gem5/kern/linux/linux_system.cc

279 lines
9.8 KiB
C++
Raw Normal View History

/*
* Copyright (c) 2004 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.
*/
/**
* @file
* This code loads the linux kernel, console, pal and patches certain
* functions. The symbol tables are loaded so that traces can show
* the executing function and we can skip functions. Various delay
* loops are skipped and their final values manually computed to speed
* up boot time.
*/
#include "base/loader/symtab.hh"
#include "base/trace.hh"
#include "cpu/exec_context.hh"
#include "cpu/base_cpu.hh"
#include "kern/linux/linux_events.hh"
#include "kern/linux/linux_system.hh"
#include "kern/system_events.hh"
#include "mem/functional_mem/memory_control.hh"
#include "mem/functional_mem/physical_memory.hh"
#include "sim/builder.hh"
#include "dev/platform.hh"
#include "targetarch/isa_traits.hh"
#include "targetarch/vtophys.hh"
#include "sim/debug.hh"
using namespace std;
- Clean up and factor out all of the binning code into a single place so it's easier to work with. - Add support for binning kernel/user/idle time separately from lisa's binning stuff, but make the two compatible. - KernelStats used to directly implement the pImpl idiom, but it makes more sense to just remove the level of indirection and make the exec context have a pointer to the stats. - Factor common code out of LinuxSystem and Tru64System and put it into the System base class. While doing that, make all constructors take a pointer to a parameter struct instead of naming the parameters individually to make it much easier to add parameters to these classes. SConscript: Move the function tracking and binning stuff around. arch/alpha/ev5.cc: kernelStats is now a pointer arch/alpha/pseudo_inst.cc: kernelStats is now a pointer the parameters to the system have been moved into their own struct base/trace.hh: provide a little functor class for wrapping a string that can allow you to define name() in any scope very simply for use with DPRINTF cpu/base_cpu.cc: New order of arguments for consistency. cpu/exec_context.cc: kernelStats no longer has the level of indirection in it, execContext has the indirection now. so, kernelStats is a pointer. We also need a pointer to the kernelBinning stuff from the system and we need to figure out if we want to do binning or not. Move a whole bunch of code into kern_binning.cc so it's all in the same place. cpu/exec_context.hh: We want pointers to the kernel binning/stats stuff and we'll have the exec_context and system have the level of indirection instead of having the extra layer in the kernel stats class. cpu/simple_cpu/simple_cpu.cc: call through the exec context to do the special binning stuff. kern/kernel_stats.cc: kern/kernel_stats.hh: Re-organize the stats stuff and remove the level of indirection (that was there to simplify building) and move the binning stuff into its own class/file. kern/linux/linux_system.cc: kern/linux/linux_system.hh: kern/tru64/tru64_system.cc: kern/tru64/tru64_system.hh: sim/system.cc: sim/system.hh: move lots of common system code into the base system class so that it can be shared between linux, tru64, and whatever else we decide to support in the future. Make the constructor take a pointer to a parameter struct so that it is easier to pass parameters to the parent. kern/system_events.cc: move the majority of the binning code into the Kernel::Binning class in the kern_binning file kern/system_events.hh: FnEvents only need to know the bin create the Idle start event to find the PCBB of the idle process when it starts. kern/tru64/tru64_events.cc: memCtrl -> memctrl sim/process.cc: sim/process.hh: re-order args for consistency --HG-- extra : convert_revision : 86cb39738c41fcd680f2aad125c9dde000227b2b
2004-08-20 17:35:31 +02:00
LinuxSystem::LinuxSystem(Params *p)
: System(p)
{
Addr addr = 0;
/**
- Clean up and factor out all of the binning code into a single place so it's easier to work with. - Add support for binning kernel/user/idle time separately from lisa's binning stuff, but make the two compatible. - KernelStats used to directly implement the pImpl idiom, but it makes more sense to just remove the level of indirection and make the exec context have a pointer to the stats. - Factor common code out of LinuxSystem and Tru64System and put it into the System base class. While doing that, make all constructors take a pointer to a parameter struct instead of naming the parameters individually to make it much easier to add parameters to these classes. SConscript: Move the function tracking and binning stuff around. arch/alpha/ev5.cc: kernelStats is now a pointer arch/alpha/pseudo_inst.cc: kernelStats is now a pointer the parameters to the system have been moved into their own struct base/trace.hh: provide a little functor class for wrapping a string that can allow you to define name() in any scope very simply for use with DPRINTF cpu/base_cpu.cc: New order of arguments for consistency. cpu/exec_context.cc: kernelStats no longer has the level of indirection in it, execContext has the indirection now. so, kernelStats is a pointer. We also need a pointer to the kernelBinning stuff from the system and we need to figure out if we want to do binning or not. Move a whole bunch of code into kern_binning.cc so it's all in the same place. cpu/exec_context.hh: We want pointers to the kernel binning/stats stuff and we'll have the exec_context and system have the level of indirection instead of having the extra layer in the kernel stats class. cpu/simple_cpu/simple_cpu.cc: call through the exec context to do the special binning stuff. kern/kernel_stats.cc: kern/kernel_stats.hh: Re-organize the stats stuff and remove the level of indirection (that was there to simplify building) and move the binning stuff into its own class/file. kern/linux/linux_system.cc: kern/linux/linux_system.hh: kern/tru64/tru64_system.cc: kern/tru64/tru64_system.hh: sim/system.cc: sim/system.hh: move lots of common system code into the base system class so that it can be shared between linux, tru64, and whatever else we decide to support in the future. Make the constructor take a pointer to a parameter struct so that it is easier to pass parameters to the parent. kern/system_events.cc: move the majority of the binning code into the Kernel::Binning class in the kern_binning file kern/system_events.hh: FnEvents only need to know the bin create the Idle start event to find the PCBB of the idle process when it starts. kern/tru64/tru64_events.cc: memCtrl -> memctrl sim/process.cc: sim/process.hh: re-order args for consistency --HG-- extra : convert_revision : 86cb39738c41fcd680f2aad125c9dde000227b2b
2004-08-20 17:35:31 +02:00
* find the address of the est_cycle_freq variable and insert it
* so we don't through the lengthly process of trying to
* calculated it by using the PIT, RTC, etc.
*/
if (kernelSymtab->findAddress("est_cycle_freq", addr)) {
Addr paddr = vtophys(physmem, addr);
uint8_t *est_cycle_frequency =
physmem->dma_addr(paddr, sizeof(uint64_t));
if (est_cycle_frequency)
*(uint64_t *)est_cycle_frequency = htoa(ticksPerSecond);
}
/**
* Since we aren't using a bootloader, we have to copy the kernel arguments
* directly into the kernels memory.
*/
{
Addr paddr = vtophys(physmem, PARAM_ADDR);
char *commandline = (char*)physmem->dma_addr(paddr, sizeof(uint64_t));
if (commandline)
- Clean up and factor out all of the binning code into a single place so it's easier to work with. - Add support for binning kernel/user/idle time separately from lisa's binning stuff, but make the two compatible. - KernelStats used to directly implement the pImpl idiom, but it makes more sense to just remove the level of indirection and make the exec context have a pointer to the stats. - Factor common code out of LinuxSystem and Tru64System and put it into the System base class. While doing that, make all constructors take a pointer to a parameter struct instead of naming the parameters individually to make it much easier to add parameters to these classes. SConscript: Move the function tracking and binning stuff around. arch/alpha/ev5.cc: kernelStats is now a pointer arch/alpha/pseudo_inst.cc: kernelStats is now a pointer the parameters to the system have been moved into their own struct base/trace.hh: provide a little functor class for wrapping a string that can allow you to define name() in any scope very simply for use with DPRINTF cpu/base_cpu.cc: New order of arguments for consistency. cpu/exec_context.cc: kernelStats no longer has the level of indirection in it, execContext has the indirection now. so, kernelStats is a pointer. We also need a pointer to the kernelBinning stuff from the system and we need to figure out if we want to do binning or not. Move a whole bunch of code into kern_binning.cc so it's all in the same place. cpu/exec_context.hh: We want pointers to the kernel binning/stats stuff and we'll have the exec_context and system have the level of indirection instead of having the extra layer in the kernel stats class. cpu/simple_cpu/simple_cpu.cc: call through the exec context to do the special binning stuff. kern/kernel_stats.cc: kern/kernel_stats.hh: Re-organize the stats stuff and remove the level of indirection (that was there to simplify building) and move the binning stuff into its own class/file. kern/linux/linux_system.cc: kern/linux/linux_system.hh: kern/tru64/tru64_system.cc: kern/tru64/tru64_system.hh: sim/system.cc: sim/system.hh: move lots of common system code into the base system class so that it can be shared between linux, tru64, and whatever else we decide to support in the future. Make the constructor take a pointer to a parameter struct so that it is easier to pass parameters to the parent. kern/system_events.cc: move the majority of the binning code into the Kernel::Binning class in the kern_binning file kern/system_events.hh: FnEvents only need to know the bin create the Idle start event to find the PCBB of the idle process when it starts. kern/tru64/tru64_events.cc: memCtrl -> memctrl sim/process.cc: sim/process.hh: re-order args for consistency --HG-- extra : convert_revision : 86cb39738c41fcd680f2aad125c9dde000227b2b
2004-08-20 17:35:31 +02:00
strcpy(commandline, params->boot_osflags.c_str());
}
/**
* EV5 only supports 127 ASNs so we are going to tell the kernel that the
* paritiuclar EV6 we have only supports 127 asns.
* @todo At some point we should change ev5.hh and the palcode to support
* 255 ASNs.
