gem5/configs/common/FSConfig.py

707 lines
26 KiB
Python
Raw Normal View History

# Copyright (c) 2010-2012, 2015 ARM Limited
# 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.
#
2011-02-07 07:14:18 +01:00
# Copyright (c) 2010-2011 Advanced Micro Devices, Inc.
# Copyright (c) 2006-2008 The Regents of The University of Michigan
# All rights reserved.
#
# 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: Kevin Lim
from m5.objects import *
from Benchmarks import *
from m5.util import *
import PlatformConfig
# Populate to reflect supported os types per target ISA
os_types = { 'alpha' : [ 'linux' ],
'mips' : [ 'linux' ],
'sparc' : [ 'linux' ],
'x86' : [ 'linux' ],
'arm' : [ 'linux',
'android-gingerbread',
'android-ics',
'android-jellybean',
'android-kitkat' ],
}
class CowIdeDisk(IdeDisk):
image = CowDiskImage(child=RawDiskImage(read_only=True),
read_only=False)
def childImage(self, ci):
self.image.child.image_file = ci
class MemBus(SystemXBar):
badaddr_responder = BadAddr()
default = Self.badaddr_responder.pio
def fillInCmdline(mdesc, template, **kwargs):
kwargs.setdefault('disk', mdesc.disk())
kwargs.setdefault('rootdev', mdesc.rootdev())
kwargs.setdefault('mem', mdesc.mem())
kwargs.setdefault('script', mdesc.script())
return template % kwargs
def makeLinuxAlphaSystem(mem_mode, mdesc=None, ruby=False, cmdline=None):
class BaseTsunami(Tsunami):
ethernet = NSGigE(pci_bus=0, pci_dev=1, pci_func=0)
ide = IdeController(disks=[Parent.disk0, Parent.disk2],
pci_func=0, pci_dev=0, pci_bus=0)
self = LinuxAlphaSystem()
Finish test clean-up & reorg. configs/common/FSConfig.py: Add default Machine() param configs/example/fs.py: configs/example/se.py: make it work again src/python/m5/objects/BaseCPU.py: Make mem PhysicalMemory so that a Parent.any proxy works well src/sim/process.cc: Increase default stack size so we don't get an 'increasing stack' message on 'hello world' tests/SConscript: Add full list of current configs. tests/configs/simple-atomic.py: tests/configs/simple-timing.py: don't need SEConfig anymore tests/quick/00.hello/test.py: tests/quick/20.eio-short/test.py: fix tests/run.py: move configs to separate dir --HG-- rename : configs/test/fs.py => configs/example/fs.py rename : configs/test/test.py => configs/example/se.py rename : tests/simple-atomic.py => tests/configs/simple-atomic.py rename : tests/simple-timing.py => tests/configs/simple-timing.py rename : tests/linux-mpboot/ref/alpha/atomic/config.ini => tests/quick/10.linux-boot/ref/alpha/linux/tsunami-simple-atomic-dual/config.ini rename : tests/linux-mpboot/ref/alpha/atomic/config.out => tests/quick/10.linux-boot/ref/alpha/linux/tsunami-simple-atomic-dual/config.out rename : tests/linux-mpboot/ref/alpha/atomic/console.system.sim_console => tests/quick/10.linux-boot/ref/alpha/linux/tsunami-simple-atomic-dual/console.system.sim_console rename : tests/linux-mpboot/ref/alpha/atomic/m5stats.txt => tests/quick/10.linux-boot/ref/alpha/linux/tsunami-simple-atomic-dual/m5stats.txt rename : tests/linux-mpboot/ref/alpha/atomic/stderr => tests/quick/10.linux-boot/ref/alpha/linux/tsunami-simple-atomic-dual/stderr rename : tests/linux-mpboot/ref/alpha/atomic/stdout => tests/quick/10.linux-boot/ref/alpha/linux/tsunami-simple-atomic-dual/stdout rename : tests/linux-boot/ref/alpha/atomic/config.ini => tests/quick/10.linux-boot/ref/alpha/linux/tsunami-simple-atomic/config.ini rename : tests/linux-boot/ref/alpha/atomic/config.out => tests/quick/10.linux-boot/ref/alpha/linux/tsunami-simple-atomic/config.out rename : tests/linux-boot/ref/alpha/atomic/console.system.sim_console => tests/quick/10.linux-boot/ref/alpha/linux/tsunami-simple-atomic/console.system.sim_console rename : tests/linux-boot/ref/alpha/atomic/m5stats.txt => tests/quick/10.linux-boot/ref/alpha/linux/tsunami-simple-atomic/m5stats.txt rename : tests/linux-boot/ref/alpha/atomic/stderr => tests/quick/10.linux-boot/ref/alpha/linux/tsunami-simple-atomic/stderr rename : tests/linux-boot/ref/alpha/atomic/stdout => tests/quick/10.linux-boot/ref/alpha/linux/tsunami-simple-atomic/stdout rename : tests/linux-mpboot/ref/alpha/timing/config.ini => tests/quick/10.linux-boot/ref/alpha/linux/tsunami-simple-timing-dual/config.ini rename : tests/linux-mpboot/ref/alpha/timing/config.out => tests/quick/10.linux-boot/ref/alpha/linux/tsunami-simple-timing-dual/config.out rename : tests/linux-mpboot/ref/alpha/timing/console.system.sim_console => tests/quick/10.linux-boot/ref/alpha/linux/tsunami-simple-timing-dual/console.system.sim_console rename : tests/linux-mpboot/ref/alpha/timing/m5stats.txt => tests/quick/10.linux-boot/ref/alpha/linux/tsunami-simple-timing-dual/m5stats.txt rename : tests/linux-mpboot/ref/alpha/timing/stderr => tests/quick/10.linux-boot/ref/alpha/linux/tsunami-simple-timing-dual/stderr rename : tests/linux-mpboot/ref/alpha/timing/stdout => tests/quick/10.linux-boot/ref/alpha/linux/tsunami-simple-timing-dual/stdout rename : tests/test-progs/hello/bin/mips/linux/hello_mips => tests/test-progs/hello/bin/mips/linux/hello rename : tests/test-progs/hello/bin/sparc/bin => tests/test-progs/hello/bin/sparc/linux/hello extra : convert_revision : d68ee6d7eefa7ba57370f3fb3c3589f86a6ea6b4
2006-08-16 20:42:44 +02:00
if not mdesc:
# generic system
mdesc = SysConfig()
self.readfile = mdesc.script()
self.tsunami = BaseTsunami()
# Create the io bus to connect all device ports
self.iobus = IOXBar()
self.tsunami.attachIO(self.iobus)
self.tsunami.ide.pio = self.iobus.master
self.tsunami.ethernet.pio = self.iobus.master
if ruby:
