the GPUISA class is meant to encapsulate any ISA-specific behavior - special
register accesses, isa-specific WF/kernel state, etc. - in a generic enough
way so that it may be used in ISA-agnostic code.
gpu-compute: use the GPUISA object to advance the PC
the GPU model treats the PC as a pointer to individual instruction objects -
which are store in a contiguous array - and not a byte address to be fetched
from the real memory system. this is ok for HSAIL because all instructions
are considered by the model to be the same size.
in machine ISA, however, instructions may be 32b or 64b, and branches are
calculated by advancing the PC by the number of words (4 byte chunks) it
needs to advance in the real instruction stream. because of this there is
a mismatch between the PC we use to index into the instruction array, and
the actual byte address PC the ISA expects. here we move the PC advance
calculation to the ISA so that differences in the instrucion sizes may be
accounted for in generic way.
This adds a vector register type. The type is defined as a std::array of a
fixed number of uint64_ts. The isa_parser.py has been modified to parse vector
register operands and generate the required code. Different cpus have vector
register files now.
This patch adds methods in KvmCPU model to handle KVM exits caused by syscall
instructions and page faults. These types of exits will be encountered if
KvmCPU is run in SE mode.
We currently generate and compile one version of the ISA code per CPU
model. This is obviously wasting a lot of resources at compile
time. This changeset factors out the interface into a separate
ExecContext class, which also serves as documentation for the
interface between CPUs and the ISA code. While doing so, this
changeset also fixes up interface inconsistencies between the
different CPU models.
The main argument for using one set of ISA code per CPU model has
always been performance as this avoid indirect branches in the
generated code. However, this argument does not hold water. Booting
Linux on a simulated ARM system running in atomic mode
(opt/10.linux-boot/realview-simple-atomic) is actually 2% faster
(compiled using clang 3.4) after applying this patch. Additionally,
compilation time is decreased by 35%.
This patch encompasses several interrelated and interdependent changes
to the ISA generation step. The end goal is to reduce the size of the
generated compilation units for instruction execution and decoding so
that batch compilation can proceed with all CPUs active without
exhausting physical memory.
The ISA parser (src/arch/isa_parser.py) has been improved so that it can
accept 'split [output_type];' directives at the top level of the grammar
and 'split(output_type)' python calls within 'exec {{ ... }}' blocks.
This has the effect of "splitting" the files into smaller compilation
units. I use air-quotes around "splitting" because the files themselves
are not split, but preprocessing directives are inserted to have the same
effect.
Architecturally, the ISA parser has had some changes in how it works.
In general, it emits code sooner. It doesn't generate per-CPU files,
and instead defers to the C preprocessor to create the duplicate copies
for each CPU type. Likewise there are more files emitted and the C
preprocessor does more substitution that used to be done by the ISA parser.
Finally, the build system (SCons) needs to be able to cope with a
dynamic list of source files coming out of the ISA parser. The changes
to the SCons{cript,truct} files support this. In broad strokes, the
targets requested on the command line are hidden from SCons until all
the build dependencies are determined, otherwise it would try, realize
it can't reach the goal, and terminate in failure. Since build steps
(i.e. running the ISA parser) must be taken to determine the file list,
several new build stages have been inserted at the very start of the
build. First, the build dependencies from the ISA parser will be emitted
to arch/$ISA/generated/inc.d, which is then read by a new SCons builder
to finalize the dependencies. (Once inc.d exists, the ISA parser will not
need to be run to complete this step.) Once the dependencies are known,
the 'Environments' are made by the makeEnv() function. This function used
to be called before the build began but now happens during the build.
It is easy to see that this step is quite slow; this is a known issue
and it's important to realize that it was already slow, but there was
no obvious cause to attribute it to since nothing was displayed to the
terminal. Since new steps that used to be performed serially are now in a
potentially-parallel build phase, the pathname handling in the SCons scripts
has been tightened up to deal with chdir() race conditions. In general,
pathnames are computed earlier and more likely to be stored, passed around,
and processed as absolute paths rather than relative paths. In the end,
some of these issues had to be fixed by inserting serializing dependencies
in the build.
