Commit graph

111 commits

Author SHA1 Message Date
Ali Saidi
79170b1be5 random mix of tidbits
configs/test/fs.py:
    update fs.py to use a bus bridge
cpu/simple/cpu.hh:
    cpu should just return that it doesn't snoop any address ranges
python/m5/objects/System.py:
    move boot_osflags to system

--HG--
extra : convert_revision : b4256df7eada7e65b69513361de8bffc3fdd680b
2006-04-28 15:40:45 -04:00
Ali Saidi
53d93ef918 add a bridge object, modify bus object to be able to connect to other buses or bridges without panicing
SConscript:
    add new cc files to scons
mem/bus.cc:
mem/bus.hh:
    implement addressRanges() on the bus.
    propigate address ranges to anyone who is interested stripping out ranges of who your propigating to (to avoid livelock)
mem/packet.hh:
    add intersect function that returns true if two packets touch at least one byte of the same data (for functional access)
    add fixPacket() that will eventually take the correct action giving a timing and functional packet, right now it panics
mem/physical.cc:
    Don't panic if the physical memory recieves a status change, just ignore.

--HG--
extra : convert_revision : d470d51f2fb1db2700ad271e09792315ef33ba01
2006-04-28 15:37:48 -04:00
Ali Saidi
8f8d09538f Mostly done with all device models for new memory system. Still need to get timing packets working and get sinic working
after merge from head. Checkpointing may need some work now. Endian-happiness still not complete.

SConscript:
    add all devices back into make file
base/inet.hh:
dev/etherbus.cc:
dev/etherbus.hh:
dev/etherdump.cc:
dev/etherdump.hh:
dev/etherint.hh:
dev/etherlink.cc:
dev/etherlink.hh:
dev/etherpkt.cc:
dev/etherpkt.hh:
dev/ethertap.cc:
dev/ethertap.hh:
dev/pktfifo.cc:
dev/pktfifo.hh:
    rename PacketPtr EthPacketPtr so it doesn't conflict with the PacketPtr type in the memory system
configs/test/fs.py:
    add nics to fs.py
cpu/cpu_exec_context.cc:
    remove this check, as it's not valid. We may want to add something else back in to make sure that no one can delete the
    static virtual ports in the exec context
cpu/simple/cpu.cc:
cpu/simple/cpu.hh:
dev/alpha_console.cc:
dev/ide_ctrl.cc:
    use new methods for accessing packet data
dev/ide_disk.cc:
    add some more dprintfs
dev/io_device.cc:
    delete packets when we are done with them. Update for new packet methods to access data
dev/isa_fake.cc:
dev/pciconfigall.cc:
dev/tsunami_cchip.cc:
dev/tsunami_io.cc:
dev/tsunami_pchip.cc:
dev/uart8250.cc:
dev/uart8250.hh:
mem/physical.cc:
mem/port.cc:
    dUpdate for new packet methods to access data
dev/ns_gige.cc:
    Update for new memory system
dev/ns_gige.hh:
python/m5/objects/Ethernet.py:
    update for new memory system
dev/sinic.cc:
dev/sinic.hh:
    Update for new memory system. Untested as need to merge in head because of kernel driver differences between versions
mem/packet.hh:
    Add methods to access data instead of accessing it directly.

--HG--
extra : convert_revision : 223f43876afd404e68337270cd9a5e44d0bf553e
2006-04-24 19:31:50 -04:00
Ali Saidi
6dc3b2fa39 make ide disk work for newmem
SConscript:
    compile ide devices
base/chunk_generator.hh:
    add another parameter to the chuck generator called complete() which
    returns the number of bytes transfered so far. Very useful for
    adding to a pointer.
configs/test/fs.py:
    Add ide disk to fs test configuration
dev/ide_ctrl.cc:
dev/ide_ctrl.hh:
dev/ide_disk.cc:
dev/ide_disk.hh:
dev/io_device.cc:
dev/io_device.hh:
dev/pciconfigall.cc:
dev/pciconfigall.hh:
dev/pcidev.cc:
dev/pcidev.hh:
    update for new memory system
mem/bus.cc:
    support devices that return multiple ranges
    remove old ranges before using new info
mem/packet.hh:
    make senderstate void* per steve's request that we use every
    construct possible in C++
mem/physical.cc:
    have memory stamp the packet with the time.
mem/physical.hh:
    actually set the memory latency variable
python/m5/objects/Device.py:
    Add DmaDevice
python/m5/objects/Ide.py:
    Ide disk no longer has a physmem pointer
python/m5/objects/Pci.py:
    update pci device for newmem
python/m5/objects/PhysicalMemory.py:
    add latency parameter for physical memory
sim/byteswap.hh:
    use fast architecture dependent byteswap calls if they exist