# Store the dma devices for later connection to dma ruby ports.
# Append an underscore to dma_ports to avoid the SimObjectVector check.
self._dma_ports = [self.tsunami.ide.dma, self.tsunami.ethernet.dma]
else:
self.membus = MemBus()
# By default the bridge responds to all addresses above the I/O
# base address (including the PCI config space)
IO_address_space_base = 0x80000000000
self.bridge = Bridge(delay='50ns',
ranges = [AddrRange(IO_address_space_base, Addr.max)])
self.bridge.master = self.iobus.slave
self.bridge.slave = self.membus.master
self.tsunami.ide.dma = self.iobus.slave
self.tsunami.ethernet.dma = self.iobus.slave
self.system_port = self.membus.slave
self.mem_ranges = [AddrRange(mdesc.mem())]
self.disk0 = CowIdeDisk(driveID='master')
self.disk2 = CowIdeDisk(driveID='master')
self.disk0.childImage(mdesc.disk())
self.disk2.childImage(disk('linux-bigswap2.img'))
self.simple_disk = SimpleDisk(disk=RawDiskImage(image_file = mdesc.disk(),
read_only = True))
self.intrctrl = IntrControl()
self.mem_mode = mem_mode
self.terminal = Terminal()
self.kernel = binary('vmlinux')
self.pal = binary('ts_osfpal')
self.console = binary('console')
if not cmdline:
cmdline = 'root=/dev/hda1 console=ttyS0'
self.boot_osflags = fillInCmdline(mdesc, cmdline)
return self
def makeSparcSystem(mem_mode, mdesc=None, cmdline=None):
# Constants from iob.cc and uart8250.cc
iob_man_addr = 0x9800000000
uart_pio_size = 8
Move SimObject python files alongside the C++ and fix the SConscript files so that only the objects that are actually available in a given build are compiled in. Remove a bunch of files that aren't used anymore. --HG-- rename : src/python/m5/objects/AlphaTLB.py => src/arch/alpha/AlphaTLB.py rename : src/python/m5/objects/SparcTLB.py => src/arch/sparc/SparcTLB.py rename : src/python/m5/objects/BaseCPU.py => src/cpu/BaseCPU.py rename : src/python/m5/objects/FuncUnit.py => src/cpu/FuncUnit.py rename : src/python/m5/objects/IntrControl.py => src/cpu/IntrControl.py rename : src/python/m5/objects/MemTest.py => src/cpu/memtest/MemTest.py rename : src/python/m5/objects/FUPool.py => src/cpu/o3/FUPool.py rename : src/python/m5/objects/FuncUnitConfig.py => src/cpu/o3/FuncUnitConfig.py rename : src/python/m5/objects/O3CPU.py => src/cpu/o3/O3CPU.py rename : src/python/m5/objects/OzoneCPU.py => src/cpu/ozone/OzoneCPU.py rename : src/python/m5/objects/SimpleOzoneCPU.py => src/cpu/ozone/SimpleOzoneCPU.py rename : src/python/m5/objects/BadDevice.py => src/dev/BadDevice.py rename : src/python/m5/objects/Device.py => src/dev/Device.py rename : src/python/m5/objects/DiskImage.py => src/dev/DiskImage.py rename : src/python/m5/objects/Ethernet.py => src/dev/Ethernet.py rename : src/python/m5/objects/Ide.py => src/dev/Ide.py rename : src/python/m5/objects/Pci.py => src/dev/Pci.py rename : src/python/m5/objects/Platform.py => src/dev/Platform.py rename : src/python/m5/objects/SimConsole.py => src/dev/SimConsole.py rename : src/python/m5/objects/SimpleDisk.py => src/dev/SimpleDisk.py rename : src/python/m5/objects/Uart.py => src/dev/Uart.py rename : src/python/m5/objects/AlphaConsole.py => src/dev/alpha/AlphaConsole.py rename : src/python/m5/objects/Tsunami.py => src/dev/alpha/Tsunami.py rename : src/python/m5/objects/T1000.py => src/dev/sparc/T1000.py rename : src/python/m5/objects/Bridge.py => src/mem/Bridge.py rename : src/python/m5/objects/Bus.py => src/mem/Bus.py rename : src/python/m5/objects/MemObject.py => src/mem/MemObject.py rename : src/python/m5/objects/PhysicalMemory.py => src/mem/PhysicalMemory.py rename : src/python/m5/objects/BaseCache.py => src/mem/cache/BaseCache.py rename : src/python/m5/objects/CoherenceProtocol.py => src/mem/cache/coherence/CoherenceProtocol.py rename : src/python/m5/objects/Repl.py => src/mem/cache/tags/Repl.py rename : src/python/m5/objects/Process.py => src/sim/Process.py rename : src/python/m5/objects/Root.py => src/sim/Root.py rename : src/python/m5/objects/System.py => src/sim/System.py extra : convert_revision : 173f8764bafa8ef899198438fa5573874e407321
2007-05-28 04:21:17 +02:00
class CowMmDisk(MmDisk):
image = CowDiskImage(child=RawDiskImage(read_only=True),
read_only=False)
def childImage(self, ci):
self.image.child.image_file = ci
self = SparcSystem()
if not mdesc:
# generic system
mdesc = SysConfig()
self.readfile = mdesc.script()
self.iobus = IOXBar()
Bus: Split the bus into a non-coherent and coherent bus This patch introduces a class hierarchy of buses, a non-coherent one, and a coherent one, splitting the existing bus functionality. By doing so it also enables further specialisation of the two types of buses. A non-coherent bus connects a number of non-snooping masters and slaves, and routes the request and response packets based on the address. The request packets issued by the master connected to a non-coherent bus could still snoop in caches attached to a coherent bus, as is the case with the I/O bus and memory bus in most system configurations. No snoops will, however, reach any master on the non-coherent bus itself. The non-coherent bus can be used as a template for modelling PCI, PCIe, and non-coherent AMBA and OCP buses, and is typically used for the I/O buses. A coherent bus connects a number of (potentially) snooping masters and slaves, and routes the request and response packets based on the address, and also forwards all requests to the snoopers and deals with the snoop responses. The coherent bus can be used as a template for modelling QPI, HyperTransport, ACE and coherent OCP buses, and is typically used for the L1-to-L2 buses and as the main system interconnect. The configuration scripts are updated to use a NoncoherentBus for all peripheral and I/O buses. A bit of minor tidying up has also been done. --HG-- rename : src/mem/bus.cc => src/mem/coherent_bus.cc rename : src/mem/bus.hh => src/mem/coherent_bus.hh rename : src/mem/bus.cc => src/mem/noncoherent_bus.cc rename : src/mem/bus.hh => src/mem/noncoherent_bus.hh