*/
if (kernelSymtab->findAddress("dp264_mv", addr)) {
Addr paddr = vtophys(physmem, addr);
char *dp264_mv = (char *)physmem->dma_addr(paddr, sizeof(uint64_t));
if (dp264_mv) {
*(uint32_t*)(dp264_mv+0x18) = htoa((uint32_t)127);
} else
- Clean up and factor out all of the binning code into a single place so it's easier to work with. - Add support for binning kernel/user/idle time separately from lisa's binning stuff, but make the two compatible. - KernelStats used to directly implement the pImpl idiom, but it makes more sense to just remove the level of indirection and make the exec context have a pointer to the stats. - Factor common code out of LinuxSystem and Tru64System and put it into the System base class. While doing that, make all constructors take a pointer to a parameter struct instead of naming the parameters individually to make it much easier to add parameters to these classes. SConscript: Move the function tracking and binning stuff around. arch/alpha/ev5.cc: kernelStats is now a pointer arch/alpha/pseudo_inst.cc: kernelStats is now a pointer the parameters to the system have been moved into their own struct base/trace.hh: provide a little functor class for wrapping a string that can allow you to define name() in any scope very simply for use with DPRINTF cpu/base_cpu.cc: New order of arguments for consistency. cpu/exec_context.cc: kernelStats no longer has the level of indirection in it, execContext has the indirection now. so, kernelStats is a pointer. We also need a pointer to the kernelBinning stuff from the system and we need to figure out if we want to do binning or not. Move a whole bunch of code into kern_binning.cc so it's all in the same place. cpu/exec_context.hh: We want pointers to the kernel binning/stats stuff and we'll have the exec_context and system have the level of indirection instead of having the extra layer in the kernel stats class. cpu/simple_cpu/simple_cpu.cc: call through the exec context to do the special binning stuff. kern/kernel_stats.cc: kern/kernel_stats.hh: Re-organize the stats stuff and remove the level of indirection (that was there to simplify building) and move the binning stuff into its own class/file. kern/linux/linux_system.cc: kern/linux/linux_system.hh: kern/tru64/tru64_system.cc: kern/tru64/tru64_system.hh: sim/system.cc: sim/system.hh: move lots of common system code into the base system class so that it can be shared between linux, tru64, and whatever else we decide to support in the future. Make the constructor take a pointer to a parameter struct so that it is easier to pass parameters to the parent. kern/system_events.cc: move the majority of the binning code into the Kernel::Binning class in the kern_binning file kern/system_events.hh: FnEvents only need to know the bin create the Idle start event to find the PCBB of the idle process when it starts. kern/tru64/tru64_events.cc: memCtrl -> memctrl sim/process.cc: sim/process.hh: re-order args for consistency --HG-- extra : convert_revision : 86cb39738c41fcd680f2aad125c9dde000227b2b
2004-08-20 17:35:31 +02:00
panic("could not translate dp264_mv addr\n");
} else
- Clean up and factor out all of the binning code into a single place so it's easier to work with. - Add support for binning kernel/user/idle time separately from lisa's binning stuff, but make the two compatible. - KernelStats used to directly implement the pImpl idiom, but it makes more sense to just remove the level of indirection and make the exec context have a pointer to the stats. - Factor common code out of LinuxSystem and Tru64System and put it into the System base class. While doing that, make all constructors take a pointer to a parameter struct instead of naming the parameters individually to make it much easier to add parameters to these classes. SConscript: Move the function tracking and binning stuff around. arch/alpha/ev5.cc: kernelStats is now a pointer arch/alpha/pseudo_inst.cc: kernelStats is now a pointer the parameters to the system have been moved into their own struct base/trace.hh: provide a little functor class for wrapping a string that can allow you to define name() in any scope very simply for use with DPRINTF cpu/base_cpu.cc: New order of arguments for consistency. cpu/exec_context.cc: kernelStats no longer has the level of indirection in it, execContext has the indirection now. so, kernelStats is a pointer. We also need a pointer to the kernelBinning stuff from the system and we need to figure out if we want to do binning or not. Move a whole bunch of code into kern_binning.cc so it's all in the same place. cpu/exec_context.hh: We want pointers to the kernel binning/stats stuff and we'll have the exec_context and system have the level of indirection instead of having the extra layer in the kernel stats class. cpu/simple_cpu/simple_cpu.cc: call through the exec context to do the special binning stuff. kern/kernel_stats.cc: kern/kernel_stats.hh: Re-organize the stats stuff and remove the level of indirection (that was there to simplify building) and move the binning stuff into its own class/file. kern/linux/linux_system.cc: kern/linux/linux_system.hh: kern/tru64/tru64_system.cc: kern/tru64/tru64_system.hh: sim/system.cc: sim/system.hh: move lots of common system code into the base system class so that it can be shared between linux, tru64, and whatever else we decide to support in the future. Make the constructor take a pointer to a parameter struct so that it is easier to pass parameters to the parent. kern/system_events.cc: move the majority of the binning code into the Kernel::Binning class in the kern_binning file kern/system_events.hh: FnEvents only need to know the bin create the Idle start event to find the PCBB of the idle process when it starts. kern/tru64/tru64_events.cc: memCtrl -> memctrl sim/process.cc: sim/process.hh: re-order args for consistency --HG-- extra : convert_revision : 86cb39738c41fcd680f2aad125c9dde000227b2b
2004-08-20 17:35:31 +02:00
panic("could not find dp264_mv\n");
#ifndef NDEBUG
- Clean up and factor out all of the binning code into a single place so it's easier to work with. - Add support for binning kernel/user/idle time separately from lisa's binning stuff, but make the two compatible. - KernelStats used to directly implement the pImpl idiom, but it makes more sense to just remove the level of indirection and make the exec context have a pointer to the stats. - Factor common code out of LinuxSystem and Tru64System and put it into the System base class. While doing that, make all constructors take a pointer to a parameter struct instead of naming the parameters individually to make it much easier to add parameters to these classes. SConscript: Move the function tracking and binning stuff around. arch/alpha/ev5.cc: kernelStats is now a pointer arch/alpha/pseudo_inst.cc: kernelStats is now a pointer the parameters to the system have been moved into their own struct base/trace.hh: provide a little functor class for wrapping a string that can allow you to define name() in any scope very simply for use with DPRINTF cpu/base_cpu.cc: New order of arguments for consistency. cpu/exec_context.cc: kernelStats no longer has the level of indirection in it, execContext has the indirection now. so, kernelStats is a pointer. We also need a pointer to the kernelBinning stuff from the system and we need to figure out if we want to do binning or not. Move a whole bunch of code into kern_binning.cc so it's all in the same place. cpu/exec_context.hh: We want pointers to the kernel binning/stats stuff and we'll have the exec_context and system have the level of indirection instead of having the extra layer in the kernel stats class. cpu/simple_cpu/simple_cpu.cc: call through the exec context to do the special binning stuff. kern/kernel_stats.cc: kern/kernel_stats.hh: Re-organize the stats stuff and remove the level of indirection (that was there to simplify building) and move the binning stuff into its own class/file. kern/linux/linux_system.cc: kern/linux/linux_system.hh: kern/tru64/tru64_system.cc: kern/tru64/tru64_system.hh: sim/system.cc: sim/system.hh: move lots of common system code into the base system class so that it can be shared between linux, tru64, and whatever else we decide to support in the future. Make the constructor take a pointer to a parameter struct so that it is easier to pass parameters to the parent. kern/system_events.cc: move the majority of the binning code into the Kernel::Binning class in the kern_binning file kern/system_events.hh: FnEvents only need to know the bin create the Idle start event to find the PCBB of the idle process when it starts. kern/tru64/tru64_events.cc: memCtrl -> memctrl sim/process.cc: sim/process.hh: re-order args for consistency --HG-- extra : convert_revision : 86cb39738c41fcd680f2aad125c9dde000227b2b
2004-08-20 17:35:31 +02:00
kernelPanicEvent = new BreakPCEvent(&pcEventQueue, "kernel panic");
if (kernelSymtab->findAddress("panic", addr))
kernelPanicEvent->schedule(addr);
else
panic("could not find kernel symbol \'panic\'");
kernelDieEvent = new BreakPCEvent(&pcEventQueue, "die if kernel");
if (kernelSymtab->findAddress("die_if_kernel", addr))
kernelDieEvent->schedule(addr);
else
panic("could not find kernel symbol \'die_if_kernel\'");
#endif
/**
- Clean up and factor out all of the binning code into a single place so it's easier to work with. - Add support for binning kernel/user/idle time separately from lisa's binning stuff, but make the two compatible. - KernelStats used to directly implement the pImpl idiom, but it makes more sense to just remove the level of indirection and make the exec context have a pointer to the stats. - Factor common code out of LinuxSystem and Tru64System and put it into the System base class. While doing that, make all constructors take a pointer to a parameter struct instead of naming the parameters individually to make it much easier to add parameters to these classes. SConscript: Move the function tracking and binning stuff around. arch/alpha/ev5.cc: kernelStats is now a pointer arch/alpha/pseudo_inst.cc: kernelStats is now a pointer the parameters to the system have been moved into their own struct base/trace.hh: provide a little functor class for wrapping a string that can allow you to define name() in any scope very simply for use with DPRINTF cpu/base_cpu.cc: New order of arguments for consistency. cpu/exec_context.cc: kernelStats no longer has the level of indirection in it, execContext has the indirection now. so, kernelStats is a pointer. We also need a pointer to the kernelBinning stuff from the system and we need to figure out if we want to do binning or not. Move a whole bunch of code into kern_binning.cc so it's all in the same place. cpu/exec_context.hh: We want pointers to the kernel binning/stats stuff and we'll have the exec_context and system have the level of indirection instead of having the extra layer in the kernel stats class. cpu/simple_cpu/simple_cpu.cc: call through the exec context to do the special binning stuff. kern/kernel_stats.cc: kern/kernel_stats.hh: Re-organize the stats stuff and remove the level of indirection (that was there to simplify building) and move the binning stuff into its own class/file. kern/linux/linux_system.cc: kern/linux/linux_system.hh: kern/tru64/tru64_system.cc: kern/tru64/tru64_system.hh: sim/system.cc: sim/system.hh: move lots of common system code into the base system class so that it can be shared between linux, tru64, and whatever else we decide to support in the future. Make the constructor take a pointer to a parameter struct so that it is easier to pass parameters to the parent. kern/system_events.cc: move the majority of the binning code into the Kernel::Binning class in the kern_binning file kern/system_events.hh: FnEvents only need to know the bin create the Idle start event to find the PCBB of the idle process when it starts. kern/tru64/tru64_events.cc: memCtrl -> memctrl sim/process.cc: sim/process.hh: re-order args for consistency --HG-- extra : convert_revision : 86cb39738c41fcd680f2aad125c9dde000227b2b
2004-08-20 17:35:31 +02:00
* Any time ide_delay_50ms, calibarte_delay or
* determine_cpu_caches is called just skip the
* function. Currently determine_cpu_caches only is used put
* information in proc, however if that changes in the future we
* will have to fill in the cache size variables appropriately.