--HG--
extra : convert_revision : e3cf2e8f61064ad302d94bc22010a00c59f3f793
2006-04-20 17:14:30 -04:00
Ali Saidi
f6fc18f03d fullsys now builds and runs for about one cycle
SConscript:
    easier to fix than temporarily remove
cpu/simple/cpu.cc:
cpu/simple/cpu.hh:
    mem needed for both fullsys and syscall
dev/baddev.cc:
    fix for new mem system
dev/io_device.cc:
    fix typo
dev/io_device.hh:
    PioDevice needs to be a memobject
dev/isa_fake.cc:
dev/pciconfigall.cc:
dev/pciconfigall.hh:
    fix for new mem systems
dev/platform.cc:
dev/platform.hh:
dev/tsunami.cc:
dev/tsunami.hh:
    rather than the platform have a pointer to pciconfig, go the other
    way so all devices are the same and can have a platform pointer
dev/tsunami_cchip.cc:
dev/tsunami_io.cc:
dev/tsunami_io.hh:
dev/tsunami_pchip.cc:
dev/tsunami_pchip.hh:
dev/uart8250.cc:
python/m5/objects/AlphaConsole.py:
python/m5/objects/BadDevice.py:
python/m5/objects/BaseCPU.py:
python/m5/objects/Device.py:
python/m5/objects/Pci.py:
python/m5/objects/PhysicalMemory.py:
python/m5/objects/System.py:
python/m5/objects/Tsunami.py:
python/m5/objects/Uart.py:
    fixes for newmem

--HG--
extra : convert_revision : b7b67e19095cca64889f6307725aa2f3d84c7105
2006-04-11 13:42:47 -04:00
Ali Saidi
6240f8c4bc fixes for newmem
ALPHA_FS finally compiles again

SConscript:
    Use a couple more FS sources, still don't compile that much
arch/alpha/faults.hh:
    the unimp fault should probably exist in nonfs too.
dev/alpha_console.cc:
dev/alpha_console.hh:
dev/simconsole.cc:
dev/simple_disk.cc:
dev/simple_disk.hh:
dev/uart.cc:
dev/uart.hh:
dev/uart8250.cc:
dev/uart8250.hh:
sim/process.cc:
sim/system.cc:
    fixes for newmem
dev/io_device.hh:
    a system pointer is probably useful for every device to have
mem/bus.hh:
mem/physical.cc:
    new address ranges function
python/m5/objects/SimpleDisk.py:
    simple disk now has a system pointer rather than physmem directly

--HG--
extra : convert_revision : d8c0a5c6510a6210aec5e8adfb0a4a06ec0dcebf
2006-04-06 14:57:51 -04:00
Ali Saidi
c27c122afc Add the bus and connector objects to scons
change getPort parameter from char* to string
Add an extra phase between construction and init called connect

SConscript:
    Add the bus and connector objects to scons
cpu/simple/cpu.cc:
cpu/simple/cpu.hh:
    the connection to memory shouldn't be made until we know the memory
    object exists (e.g. after construction)
dev/io_device.hh:
    change to const string
mem/bus.hh:
    change getPort parameter from char* to string
    initialize num_interfaces
mem/mem_object.hh:
    change getPort parameter from char* to string
mem/physical.cc:
mem/physical.hh:
    change getPort parameter from char* to string
    get rid of the bus object I created last time
python/m5/objects/PhysicalMemory.py:
    get rid of the bus object I created last time
sim/main.cc:
sim/sim_object.cc:
sim/sim_object.hh:
    Add an extra phase between construction and init called connect

--HG--
extra : convert_revision : 0e994f93374fa72a06d291655c440ff1b8e155a9
2006-03-26 21:44:22 -05:00
Ali Saidi
b38f67d5b7 Implement a very very simple bus
requestTime -> time
responseTime -> packet.time

Make CPU and memory able to connect to the bus

dev/io_device.cc:
    update for request and packet both having a time
    hand platform off to port for eventual selection of request modes
dev/io_device.hh:
    update for request and packet both havig a time
    hand platform off to port for eventual selection of request modes
mem/bus.hh:
    Add a device map struct that maps a range to a portId
    - Which needs work it theory it should be an interval tree
    - but it is a list and works fine right now