2012-05-31 19:30:04 +02:00
self.membus = MemBus()
Bridge: Remove NACKs in the bridge and unify with packet queue This patch removes the NACKing in the bridge, as the split request/response busses now ensure that protocol deadlocks do not occur, i.e. the message-dependency chain is broken by always allowing responses to make progress without being stalled by requests. The NACKs had limited support in the system with most components ignoring their use (with a suitable call to panic), and as the NACKs are no longer needed to avoid protocol deadlocks, the cleanest way is to simply remove them. The bridge is the starting point as this is the only place where the NACKs are created. A follow-up patch will remove the code that deals with NACKs in the endpoints, e.g. the X86 table walker and DMA port. Ultimately the type of packet can be complete removed (until someone sees a need for modelling more complex protocols, which can now be done in parts of the system since the port and interface is split). As a consequence of the NACK removal, the bridge now has to send a retry to a master if the request or response queue was full on the first attempt. This change also makes the bridge ports very similar to QueuedPorts, and a later patch will change the bridge to use these. A first step in this direction is taken by aligning the name of the member functions, as done by this patch. A bit of tidying up has also been done as part of the simplifications. Surprisingly, this patch has no impact on any of the regressions. Hence, there was never any NACKs issued. In a follow-up patch I would suggest changing the size of the bridge buffers set in FSConfig.py to also test the situation where the bridge fills up.
2012-08-22 17:39:58 +02:00
self.bridge = Bridge(delay='50ns')
self.t1000 = T1000()
self.t1000.attachOnChipIO(self.membus)
self.t1000.attachIO(self.iobus)
self.mem_ranges = [AddrRange(Addr('1MB'), size = '64MB'),
AddrRange(Addr('2GB'), size ='256MB')]
self.bridge.master = self.iobus.slave
self.bridge.slave = self.membus.master
self.rom.port = self.membus.master
self.nvram.port = self.membus.master
self.hypervisor_desc.port = self.membus.master
self.partition_desc.port = self.membus.master
self.intrctrl = IntrControl()
self.disk0 = CowMmDisk()
self.disk0.childImage(disk('disk.s10hw2'))
self.disk0.pio = self.iobus.master
# The puart0 and hvuart are placed on the IO bus, so create ranges
# for them. The remaining IO range is rather fragmented, so poke
# holes for the iob and partition descriptors etc.
self.bridge.ranges = \
[
AddrRange(self.t1000.puart0.pio_addr,
self.t1000.puart0.pio_addr + uart_pio_size - 1),
AddrRange(self.disk0.pio_addr,
self.t1000.fake_jbi.pio_addr +
self.t1000.fake_jbi.pio_size - 1),
AddrRange(self.t1000.fake_clk.pio_addr,
iob_man_addr - 1),
AddrRange(self.t1000.fake_l2_1.pio_addr,
self.t1000.fake_ssi.pio_addr +
self.t1000.fake_ssi.pio_size - 1),
AddrRange(self.t1000.hvuart.pio_addr,
self.t1000.hvuart.pio_addr + uart_pio_size - 1)
]
Implement Niagara I/O interface and rework interrupts configs/common/FSConfig.py: Use binaries we've compiled instead of the ones that come with Legion src/arch/alpha/interrupts.hh: get rid of post(int int_type) and add a get_vec function that gets the interrupt vector for an interrupt number src/arch/sparc/asi.cc: Add AsiIsInterrupt() to AsiIsMmu() src/arch/sparc/faults.cc: src/arch/sparc/faults.hh: Add InterruptVector type src/arch/sparc/interrupts.hh: rework interrupts. They are no longer cleared when created... A I/O or ASI read/write needs to happen before they are cleared src/arch/sparc/isa_traits.hh: Add the "interrupt" trap types to isa traits src/arch/sparc/miscregfile.cc: add names for all the misc registers and possible post an interrupt when TL is changed. src/arch/sparc/miscregfile.hh: Add a helper function to post an interrupt when pil < some set softint src/arch/sparc/regfile.cc: src/arch/sparc/regfile.hh: InterruptLevel shouldn't really live here, moved to interrupt.hh src/arch/sparc/tlb.cc: Add interrupt ASIs to TLB src/arch/sparc/ua2005.cc: Add checkSoftInt to check if a softint needs to be posted Check that a tickCompare isn't scheduled before scheduling one Post and clear interrupts on queue writes and what not src/base/bitfield.hh: Add an helper function to return the msb that is set src/cpu/base.cc: src/cpu/base.hh: get rid of post_interrupt(type) since it's no longer needed.. Add a way to see what interrupts are pending src/cpu/intr_control.cc: src/cpu/intr_control.hh: src/dev/alpha/tsunami_cchip.cc: src/python/m5/objects/IntrControl.py: Make IntrControl have a system pointer rather than using a cpu pointer to get one src/dev/sparc/SConscript: add iob to SConsscrip tests/quick/10.linux-boot/ref/alpha/linux/tsunami-simple-atomic-dual/config.ini: tests/quick/10.linux-boot/ref/alpha/linux/tsunami-simple-atomic-dual/config.out: tests/quick/10.linux-boot/ref/alpha/linux/tsunami-simple-atomic/config.ini: tests/quick/10.linux-boot/ref/alpha/linux/tsunami-simple-atomic/config.out: tests/quick/10.linux-boot/ref/alpha/linux/tsunami-simple-timing-dual/config.ini: tests/quick/10.linux-boot/ref/alpha/linux/tsunami-simple-timing-dual/config.out: tests/quick/10.linux-boot/ref/alpha/linux/tsunami-simple-timing/config.ini: tests/quick/10.linux-boot/ref/alpha/linux/tsunami-simple-timing/config.out: tests/quick/80.netperf-stream/ref/alpha/linux/twosys-tsunami-simple-atomic/config.ini: tests/quick/80.netperf-stream/ref/alpha/linux/twosys-tsunami-simple-atomic/config.out: update config.ini/out for intrcntrl not having a cpu pointer anymore --HG-- extra : convert_revision : 38614f6b9ffc8f3c93949a94ff04b7d2987168dd
2007-03-03 23:22:47 +01:00
self.reset_bin = binary('reset_new.bin')
self.hypervisor_bin = binary('q_new.bin')
self.openboot_bin = binary('openboot_new.bin')
self.nvram_bin = binary('nvram1')
self.hypervisor_desc_bin = binary('1up-hv.bin')
self.partition_desc_bin = binary('1up-md.bin')
self.system_port = self.membus.slave
return self
def makeArmSystem(mem_mode, machine_type, num_cpus=1, mdesc=None,
dtb_filename=None, bare_metal=False, cmdline=None,
external_memory=""):
assert machine_type
default_dtbs = {
"RealViewEB": None,
"RealViewPBX": None,
"VExpress_EMM": "vexpress.aarch32.ll_20131205.0-gem5.%dcpu.dtb" % num_cpus,
"VExpress_EMM64": "vexpress.aarch64.20140821.dtb",
}
default_kernels = {
"RealViewEB": "vmlinux.arm.smp.fb.2.6.38.8",
"RealViewPBX": "vmlinux.arm.smp.fb.2.6.38.8",
"VExpress_EMM": "vmlinux.aarch32.ll_20131205.0-gem5",
"VExpress_EMM64": "vmlinux.aarch64.20140821",
}
pci_devices = []
if bare_metal:
self = ArmSystem()
else:
self = LinuxArmSystem()
if not mdesc:
# generic system
mdesc = SysConfig()
self.readfile = mdesc.script()
self.iobus = IOXBar()
Bus: Split the bus into a non-coherent and coherent bus This patch introduces a class hierarchy of buses, a non-coherent one, and a coherent one, splitting the existing bus functionality. By doing so it also enables further specialisation of the two types of buses. A non-coherent bus connects a number of non-snooping masters and slaves, and routes the request and response packets based on the address. The request packets issued by the master connected to a non-coherent bus could still snoop in caches attached to a coherent bus, as is the case with the I/O bus and memory bus in most system configurations. No snoops will, however, reach any master on the non-coherent bus itself. The non-coherent bus can be used as a template for modelling PCI, PCIe, and non-coherent AMBA and OCP buses, and is typically used for the I/O buses. A coherent bus connects a number of (potentially) snooping masters and slaves, and routes the request and response packets based on the address, and also forwards all requests to the snoopers and deals with the snoop responses. The coherent bus can be used as a template for modelling QPI, HyperTransport, ACE and coherent OCP buses, and is typically used for the L1-to-L2 buses and as the main system interconnect. The configuration scripts are updated to use a NoncoherentBus for all peripheral and I/O buses. A bit of minor tidying up has also been done. --HG-- rename : src/mem/bus.cc => src/mem/coherent_bus.cc rename : src/mem/bus.hh => src/mem/coherent_bus.hh rename : src/mem/bus.cc => src/mem/noncoherent_bus.cc rename : src/mem/bus.hh => src/mem/noncoherent_bus.hh
2012-05-31 19:30:04 +02:00
self.membus = MemBus()
self.membus.badaddr_responder.warn_access = "warn"
Bridge: Remove NACKs in the bridge and unify with packet queue This patch removes the NACKing in the bridge, as the split request/response busses now ensure that protocol deadlocks do not occur, i.e. the message-dependency chain is broken by always allowing responses to make progress without being stalled by requests. The NACKs had limited support in the system with most components ignoring their use (with a suitable call to panic), and as the NACKs are no longer needed to avoid protocol deadlocks, the cleanest way is to simply remove them. The bridge is the starting point as this is the only place where the NACKs are created. A follow-up patch will remove the code that deals with NACKs in the endpoints, e.g. the X86 table walker and DMA port. Ultimately the type of packet can be complete removed (until someone sees a need for modelling more complex protocols, which can now be done in parts of the system since the port and interface is split). As a consequence of the NACK removal, the bridge now has to send a retry to a master if the request or response queue was full on the first attempt. This change also makes the bridge ports very similar to QueuedPorts, and a later patch will change the bridge to use these. A first step in this direction is taken by aligning the name of the member functions, as done by this patch. A bit of tidying up has also been done as part of the simplifications. Surprisingly, this patch has no impact on any of the regressions. Hence, there was never any NACKs issued. In a follow-up patch I would suggest changing the size of the bridge buffers set in FSConfig.py to also test the situation where the bridge fills up.
