;
|
2010-06-02 19:58:16 +02:00
|
|
|
case MISCREG_FPEXC:
|
|
|
|
{
|
2011-04-04 18:42:28 +02:00
|
|
|
// vfpv3 architecture, section B.6.1 of DDI04068
|
|
|
|
// bit 29 - valid only if fpexc[31] is 0
|
2010-06-02 19:58:16 +02:00
|
|
|
const uint32_t fpexcMask = 0x60000000;
|
|
|
|
newVal = (newVal & fpexcMask) |
|
|
|
|
(miscRegs[MISCREG_FPEXC] & ~fpexcMask);
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
case MISCREG_SCTLR:
|
|
|
|
{
|
|
|
|
DPRINTF(MiscRegs, "Writing SCTLR: %#x\n", newVal);
|
|
|
|
SCTLR sctlr = miscRegs[MISCREG_SCTLR];
|
|
|
|
SCTLR new_sctlr = newVal;
|
|
|
|
new_sctlr.nmfi = (bool)sctlr.nmfi;
|
|
|
|
miscRegs[MISCREG_SCTLR] = (MiscReg)new_sctlr;
|
2010-11-15 21:04:03 +01:00
|
|
|
tc->getITBPtr()->invalidateMiscReg();
|
|
|
|
tc->getDTBPtr()->invalidateMiscReg();
|
2011-08-19 22:08:08 +02:00
|
|
|
|
|
|
|
// Check if all CPUs are booted with caches enabled
|
|
|
|
// so we can stop enforcing coherency of some kernel
|
|
|
|
// structures manually.
|
|
|
|
sys = tc->getSystemPtr();
|
|
|
|
for (x = 0; x < sys->numContexts(); x++) {
|
|
|
|
oc = sys->getThreadContext(x);
|
|
|
|
SCTLR other_sctlr = oc->readMiscRegNoEffect(MISCREG_SCTLR);
|
|
|
|
if (!other_sctlr.c && oc->status() != ThreadContext::Halted)
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
|
|
|
|
for (x = 0; x < sys->numContexts(); x++) {
|
|
|
|
oc = sys->getThreadContext(x);
|
|
|
|
oc->getDTBPtr()->allCpusCaching();
|
|
|
|
oc->getITBPtr()->allCpusCaching();
|
2012-01-31 16:46:03 +01:00
|
|
|
#if USE_CHECKER
|
|
|
|
CheckerCPU *checker =
|
|
|
|
dynamic_cast<CheckerCPU*>(oc->getCheckerCpuPtr());
|
|
|
|
if (checker) {
|
|
|
|
checker->getDTBPtr()->allCpusCaching();
|
|
|
|
checker->getITBPtr()->allCpusCaching();
|
|
|
|
}
|
|
|
|
#endif
|
2011-08-19 22:08:08 +02:00
|
|
|
}
|
2010-06-02 19:58:16 +02:00
|
|
|
return;
|
|
|
|
}
|
|
|
|
case MISCREG_TLBTR:
|
|
|
|
case MISCREG_MVFR0:
|
|
|
|
case MISCREG_MVFR1:
|
|
|
|
case MISCREG_MPIDR:
|
|
|
|
case MISCREG_FPSID:
|
|
|
|
return;
|
|
|
|
case MISCREG_TLBIALLIS:
|
|
|
|
case MISCREG_TLBIALL:
|
2011-05-05 03:38:28 +02:00
|
|
|
sys = tc->getSystemPtr();
|
|
|
|
for (x = 0; x < sys->numContexts(); x++) {
|
|
|
|
oc = sys->getThreadContext(x);
|
|
|
|
assert(oc->getITBPtr() && oc->getDTBPtr());
|
|
|
|
oc->getITBPtr()->flushAll();
|
|
|
|
oc->getDTBPtr()->flushAll();
|