Minor note:
For the null ISA, we just provide a dummy inc.d so SCons is never
compelled to try to generate it. While it seems slightly wrong to have
anything in src/arch/*/generated (i.e. a non-generated 'generated' file),
it's by far the simplest solution.
These classes are always used together, and merging them will give the ISAs
more flexibility in how they cache things and manage the process.
--HG--
rename : src/arch/x86/predecoder_tables.cc => src/arch/x86/decoder_tables.cc
Enables the CheckerCPU to be selected at runtime with the --checker option
from the configs/example/fs.py and configs/example/se.py configuration
files. Also merges with the SE/FS changes.
Because there are no longer architecture independent but specialized functions
in arch/XXX/faults.hh, code that isn't using the faults from a particular ISA
no longer needs to be able to include them through the switching header file
arch/faults.hh. By removing that header file (arch/faults.hh), the potential
interface between ISA code and non ISA code is narrowed.
There was a change a while ago that refactored some scons stuff which got rid
of cpu_models.py but also accidentally got rid of the ISA parser as a source
for its target files. That meant that changes which affected the parser
wouldn't cause a rebuild unless they also changed one of the description
files. This change fixes that.
There may not be a formally correct spelling for the past tense of mmap, but
mmapped is the spelling Google doesn't try to autocorrect. This makes sense
because it mirrors the past tense of map->mapped and not the past tense of
cape->caped.
--HG--
rename : src/arch/alpha/mmaped_ipr.hh => src/arch/alpha/mmapped_ipr.hh
rename : src/arch/arm/mmaped_ipr.hh => src/arch/arm/mmapped_ipr.hh
rename : src/arch/mips/mmaped_ipr.hh => src/arch/mips/mmapped_ipr.hh
rename : src/arch/power/mmaped_ipr.hh => src/arch/power/mmapped_ipr.hh
rename : src/arch/sparc/mmaped_ipr.hh => src/arch/sparc/mmapped_ipr.hh
rename : src/arch/x86/mmaped_ipr.hh => src/arch/x86/mmapped_ipr.hh
I like the brevity of Ali's recent change, but the ambiguity of
sometimes showing the source and sometimes the target is a little
confusing. This patch makes scons typically list all sources and
all targets for each action, with the common path prefix factored
out for brevity. It's a little more verbose now but also more
informative.
Somehow Ali talked me into adding colors too, which is a whole
'nother story.
Edit AlphaISA to support the inorder model. Mostly alternate constructor functions and also a few skeleton multithreaded support functions
* * *
Remove namespace from header file. Causes compiler issues that are hard to find
* * *
Separate the TLB from the CPU and allow it to live in the TLBUnit resource. Give CPU accessor functions for access and also bind at construction time
* * *
Expose memory access size and flags through instruction object
(temporarily memAccSize and memFlags to get TLB stuff working.)
into ahchoo.blinky.homelinux.org:/home/gblack/m5/newmem-x86
src/arch/mips/utility.hh:
src/arch/x86/SConscript:
Hand merge
--HG--
extra : convert_revision : 0ba457aab52bf6ffc9191fd1fe1006ca7704b5b0
automatic. The point is that now a subdirectory can be added
to the build process just by creating a SConscript file in it.
The process has two passes. On the first pass, all subdirs
of the root of the tree are searched for SConsopts files.
These files contain any command line options that ought to be
added for a particular subdirectory. On the second pass,
all subdirs of the src directory are searched for SConscript
files. These files describe how to build any given subdirectory.
I have added a Source() function. Any file (relative to the
directory in which the SConscript resides) passed to that
function is added to the build. Clean up everything to take
advantage of Source().
function is added to the list of files to be built.