2012-08-22 17:39:58 +02:00
self.bridge = Bridge(delay='50ns')
self.bridge.master = self.iobus.slave
self.bridge.slave = self.membus.master
self.mem_mode = mem_mode
platform_class = PlatformConfig.get(machine_type)
# Resolve the real platform name, the original machine_type
# variable might have been an alias.
machine_type = platform_class.__name__
self.realview = platform_class()
if not dtb_filename and not bare_metal:
try:
dtb_filename = default_dtbs[machine_type]
except KeyError:
fatal("No DTB specified and no default DTB known for '%s'" % \
machine_type)
if isinstance(self.realview, VExpress_EMM64):
if os.path.split(mdesc.disk())[-1] == 'linux-aarch32-ael.img':
print "Selected 64-bit ARM architecture, updating default disk image..."
mdesc.diskname = 'linaro-minimal-aarch64.img'
# Attach any PCI devices this platform supports
self.realview.attachPciDevices()
self.cf0 = CowIdeDisk(driveID='master')
self.cf0.childImage(mdesc.disk())
# Old platforms have a built-in IDE or CF controller. Default to
# the IDE controller if both exist. New platforms expect the
# storage controller to be added from the config script.
if hasattr(self.realview, "ide"):
self.realview.ide.disks = [self.cf0]
elif hasattr(self.realview, "cf_ctrl"):
self.realview.cf_ctrl.disks = [self.cf0]
else:
self.pci_ide = IdeController(disks=[self.cf0])
pci_devices.append(self.pci_ide)
self.mem_ranges = []
size_remain = long(Addr(mdesc.mem()))
for region in self.realview._mem_regions:
if size_remain > long(region[1]):
self.mem_ranges.append(AddrRange(region[0], size=region[1]))
size_remain = size_remain - long(region[1])
else:
self.mem_ranges.append(AddrRange(region[0], size=size_remain))
size_remain = 0
break
warn("Memory size specified spans more than one region. Creating" \
" another memory controller for that range.")
if size_remain > 0:
fatal("The currently selected ARM platforms doesn't support" \
" the amount of DRAM you've selected. Please try" \
" another platform")
if bare_metal:
# EOT character on UART will end the simulation
self.realview.uart.end_on_eot = True
else:
if machine_type in default_kernels:
self.kernel = binary(default_kernels[machine_type])
if dtb_filename:
self.dtb_filename = binary(dtb_filename)
self.machine_type = machine_type if machine_type in ArmMachineType.map \
else "DTOnly"
# Ensure that writes to the UART actually go out early in the boot
if not cmdline:
cmdline = 'earlyprintk=pl011,0x1c090000 console=ttyAMA0 ' + \
'lpj=19988480 norandmaps rw loglevel=8 ' + \
'mem=%(mem)s root=%(rootdev)s'
# When using external memory, gem5 writes the boot loader to nvmem
# and then SST will read from it, but SST can only get to nvmem from
# iobus, as gem5's membus is only used for initialization and
# SST doesn't use it. Attaching nvmem to iobus solves this issue.
# During initialization, system_port -> membus -> iobus -> nvmem.
if external_memory:
self.realview.setupBootLoader(self.iobus, self, binary)
else:
self.realview.setupBootLoader(self.membus, self, binary)
self.gic_cpu_addr = self.realview.gic.cpu_addr
self.flags_addr = self.realview.realview_io.pio_addr + 0x30
2011-05-05 03:38:28 +02:00
# This check is for users who have previously put 'android' in
# the disk image filename to tell the config scripts to
# prepare the kernel with android-specific boot options. That
# behavior has been replaced with a more explicit option per
# the error message below. The disk can have any name now and
# doesn't need to include 'android' substring.
if (os.path.split(mdesc.disk())[-1]).lower().count('android'):
if 'android' not in mdesc.os_type():
fatal("It looks like you are trying to boot an Android " \
"platform. To boot Android, you must specify " \
"--os-type with an appropriate Android release on " \
"the command line.")
# android-specific tweaks
if 'android' in mdesc.os_type():
# generic tweaks
cmdline += " init=/init"
# release-specific tweaks
if 'kitkat' in mdesc.os_type():
cmdline += " androidboot.hardware=gem5 qemu=1 qemu.gles=0 " + \
"android.bootanim=0"
self.boot_osflags = fillInCmdline(mdesc, cmdline)
if external_memory:
# I/O traffic enters iobus
self.external_io = ExternalMaster(port_data="external_io",
port_type=external_memory)
self.external_io.port = self.iobus.slave
# Ensure iocache only receives traffic destined for (actual) memory.
self.iocache = ExternalSlave(port_data="iocache",
port_type=external_memory,
addr_ranges=self.mem_ranges)