*/
- Clean up and factor out all of the binning code into a single place so it's easier to work with. - Add support for binning kernel/user/idle time separately from lisa's binning stuff, but make the two compatible. - KernelStats used to directly implement the pImpl idiom, but it makes more sense to just remove the level of indirection and make the exec context have a pointer to the stats. - Factor common code out of LinuxSystem and Tru64System and put it into the System base class. While doing that, make all constructors take a pointer to a parameter struct instead of naming the parameters individually to make it much easier to add parameters to these classes. SConscript: Move the function tracking and binning stuff around. arch/alpha/ev5.cc: kernelStats is now a pointer arch/alpha/pseudo_inst.cc: kernelStats is now a pointer the parameters to the system have been moved into their own struct base/trace.hh: provide a little functor class for wrapping a string that can allow you to define name() in any scope very simply for use with DPRINTF cpu/base_cpu.cc: New order of arguments for consistency. cpu/exec_context.cc: kernelStats no longer has the level of indirection in it, execContext has the indirection now. so, kernelStats is a pointer. We also need a pointer to the kernelBinning stuff from the system and we need to figure out if we want to do binning or not. Move a whole bunch of code into kern_binning.cc so it's all in the same place. cpu/exec_context.hh: We want pointers to the kernel binning/stats stuff and we'll have the exec_context and system have the level of indirection instead of having the extra layer in the kernel stats class. cpu/simple_cpu/simple_cpu.cc: call through the exec context to do the special binning stuff. kern/kernel_stats.cc: kern/kernel_stats.hh: Re-organize the stats stuff and remove the level of indirection (that was there to simplify building) and move the binning stuff into its own class/file. kern/linux/linux_system.cc: kern/linux/linux_system.hh: kern/tru64/tru64_system.cc: kern/tru64/tru64_system.hh: sim/system.cc: sim/system.hh: move lots of common system code into the base system class so that it can be shared between linux, tru64, and whatever else we decide to support in the future. Make the constructor take a pointer to a parameter struct so that it is easier to pass parameters to the parent. kern/system_events.cc: move the majority of the binning code into the Kernel::Binning class in the kern_binning file kern/system_events.hh: FnEvents only need to know the bin create the Idle start event to find the PCBB of the idle process when it starts. kern/tru64/tru64_events.cc: memCtrl -> memctrl sim/process.cc: sim/process.hh: re-order args for consistency --HG-- extra : convert_revision : 86cb39738c41fcd680f2aad125c9dde000227b2b
2004-08-20 17:35:31 +02:00
skipIdeDelay50msEvent = new SkipFuncEvent(&pcEventQueue, "ide_delay_50ms");
if (kernelSymtab->findAddress("ide_delay_50ms", addr))
skipIdeDelay50msEvent->schedule(addr+sizeof(MachInst));
- Clean up and factor out all of the binning code into a single place so it's easier to work with. - Add support for binning kernel/user/idle time separately from lisa's binning stuff, but make the two compatible. - KernelStats used to directly implement the pImpl idiom, but it makes more sense to just remove the level of indirection and make the exec context have a pointer to the stats. - Factor common code out of LinuxSystem and Tru64System and put it into the System base class. While doing that, make all constructors take a pointer to a parameter struct instead of naming the parameters individually to make it much easier to add parameters to these classes. SConscript: Move the function tracking and binning stuff around. arch/alpha/ev5.cc: kernelStats is now a pointer arch/alpha/pseudo_inst.cc: kernelStats is now a pointer the parameters to the system have been moved into their own struct base/trace.hh: provide a little functor class for wrapping a string that can allow you to define name() in any scope very simply for use with DPRINTF cpu/base_cpu.cc: New order of arguments for consistency. cpu/exec_context.cc: kernelStats no longer has the level of indirection in it, execContext has the indirection now. so, kernelStats is a pointer. We also need a pointer to the kernelBinning stuff from the system and we need to figure out if we want to do binning or not. Move a whole bunch of code into kern_binning.cc so it's all in the same place. cpu/exec_context.hh: We want pointers to the kernel binning/stats stuff and we'll have the exec_context and system have the level of indirection instead of having the extra layer in the kernel stats class. cpu/simple_cpu/simple_cpu.cc: call through the exec context to do the special binning stuff. kern/kernel_stats.cc: kern/kernel_stats.hh: Re-organize the stats stuff and remove the level of indirection (that was there to simplify building) and move the binning stuff into its own class/file. kern/linux/linux_system.cc: kern/linux/linux_system.hh: kern/tru64/tru64_system.cc: kern/tru64/tru64_system.hh: sim/system.cc: sim/system.hh: move lots of common system code into the base system class so that it can be shared between linux, tru64, and whatever else we decide to support in the future. Make the constructor take a pointer to a parameter struct so that it is easier to pass parameters to the parent. kern/system_events.cc: move the majority of the binning code into the Kernel::Binning class in the kern_binning file kern/system_events.hh: FnEvents only need to know the bin create the Idle start event to find the PCBB of the idle process when it starts. kern/tru64/tru64_events.cc: memCtrl -> memctrl sim/process.cc: sim/process.hh: re-order args for consistency --HG-- extra : convert_revision : 86cb39738c41fcd680f2aad125c9dde000227b2b
2004-08-20 17:35:31 +02:00
skipDelayLoopEvent = new LinuxSkipDelayLoopEvent(&pcEventQueue,
"calibrate_delay");
if (kernelSymtab->findAddress("calibrate_delay", addr))
skipDelayLoopEvent->schedule(addr+sizeof(MachInst));
- Clean up and factor out all of the binning code into a single place so it's easier to work with. - Add support for binning kernel/user/idle time separately from lisa's binning stuff, but make the two compatible. - KernelStats used to directly implement the pImpl idiom, but it makes more sense to just remove the level of indirection and make the exec context have a pointer to the stats. - Factor common code out of LinuxSystem and Tru64System and put it into the System base class. While doing that, make all constructors take a pointer to a parameter struct instead of naming the parameters individually to make it much easier to add parameters to these classes. SConscript: Move the function tracking and binning stuff around. arch/alpha/ev5.cc: kernelStats is now a pointer arch/alpha/pseudo_inst.cc: kernelStats is now a pointer the parameters to the system have been moved into their own struct base/trace.hh: provide a little functor class for wrapping a string that can allow you to define name() in any scope very simply for use with DPRINTF cpu/base_cpu.cc: New order of arguments for consistency. cpu/exec_context.cc: kernelStats no longer has the level of indirection in it, execContext has the indirection now. so, kernelStats is a pointer. We also need a pointer to the kernelBinning stuff from the system and we need to figure out if we want to do binning or not. Move a whole bunch of code into kern_binning.cc so it's all in the same place. cpu/exec_context.hh: We want pointers to the kernel binning/stats stuff and we'll have the exec_context and system have the level of indirection instead of having the extra layer in the kernel stats class. cpu/simple_cpu/simple_cpu.cc: call through the exec context to do the special binning stuff. kern/kernel_stats.cc: kern/kernel_stats.hh: Re-organize the stats stuff and remove the level of indirection (that was there to simplify building) and move the binning stuff into its own class/file. kern/linux/linux_system.cc: kern/linux/linux_system.hh: kern/tru64/tru64_system.cc: kern/tru64/tru64_system.hh: sim/system.cc: sim/system.hh: move lots of common system code into the base system class so that it can be shared between linux, tru64, and whatever else we decide to support in the future. Make the constructor take a pointer to a parameter struct so that it is easier to pass parameters to the parent. kern/system_events.cc: move the majority of the binning code into the Kernel::Binning class in the kern_binning file kern/system_events.hh: FnEvents only need to know the bin create the Idle start event to find the PCBB of the idle process when it starts. kern/tru64/tru64_events.cc: memCtrl -> memctrl sim/process.cc: sim/process.hh: re-order args for consistency --HG-- extra : convert_revision : 86cb39738c41fcd680f2aad125c9dde000227b2b
2004-08-20 17:35:31 +02:00
skipCacheProbeEvent = new SkipFuncEvent(&pcEventQueue,
"determine_cpu_caches");
if (kernelSymtab->findAddress("determine_cpu_caches", addr))
skipCacheProbeEvent->schedule(addr+sizeof(MachInst));
- Clean up and factor out all of the binning code into a single place so it's easier to work with. - Add support for binning kernel/user/idle time separately from lisa's binning stuff, but make the two compatible. - KernelStats used to directly implement the pImpl idiom, but it makes more sense to just remove the level of indirection and make the exec context have a pointer to the stats. - Factor common code out of LinuxSystem and Tru64System and put it into the System base class. While doing that, make all constructors take a pointer to a parameter struct instead of naming the parameters individually to make it much easier to add parameters to these classes. SConscript: Move the function tracking and binning stuff around. arch/alpha/ev5.cc: kernelStats is now a pointer arch/alpha/pseudo_inst.cc: kernelStats is now a pointer the parameters to the system have been moved into their own struct base/trace.hh: provide a little functor class for wrapping a string that can allow you to define name() in any scope very simply for use with DPRINTF cpu/base_cpu.cc: New order of arguments for consistency. cpu/exec_context.cc: kernelStats no longer has the level of indirection in it, execContext has the indirection now. so, kernelStats is a pointer. We also need a pointer to the kernelBinning stuff from the system and we need to figure out if we want to do binning or not. Move a whole bunch of code into kern_binning.cc so it's all in the same place. cpu/exec_context.hh: We want pointers to the kernel binning/stats stuff and we'll have the exec_context and system have the level of indirection instead of having the extra layer in the kernel stats class. cpu/simple_cpu/simple_cpu.cc: call through the exec context to do the special binning stuff. kern/kernel_stats.cc: kern/kernel_stats.hh: Re-organize the stats stuff and remove the level of indirection (that was there to simplify building) and move the binning stuff into its own class/file. kern/linux/linux_system.cc: kern/linux/linux_system.hh: kern/tru64/tru64_system.cc: kern/tru64/tru64_system.hh: sim/system.cc: sim/system.hh: move lots of common system code into the base system class so that it can be shared between linux, tru64, and whatever else we decide to support in the future. Make the constructor take a pointer to a parameter struct so that it is easier to pass parameters to the parent. kern/system_events.cc: move the majority of the binning code into the Kernel::Binning class in the kern_binning file kern/system_events.hh: FnEvents only need to know the bin create the Idle start event to find the PCBB of the idle process when it starts. kern/tru64/tru64_events.cc: memCtrl -> memctrl sim/process.cc: sim/process.hh: re-order args for consistency --HG-- extra : convert_revision : 86cb39738c41fcd680f2aad125c9dde000227b2b
2004-08-20 17:35:31 +02:00
debugPrintkEvent = new DebugPrintkEvent(&pcEventQueue, "dprintk");
if (kernelSymtab->findAddress("dprintk", addr))