--HG--
extra : convert_revision : 103f6b490d2eb224436688c89cdc015211c4fd30
src/arch/SConscript:
add mmaped_ipr.hh to switch headers
src/arch/sparc/asi.hh:
make ASI_IMPLICT=0 so by default nothing needs to be done
src/arch/sparc/miscregfile.hh:
miscregfile no longer needs to include asi.hh
src/arch/sparc/tlb.cc:
src/arch/sparc/tlb.hh:
implement panic instructions for mmaped ipr reads
src/cpu/simple/atomic.cc:
add check for mmaped iprs and handle them if it exists
src/mem/request.hh:
allocate space in the flags for mmaped iprs. Put in in the first 8 bits so that by default its fast. Move the other flags up 8 bits
--HG--
extra : convert_revision : 31255b0494588c4d06a727fe35241121d741b115
SConstruct:
Put the code to make a switching header directory into a function so they are easy to make.
src/arch/SConscript:
Replace switching header code with the new function call.
src/kern/SConscript:
Created a new switching header directory in kern, and moved the declaration of some source files here.
--HG--
rename : src/kern/kernel_stats.cc => src/kern/base_kernel_stats.cc
rename : src/kern/kernel_stats.hh => src/kern/base_kernel_stats.hh
extra : convert_revision : 98f5320a5ade567c3e4f67fef123dfb0c5122545
and PhysicalMemory. *No* support for caches or O3CPU.
Note that properly setting cpu_id on all CPUs is now required
for correct operation.
src/arch/SConscript:
src/base/traceflags.py:
src/cpu/base.hh:
src/cpu/simple/atomic.cc:
src/cpu/simple/timing.cc:
src/cpu/simple/timing.hh:
src/mem/physical.cc:
src/mem/physical.hh:
src/mem/request.hh:
src/python/m5/objects/BaseCPU.py:
tests/configs/simple-atomic.py:
tests/configs/simple-timing.py:
tests/configs/tsunami-simple-atomic-dual.py:
tests/configs/tsunami-simple-atomic.py:
tests/configs/tsunami-simple-timing-dual.py:
tests/configs/tsunami-simple-timing.py:
Implement Alpha LL/SC support for SimpleCPU (Atomic & Timing)
and PhysicalMemory. *No* support for caches or O3CPU.
--HG--
extra : convert_revision : 6ce982d44924cc477e049b9adf359818908e72be
src/arch/SConscript:
Fixes for including the CheckerCPU if it's specified via command line. Previously the env variable was actually being modified.
src/cpu/SConscript:
Copy the CPU_MODELS from the env, don't create a proxy to it.
--HG--
extra : convert_revision : 7d069bd93a6834ccaa1c378b2bc76dce76745c19
SConstruct:
Remove check for Checker from this SConstruct
src/arch/SConscript:
Specific check if CheckerCPU is being used. Not the cleanest, but works for now.
src/cpu/SConscript:
Code to handle using the CheckerCPU a little better. Allows -c to be used normally.
--HG--
extra : convert_revision : 0a82f16db0f38e5ce114d08368477bd211331fa3
Also delete no-longer-needed SPARC test.py files (should
have gone with my last chnageset where LiveProcess
became ISA-independent).
src/arch/SConscript:
Mistakenly committed hard-coded python binary name.
Should use same python to run isa_parser that was
used to run scons.
--HG--
extra : convert_revision : a3acd5496f3b930d57bb59ae164b8a4a8065bbf8
User script now invokes initialization and
simulation loop after building configuration.
These functions are exported from C++ to Python
using SWIG.
SConstruct:
Set up SWIG builder & scanner.
Set up symlinking of source files into build directory
(by not disabling the default behavior).
configs/test/test.py:
Rewrite to use new script-driven interface.
Include a sample option.
src/SConscript:
Set up symlinking of source files into build directory
(by not disabling the default behavior).
Add SWIG-generated main_wrap.cc to source list.
src/arch/SConscript:
Set up symlinking of source files into build directory
(by not disabling the default behavior).
src/arch/alpha/ev5.cc:
src/arch/alpha/isa/decoder.isa:
src/cpu/o3/alpha_cpu_impl.hh:
src/cpu/trace/opt_cpu.cc:
src/cpu/trace/trace_cpu.cc:
src/sim/pseudo_inst.cc:
src/sim/root.cc:
src/sim/serialize.cc:
src/sim/syscall_emul.cc:
SimExit() is now exitSimLoop().
src/cpu/base.cc:
SimExitEvent is now SimLoopExitEvent
src/python/SConscript:
Add SWIG build command for main.i.