2012-01-31 16:46:03 +01:00
|
|
|
#if USE_CHECKER
|
|
|
|
CheckerCPU *checker =
|
|
|
|
dynamic_cast<CheckerCPU*>(oc->getCheckerCpuPtr());
|
|
|
|
if (checker) {
|
|
|
|
checker->getITBPtr()->flushAll();
|
|
|
|
checker->getDTBPtr()->flushAll();
|
|
|
|
}
|
|
|
|
#endif
|
2011-05-05 03:38:28 +02:00
|
|
|
}
|
2010-06-02 19:58:16 +02:00
|
|
|
return;
|
|
|
|
case MISCREG_ITLBIALL:
|
|
|
|
tc->getITBPtr()->flushAll();
|
|
|
|
return;
|
|
|
|
case MISCREG_DTLBIALL:
|
|
|
|
tc->getDTBPtr()->flushAll();
|
|
|
|
return;
|
|
|
|
case MISCREG_TLBIMVAIS:
|
|
|
|
case MISCREG_TLBIMVA:
|
2011-05-05 03:38:28 +02:00
|
|
|
sys = tc->getSystemPtr();
|
|
|
|
for (x = 0; x < sys->numContexts(); x++) {
|
|
|
|
oc = sys->getThreadContext(x);
|
|
|
|
assert(oc->getITBPtr() && oc->getDTBPtr());
|
|
|
|
oc->getITBPtr()->flushMvaAsid(mbits(newVal, 31, 12),
|
|
|
|
bits(newVal, 7,0));
|
|
|
|
oc->getDTBPtr()->flushMvaAsid(mbits(newVal, 31, 12),
|
|
|
|
bits(newVal, 7,0));
|
2012-01-31 16:46:03 +01:00
|
|
|
#if USE_CHECKER
|
|
|
|
CheckerCPU *checker =
|
|
|
|
dynamic_cast<CheckerCPU*>(oc->getCheckerCpuPtr());
|
|
|
|
if (checker) {
|
|
|
|
checker->getITBPtr()->flushMvaAsid(mbits(newVal, 31, 12),
|
|
|
|
bits(newVal, 7,0));
|
|
|
|
checker->getDTBPtr()->flushMvaAsid(mbits(newVal, 31, 12),
|
|
|
|
bits(newVal, 7,0));
|
|
|
|
}
|
|
|
|
#endif
|
2011-05-05 03:38:28 +02:00
|
|
|
}
|
2010-06-02 19:58:16 +02:00
|
|
|
return;
|
|
|
|
case MISCREG_TLBIASIDIS:
|
|
|
|
case MISCREG_TLBIASID:
|
2011-05-05 03:38:28 +02:00
|
|
|
sys = tc->getSystemPtr();
|
|
|
|
for (x = 0; x < sys->numContexts(); x++) {
|
|
|
|
oc = sys->getThreadContext(x);
|
|
|
|
assert(oc->getITBPtr() && oc->getDTBPtr());
|
|
|
|
oc->getITBPtr()->flushAsid(bits(newVal, 7,0));
|
|
|
|
oc->getDTBPtr()->flushAsid(bits(newVal, 7,0));
|
2012-01-31 16:46:03 +01:00
|
|
|
#if USE_CHECKER
|
|
|
|
CheckerCPU *checker =
|
|
|
|
dynamic_cast<CheckerCPU*>(oc->getCheckerCpuPtr());
|
|
|
|
if (checker) {
|
|
|
|
checker->getITBPtr()->flushAsid(bits(newVal, 7,0));
|
|
|
|
checker->getDTBPtr()->flushAsid(bits(newVal, 7,0));
|
|
|
|
}
|
|
|
|
#endif
|
2011-05-05 03:38:28 +02:00
|
|
|
}
|
2010-06-02 19:58:16 +02:00
|
|
|
return;
|
|
|
|
case MISCREG_TLBIMVAAIS:
|
|
|
|
case MISCREG_TLBIMVAA:
|
2011-05-05 03:38:28 +02:00
|
|
|
sys = tc->getSystemPtr();
|
|
|
|
for (x = 0; x < sys->numContexts(); x++) {