- Clean up and factor out all of the binning code into a single place so it's easier to work with. - Add support for binning kernel/user/idle time separately from lisa's binning stuff, but make the two compatible. - KernelStats used to directly implement the pImpl idiom, but it makes more sense to just remove the level of indirection and make the exec context have a pointer to the stats. - Factor common code out of LinuxSystem and Tru64System and put it into the System base class. While doing that, make all constructors take a pointer to a parameter struct instead of naming the parameters individually to make it much easier to add parameters to these classes. SConscript: Move the function tracking and binning stuff around. arch/alpha/ev5.cc: kernelStats is now a pointer arch/alpha/pseudo_inst.cc: kernelStats is now a pointer the parameters to the system have been moved into their own struct base/trace.hh: provide a little functor class for wrapping a string that can allow you to define name() in any scope very simply for use with DPRINTF cpu/base_cpu.cc: New order of arguments for consistency. cpu/exec_context.cc: kernelStats no longer has the level of indirection in it, execContext has the indirection now. so, kernelStats is a pointer. We also need a pointer to the kernelBinning stuff from the system and we need to figure out if we want to do binning or not. Move a whole bunch of code into kern_binning.cc so it's all in the same place. cpu/exec_context.hh: We want pointers to the kernel binning/stats stuff and we'll have the exec_context and system have the level of indirection instead of having the extra layer in the kernel stats class. cpu/simple_cpu/simple_cpu.cc: call through the exec context to do the special binning stuff. kern/kernel_stats.cc: kern/kernel_stats.hh: Re-organize the stats stuff and remove the level of indirection (that was there to simplify building) and move the binning stuff into its own class/file. kern/linux/linux_system.cc: kern/linux/linux_system.hh: kern/tru64/tru64_system.cc: kern/tru64/tru64_system.hh: sim/system.cc: sim/system.hh: move lots of common system code into the base system class so that it can be shared between linux, tru64, and whatever else we decide to support in the future. Make the constructor take a pointer to a parameter struct so that it is easier to pass parameters to the parent. kern/system_events.cc: move the majority of the binning code into the Kernel::Binning class in the kern_binning file kern/system_events.hh: FnEvents only need to know the bin create the Idle start event to find the PCBB of the idle process when it starts. kern/tru64/tru64_events.cc: memCtrl -> memctrl sim/process.cc: sim/process.hh: re-order args for consistency --HG-- extra : convert_revision : 86cb39738c41fcd680f2aad125c9dde000227b2b
2004-08-20 17:35:31 +02:00
debugPrintkEvent->schedule(addr+8);
idleStartEvent = new IdleStartEvent(&pcEventQueue, "cpu_idle", this);
if (kernelSymtab->findAddress("cpu_idle", addr))
idleStartEvent->schedule(addr);
- Clean up and factor out all of the binning code into a single place so it's easier to work with. - Add support for binning kernel/user/idle time separately from lisa's binning stuff, but make the two compatible. - KernelStats used to directly implement the pImpl idiom, but it makes more sense to just remove the level of indirection and make the exec context have a pointer to the stats. - Factor common code out of LinuxSystem and Tru64System and put it into the System base class. While doing that, make all constructors take a pointer to a parameter struct instead of naming the parameters individually to make it much easier to add parameters to these classes. SConscript: Move the function tracking and binning stuff around. arch/alpha/ev5.cc: kernelStats is now a pointer arch/alpha/pseudo_inst.cc: kernelStats is now a pointer the parameters to the system have been moved into their own struct base/trace.hh: provide a little functor class for wrapping a string that can allow you to define name() in any scope very simply for use with DPRINTF cpu/base_cpu.cc: New order of arguments for consistency. cpu/exec_context.cc: kernelStats no longer has the level of indirection in it, execContext has the indirection now. so, kernelStats is a pointer. We also need a pointer to the kernelBinning stuff from the system and we need to figure out if we want to do binning or not. Move a whole bunch of code into kern_binning.cc so it's all in the same place. cpu/exec_context.hh: We want pointers to the kernel binning/stats stuff and we'll have the exec_context and system have the level of indirection instead of having the extra layer in the kernel stats class. cpu/simple_cpu/simple_cpu.cc: call through the exec context to do the special binning stuff. kern/kernel_stats.cc: kern/kernel_stats.hh: Re-organize the stats stuff and remove the level of indirection (that was there to simplify building) and move the binning stuff into its own class/file. kern/linux/linux_system.cc: kern/linux/linux_system.hh: kern/tru64/tru64_system.cc: kern/tru64/tru64_system.hh: sim/system.cc: sim/system.hh: move lots of common system code into the base system class so that it can be shared between linux, tru64, and whatever else we decide to support in the future. Make the constructor take a pointer to a parameter struct so that it is easier to pass parameters to the parent. kern/system_events.cc: move the majority of the binning code into the Kernel::Binning class in the kern_binning file kern/system_events.hh: FnEvents only need to know the bin create the Idle start event to find the PCBB of the idle process when it starts. kern/tru64/tru64_events.cc: memCtrl -> memctrl sim/process.cc: sim/process.hh: re-order args for consistency --HG-- extra : convert_revision : 86cb39738c41fcd680f2aad125c9dde000227b2b
2004-08-20 17:35:31 +02:00
printThreadEvent = new PrintThreadInfo(&pcEventQueue, "threadinfo");
if (kernelSymtab->findAddress("alpha_switch_to", addr) && DTRACE(Thread))
printThreadEvent->schedule(addr + sizeof(MachInst) * 6);
intStartEvent = new InterruptStartEvent(&pcEventQueue, "intStartEvent");
if (params->bin_int) {
if (palSymtab->findAddress("sys_int_21", addr))
intStartEvent->schedule(addr + sizeof(MachInst) * 2);
else
panic("could not find symbol: sys_int_21\n");
intEndEvent = new InterruptEndEvent(&pcEventQueue, "intEndEvent");
if (palSymtab->findAddress("rti_to_kern", addr))
intEndEvent->schedule(addr) ;
else
panic("could not find symbol: rti_to_kern\n");
intEndEvent2 = new InterruptEndEvent(&pcEventQueue, "intEndEvent2");
if (palSymtab->findAddress("rti_to_user", addr))
intEndEvent2->schedule(addr);
else
panic("could not find symbol: rti_to_user\n");
intEndEvent3 = new InterruptEndEvent(&pcEventQueue, "intEndEvent3");
if (kernelSymtab->findAddress("do_softirq", addr))
intEndEvent3->schedule(addr + sizeof(MachInst) * 2);
else
panic("could not find symbol: do_softirq\n");
}
}
LinuxSystem::~LinuxSystem()
{
#ifndef NDEBUG
delete kernelPanicEvent;
- Clean up and factor out all of the binning code into a single place so it's easier to work with. - Add support for binning kernel/user/idle time separately from lisa's binning stuff, but make the two compatible. - KernelStats used to directly implement the pImpl idiom, but it makes more sense to just remove the level of indirection and make the exec context have a pointer to the stats. - Factor common code out of LinuxSystem and Tru64System and put it into the System base class. While doing that, make all constructors take a pointer to a parameter struct instead of naming the parameters individually to make it much easier to add parameters to these classes. SConscript: Move the function tracking and binning stuff around. arch/alpha/ev5.cc: kernelStats is now a pointer arch/alpha/pseudo_inst.cc: kernelStats is now a pointer the parameters to the system have been moved into their own struct base/trace.hh: provide a little functor class for wrapping a string that can allow you to define name() in any scope very simply for use with DPRINTF cpu/base_cpu.cc: New order of arguments for consistency. cpu/exec_context.cc: kernelStats no longer has the level of indirection in it, execContext has the indirection now. so, kernelStats is a pointer. We also need a pointer to the kernelBinning stuff from the system and we need to figure out if we want to do binning or not. Move a whole bunch of code into kern_binning.cc so it's all in the same place. cpu/exec_context.hh: We want pointers to the kernel binning/stats stuff and we'll have the exec_context and system have the level of indirection instead of having the extra layer in the kernel stats class. cpu/simple_cpu/simple_cpu.cc: call through the exec context to do the special binning stuff. kern/kernel_stats.cc: kern/kernel_stats.hh: Re-organize the stats stuff and remove the level of indirection (that was there to simplify building) and move the binning stuff into its own class/file. kern/linux/linux_system.cc: kern/linux/linux_system.hh: kern/tru64/tru64_system.cc: kern/tru64/tru64_system.hh: sim/system.cc: sim/system.hh: move lots of common system code into the base system class so that it can be shared between linux, tru64, and whatever else we decide to support in the future. Make the constructor take a pointer to a parameter struct so that it is easier to pass parameters to the parent. kern/system_events.cc: move the majority of the binning code into the Kernel::Binning class in the kern_binning file kern/system_events.hh: FnEvents only need to know the bin create the Idle start event to find the PCBB of the idle process when it starts. kern/tru64/tru64_events.cc: memCtrl -> memctrl sim/process.cc: sim/process.hh: re-order args for consistency --HG-- extra : convert_revision : 86cb39738c41fcd680f2aad125c9dde000227b2b
2004-08-20 17:35:31 +02:00
#endif
delete skipIdeDelay50msEvent;
delete skipDelayLoopEvent;
delete skipCacheProbeEvent;
- Clean up and factor out all of the binning code into a single place so it's easier to work with. - Add support for binning kernel/user/idle time separately from lisa's binning stuff, but make the two compatible. - KernelStats used to directly implement the pImpl idiom, but it makes more sense to just remove the level of indirection and make the exec context have a pointer to the stats. - Factor common code out of LinuxSystem and Tru64System and put it into the System base class. While doing that, make all constructors take a pointer to a parameter struct instead of naming the parameters individually to make it much easier to add parameters to these classes. SConscript: Move the function tracking and binning stuff around. arch/alpha/ev5.cc: kernelStats is now a pointer arch/alpha/pseudo_inst.cc: kernelStats is now a pointer the parameters to the system have been moved into their own struct base/trace.hh: provide a little functor class for wrapping a string that can allow you to define name() in any scope very simply for use with DPRINTF cpu/base_cpu.cc: New order of arguments for consistency. cpu/exec_context.cc: kernelStats no longer has the level of indirection in it, execContext has the indirection now. so, kernelStats is a pointer. We also need a pointer to the kernelBinning stuff from the system and we need to figure out if we want to do binning or not. Move a whole bunch of code into kern_binning.cc so it's all in the same place. cpu/exec_context.hh: We want pointers to the kernel binning/stats stuff and we'll have the exec_context and system have the level of indirection instead of having the extra layer in the kernel stats class. cpu/simple_cpu/simple_cpu.cc: call through the exec context to do the special binning stuff. kern/kernel_stats.cc: kern/kernel_stats.hh: Re-organize the stats stuff and remove the level of indirection (that was there to simplify building) and move the binning stuff into its own class/file. kern/linux/linux_system.cc: kern/linux/linux_system.hh: kern/tru64/tru64_system.cc: kern/tru64/tru64_system.hh: sim/system.cc: sim/system.hh: move lots of common system code into the base system class so that it can be shared between linux, tru64, and whatever else we decide to support in the future. Make the constructor take a pointer to a parameter struct so that it is easier to pass parameters to the parent. kern/system_events.cc: move the majority of the binning code into the Kernel::Binning class in the kern_binning file kern/system_events.hh: FnEvents only need to know the bin create the Idle start event to find the PCBB of the idle process when it starts. kern/tru64/tru64_events.cc: memCtrl -> memctrl sim/process.cc: sim/process.hh: re-order args for consistency --HG-- extra : convert_revision : 86cb39738c41fcd680f2aad125c9dde000227b2b