Use python/m5 in build dir as source for zip archive...
easy now with file duplication enabled.
src/python/m5/__init__.py:
- Move copyright notice back to C++ so we can print
it right away, even for interactive sessions.
- Get rid of argument parsing code; just provide default
option descriptors for user script to call optparse with.
- Don't clutter m5 namespace by sucking in all of m5.config
and m5.objects.
- Move instantiate() function here from config.py.
src/python/m5/config.py:
- Move instantiate() function to __init__.py.
- Param.Foo deferred type lookups must use m5.objects
namespace now (not m5).
src/python/m5/objects/AlphaConsole.py:
src/python/m5/objects/AlphaFullCPU.py:
src/python/m5/objects/AlphaTLB.py:
src/python/m5/objects/BadDevice.py:
src/python/m5/objects/BaseCPU.py:
src/python/m5/objects/BaseCache.py:
src/python/m5/objects/Bridge.py:
src/python/m5/objects/Bus.py:
src/python/m5/objects/CoherenceProtocol.py:
src/python/m5/objects/Device.py:
src/python/m5/objects/DiskImage.py:
src/python/m5/objects/Ethernet.py:
src/python/m5/objects/Ide.py:
src/python/m5/objects/IntrControl.py:
src/python/m5/objects/MemObject.py:
src/python/m5/objects/MemTest.py:
src/python/m5/objects/Pci.py:
src/python/m5/objects/PhysicalMemory.py:
src/python/m5/objects/Platform.py:
src/python/m5/objects/Process.py:
src/python/m5/objects/Repl.py:
src/python/m5/objects/Root.py:
src/python/m5/objects/SimConsole.py:
src/python/m5/objects/SimpleDisk.py:
src/python/m5/objects/System.py:
src/python/m5/objects/Tsunami.py:
src/python/m5/objects/Uart.py:
Fix up imports (m5 namespace no longer includes m5.config).
src/sim/eventq.cc:
src/sim/eventq.hh:
Support for Python-called simulate() function:
- Use IsExitEvent flag to signal events that want
to exit the simulation loop gracefully (instead of
calling exit() to terminate the process).
- Modify interface to hand exit event object back to
caller so it can be inspected for cause.
src/sim/host.hh:
Add MaxTick constant.
src/sim/main.cc:
Move copyright notice back to C++ so we can print
it right away, even for interactive sessions.
Use PYTHONPATH environment var to set module path
(instead of clunky code injection method).
Move main control from here into Python:
- Separate initialization code and simulation loop
into separate functions callable from Python.
- Make Python interpreter invocation more pure (more
like directly invoking interpreter).
Add -i and -p flags (only options on binary itself;
other options processed by Python).
Import readline package when using interactive mode.
src/sim/sim_events.cc:
SimExitEvent is now SimLoopExitEvent, and uses
IsSimExit flag to terminate loop (instead of
exiting simulator process).
src/sim/sim_events.hh:
SimExitEvent is now SimLoopExitEvent, and uses
IsSimExit flag to terminate loop (instead of
exiting simulator process).
Get rid of a few unused constructors.
src/sim/sim_exit.hh:
SimExit() is now exitSimLoop().
Get rid of unused functions.
Add comments.
--HG--
extra : convert_revision : 280b0d671516b25545a6f24cefa64a68319ff3d4
It doesn't compile, and if we really want this we should
start over from scratch and see if we can reuse parts from
BaseSimpleCPU (e.g., derive a FastSimpleCPU).
SConstruct:
src/arch/SConscript:
src/cpu/cpu_models.py:
Get rid of FastCPU model.
--HG--
extra : convert_revision : f3f0362f7292bf21e8e0d4d84c475be8a9789a0c