|
|
|
|
oc = sys->getThreadContext(x);
|
|
|
|
assert(oc->getITBPtr() && oc->getDTBPtr());
|
|
|
|
oc->getITBPtr()->flushMva(mbits(newVal, 31,12));
|
|
|
|
oc->getDTBPtr()->flushMva(mbits(newVal, 31,12));
|
2012-01-31 16:46:03 +01:00
|
|
|
#if USE_CHECKER
|
|
|
|
CheckerCPU *checker =
|
|
|
|
dynamic_cast<CheckerCPU*>(oc->getCheckerCpuPtr());
|
|
|
|
if (checker) {
|
|
|
|
checker->getITBPtr()->flushMva(mbits(newVal, 31,12));
|
|
|
|
checker->getDTBPtr()->flushMva(mbits(newVal, 31,12));
|
|
|
|
}
|
|
|
|
#endif
|
2011-05-05 03:38:28 +02:00
|
|
|
}
|
2010-06-02 19:58:16 +02:00
|
|
|
return;
|
|
|
|
case MISCREG_ITLBIMVA:
|
|
|
|
tc->getITBPtr()->flushMvaAsid(mbits(newVal, 31, 12),
|
|
|
|
bits(newVal, 7,0));
|
|
|
|
return;
|
|
|
|
case MISCREG_DTLBIMVA:
|
|
|
|
tc->getDTBPtr()->flushMvaAsid(mbits(newVal, 31, 12),
|
|
|
|
bits(newVal, 7,0));
|
|
|
|
return;
|
|
|
|
case MISCREG_ITLBIASID:
|
|
|
|
tc->getITBPtr()->flushAsid(bits(newVal, 7,0));
|
|
|
|
return;
|
|
|
|
case MISCREG_DTLBIASID:
|
|
|
|
tc->getDTBPtr()->flushAsid(bits(newVal, 7,0));
|
2010-06-02 19:58:16 +02:00
|
|
|
return;
|
2010-08-23 18:18:40 +02:00
|
|
|
case MISCREG_ACTLR:
|
|
|
|
warn("Not doing anything for write of miscreg ACTLR\n");
|
|
|
|
break;
|
|
|
|
case MISCREG_PMCR:
|
2011-02-23 22:10:48 +01:00
|
|
|
{
|
|
|
|
// Performance counters not implemented. Instead, interpret
|
|
|
|
// a reset command to this register to reset the simulator
|
|
|
|
// statistics.
|
|
|
|
// PMCR_E | PMCR_P | PMCR_C
|
|
|
|
const int ResetAndEnableCounters = 0x7;
|
|
|
|
if (newVal == ResetAndEnableCounters) {
|
|
|
|
inform("Resetting all simobject stats\n");
|
|
|
|
Stats::schedStatEvent(false, true);
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
2010-08-23 18:18:40 +02:00
|
|
|
case MISCREG_PMCCNTR:
|
|
|
|
case MISCREG_PMSELR:
|
|
|
|
warn("Not doing anything for write to miscreg %s\n",
|
|
|
|
miscRegName[misc_reg]);
|
|
|
|
break;
|
2010-06-02 19:58:18 +02:00
|
|
|
case MISCREG_V2PCWPR:
|
|
|
|
case MISCREG_V2PCWPW:
|
|
|
|
case MISCREG_V2PCWUR:
|
|
|
|
case MISCREG_V2PCWUW:
|
|
|
|
case MISCREG_V2POWPR:
|
|
|
|
case MISCREG_V2POWPW:
|
|
|
|
case MISCREG_V2POWUR:
|
|
|
|
case MISCREG_V2POWUW:
|
|
|
|
{
|
|
|
|
RequestPtr req = new Request;
|
|
|
|
unsigned flags;
|
|
|
|
BaseTLB::Mode mode;
|
|
|
|
Fault fault;
|
|
|
|
switch(misc_reg) {