2004-08-20 17:35:31 +02:00
delete debugPrintkEvent;
delete idleStartEvent;
delete printThreadEvent;
delete intStartEvent;
delete intEndEvent;
delete intEndEvent2;
}
void
LinuxSystem::setDelayLoop(ExecContext *xc)
{
Addr addr = 0;
if (kernelSymtab->findAddress("loops_per_jiffy", addr)) {
Addr paddr = vtophys(physmem, addr);
uint8_t *loops_per_jiffy =
physmem->dma_addr(paddr, sizeof(uint32_t));
Tick cpuFreq = xc->cpu->getFreq();
Tick intrFreq = platform->interrupt_frequency;
*(uint32_t *)loops_per_jiffy =
(uint32_t)((cpuFreq / intrFreq) * 0.9988);
}
}
BEGIN_DECLARE_SIM_OBJECT_PARAMS(LinuxSystem)
New and improved configuration mechanism. No more writing of wierd ini files. The ini files are still used as an intermediate step, but a sophisticated python library exists to help build them more easily. SConscript: add the new embedded file stuff remove all of the old object description junk base/inifile.cc: base/inifile.hh: get rid of findDefault and findAppend since they were the source of much evil. base/trace.cc: For now, if we don't have the dprintf_stream set up, dump to standard out. We probably want a command line option for this. dev/alpha_console.cc: PioDevice now takes a platform parameter. All PioDevices must have a pio_latency parameter. We stick a dummy parameter in here for now until we get rid of the builder stuff. dev/alpha_console.hh: don't need Platform anymore dev/baddev.cc: PioDevice now takes a platform parameter. All PioDevices must have a pio_latency parameter. We stick a dummy parameter in here for now until we get rid of the builder stuff. Same for the platform parameter, though we just pass the PioDevice a null parameter since it isn't used by this device and it's quicker. dev/baddev.hh: fix #include guards dev/etherlink.cc: rename parameters. dev/ethertap.cc: rename parameters dev/ide_ctrl.cc: All devices need an address even if it will get overwritten later. dev/ide_disk.cc: use an enum for the drive ID stuff. rename disk_delay -> delay Actually, I think that we should implement "cable select" and have the controller tell the drive what it is. dev/io_device.cc: dev/io_device.hh: All IO devices take a Platform * dev/ns_gige.cc: all devices need an io_bus. rename header_bus to io_bus We don't need stuff for the interrupt controller since it's all in the platform now. dev/ns_gige.hh: We don't need stuff for the interrupt controller now since it's all in the platform. dev/pciconfigall.cc: Pass a dummy NULL to the PioDevice for the platform since we don't need one. dev/pcidev.cc: Move a bunch of common functionality into the PciDev dev/platform.hh: remove unneeded code dev/tsunami.cc: remove unused param dev/tsunami_cchip.cc: pass platform pointer dev/tsunami_io.cc: dev/tsunami_pchip.cc: dev/uart.cc: pass platform variable dev/uart.hh: don't need to keep a platform pointer. it's in the base class kern/linux/linux_system.cc: kern/tru64/tru64_system.cc: rename some parameters sim/builder.cc: clean up builder code. use more parameters from the config node. all sections with a type= are now created, the old mechanisms no longer work sim/builder.hh: remove some extra variables since they are found in the ConfigNode sim/main.cc: add a quick hack command line argument -X to dump out the embedded files. (probably should be fixed up a little.) accept .mpy files printing to the streams has to happen after the hierarchy is built since we're moving away from param contexts sim/param.cc: add parsing support for ranges sim/process.cc: isValid isn't very useful anymore. interpret the names stdout, stderr, cout, cerr for the file descriptors sim/pyconfig/SConscript: Add Action handlers for creating an embedded python file and for creating an embedded C file. use these action handlers to embed all objects found in the objects tree into the binary along with the importer and the m5config stuff sim/pyconfig/m5config.py: Major changes to the original configuration file generator. These changes largely involve implementing copy-on-write like semantics for all of the SimObjects. Real documentation must be written. sim/universe.cc: Universe becomes a SimObject since we don't really have the notion of param contexts in the python code. --HG-- rename : sim/pyconfig/m5configbase.py => sim/pyconfig/m5config.py extra : convert_revision : c353453e5beb91c37f15755998fc0d8858c6829a
2005-01-15 10:12:25 +01:00
SimObjectParam<MemoryController *> memctrl;
SimObjectParam<PhysicalMemory *> physmem;
New and improved configuration mechanism. No more writing of wierd ini files. The ini files are still used as an intermediate step, but a sophisticated python library exists to help build them more easily. SConscript: add the new embedded file stuff remove all of the old object description junk base/inifile.cc: base/inifile.hh: get rid of findDefault and findAppend since they were the source of much evil. base/trace.cc: For now, if we don't have the dprintf_stream set up, dump to standard out. We probably want a command line option for this. dev/alpha_console.cc: PioDevice now takes a platform parameter. All PioDevices must have a pio_latency parameter. We stick a dummy parameter in here for now until we get rid of the builder stuff. dev/alpha_console.hh: don't need Platform anymore dev/baddev.cc: PioDevice now takes a platform parameter. All PioDevices must have a pio_latency parameter. We stick a dummy parameter in here for now until we get rid of the builder stuff. Same for the platform parameter, though we just pass the PioDevice a null parameter since it isn't used by this device and it's quicker. dev/baddev.hh: fix #include guards dev/etherlink.cc: rename parameters. dev/ethertap.cc: rename parameters dev/ide_ctrl.cc: All devices need an address even if it will get overwritten later. dev/ide_disk.cc: use an enum for the drive ID stuff. rename disk_delay -> delay Actually, I think that we should implement "cable select" and have the controller tell the drive what it is. dev/io_device.cc: dev/io_device.hh: All IO devices take a Platform * dev/ns_gige.cc: all devices need an io_bus. rename header_bus to io_bus We don't need stuff for the interrupt controller since it's all in the platform now. dev/ns_gige.hh: We don't need stuff for the interrupt controller now since it's all in the platform. dev/pciconfigall.cc: Pass a dummy NULL to the PioDevice for the platform since we don't need one. dev/pcidev.cc: Move a bunch of common functionality into the PciDev dev/platform.hh: remove unneeded code dev/tsunami.cc: remove unused param dev/tsunami_cchip.cc: pass platform pointer dev/tsunami_io.cc: dev/tsunami_pchip.cc: dev/uart.cc: pass platform variable dev/uart.hh: don't need to keep a platform pointer. it's in the base class kern/linux/linux_system.cc: kern/tru64/tru64_system.cc: rename some parameters sim/builder.cc: clean up builder code. use more parameters from the config node. all sections with a type= are now created, the old mechanisms no longer work sim/builder.hh: remove some extra variables since they are found in the ConfigNode sim/main.cc: add a quick hack command line argument -X to dump out the embedded files. (probably should be fixed up a little.) accept .mpy files printing to the streams has to happen after the hierarchy is built since we're moving away from param contexts sim/param.cc: add parsing support for ranges sim/process.cc: isValid isn't very useful anymore. interpret the names stdout, stderr, cout, cerr for the file descriptors sim/pyconfig/SConscript: Add Action handlers for creating an embedded python file and for creating an embedded C file. use these action handlers to embed all objects found in the objects tree into the binary along with the importer and the m5config stuff sim/pyconfig/m5config.py: Major changes to the original configuration file generator. These changes largely involve implementing copy-on-write like semantics for all of the SimObjects. Real documentation must be written. sim/universe.cc: Universe becomes a SimObject since we don't really have the notion of param contexts in the python code. --HG-- rename : sim/pyconfig/m5configbase.py => sim/pyconfig/m5config.py extra : convert_revision : c353453e5beb91c37f15755998fc0d8858c6829a
2005-01-15 10:12:25 +01:00
Param<string> kernel;
Param<string> console;
Param<string> pal;
- Clean up and factor out all of the binning code into a single place so it's easier to work with. - Add support for binning kernel/user/idle time separately from lisa's binning stuff, but make the two compatible. - KernelStats used to directly implement the pImpl idiom, but it makes more sense to just remove the level of indirection and make the exec context have a pointer to the stats. - Factor common code out of LinuxSystem and Tru64System and put it into the System base class. While doing that, make all constructors take a pointer to a parameter struct instead of naming the parameters individually to make it much easier to add parameters to these classes. SConscript: Move the function tracking and binning stuff around. arch/alpha/ev5.cc: kernelStats is now a pointer arch/alpha/pseudo_inst.cc: kernelStats is now a pointer the parameters to the system have been moved into their own struct base/trace.hh: provide a little functor class for wrapping a string that can allow you to define name() in any scope very simply for use with DPRINTF cpu/base_cpu.cc: New order of arguments for consistency. cpu/exec_context.cc: kernelStats no longer has the level of indirection in it, execContext has the indirection now. so, kernelStats is a pointer. We also need a pointer to the kernelBinning stuff from the system and we need to figure out if we want to do binning or not. Move a whole bunch of code into kern_binning.cc so it's all in the same place. cpu/exec_context.hh: We want pointers to the kernel binning/stats stuff and we'll have the exec_context and system have the level of indirection instead of having the extra layer in the kernel stats class. cpu/simple_cpu/simple_cpu.cc: call through the exec context to do the special binning stuff. kern/kernel_stats.cc: kern/kernel_stats.hh: Re-organize the stats stuff and remove the level of indirection (that was there to simplify building) and move the binning stuff into its own class/file. kern/linux/linux_system.cc: kern/linux/linux_system.hh: kern/tru64/tru64_system.cc: kern/tru64/tru64_system.hh: sim/system.cc: sim/system.hh: move lots of common system code into the base system class so that it can be shared between linux, tru64, and whatever else we decide to support in the future. Make the constructor take a pointer to a parameter struct so that it is easier to pass parameters to the parent. kern/system_events.cc: move the majority of the binning code into the Kernel::Binning class in the kern_binning file kern/system_events.hh: FnEvents only need to know the bin create the Idle start event to find the PCBB of the idle process when it starts. kern/tru64/tru64_events.cc: memCtrl -> memctrl sim/process.cc: sim/process.hh: re-order args for consistency --HG-- extra : convert_revision : 86cb39738c41fcd680f2aad125c9dde000227b2b
2004-08-20 17:35:31 +02:00
Param<string> boot_osflags;
Param<string> readfile;