    Add a function called findPort which returns port for an addr range

    Add a deviceBlockSize function that really shouldn't exist, but it
    was easier than fixing the translating port
mem/packet.hh:
    add a time to each packet
mem/physical.cc:
mem/physical.hh:
python/m5/objects/PhysicalMemory.py:
    Make physical memory take a MemObject parameter of what to connect to
mem/request.hh:
    remove requestTime/responseTime for just time in request which
    is requset time and the time in the packet which is responsetime
python/m5/objects/BaseCPU.py:
    Instead of memory cpu connects to any memory object
python/m5/objects/Bus.py:
    Fix for new bus object

--HG--
extra : convert_revision : 72605e8a3fcdd9e80a41f439909ee7feb3f1fe1d
2006-03-25 18:31:20 -05:00
Ali Saidi
97e424982a add translations for new sections that are mmapped or when the brk
is changed
Add a default machine width parameter
Arch based live processes

arch/alpha/linux/process.cc:
arch/alpha/linux/process.hh:
arch/alpha/process.cc:
arch/alpha/process.hh:
arch/alpha/tru64/process.cc:
arch/alpha/tru64/process.hh:
arch/mips/linux_process.cc:
arch/mips/process.cc:
arch/mips/process.hh:
arch/sparc/linux/process.cc:
arch/sparc/linux/process.hh:
arch/sparc/process.cc:
arch/sparc/process.hh:
configs/test/test.py:
python/m5/objects/Process.py:
sim/process.cc:
sim/process.hh:
    Architecture based live processes
arch/mips/isa_traits.hh:
arch/sparc/isa_traits.hh:
    Add a default machine width parameter
mem/port.hh:
    gcc 4 really wants  a virtual destructor
sim/byteswap.hh:
    remove the comment around long and unsigned long even though uint32_t
    and int32_t are defined. Seems to work with gcc 4 and 3.4.3.
sim/syscall_emul.cc:
sim/syscall_emul.hh:
    add translations for new sections that are mmapped or when the brk
    is changed

--HG--
extra : convert_revision : e2f9f228113c7127c87ef2358209a399c30ed5c6
2006-03-15 17:04:50 -05:00
Steve Reinhardt
e3d96aa889 Compiles now (with CPU_MODELS=SimpleCPU), but hangs
on execution.

configs/test/test.py:
    Move test binary out of m5-test, don't depend on
    m5-test/Benchmarks.
python/m5/objects/System.py:
    Split out full-system-only parameters (lost in merge).
sim/system.cc:
    Need to be able to instantiate System directly in SE mode
    (lost in merge).
sim/system.hh:
    A few more functions here are FS-only.
configs/test/hello:
    Add in binary.

--HG--
rename : configs/test.py => configs/test/test.py
extra : convert_revision : 4051b18772e0a0dcb97eb591d4373683be9f4395
2006-03-10 10:01:29 -05:00
Gabe Black
872bbdfc33 Hand merge. Stuff probably doesn't compile.
--HG--
rename : arch/alpha/isa_desc => arch/alpha/isa/main.isa
rename : arch/alpha/alpha_linux_process.cc => arch/alpha/linux/process.cc
rename : arch/alpha/alpha_linux_process.hh => arch/alpha/linux/process.hh
rename : arch/alpha/alpha_tru64_process.cc => arch/alpha/tru64/process.cc
rename : arch/alpha/alpha_tru64_process.hh => arch/alpha/tru64/process.hh
rename : cpu/exec_context.cc => cpu/cpu_exec_context.cc
rename : cpu/exec_context.hh => cpu/cpu_exec_context.hh
extra : convert_revision : 7d1efcedd708815d985a951f6f010fbd83dc27e8
2006-03-09 18:35:28 -05:00
Ali Saidi
d01a593433 move alpha specific code into arch/alpha
System -> AlphaSystem

SConscript:
    Move some more files around
arch/alpha/freebsd/system.cc:
arch/alpha/freebsd/system.hh:
    System -> AlphaSystem
arch/alpha/linux/aligned.hh:
arch/alpha/linux/hwrpb.hh:
arch/alpha/linux/system.cc:
arch/alpha/linux/thread_info.hh:
arch/alpha/linux/threadinfo.hh:
    move alpha specific code into arch/alpha
python/m5/objects/System.py:
    Split off arch specific parts of system into AlphaSystem