|
|
|
|
case MISCREG_V2PCWPR:
|
|
|
|
flags = TLB::MustBeOne;
|
|
|
|
mode = BaseTLB::Read;
|
|
|
|
break;
|
|
|
|
case MISCREG_V2PCWPW:
|
|
|
|
flags = TLB::MustBeOne;
|
|
|
|
mode = BaseTLB::Write;
|
|
|
|
break;
|
|
|
|
case MISCREG_V2PCWUR:
|
|
|
|
flags = TLB::MustBeOne | TLB::UserMode;
|
|
|
|
mode = BaseTLB::Read;
|
|
|
|
break;
|
|
|
|
case MISCREG_V2PCWUW:
|
|
|
|
flags = TLB::MustBeOne | TLB::UserMode;
|
|
|
|
mode = BaseTLB::Write;
|
|
|
|
break;
|
2010-06-03 18:20:49 +02:00
|
|
|
default:
|
2010-06-02 19:58:18 +02:00
|
|
|
panic("Security Extensions not implemented!");
|
|
|
|
}
|
2011-04-04 18:42:28 +02:00
|
|
|
warn("Translating via MISCREG in atomic mode! Fix Me!\n");
|
2012-02-12 23:07:38 +01:00
|
|
|
req->setVirt(0, val, 1, flags, tc->pcState().pc(),
|
|
|
|
Request::funcMasterId);
|
2010-06-02 19:58:18 +02:00
|
|
|
fault = tc->getDTBPtr()->translateAtomic(req, tc, mode);
|
|
|
|
if (fault == NoFault) {
|
|
|
|
miscRegs[MISCREG_PAR] =
|
|
|
|
(req->getPaddr() & 0xfffff000) |
|
|
|
|
(tc->getDTBPtr()->getAttr() );
|
|
|
|
DPRINTF(MiscRegs,
|
|
|
|
"MISCREG: Translated addr 0x%08x: PAR: 0x%08x\n",
|
|
|
|
val, miscRegs[MISCREG_PAR]);
|
|
|
|
}
|
|
|
|
else {
|
|
|
|
// Set fault bit and FSR
|
|
|
|
FSR fsr = miscRegs[MISCREG_DFSR];
|
|
|
|
miscRegs[MISCREG_PAR] =
|
|
|
|
(fsr.ext << 6) |
|
|
|
|
(fsr.fsHigh << 5) |
|
|
|
|
(fsr.fsLow << 1) |
|
|
|
|
0x1; // F bit
|
|
|
|
}
|
|
|
|
return;
|
|
|
|
}
|
2010-11-15 21:04:03 +01:00
|
|
|
case MISCREG_CONTEXTIDR:
|
|
|
|
case MISCREG_PRRR:
|
|
|
|
case MISCREG_NMRR:
|
|
|
|
case MISCREG_DACR:
|
|
|
|
tc->getITBPtr()->invalidateMiscReg();
|
|
|
|
tc->getDTBPtr()->invalidateMiscReg();
|
|
|
|
break;
|
2011-04-04 18:42:28 +02:00
|
|
|
case MISCREG_CPSR_MODE:
|
|
|
|
// This miscreg is used by copy*Regs to set the CPSR mode
|
|
|
|
// without updating other CPSR variables. It's used to
|
|
|
|
// make sure the register map is in such a state that we can
|
|
|
|
// see all of the registers for the copy.
|
|
|
|
updateRegMap(val);
|
|
|
|
return;
|
2011-09-13 19:06:13 +02:00
|
|
|
case MISCREG_L2CTLR:
|
|
|
|
warn("miscreg L2CTLR (%s) written with %#x. ignored...\n",
|
|
|
|
miscRegName[misc_reg], uint32_t(val));
|
2010-06-02 19:58:16 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
setMiscRegNoEffect(misc_reg, newVal);
|
|
|
|
}
|
|
|
|
|
|
|
|
}
|