- Clean up and factor out all of the binning code into a single place so it's easier to work with. - Add support for binning kernel/user/idle time separately from lisa's binning stuff, but make the two compatible. - KernelStats used to directly implement the pImpl idiom, but it makes more sense to just remove the level of indirection and make the exec context have a pointer to the stats. - Factor common code out of LinuxSystem and Tru64System and put it into the System base class. While doing that, make all constructors take a pointer to a parameter struct instead of naming the parameters individually to make it much easier to add parameters to these classes. SConscript: Move the function tracking and binning stuff around. arch/alpha/ev5.cc: kernelStats is now a pointer arch/alpha/pseudo_inst.cc: kernelStats is now a pointer the parameters to the system have been moved into their own struct base/trace.hh: provide a little functor class for wrapping a string that can allow you to define name() in any scope very simply for use with DPRINTF cpu/base_cpu.cc: New order of arguments for consistency. cpu/exec_context.cc: kernelStats no longer has the level of indirection in it, execContext has the indirection now. so, kernelStats is a pointer. We also need a pointer to the kernelBinning stuff from the system and we need to figure out if we want to do binning or not. Move a whole bunch of code into kern_binning.cc so it's all in the same place. cpu/exec_context.hh: We want pointers to the kernel binning/stats stuff and we'll have the exec_context and system have the level of indirection instead of having the extra layer in the kernel stats class. cpu/simple_cpu/simple_cpu.cc: call through the exec context to do the special binning stuff. kern/kernel_stats.cc: kern/kernel_stats.hh: Re-organize the stats stuff and remove the level of indirection (that was there to simplify building) and move the binning stuff into its own class/file. kern/linux/linux_system.cc: kern/linux/linux_system.hh: kern/tru64/tru64_system.cc: kern/tru64/tru64_system.hh: sim/system.cc: sim/system.hh: move lots of common system code into the base system class so that it can be shared between linux, tru64, and whatever else we decide to support in the future. Make the constructor take a pointer to a parameter struct so that it is easier to pass parameters to the parent. kern/system_events.cc: move the majority of the binning code into the Kernel::Binning class in the kern_binning file kern/system_events.hh: FnEvents only need to know the bin create the Idle start event to find the PCBB of the idle process when it starts. kern/tru64/tru64_events.cc: memCtrl -> memctrl sim/process.cc: sim/process.hh: re-order args for consistency --HG-- extra : convert_revision : 86cb39738c41fcd680f2aad125c9dde000227b2b
2004-08-20 17:35:31 +02:00
Param<unsigned int> init_param;
Param<uint64_t> system_type;
Param<uint64_t> system_rev;
Param<bool> bin;
VectorParam<string> binned_fns;
Param<bool> bin_int;
END_DECLARE_SIM_OBJECT_PARAMS(LinuxSystem)
BEGIN_INIT_SIM_OBJECT_PARAMS(LinuxSystem)
New and improved configuration mechanism. No more writing of wierd ini files. The ini files are still used as an intermediate step, but a sophisticated python library exists to help build them more easily. SConscript: add the new embedded file stuff remove all of the old object description junk base/inifile.cc: base/inifile.hh: get rid of findDefault and findAppend since they were the source of much evil. base/trace.cc: For now, if we don't have the dprintf_stream set up, dump to standard out. We probably want a command line option for this. dev/alpha_console.cc: PioDevice now takes a platform parameter. All PioDevices must have a pio_latency parameter. We stick a dummy parameter in here for now until we get rid of the builder stuff. dev/alpha_console.hh: don't need Platform anymore dev/baddev.cc: PioDevice now takes a platform parameter. All PioDevices must have a pio_latency parameter. We stick a dummy parameter in here for now until we get rid of the builder stuff. Same for the platform parameter, though we just pass the PioDevice a null parameter since it isn't used by this device and it's quicker. dev/baddev.hh: fix #include guards dev/etherlink.cc: rename parameters. dev/ethertap.cc: rename parameters dev/ide_ctrl.cc: All devices need an address even if it will get overwritten later. dev/ide_disk.cc: use an enum for the drive ID stuff. rename disk_delay -> delay Actually, I think that we should implement "cable select" and have the controller tell the drive what it is. dev/io_device.cc: dev/io_device.hh: All IO devices take a Platform * dev/ns_gige.cc: all devices need an io_bus. rename header_bus to io_bus We don't need stuff for the interrupt controller since it's all in the platform now. dev/ns_gige.hh: We don't need stuff for the interrupt controller now since it's all in the platform. dev/pciconfigall.cc: Pass a dummy NULL to the PioDevice for the platform since we don't need one. dev/pcidev.cc: Move a bunch of common functionality into the PciDev dev/platform.hh: remove unneeded code dev/tsunami.cc: remove unused param dev/tsunami_cchip.cc: pass platform pointer dev/tsunami_io.cc: dev/tsunami_pchip.cc: dev/uart.cc: pass platform variable dev/uart.hh: don't need to keep a platform pointer. it's in the base class kern/linux/linux_system.cc: kern/tru64/tru64_system.cc: rename some parameters sim/builder.cc: clean up builder code. use more parameters from the config node. all sections with a type= are now created, the old mechanisms no longer work sim/builder.hh: remove some extra variables since they are found in the ConfigNode sim/main.cc: add a quick hack command line argument -X to dump out the embedded files. (probably should be fixed up a little.) accept .mpy files printing to the streams has to happen after the hierarchy is built since we're moving away from param contexts sim/param.cc: add parsing support for ranges sim/process.cc: isValid isn't very useful anymore. interpret the names stdout, stderr, cout, cerr for the file descriptors sim/pyconfig/SConscript: Add Action handlers for creating an embedded python file and for creating an embedded C file. use these action handlers to embed all objects found in the objects tree into the binary along with the importer and the m5config stuff sim/pyconfig/m5config.py: Major changes to the original configuration file generator. These changes largely involve implementing copy-on-write like semantics for all of the SimObjects. Real documentation must be written. sim/universe.cc: Universe becomes a SimObject since we don't really have the notion of param contexts in the python code. --HG-- rename : sim/pyconfig/m5configbase.py => sim/pyconfig/m5config.py extra : convert_revision : c353453e5beb91c37f15755998fc0d8858c6829a
2005-01-15 10:12:25 +01:00
INIT_PARAM(memctrl, "memory controller"),
INIT_PARAM(physmem, "phsyical memory"),
New and improved configuration mechanism. No more writing of wierd ini files. The ini files are still used as an intermediate step, but a sophisticated python library exists to help build them more easily. SConscript: add the new embedded file stuff remove all of the old object description junk base/inifile.cc: base/inifile.hh: get rid of findDefault and findAppend since they were the source of much evil. base/trace.cc: For now, if we don't have the dprintf_stream set up, dump to standard out. We probably want a command line option for this. dev/alpha_console.cc: PioDevice now takes a platform parameter. All PioDevices must have a pio_latency parameter. We stick a dummy parameter in here for now until we get rid of the builder stuff. dev/alpha_console.hh: don't need Platform anymore dev/baddev.cc: PioDevice now takes a platform parameter. All PioDevices must have a pio_latency parameter. We stick a dummy parameter in here for now until we get rid of the builder stuff. Same for the platform parameter, though we just pass the PioDevice a null parameter since it isn't used by this device and it's quicker. dev/baddev.hh: fix #include guards dev/etherlink.cc: rename parameters. dev/ethertap.cc: rename parameters dev/ide_ctrl.cc: All devices need an address even if it will get overwritten later. dev/ide_disk.cc: use an enum for the drive ID stuff. rename disk_delay -> delay Actually, I think that we should implement "cable select" and have the controller tell the drive what it is. dev/io_device.cc: dev/io_device.hh: All IO devices take a Platform * dev/ns_gige.cc: all devices need an io_bus. rename header_bus to io_bus We don't need stuff for the interrupt controller since it's all in the platform now. dev/ns_gige.hh: We don't need stuff for the interrupt controller now since it's all in the platform. dev/pciconfigall.cc: Pass a dummy NULL to the PioDevice for the platform since we don't need one. dev/pcidev.cc: Move a bunch of common functionality into the PciDev dev/platform.hh: remove unneeded code dev/tsunami.cc: remove unused param dev/tsunami_cchip.cc: pass platform pointer dev/tsunami_io.cc: dev/tsunami_pchip.cc: dev/uart.cc: pass platform variable dev/uart.hh: don't need to keep a platform pointer. it's in the base class kern/linux/linux_system.cc: kern/tru64/tru64_system.cc: rename some parameters sim/builder.cc: clean up builder code. use more parameters from the config node. all sections with a type= are now created, the old mechanisms no longer work sim/builder.hh: remove some extra variables since they are found in the ConfigNode sim/main.cc: add a quick hack command line argument -X to dump out the embedded files. (probably should be fixed up a little.) accept .mpy files printing to the streams has to happen after the hierarchy is built since we're moving away from param contexts sim/param.cc: add parsing support for ranges sim/process.cc: isValid isn't very useful anymore. interpret the names stdout, stderr, cout, cerr for the file descriptors sim/pyconfig/SConscript: Add Action handlers for creating an embedded python file and for creating an embedded C file. use these action handlers to embed all objects found in the objects tree into the binary along with the importer and the m5config stuff sim/pyconfig/m5config.py: Major changes to the original configuration file generator. These changes largely involve implementing copy-on-write like semantics for all of the SimObjects. Real documentation must be written. sim/universe.cc: Universe becomes a SimObject since we don't really have the notion of param contexts in the python code. --HG-- rename : sim/pyconfig/m5configbase.py => sim/pyconfig/m5config.py extra : convert_revision : c353453e5beb91c37f15755998fc0d8858c6829a
2005-01-15 10:12:25 +01:00
INIT_PARAM(kernel, "file that contains the kernel code"),
INIT_PARAM(console, "file that contains the console code"),
INIT_PARAM(pal, "file that contains palcode"),
INIT_PARAM_DFLT(boot_osflags, "flags to pass to the kernel during boot",
- Clean up and factor out all of the binning code into a single place so it's easier to work with. - Add support for binning kernel/user/idle time separately from lisa's binning stuff, but make the two compatible. - KernelStats used to directly implement the pImpl idiom, but it makes more sense to just remove the level of indirection and make the exec context have a pointer to the stats. - Factor common code out of LinuxSystem and Tru64System and put it into the System base class. While doing that, make all constructors take a pointer to a parameter struct instead of naming the parameters individually to make it much easier to add parameters to these classes. SConscript: Move the function tracking and binning stuff around. arch/alpha/ev5.cc: kernelStats is now a pointer arch/alpha/pseudo_inst.cc: kernelStats is now a pointer the parameters to the system have been moved into their own struct base/trace.hh: provide a little functor class for wrapping a string that can allow you to define name() in any scope very simply for use with DPRINTF cpu/base_cpu.cc: New order of arguments for consistency. cpu/exec_context.cc: kernelStats no longer has the level of indirection in it, execContext has the indirection now. so, kernelStats is a pointer. We also need a pointer to the kernelBinning stuff from the system and we need to figure out if we want to do binning or not. Move a whole bunch of code into kern_binning.cc so it's all in the same place. cpu/exec_context.hh: We want pointers to the kernel binning/stats stuff and we'll have the exec_context and system have the level of indirection instead of having the extra layer in the kernel stats class. cpu/simple_cpu/simple_cpu.cc: call through the exec context to do the special binning stuff. kern/kernel_stats.cc: kern/kernel_stats.hh: Re-organize the stats stuff and remove the level of indirection (that was there to simplify building) and move the binning stuff into its own class/file. kern/linux/linux_system.cc: kern/linux/linux_system.hh: kern/tru64/tru64_system.cc: kern/tru64/tru64_system.hh: sim/system.cc: sim/system.hh: move lots of common system code into the base system class so that it can be shared between linux, tru64, and whatever else we decide to support in the future. Make the constructor take a pointer to a parameter struct so that it is easier to pass parameters to the parent. kern/system_events.cc: move the majority of the binning code into the Kernel::Binning class in the kern_binning file kern/system_events.hh: FnEvents only need to know the bin create the Idle start event to find the PCBB of the idle process when it starts. kern/tru64/tru64_events.cc: memCtrl -> memctrl sim/process.cc: sim/process.hh: re-order args for consistency --HG-- extra : convert_revision : 86cb39738c41fcd680f2aad125c9dde000227b2b