--HG--
rename : kern/freebsd/freebsd_system.cc => arch/alpha/freebsd/system.cc
rename : kern/freebsd/freebsd_system.hh => arch/alpha/freebsd/system.hh
rename : kern/linux/aligned.hh => arch/alpha/linux/aligned.hh
rename : kern/linux/hwrpb.hh => arch/alpha/linux/hwrpb.hh
rename : kern/linux/linux_system.cc => arch/alpha/linux/system.cc
rename : kern/linux/linux_system.hh => arch/alpha/linux/system.hh
rename : kern/linux/thread_info.hh => arch/alpha/linux/thread_info.hh
rename : kern/linux/linux_threadinfo.hh => arch/alpha/linux/threadinfo.hh
rename : kern/tru64/tru64_system.cc => arch/alpha/tru64/system.cc
rename : kern/tru64/tru64_system.hh => arch/alpha/tru64/system.hh
extra : convert_revision : 1c496efb823326a4b629574bd9b5f30e99a0aa40
2006-03-04 20:45:01 -05:00
Nathan Binkert
7546fabe68 Ethernet devices have an RSS option to tell the driver to
use Receive side scaling

dev/ns_gige.cc:
dev/ns_gige.hh:
dev/ns_gige_reg.h:
dev/sinic.cc:
dev/sinic.hh:
dev/sinicreg.hh:
    add support for setting the RSS flag to notify the driver
    to use RSS

--HG--
extra : convert_revision : 5f0c11668ae976634b3bf0caad669a9464a4c041
2006-03-03 14:17:48 -05:00
Steve Reinhardt
22504f8b48 More progress toward actually running a program.
See configs/test.py for test config (using simple
binary in my home directory on zizzer).

base/chunk_generator.hh:
    Fix assertion for chunkSize == 0 (not a power of 2)
base/intmath.hh:
    Fix roundDown to take integer alignments.
cpu/base.cc:
    Register exec contexts regardless of state (not sure why
    this check was in here in the first place).
mem/physical.cc:
    Add breaks to switch.
python/m5/objects/BaseCPU.py:
    Default mem to Parent.any (e.g. get from System).
python/m5/objects/Ethernet.py:
python/m5/objects/Root.py:
    HierParams is gone.
python/m5/objects/PhysicalMemory.py:
    mmu param is full-system only.
sim/process.cc:
    Stack mapping request must be page-aligned and page-sized.
    Don't delete objFile object in create since we are counting
    on it being around for startup().

--HG--
extra : convert_revision : 90c43ee927e7d82a045d6e10302d965797d006f7
2006-03-01 18:45:50 -05:00
Ron Dreslinski
31fc398f06 Fixes so that it compiles properly. Still working on .py file issues.
SConscript:
    Add Back memory to be built
mem/physical.hh:
    Fix function declerations
python/m5/objects/BaseCPU.py:
    Remove IL1 and DL1 params from the cpu object

--HG--
extra : convert_revision : 2f285dc626bc8d84d095def68e986fe7e6f3d8e9
2006-02-27 16:33:11 -05:00
Ron Dreslinski
b6247c9ea7 Add support for multiple ports on the memory. Hook up simple cpu to memory.
Ready to start testing if I could fix the linking errors I can't ever seem to fix.

cpu/simple/cpu.cc:
cpu/simple/cpu.hh:
    Add connecting of ports until builder can handle it.
mem/physical.cc:
    Add function to allocate a port in the object

    Remove some full_sys stuff untill needed
mem/physical.hh:
    Add function to allocate a port in the object
python/m5/objects/BaseCPU.py:
    Update the params
sim/process.cc:
    Make sure to use the right name (hopefully CPU constructor already called)

--HG--
extra : convert_revision : 4089caf20d7eb53e5463c8ac93ddce5e43ea5d85
2006-02-23 17:02:34 -05:00
Ron Dreslinski
8fc06589cb Update functional memory to have a response event
Clean out old memory python files, move them into old_mem directory.  Maybe we should just delete them, they are under revision control.

Add new py files for new objects.

SConscript:
    Update because memory is just a header file now
base/chunk_generator.hh:
    Make Chunk Generator return the entire size if the chunk_size is set to zero.  Useful when trying to chunck on blocksize of memory, which can write large pieces of data.
cpu/simple/cpu.cc:
    Make sure to delete the pkt.
mem/physical.cc:
mem/physical.hh:
    Set up response event.
mem/port.cc:
    Rename rqst to req to conform to same standard naming convention.
python/m5/objects/PhysicalMemory.py:
    Update the params, inheritence

--HG--
extra : convert_revision : 857154ec256522baf423b715833930497999549b
2006-02-23 13:51:54 -05:00
Nathan Binkert
3a0102536b Get rid of the code that delays PIO write accesses
until the cache access occurs.  The fundamental problem
is that a subsequent read that occurs functionally will
get a functionally incorrect result that can break
driver code.

dev/ns_gige.cc:
dev/ns_gige.hh:
dev/sinic.cc:
dev/sinic.hh:
    get rid of pio_delay write and the associated code to move
    the write to the cache access function
dev/sinicreg.hh:
    no more write delays
python/m5/objects/Ethernet.py:
    get rid of pio_delay write