self.iocache.port = self.iobus.master
# Let system_port get to nvmem and nothing else.
self.bridge.ranges = [self.realview.nvmem.range]
self.realview.attachOnChipIO(self.iobus)
else:
self.realview.attachOnChipIO(self.membus, self.bridge)
# Attach off-chip devices
self.realview.attachIO(self.iobus)
for dev_id, dev in enumerate(pci_devices):
dev.pci_bus, dev.pci_dev, dev.pci_func = (0, dev_id + 1, 0)
self.realview.attachPciDevice(dev, self.iobus)
self.intrctrl = IntrControl()
self.terminal = Terminal()
2011-02-12 01:29:35 +01:00
self.vncserver = VncServer()
self.system_port = self.membus.slave
return self
def makeLinuxMipsSystem(mem_mode, mdesc=None, cmdline=None):
class BaseMalta(Malta):
ethernet = NSGigE(pci_bus=0, pci_dev=1, pci_func=0)
ide = IdeController(disks=[Parent.disk0, Parent.disk2],
pci_func=0, pci_dev=0, pci_bus=0)
self = LinuxMipsSystem()
if not mdesc:
# generic system
mdesc = SysConfig()
self.readfile = mdesc.script()
self.iobus = IOXBar()
Bus: Split the bus into a non-coherent and coherent bus This patch introduces a class hierarchy of buses, a non-coherent one, and a coherent one, splitting the existing bus functionality. By doing so it also enables further specialisation of the two types of buses. A non-coherent bus connects a number of non-snooping masters and slaves, and routes the request and response packets based on the address. The request packets issued by the master connected to a non-coherent bus could still snoop in caches attached to a coherent bus, as is the case with the I/O bus and memory bus in most system configurations. No snoops will, however, reach any master on the non-coherent bus itself. The non-coherent bus can be used as a template for modelling PCI, PCIe, and non-coherent AMBA and OCP buses, and is typically used for the I/O buses. A coherent bus connects a number of (potentially) snooping masters and slaves, and routes the request and response packets based on the address, and also forwards all requests to the snoopers and deals with the snoop responses. The coherent bus can be used as a template for modelling QPI, HyperTransport, ACE and coherent OCP buses, and is typically used for the L1-to-L2 buses and as the main system interconnect. The configuration scripts are updated to use a NoncoherentBus for all peripheral and I/O buses. A bit of minor tidying up has also been done. --HG-- rename : src/mem/bus.cc => src/mem/coherent_bus.cc rename : src/mem/bus.hh => src/mem/coherent_bus.hh rename : src/mem/bus.cc => src/mem/noncoherent_bus.cc rename : src/mem/bus.hh => src/mem/noncoherent_bus.hh
2012-05-31 19:30:04 +02:00
self.membus = MemBus()
Bridge: Remove NACKs in the bridge and unify with packet queue This patch removes the NACKing in the bridge, as the split request/response busses now ensure that protocol deadlocks do not occur, i.e. the message-dependency chain is broken by always allowing responses to make progress without being stalled by requests. The NACKs had limited support in the system with most components ignoring their use (with a suitable call to panic), and as the NACKs are no longer needed to avoid protocol deadlocks, the cleanest way is to simply remove them. The bridge is the starting point as this is the only place where the NACKs are created. A follow-up patch will remove the code that deals with NACKs in the endpoints, e.g. the X86 table walker and DMA port. Ultimately the type of packet can be complete removed (until someone sees a need for modelling more complex protocols, which can now be done in parts of the system since the port and interface is split). As a consequence of the NACK removal, the bridge now has to send a retry to a master if the request or response queue was full on the first attempt. This change also makes the bridge ports very similar to QueuedPorts, and a later patch will change the bridge to use these. A first step in this direction is taken by aligning the name of the member functions, as done by this patch. A bit of tidying up has also been done as part of the simplifications. Surprisingly, this patch has no impact on any of the regressions. Hence, there was never any NACKs issued. In a follow-up patch I would suggest changing the size of the bridge buffers set in FSConfig.py to also test the situation where the bridge fills up.
2012-08-22 17:39:58 +02:00
self.bridge = Bridge(delay='50ns')
self.mem_ranges = [AddrRange('1GB')]
self.bridge.master = self.iobus.slave
self.bridge.slave = self.membus.master
self.disk0 = CowIdeDisk(driveID='master')
self.disk2 = CowIdeDisk(driveID='master')
self.disk0.childImage(mdesc.disk())
self.disk2.childImage(disk('linux-bigswap2.img'))
self.malta = BaseMalta()
self.malta.attachIO(self.iobus)
self.malta.ide.pio = self.iobus.master
self.malta.ide.dma = self.iobus.slave
self.malta.ethernet.pio = self.iobus.master
self.malta.ethernet.dma = self.iobus.slave
self.simple_disk = SimpleDisk(disk=RawDiskImage(image_file = mdesc.disk(),
read_only = True))
self.intrctrl = IntrControl()
self.mem_mode = mem_mode
self.terminal = Terminal()
self.kernel = binary('mips/vmlinux')
self.console = binary('mips/console')
if not cmdline:
cmdline = 'root=/dev/hda1 console=ttyS0'
self.boot_osflags = fillInCmdline(mdesc, cmdline)
self.system_port = self.membus.slave
return self
def x86IOAddress(port):
IO_address_space_base = 0x8000000000000000
return IO_address_space_base + port
def connectX86ClassicSystem(x86_sys, numCPUs):
# Constants similar to x86_traits.hh
IO_address_space_base = 0x8000000000000000
pci_config_address_space_base = 0xc000000000000000
interrupts_address_space_base = 0xa000000000000000
APIC_range_size = 1 << 12;
Bus: Split the bus into a non-coherent and coherent bus This patch introduces a class hierarchy of buses, a non-coherent one, and a coherent one, splitting the existing bus functionality. By doing so it also enables further specialisation of the two types of buses. A non-coherent bus connects a number of non-snooping masters and slaves, and routes the request and response packets based on the address. The request packets issued by the master connected to a non-coherent bus could still snoop in caches attached to a coherent bus, as is the case with the I/O bus and memory bus in most system configurations. No snoops will, however, reach any master on the non-coherent bus itself. The non-coherent bus can be used as a template for modelling PCI, PCIe, and non-coherent AMBA and OCP buses, and is typically used for the I/O buses. A coherent bus connects a number of (potentially) snooping masters and slaves, and routes the request and response packets based on the address, and also forwards all requests to the snoopers and deals with the snoop responses. The coherent bus can be used as a template for modelling QPI, HyperTransport, ACE and coherent OCP buses, and is typically used for the L1-to-L2 buses and as the main system interconnect. The configuration scripts are updated to use a NoncoherentBus for all peripheral and I/O buses. A bit of minor tidying up has also been done. --HG-- rename : src/mem/bus.cc => src/mem/coherent_bus.cc rename : src/mem/bus.hh => src/mem/coherent_bus.hh rename : src/mem/bus.cc => src/mem/noncoherent_bus.cc rename : src/mem/bus.hh => src/mem/noncoherent_bus.hh
2012-05-31 19:30:04 +02:00
x86_sys.membus = MemBus()
2011-02-07 07:14:18 +01:00
# North Bridge
x86_sys.iobus = IOXBar()
Bridge: Remove NACKs in the bridge and unify with packet queue This patch removes the NACKing in the bridge, as the split request/response busses now ensure that protocol deadlocks do not occur, i.e. the message-dependency chain is broken by always allowing responses to make progress without being stalled by requests. The NACKs had limited support in the system with most components ignoring their use (with a suitable call to panic), and as the NACKs are no longer needed to avoid protocol deadlocks, the cleanest way is to simply remove them. The bridge is the starting point as this is the only place where the NACKs are created. A follow-up patch will remove the code that deals with NACKs in the endpoints, e.g. the X86 table walker and DMA port. Ultimately the type of packet can be complete removed (until someone sees a need for modelling more complex protocols, which can now be done in parts of the system since the port and interface is split). As a consequence of the NACK removal, the bridge now has to send a retry to a master if the request or response queue was full on the first attempt. This change also makes the bridge ports very similar to QueuedPorts, and a later patch will change the bridge to use these. A first step in this direction is taken by aligning the name of the member functions, as done by this patch. A bit of tidying up has also been done as part of the simplifications. Surprisingly, this patch has no impact on any of the regressions. Hence, there was never any NACKs issued. In a follow-up patch I would suggest changing the size of the bridge buffers set in FSConfig.py to also test the situation where the bridge fills up.