2004-08-20 17:35:31 +02:00
"a"),
INIT_PARAM_DFLT(readfile, "file to read startup script from", ""),
INIT_PARAM_DFLT(init_param, "numerical value to pass into simulator", 0),
INIT_PARAM_DFLT(system_type, "Type of system we are emulating", 34),
INIT_PARAM_DFLT(system_rev, "Revision of system we are emulating", 1<<10),
INIT_PARAM_DFLT(bin, "is this system to be binned", false),
INIT_PARAM(binned_fns, "functions to be broken down and binned"),
INIT_PARAM_DFLT(bin_int, "is interrupt code binned seperately?", true)
END_INIT_SIM_OBJECT_PARAMS(LinuxSystem)
CREATE_SIM_OBJECT(LinuxSystem)
{
- Clean up and factor out all of the binning code into a single place so it's easier to work with. - Add support for binning kernel/user/idle time separately from lisa's binning stuff, but make the two compatible. - KernelStats used to directly implement the pImpl idiom, but it makes more sense to just remove the level of indirection and make the exec context have a pointer to the stats. - Factor common code out of LinuxSystem and Tru64System and put it into the System base class. While doing that, make all constructors take a pointer to a parameter struct instead of naming the parameters individually to make it much easier to add parameters to these classes. SConscript: Move the function tracking and binning stuff around. arch/alpha/ev5.cc: kernelStats is now a pointer arch/alpha/pseudo_inst.cc: kernelStats is now a pointer the parameters to the system have been moved into their own struct base/trace.hh: provide a little functor class for wrapping a string that can allow you to define name() in any scope very simply for use with DPRINTF cpu/base_cpu.cc: New order of arguments for consistency. cpu/exec_context.cc: kernelStats no longer has the level of indirection in it, execContext has the indirection now. so, kernelStats is a pointer. We also need a pointer to the kernelBinning stuff from the system and we need to figure out if we want to do binning or not. Move a whole bunch of code into kern_binning.cc so it's all in the same place. cpu/exec_context.hh: We want pointers to the kernel binning/stats stuff and we'll have the exec_context and system have the level of indirection instead of having the extra layer in the kernel stats class. cpu/simple_cpu/simple_cpu.cc: call through the exec context to do the special binning stuff. kern/kernel_stats.cc: kern/kernel_stats.hh: Re-organize the stats stuff and remove the level of indirection (that was there to simplify building) and move the binning stuff into its own class/file. kern/linux/linux_system.cc: kern/linux/linux_system.hh: kern/tru64/tru64_system.cc: kern/tru64/tru64_system.hh: sim/system.cc: sim/system.hh: move lots of common system code into the base system class so that it can be shared between linux, tru64, and whatever else we decide to support in the future. Make the constructor take a pointer to a parameter struct so that it is easier to pass parameters to the parent. kern/system_events.cc: move the majority of the binning code into the Kernel::Binning class in the kern_binning file kern/system_events.hh: FnEvents only need to know the bin create the Idle start event to find the PCBB of the idle process when it starts. kern/tru64/tru64_events.cc: memCtrl -> memctrl sim/process.cc: sim/process.hh: re-order args for consistency --HG-- extra : convert_revision : 86cb39738c41fcd680f2aad125c9dde000227b2b
2004-08-20 17:35:31 +02:00
System::Params *p = new System::Params;
p->name = getInstanceName();
New and improved configuration mechanism. No more writing of wierd ini files. The ini files are still used as an intermediate step, but a sophisticated python library exists to help build them more easily. SConscript: add the new embedded file stuff remove all of the old object description junk base/inifile.cc: base/inifile.hh: get rid of findDefault and findAppend since they were the source of much evil. base/trace.cc: For now, if we don't have the dprintf_stream set up, dump to standard out. We probably want a command line option for this. dev/alpha_console.cc: PioDevice now takes a platform parameter. All PioDevices must have a pio_latency parameter. We stick a dummy parameter in here for now until we get rid of the builder stuff. dev/alpha_console.hh: don't need Platform anymore dev/baddev.cc: PioDevice now takes a platform parameter. All PioDevices must have a pio_latency parameter. We stick a dummy parameter in here for now until we get rid of the builder stuff. Same for the platform parameter, though we just pass the PioDevice a null parameter since it isn't used by this device and it's quicker. dev/baddev.hh: fix #include guards dev/etherlink.cc: rename parameters. dev/ethertap.cc: rename parameters dev/ide_ctrl.cc: All devices need an address even if it will get overwritten later. dev/ide_disk.cc: use an enum for the drive ID stuff. rename disk_delay -> delay Actually, I think that we should implement "cable select" and have the controller tell the drive what it is. dev/io_device.cc: dev/io_device.hh: All IO devices take a Platform * dev/ns_gige.cc: all devices need an io_bus. rename header_bus to io_bus We don't need stuff for the interrupt controller since it's all in the platform now. dev/ns_gige.hh: We don't need stuff for the interrupt controller now since it's all in the platform. dev/pciconfigall.cc: Pass a dummy NULL to the PioDevice for the platform since we don't need one. dev/pcidev.cc: Move a bunch of common functionality into the PciDev dev/platform.hh: remove unneeded code dev/tsunami.cc: remove unused param dev/tsunami_cchip.cc: pass platform pointer dev/tsunami_io.cc: dev/tsunami_pchip.cc: dev/uart.cc: pass platform variable dev/uart.hh: don't need to keep a platform pointer. it's in the base class kern/linux/linux_system.cc: kern/tru64/tru64_system.cc: rename some parameters sim/builder.cc: clean up builder code. use more parameters from the config node. all sections with a type= are now created, the old mechanisms no longer work sim/builder.hh: remove some extra variables since they are found in the ConfigNode sim/main.cc: add a quick hack command line argument -X to dump out the embedded files. (probably should be fixed up a little.) accept .mpy files printing to the streams has to happen after the hierarchy is built since we're moving away from param contexts sim/param.cc: add parsing support for ranges sim/process.cc: isValid isn't very useful anymore. interpret the names stdout, stderr, cout, cerr for the file descriptors sim/pyconfig/SConscript: Add Action handlers for creating an embedded python file and for creating an embedded C file. use these action handlers to embed all objects found in the objects tree into the binary along with the importer and the m5config stuff sim/pyconfig/m5config.py: Major changes to the original configuration file generator. These changes largely involve implementing copy-on-write like semantics for all of the SimObjects. Real documentation must be written. sim/universe.cc: Universe becomes a SimObject since we don't really have the notion of param contexts in the python code. --HG-- rename : sim/pyconfig/m5configbase.py => sim/pyconfig/m5config.py extra : convert_revision : c353453e5beb91c37f15755998fc0d8858c6829a
2005-01-15 10:12:25 +01:00
p->memctrl = memctrl;
- Clean up and factor out all of the binning code into a single place so it's easier to work with. - Add support for binning kernel/user/idle time separately from lisa's binning stuff, but make the two compatible. - KernelStats used to directly implement the pImpl idiom, but it makes more sense to just remove the level of indirection and make the exec context have a pointer to the stats. - Factor common code out of LinuxSystem and Tru64System and put it into the System base class. While doing that, make all constructors take a pointer to a parameter struct instead of naming the parameters individually to make it much easier to add parameters to these classes. SConscript: Move the function tracking and binning stuff around. arch/alpha/ev5.cc: kernelStats is now a pointer arch/alpha/pseudo_inst.cc: kernelStats is now a pointer the parameters to the system have been moved into their own struct base/trace.hh: provide a little functor class for wrapping a string that can allow you to define name() in any scope very simply for use with DPRINTF cpu/base_cpu.cc: New order of arguments for consistency. cpu/exec_context.cc: kernelStats no longer has the level of indirection in it, execContext has the indirection now. so, kernelStats is a pointer. We also need a pointer to the kernelBinning stuff from the system and we need to figure out if we want to do binning or not. Move a whole bunch of code into kern_binning.cc so it's all in the same place. cpu/exec_context.hh: We want pointers to the kernel binning/stats stuff and we'll have the exec_context and system have the level of indirection instead of having the extra layer in the kernel stats class. cpu/simple_cpu/simple_cpu.cc: call through the exec context to do the special binning stuff. kern/kernel_stats.cc: kern/kernel_stats.hh: Re-organize the stats stuff and remove the level of indirection (that was there to simplify building) and move the binning stuff into its own class/file. kern/linux/linux_system.cc: kern/linux/linux_system.hh: kern/tru64/tru64_system.cc: kern/tru64/tru64_system.hh: sim/system.cc: sim/system.hh: move lots of common system code into the base system class so that it can be shared between linux, tru64, and whatever else we decide to support in the future. Make the constructor take a pointer to a parameter struct so that it is easier to pass parameters to the parent. kern/system_events.cc: move the majority of the binning code into the Kernel::Binning class in the kern_binning file kern/system_events.hh: FnEvents only need to know the bin create the Idle start event to find the PCBB of the idle process when it starts. kern/tru64/tru64_events.cc: memCtrl -> memctrl sim/process.cc: sim/process.hh: re-order args for consistency --HG-- extra : convert_revision : 86cb39738c41fcd680f2aad125c9dde000227b2b
2004-08-20 17:35:31 +02:00
p->physmem = physmem;