--HG--
extra : convert_revision : 1dcb51b8f4514e717bc334a782dfdf06d29ae69d
2006-02-20 23:41:50 -05:00
Steve Reinhardt
03a2aca9a9 Changes for Process object initialization in merged-memory environment.
System object now exists for both fullsys and syscall emulation, as the
latter needs it so that Process objects can find the shared PhysicalMemory
for initialization.
Changes are incomplete: still need to fix up Process (& EioProcess) memory
initialization and syscall emulation code for new mem interface.

arch/alpha/alpha_linux_process.cc:
arch/alpha/alpha_linux_process.hh:
arch/alpha/alpha_tru64_process.cc:
arch/alpha/alpha_tru64_process.hh:
cpu/base.cc:
cpu/base.hh:
    Take System argument in constructor.
cpu/exec_context.cc:
    Take System argument in constructor.
    Merge two constructors into a single one.
cpu/exec_context.hh:
    Take System argument in constructor.
    Merge two constructors into a single one.
    Replace dummy translation with lookup in Process object's page table.
python/m5/objects/Process.py:
    Add System parameter to Process object (& subobjects).
python/m5/objects/System.py:
    Segregate full-system only Process parameters (most of them!).
sim/process.cc:
    Take System argument in constructor.
    Move initialization to startup() callback to occur after system & cpus
    are initialized.
    Generate ProxyMemory object to pass to loader for transparent
    virtual page allocation.
sim/process.hh:
    Take System argument in constructor.
    Move initialization to startup() callback to occur after system & cpus
    are initialized.
sim/system.cc:
sim/system.hh:
    Enable System object for non-full-system too.
    Basically involved putting most of the existing code
    inside '#ifdef FULL_SYSTEM'.
    Key thing needed for syscall emulation at this point is
    the PhysicalMemory object (for Process initialization).

--HG--
extra : convert_revision : f0f34b47bd4f77b502191affd3d03b4d6d9bcdd8
2006-01-28 00:08:22 -05:00
Nathan Binkert
47ff0af17e Virtualize sinic
separate the rx thread and tx thread and get rid of the dedicated flag.

dev/ns_gige.cc:
dev/ns_gige.hh:
dev/ns_gige_reg.h:
python/m5/objects/Ethernet.py:
    dedicated flag goes away, we have new individual flags for
    rx thread and tx thread
dev/sinic.cc:
    Virtualize sinic
    - The io registers are replicated many times in memory, allowing the NIC to
    differentiate among several virtual interfaces.
    - On the TX side, this allows multiple CPUs to initiate transmits at the same
    time without locking in the software.  If a partial packet is transmitted,
    then the state machine blocks waiting for that virtual interface to complete
    its packet.  Then the state machine will move on to the next virtual
    interface.  The commands are kept in fifo order.
    - On the RX side, multiple partial transmits can be simultaneously done.
    Though a packet does not deallocate its fifo space until all preceeding
    packets in the fifo are deallocated.  To enable multiple receives, it
    is necessary for each virtual nic to keep its own information about its
    progress through the state machine.
dev/sinic.hh:
    Virtualize sinic
    Receive state must be virtualized since we allow the receipt of packets in
    parallel.
dev/sinicreg.hh:
    Virtualize sinic
    separate rx thread and tx thread
    create a soft interrupt and add a command to trigger it.
    pad out the reserved bits in the RxDone and TxDone regs

--HG--
extra : convert_revision : c10bb23a46a89ffd1e08866c1f1621cb98069205
2005-11-25 13:33:36 -05:00
Nathan Binkert
f806a25c9e add support for delaying pio writes until the cache access occurs
dev/ns_gige.cc:
    add support for delaying pio writes until the cache access occurs
    the only write we delay are for CR_TXE and CR_RXE
dev/sinic.cc:
dev/sinic.hh:
    the txPioRequest and rxPioRequest things were more or less bogus
    add support for delaying pio writes until the cache access occurs
dev/sinicreg.hh:
    Add delay_read and delay_write to the register information struct
    for now, we won't delay any reads, and we'll delay the writes that
    initiate DMAs
python/m5/objects/Ethernet.py:
    add a parameter to delay pio writes until the timing access
    actually occurs.

--HG--
extra : convert_revision : 79b18ea2812c2935d7d5ea6eff1f55265114d05d
2005-11-21 23:43:15 -05:00
Nathan Binkert
d3a4781ac5 Actually, you should'nt do math on Clock in the config files.
python/m5/config.py:
    Clock should not be a NumericParamValue since math on it can be
    ambiguous.  (As the comment clearly says.)