2012-08-22 17:39:58 +02:00
x86_sys.bridge = Bridge(delay='50ns')
x86_sys.bridge.master = x86_sys.iobus.slave
x86_sys.bridge.slave = x86_sys.membus.master
# Allow the bridge to pass through:
# 1) kernel configured PCI device memory map address: address range
# [0xC0000000, 0xFFFF0000). (The upper 64kB are reserved for m5ops.)
# 2) the bridge to pass through the IO APIC (two pages, already contained in 1),
# 3) everything in the IO address range up to the local APIC, and
# 4) then the entire PCI address space and beyond.
x86_sys.bridge.ranges = \
[
AddrRange(0xC0000000, 0xFFFF0000),
AddrRange(IO_address_space_base,
interrupts_address_space_base - 1),
AddrRange(pci_config_address_space_base,
Addr.max)
]
# Create a bridge from the IO bus to the memory bus to allow access to
# the local APIC (two pages)
Bridge: Remove NACKs in the bridge and unify with packet queue This patch removes the NACKing in the bridge, as the split request/response busses now ensure that protocol deadlocks do not occur, i.e. the message-dependency chain is broken by always allowing responses to make progress without being stalled by requests. The NACKs had limited support in the system with most components ignoring their use (with a suitable call to panic), and as the NACKs are no longer needed to avoid protocol deadlocks, the cleanest way is to simply remove them. The bridge is the starting point as this is the only place where the NACKs are created. A follow-up patch will remove the code that deals with NACKs in the endpoints, e.g. the X86 table walker and DMA port. Ultimately the type of packet can be complete removed (until someone sees a need for modelling more complex protocols, which can now be done in parts of the system since the port and interface is split). As a consequence of the NACK removal, the bridge now has to send a retry to a master if the request or response queue was full on the first attempt. This change also makes the bridge ports very similar to QueuedPorts, and a later patch will change the bridge to use these. A first step in this direction is taken by aligning the name of the member functions, as done by this patch. A bit of tidying up has also been done as part of the simplifications. Surprisingly, this patch has no impact on any of the regressions. Hence, there was never any NACKs issued. In a follow-up patch I would suggest changing the size of the bridge buffers set in FSConfig.py to also test the situation where the bridge fills up.
2012-08-22 17:39:58 +02:00
x86_sys.apicbridge = Bridge(delay='50ns')
x86_sys.apicbridge.slave = x86_sys.iobus.master
x86_sys.apicbridge.master = x86_sys.membus.slave
x86_sys.apicbridge.ranges = [AddrRange(interrupts_address_space_base,
interrupts_address_space_base +
numCPUs * APIC_range_size
- 1)]
2011-02-07 07:14:18 +01:00
# connect the io bus
x86_sys.pc.attachIO(x86_sys.iobus)
x86_sys.system_port = x86_sys.membus.slave
2011-02-07 07:14:18 +01:00
def connectX86RubySystem(x86_sys):
# North Bridge
x86_sys.iobus = IOXBar()
2011-02-07 07:14:18 +01:00
# add the ide to the list of dma devices that later need to attach to
# dma controllers
x86_sys._dma_ports = [x86_sys.pc.south_bridge.ide.dma]
x86_sys.pc.attachIO(x86_sys.iobus, x86_sys._dma_ports)
2011-02-07 07:14:18 +01:00
def makeX86System(mem_mode, numCPUs=1, mdesc=None, self=None, Ruby=False):
if self == None:
self = X86System()
if not mdesc:
# generic system
mdesc = SysConfig()
self.readfile = mdesc.script()
self.mem_mode = mem_mode
# Physical memory
# On the PC platform, the memory region 0xC0000000-0xFFFFFFFF is reserved
# for various devices. Hence, if the physical memory size is greater than
# 3GB, we need to split it into two parts.
excess_mem_size = \
convert.toMemorySize(mdesc.mem()) - convert.toMemorySize('3GB')
if excess_mem_size <= 0:
self.mem_ranges = [AddrRange(mdesc.mem())]
else:
warn("Physical memory size specified is %s which is greater than " \
"3GB. Twice the number of memory controllers would be " \
"created." % (mdesc.mem()))
self.mem_ranges = [AddrRange('3GB'),
AddrRange(Addr('4GB'), size = excess_mem_size)]
# Platform
self.pc = Pc()
2011-02-07 07:14:18 +01:00
# Create and connect the busses required by each memory system
if Ruby:
connectX86RubySystem(self)
else:
connectX86ClassicSystem(self, numCPUs)
self.intrctrl = IntrControl()
2009-02-01 09:24:26 +01:00
# Disks
disk0 = CowIdeDisk(driveID='master')
disk2 = CowIdeDisk(driveID='master')
disk0.childImage(mdesc.disk())
disk2.childImage(disk('linux-bigswap2.img'))
self.pc.south_bridge.ide.disks = [disk0, disk2]
# Add in a Bios information structure.
structures = [X86SMBiosBiosInformation()]
self.smbios_table.structures = structures
# Set up the Intel MP table
base_entries = []
ext_entries = []
for i in xrange(numCPUs):
bp = X86IntelMPProcessor(
local_apic_id = i,
local_apic_version = 0x14,
enable = True,
bootstrap = (i == 0))
base_entries.append(bp)
io_apic = X86IntelMPIOAPIC(
id = numCPUs,
version = 0x11,
enable = True,
address = 0xfec00000)
self.pc.south_bridge.io_apic.apic_id = io_apic.id
base_entries.append(io_apic)