New and improved configuration mechanism. No more writing of wierd ini files. The ini files are still used as an intermediate step, but a sophisticated python library exists to help build them more easily. SConscript: add the new embedded file stuff remove all of the old object description junk base/inifile.cc: base/inifile.hh: get rid of findDefault and findAppend since they were the source of much evil. base/trace.cc: For now, if we don't have the dprintf_stream set up, dump to standard out. We probably want a command line option for this. dev/alpha_console.cc: PioDevice now takes a platform parameter. All PioDevices must have a pio_latency parameter. We stick a dummy parameter in here for now until we get rid of the builder stuff. dev/alpha_console.hh: don't need Platform anymore dev/baddev.cc: PioDevice now takes a platform parameter. All PioDevices must have a pio_latency parameter. We stick a dummy parameter in here for now until we get rid of the builder stuff. Same for the platform parameter, though we just pass the PioDevice a null parameter since it isn't used by this device and it's quicker. dev/baddev.hh: fix #include guards dev/etherlink.cc: rename parameters. dev/ethertap.cc: rename parameters dev/ide_ctrl.cc: All devices need an address even if it will get overwritten later. dev/ide_disk.cc: use an enum for the drive ID stuff. rename disk_delay -> delay Actually, I think that we should implement "cable select" and have the controller tell the drive what it is. dev/io_device.cc: dev/io_device.hh: All IO devices take a Platform * dev/ns_gige.cc: all devices need an io_bus. rename header_bus to io_bus We don't need stuff for the interrupt controller since it's all in the platform now. dev/ns_gige.hh: We don't need stuff for the interrupt controller now since it's all in the platform. dev/pciconfigall.cc: Pass a dummy NULL to the PioDevice for the platform since we don't need one. dev/pcidev.cc: Move a bunch of common functionality into the PciDev dev/platform.hh: remove unneeded code dev/tsunami.cc: remove unused param dev/tsunami_cchip.cc: pass platform pointer dev/tsunami_io.cc: dev/tsunami_pchip.cc: dev/uart.cc: pass platform variable dev/uart.hh: don't need to keep a platform pointer. it's in the base class kern/linux/linux_system.cc: kern/tru64/tru64_system.cc: rename some parameters sim/builder.cc: clean up builder code. use more parameters from the config node. all sections with a type= are now created, the old mechanisms no longer work sim/builder.hh: remove some extra variables since they are found in the ConfigNode sim/main.cc: add a quick hack command line argument -X to dump out the embedded files. (probably should be fixed up a little.) accept .mpy files printing to the streams has to happen after the hierarchy is built since we're moving away from param contexts sim/param.cc: add parsing support for ranges sim/process.cc: isValid isn't very useful anymore. interpret the names stdout, stderr, cout, cerr for the file descriptors sim/pyconfig/SConscript: Add Action handlers for creating an embedded python file and for creating an embedded C file. use these action handlers to embed all objects found in the objects tree into the binary along with the importer and the m5config stuff sim/pyconfig/m5config.py: Major changes to the original configuration file generator. These changes largely involve implementing copy-on-write like semantics for all of the SimObjects. Real documentation must be written. sim/universe.cc: Universe becomes a SimObject since we don't really have the notion of param contexts in the python code. --HG-- rename : sim/pyconfig/m5configbase.py => sim/pyconfig/m5config.py extra : convert_revision : c353453e5beb91c37f15755998fc0d8858c6829a
2005-01-15 10:12:25 +01:00
p->kernel_path = kernel;
p->console_path = console;
p->palcode = pal;
- Clean up and factor out all of the binning code into a single place so it's easier to work with. - Add support for binning kernel/user/idle time separately from lisa's binning stuff, but make the two compatible. - KernelStats used to directly implement the pImpl idiom, but it makes more sense to just remove the level of indirection and make the exec context have a pointer to the stats. - Factor common code out of LinuxSystem and Tru64System and put it into the System base class. While doing that, make all constructors take a pointer to a parameter struct instead of naming the parameters individually to make it much easier to add parameters to these classes. SConscript: Move the function tracking and binning stuff around. arch/alpha/ev5.cc: kernelStats is now a pointer arch/alpha/pseudo_inst.cc: kernelStats is now a pointer the parameters to the system have been moved into their own struct base/trace.hh: provide a little functor class for wrapping a string that can allow you to define name() in any scope very simply for use with DPRINTF cpu/base_cpu.cc: New order of arguments for consistency. cpu/exec_context.cc: kernelStats no longer has the level of indirection in it, execContext has the indirection now. so, kernelStats is a pointer. We also need a pointer to the kernelBinning stuff from the system and we need to figure out if we want to do binning or not. Move a whole bunch of code into kern_binning.cc so it's all in the same place. cpu/exec_context.hh: We want pointers to the kernel binning/stats stuff and we'll have the exec_context and system have the level of indirection instead of having the extra layer in the kernel stats class. cpu/simple_cpu/simple_cpu.cc: call through the exec context to do the special binning stuff. kern/kernel_stats.cc: kern/kernel_stats.hh: Re-organize the stats stuff and remove the level of indirection (that was there to simplify building) and move the binning stuff into its own class/file. kern/linux/linux_system.cc: kern/linux/linux_system.hh: kern/tru64/tru64_system.cc: kern/tru64/tru64_system.hh: sim/system.cc: sim/system.hh: move lots of common system code into the base system class so that it can be shared between linux, tru64, and whatever else we decide to support in the future. Make the constructor take a pointer to a parameter struct so that it is easier to pass parameters to the parent. kern/system_events.cc: move the majority of the binning code into the Kernel::Binning class in the kern_binning file kern/system_events.hh: FnEvents only need to know the bin create the Idle start event to find the PCBB of the idle process when it starts. kern/tru64/tru64_events.cc: memCtrl -> memctrl sim/process.cc: sim/process.hh: re-order args for consistency --HG-- extra : convert_revision : 86cb39738c41fcd680f2aad125c9dde000227b2b
2004-08-20 17:35:31 +02:00
p->boot_osflags = boot_osflags;
p->init_param = init_param;
p->readfile = readfile;
p->system_type = system_type;
p->system_rev = system_rev;
p->bin = bin;
p->binned_fns = binned_fns;
p->bin_int = bin_int;
- Clean up and factor out all of the binning code into a single place so it's easier to work with. - Add support for binning kernel/user/idle time separately from lisa's binning stuff, but make the two compatible. - KernelStats used to directly implement the pImpl idiom, but it makes more sense to just remove the level of indirection and make the exec context have a pointer to the stats. - Factor common code out of LinuxSystem and Tru64System and put it into the System base class. While doing that, make all constructors take a pointer to a parameter struct instead of naming the parameters individually to make it much easier to add parameters to these classes. SConscript: Move the function tracking and binning stuff around. arch/alpha/ev5.cc: kernelStats is now a pointer arch/alpha/pseudo_inst.cc: kernelStats is now a pointer the parameters to the system have been moved into their own struct base/trace.hh: provide a little functor class for wrapping a string that can allow you to define name() in any scope very simply for use with DPRINTF cpu/base_cpu.cc: New order of arguments for consistency. cpu/exec_context.cc: kernelStats no longer has the level of indirection in it, execContext has the indirection now. so, kernelStats is a pointer. We also need a pointer to the kernelBinning stuff from the system and we need to figure out if we want to do binning or not. Move a whole bunch of code into kern_binning.cc so it's all in the same place. cpu/exec_context.hh: We want pointers to the kernel binning/stats stuff and we'll have the exec_context and system have the level of indirection instead of having the extra layer in the kernel stats class. cpu/simple_cpu/simple_cpu.cc: call through the exec context to do the special binning stuff. kern/kernel_stats.cc: kern/kernel_stats.hh: Re-organize the stats stuff and remove the level of indirection (that was there to simplify building) and move the binning stuff into its own class/file. kern/linux/linux_system.cc: kern/linux/linux_system.hh: kern/tru64/tru64_system.cc: kern/tru64/tru64_system.hh: sim/system.cc: sim/system.hh: move lots of common system code into the base system class so that it can be shared between linux, tru64, and whatever else we decide to support in the future. Make the constructor take a pointer to a parameter struct so that it is easier to pass parameters to the parent. kern/system_events.cc: move the majority of the binning code into the Kernel::Binning class in the kern_binning file kern/system_events.hh: FnEvents only need to know the bin create the Idle start event to find the PCBB of the idle process when it starts. kern/tru64/tru64_events.cc: memCtrl -> memctrl sim/process.cc: sim/process.hh: re-order args for consistency --HG-- extra : convert_revision : 86cb39738c41fcd680f2aad125c9dde000227b2b
2004-08-20 17:35:31 +02:00
return new LinuxSystem(p);
}
REGISTER_SIM_OBJECT("LinuxSystem", LinuxSystem)
- Clean up and factor out all of the binning code into a single place so it's easier to work with. - Add support for binning kernel/user/idle time separately from lisa's binning stuff, but make the two compatible. - KernelStats used to directly implement the pImpl idiom, but it makes more sense to just remove the level of indirection and make the exec context have a pointer to the stats. - Factor common code out of LinuxSystem and Tru64System and put it into the System base class. While doing that, make all constructors take a pointer to a parameter struct instead of naming the parameters individually to make it much easier to add parameters to these classes. SConscript: Move the function tracking and binning stuff around. arch/alpha/ev5.cc: kernelStats is now a pointer arch/alpha/pseudo_inst.cc: kernelStats is now a pointer the parameters to the system have been moved into their own struct base/trace.hh: provide a little functor class for wrapping a string that can allow you to define name() in any scope very simply for use with DPRINTF cpu/base_cpu.cc: New order of arguments for consistency. cpu/exec_context.cc: kernelStats no longer has the level of indirection in it, execContext has the indirection now. so, kernelStats is a pointer. We also need a pointer to the kernelBinning stuff from the system and we need to figure out if we want to do binning or not. Move a whole bunch of code into kern_binning.cc so it's all in the same place. cpu/exec_context.hh: We want pointers to the kernel binning/stats stuff and we'll have the exec_context and system have the level of indirection instead of having the extra layer in the kernel stats class. cpu/simple_cpu/simple_cpu.cc: call through the exec context to do the special binning stuff. kern/kernel_stats.cc: kern/kernel_stats.hh: Re-organize the stats stuff and remove the level of indirection (that was there to simplify building) and move the binning stuff into its own class/file. kern/linux/linux_system.cc: kern/linux/linux_system.hh: kern/tru64/tru64_system.cc: kern/tru64/tru64_system.hh: sim/system.cc: sim/system.hh: move lots of common system code into the base system class so that it can be shared between linux, tru64, and whatever else we decide to support in the future. Make the constructor take a pointer to a parameter struct so that it is easier to pass parameters to the parent. kern/system_events.cc: move the majority of the binning code into the Kernel::Binning class in the kern_binning file kern/system_events.hh: FnEvents only need to know the bin create the Idle start event to find the PCBB of the idle process when it starts. kern/tru64/tru64_events.cc: memCtrl -> memctrl sim/process.cc: sim/process.hh: re-order args for consistency --HG-- extra : convert_revision : 86cb39738c41fcd680f2aad125c9dde000227b2b
2004-08-20 17:35:31 +02:00