--HG--
extra : convert_revision : 74f8ec846c6a980d92e0bf4bf1c7fac73a75b923
2005-11-21 00:22:29 -05:00
Nathan Binkert
c4678ece84 BARs now of type MemorySize32
python/m5/config.py:
    Add MemorySize32 (a 32-bit value specified in bytes)

--HG--
extra : convert_revision : bfeee501f7ff1aa2567a3682da129a5770cb7bd2
2005-11-21 00:02:39 -05:00
Nathan Binkert
d733d168ca Make it so one can do math with a Clock type in the config files
python/m5/config.py:
    Make Clock a NumericParamValue so you can do math with it

--HG--
extra : convert_revision : 7fa548d1a23c604a31d3ecae3853949b064a1830
2005-11-20 23:57:26 -05:00
Nathan Binkert
3b8b838417 Allow long as a multiplier in the python config
python/m5/config.py:
    Allow long as a multiplier

--HG--
extra : convert_revision : bdc45e9afc27bb7b0f2cd6aacf92758601b95ff8
2005-11-20 23:45:45 -05:00
Nathan Binkert
12d903a650 io_bus is split out into pio_bus and dma_bus so that any device
can specify either independently.

python/m5/objects/Device.py:
    io_bus is split out into pio_bus and dma_bus so that any device
    can specify either independently.
    dma_bus defaults to point to whatever pio_bus uses.

--HG--
extra : convert_revision : d35d5374d0bf592f6b5df465c05203577b8b8763
2005-11-20 16:57:53 -05:00
Ali Saidi
cb929af345 allow conversion floats -> long so that 100e9 and such work as Ticks
--HG--
extra : convert_revision : 23511baca6153bb3aa9c57be8818ad1b65f02a71
2005-11-02 15:37:47 -05:00
Ali Saidi
3b66cb49ec Merge zizzer:/bk/m5
into  zeep.eecs.umich.edu:/z/saidi/work/m5

--HG--
extra : convert_revision : 3cc23080d19cc464a8ba7c1c93b6e5d45af7d463
2005-11-02 14:56:18 -05:00
Ali Saidi
15ccd16fd8 Add Mem/Ethernet latency variability parameter
python/m5/objects/Ethernet.py:
    Add Latency Variability Parameter

--HG--
extra : convert_revision : db5431cccffea8c7247d0f72e4770d4d58bd25aa
2005-11-02 14:44:42 -05:00
Nathan Binkert
dd46db1cb9 __init__ should not return anything
--HG--
extra : convert_revision : fb46eee741f4899d76bcf927523fa151d002decf
2005-11-02 12:14:26 -05:00
Steve Reinhardt
d238b6be9d Make vector params interact with proxies properly.
--HG--
extra : convert_revision : a4067f07d71d2adc1ccbf4512a43ceee7b5cc3de
2005-11-02 10:20:39 -05:00
Steve Reinhardt
fb4f83809f Allow math on CheckedInt-derived ParamValue classes w/o
losing type information.

python/m5/config.py:
    Allow math on CheckedInt-derived ParamValue classes w/o
    losing type information.
    - Make CheckedInt derive from NumericParamValue, and *not*
    multiply inherit from long
    - Move CheckedInt bounds check to _check() hook so we can
    call it when value is updated (not just in constructor)
python/m5/convert.py:
    - make toInteger() return a long, making toLong() unnecessary
    - toMemorySize should return long rather than float

--HG--
extra : convert_revision : c1cf5e15b9ff35d9b573dd545e076fe68afef989
2005-11-01 14:11:54 -05:00
Nathan Binkert
b7b8ffa7b7 Major changes to sinic device model. Rearrage read/write, better
interrupts.