# In gem5 Pc::calcPciConfigAddr(), it required "assert(bus==0)",
# but linux kernel cannot config PCI device if it was not connected to PCI bus,
# so we fix PCI bus id to 0, and ISA bus id to 1.
pci_bus = X86IntelMPBus(bus_id = 0, bus_type='PCI')
base_entries.append(pci_bus)
isa_bus = X86IntelMPBus(bus_id = 1, bus_type='ISA')
base_entries.append(isa_bus)
connect_busses = X86IntelMPBusHierarchy(bus_id=1,
subtractive_decode=True, parent_bus=0)
ext_entries.append(connect_busses)
pci_dev4_inta = X86IntelMPIOIntAssignment(
interrupt_type = 'INT',
polarity = 'ConformPolarity',
trigger = 'ConformTrigger',
source_bus_id = 0,
source_bus_irq = 0 + (4 << 2),
dest_io_apic_id = io_apic.id,
dest_io_apic_intin = 16)
base_entries.append(pci_dev4_inta)
def assignISAInt(irq, apicPin):
assign_8259_to_apic = X86IntelMPIOIntAssignment(
interrupt_type = 'ExtInt',
polarity = 'ConformPolarity',
trigger = 'ConformTrigger',
source_bus_id = 1,
source_bus_irq = irq,
dest_io_apic_id = io_apic.id,
dest_io_apic_intin = 0)
base_entries.append(assign_8259_to_apic)
assign_to_apic = X86IntelMPIOIntAssignment(
interrupt_type = 'INT',
polarity = 'ConformPolarity',
trigger = 'ConformTrigger',
source_bus_id = 1,
source_bus_irq = irq,
dest_io_apic_id = io_apic.id,
dest_io_apic_intin = apicPin)
base_entries.append(assign_to_apic)
assignISAInt(0, 2)
assignISAInt(1, 1)
for i in range(3, 15):
assignISAInt(i, i)
self.intel_mp_table.base_entries = base_entries
self.intel_mp_table.ext_entries = ext_entries
def makeLinuxX86System(mem_mode, numCPUs=1, mdesc=None, Ruby=False,
cmdline=None):
self = LinuxX86System()
2011-02-07 07:14:18 +01:00
# Build up the x86 system and then specialize it for Linux
makeX86System(mem_mode, numCPUs, mdesc, self, Ruby)
# We assume below that there's at least 1MB of memory. We'll require 2
# just to avoid corner cases.
phys_mem_size = sum(map(lambda r: r.size(), self.mem_ranges))
assert(phys_mem_size >= 0x200000)
assert(len(self.mem_ranges) <= 2)
entries = \
[
# Mark the first megabyte of memory as reserved
X86E820Entry(addr = 0, size = '639kB', range_type = 1),
X86E820Entry(addr = 0x9fc00, size = '385kB', range_type = 2),
# Mark the rest of physical memory as available
X86E820Entry(addr = 0x100000,
size = '%dB' % (self.mem_ranges[0].size() - 0x100000),
range_type = 1),
]
# Mark [mem_size, 3GB) as reserved if memory less than 3GB, which force
# IO devices to be mapped to [0xC0000000, 0xFFFF0000). Requests to this
# specific range can pass though bridge to iobus.
if len(self.mem_ranges) == 1:
entries.append(X86E820Entry(addr = self.mem_ranges[0].size(),
size='%dB' % (0xC0000000 - self.mem_ranges[0].size()),
range_type=2))
# Reserve the last 16kB of the 32-bit address space for the m5op interface
entries.append(X86E820Entry(addr=0xFFFF0000, size='64kB', range_type=2))
# In case the physical memory is greater than 3GB, we split it into two
# parts and add a separate e820 entry for the second part. This entry
# starts at 0x100000000, which is the first address after the space
# reserved for devices.
if len(self.mem_ranges) == 2:
entries.append(X86E820Entry(addr = 0x100000000,
size = '%dB' % (self.mem_ranges[1].size()), range_type = 1))
self.e820_table.entries = entries
# Command line
if not cmdline:
cmdline = 'earlyprintk=ttyS0 console=ttyS0 lpj=7999923 root=/dev/hda1'
self.boot_osflags = fillInCmdline(mdesc, cmdline)
self.kernel = binary('x86_64-vmlinux-2.6.22.9')
return self
def makeDualRoot(full_system, testSystem, driveSystem, dumpfile):
self = Root(full_system = full_system)
self.testsys = testSystem
self.drivesys = driveSystem
self.etherlink = EtherLink()
if hasattr(testSystem, 'realview'):
self.etherlink.int0 = Parent.testsys.realview.ethernet.interface
self.etherlink.int1 = Parent.drivesys.realview.ethernet.interface
elif hasattr(testSystem, 'tsunami'):
self.etherlink.int0 = Parent.testsys.tsunami.ethernet.interface
self.etherlink.int1 = Parent.drivesys.tsunami.ethernet.interface
else:
fatal("Don't know how to connect these system together")
if dumpfile:
self.etherdump = EtherDump(file=dumpfile)
self.etherlink.dump = Parent.etherdump
return self
def makeDistRoot(testSystem,
rank,
size,
server_name,
server_port,
sync_repeat,
sync_start,
linkspeed,
linkdelay,
dumpfile):
self = Root(full_system = True)
self.testsys = testSystem
self.etherlink = DistEtherLink(speed = linkspeed,
delay = linkdelay,
dist_rank = rank,
dist_size = size,
server_name = server_name,
server_port = server_port,
sync_start = sync_start,
sync_repeat = sync_repeat)
if hasattr(testSystem, 'realview'):
self.etherlink.int0 = Parent.testsys.realview.ethernet.interface
elif hasattr(testSystem, 'tsunami'):
self.etherlink.int0 = Parent.testsys.tsunami.ethernet.interface
else:
fatal("Don't know how to connect DistEtherLink to this system")
if dumpfile:
self.etherdump = EtherDump(file=dumpfile)
self.etherlink.dump = Parent.etherdump
return self