dev/sinic.cc:
    - The prepareRead function sets all the variables in the register
    file that depend on various state bits that change on the fly.
    Includes RxDone, RxWait, TxDone, and TxWait
    - Use the new register information accessor functions to grab
    validity and size information for the read and write functions
    - read all registers directly from the register space by offset
    and size, not by actual name (less code)
    - The side effect of reading the interrupt status (clearing it) now
    happens outside the actual chunk of code where the value is loaded.
    - Add an iprRead function for when we may want speculative access
    to device registers through an ipr or special instruction.
    - When RxData or TxData are written, their busy flag is set to
    indicate that they have an outstanding transaction.
    - The RxHigh and TxLow interrupts are special, they only interrupt
    if the rxEmpty or txFull limits were hit
    - Move reset to the command register
    - Update more registers on reset, clear rxEmpty and txFull
    - Data dumps only happen if EthernetData trace flag set
    - When a DMA completes, kick the other engine if it was waiting
    - implement all of the new interrupts
    - serialize the new stuff
dev/sinic.hh:
    - Put all registers with their proper size and alignment into
    the regs struct so that we can copy multiple at a time.
    - Provide accessor functions for accessing the registers with
    different sizes.
    - Flags to track when the rx fifo hit empty and the tx fifo became
    full.  These flags are used to determine what to do when below
    the watermarks, and are reset when crossing the watermark.
    - the txDmaEvent should actually trigger the txDmaDone function
    - Add an iprRead function for when we may want speculative access
    to device registers through an ipr or special instruction.
    - The prepareRead function sets all the variables in the register
    file that depend on various state bits that change on the fly.
    - add rx_max_intr and dedicated (for dedicated thread) config params
dev/sinicreg.hh:
    Add some new registers: Command, RxMaxIntr, RxFifoSize, TxFifoSize,
    rename XxThreshold to XxFifoMark
    Move Reset to the Command register
    Add Thread to the Config register
    New interrupts, better names
    More info in RxDone and TxDone
    Easier access to information on each register (size, read, write, name)
python/m5/objects/Ethernet.py:
    Both sinic and nsgige have the dedicated thread
    Add a parameter to configure the maximum number for receive
    packets per interrupt

--HG--
extra : convert_revision : 407c5a993b6fb17326b4c623ee5d4b25fd69ac80
2005-10-21 20:28:21 -04:00
Nathan Binkert
2badc0113c Shuffle around device names to make things easier to read.
Create EtherDevBase which both Sinic and NSGigE derive from
bump fifos
drop rx max copy size to 1514 bytes to be friendlier with linux
default interrupt delay is 10us

dev/ns_gige.cc:
    Shuffle around parameters to make it easier to find stuff
dev/sinic.cc:
    Shuffle around parameters to make it easier to find stuff
    rename cycleTime -> clock
dev/sinic.hh:
    rename cycleTime -> clock

--HG--
extra : convert_revision : a673bee875e50d083098991aea20972fa8d5b5c7
2005-10-18 22:05:05 -04:00
Nathan Binkert
4f43bc65ea use the dedicated flag, no more exposing the m5reg directly
dev/ns_gige.cc:
    stop exposing the m5reg to the configuration stuff and build it
    based on exposed flags.  Expose dedicated now.
dev/ns_gige.hh:
    goodbye m5reg hello dedicated
dev/ns_gige_reg.h:
    Flags for the M5REG

--HG--
extra : convert_revision : 11134fe67cdf5291caacf9b3041739c437b983e3
2005-10-18 21:01:05 -04:00
Steve Reinhardt
af2a600f99 Add execution trace object to Root.
python/m5/objects/Root.py:
    Add execution trace object.

--HG--
extra : convert_revision : 1d1ba18779fd4d0b5af921c5f34b7a2bfb4b75d9
2005-10-06 13:50:13 -04:00
Steve Reinhardt
93e60de656 Add executable parameter to LiveProcess. This allows the argv[0] value to
stay fixed even if the path to the binary changes, so the simulation results
are independent of that path.

--HG--
extra : convert_revision : d1109cd284466c14eddc97289908a51e771fc5db
2005-10-01 16:02:47 -04:00
Nathan Binkert
f88d7c7456 Fix the EtherDump parameters
dev/etherdump.cc:
    no default parameters anymore they should be in python
python/m5/objects/Ethernet.py:
    move the maxlen parameter for EtherDump into python

--HG--
extra : convert_revision : a796353a68907dfeb22059cd3ad536e6e8f60998
2005-09-17 10:47:16 -04:00
Steve Reinhardt
845bdb0d8e Regression tests now run under scons!
For example, 'scons ALPHA_SE/test/opt/quick' will build
ALPHA_SE/m5.opt if necessary and run all the self-identified
"quick" tests on it.  Other possibilities:
- Run just test1: scons ALPHA_SE/test/opt/test1
- Run all tests:  scons ALPHA_SE/test/opt
- Run all tests on debug build: scons ALPHA_SE/test/debug
- Update test1 reference outputs in m5-test:
scons update_ref=y ALPHA_SE/test/opt/test1
The proper tests will be selected based on the setting
of FULL_SYSTEM, ALPHA_TLASER, etc.

README:
    Update directions to use scons-based test invocation.
SConscript:
    Return list of generated build environments to SConstruct
    so it can associate tests with each of them.
    Set 'M5Binary' attribute on each env to record name of
    generated binary to be tested.
build/SConstruct:
    - Support invoking m5-test tests via scons.
    - Add new non-sticky option category, for 'update_ref'.
    - Move existing "sticky" option definitions out of
    build_dir loop.  Someday we can generate help text
    from these.
    - Make 'CC' and 'CXX' sticky options; use environment vars as
    defaults if available.
    - Make config builder more scons-y.
python/m5/__init__.py:
    Make AddToPath() correctly handle relative path arguments.
    Assumes that sys.path[0] has the directory where the current
    Python file lives; new m5execfile() function sets this up
    properly for exec'd files.

--HG--
extra : convert_revision : 48896688592e210d8e63f96c34e57474853d0e66
2005-09-05 16:31:27 -04:00
Steve Reinhardt
23a9102d01 Convert type of max_time and progress_interval parameters
from Latency to Tick, and rename max_time to max_tick.

python/m5/objects/Root.py:
sim/root.cc:
    Convert type of max_time and progress_interval
    from Latency to Tick, and rename max_time to max_tick.

--HG--
extra : convert_revision : 2f2aacf6321c3003a0ce834acd8fb726abf27ce3
2005-09-01 11:32:58 -04:00
Steve Reinhardt
c4793184bd Build options are set via a build_options file in the
build directory instead of being inferred from the name
of the build directory.
Options are passed to C++ via config/*.hh files instead of
via the command line.  Build option flags are now always
defined to 0 or 1, so checks must use '#if' rather than
'#ifdef'.

SConscript:
    MySQL detection moved to SConstruct.
    Add config/*.hh files (via ConfigFile builder).
arch/alpha/alpha_memory.cc:
arch/alpha/ev5.cc:
arch/alpha/ev5.hh:
arch/alpha/isa_traits.hh:
base/fast_alloc.hh:
base/statistics.cc:
base/statistics.hh:
base/stats/events.cc:
base/stats/events.hh:
cpu/base.cc:
cpu/base.hh:
cpu/base_dyn_inst.cc:
cpu/base_dyn_inst.hh:
cpu/exec_context.cc:
cpu/exec_context.hh:
cpu/o3/alpha_cpu.hh:
cpu/o3/alpha_cpu_builder.cc:
cpu/o3/alpha_cpu_impl.hh:
cpu/o3/alpha_dyn_inst.hh:
cpu/o3/alpha_dyn_inst_impl.hh:
cpu/o3/alpha_params.hh:
cpu/o3/commit_impl.hh:
cpu/o3/cpu.cc:
cpu/o3/cpu.hh:
cpu/o3/fetch_impl.hh:
cpu/o3/iew.hh:
cpu/o3/iew_impl.hh:
cpu/o3/regfile.hh:
cpu/o3/rename_impl.hh:
cpu/o3/rob_impl.hh:
cpu/ozone/cpu.hh:
cpu/pc_event.cc:
cpu/simple/cpu.cc:
cpu/simple/cpu.hh:
sim/process.cc:
sim/process.hh:
    Convert compile flags from def/undef to 0/1.
    Set via #include config/*.hh instead of command line.
arch/alpha/isa_desc:
    Convert compile flags from def/undef to 0/1.
    Set via #include config/*.hh instead of command line.
    Revamp fenv.h support... most of the ugliness is hidden
    in base/fenv.hh now.
base/mysql.hh:
    Fix typo in #ifndef guard.
build/SConstruct:
    Build options are set via a build_options file in the
    build directory instead of being inferred from the name
    of the build directory.
    Options are passed to C++ via config/*.hh files instead of
    via the command line.
python/SConscript:
    Generate m5_build_env directly from scons options
    instead of indirectly via CPPDEFINES.
python/m5/convert.py:
    Allow '0' and '1' for booleans.
    Rewrite toBool to use dict.
base/fenv.hh:
    Revamp <fenv.h> support to make it a compile option
    (so we can test w/o it even if it's present) and to
    make isa_desc cleaner.

--HG--
extra : convert_revision : 8f97dc11185bef5e1865b3269c7341df8525c9ad
2005-08-30 13:18:54 -04:00
Benjamin Nash
0d82e0f8b6 Merge m5.eecs.umich.edu:/bk/m5
into  zed.eecs.umich.edu:/z/benash/bk/m5

--HG--
extra : convert_revision : 36bb126381caf84d85a566579225b02c15f8f356
2005-08-16 15:27:39 -04:00
Miguel Serrano
b64eae5e52 Changes for getting FreeBSD to run.
SConscript:
    Added more files to compile: dev/pcifake.cc, dev/isa_fake.cc, kern/freebsd/freebsd_system.cc, kern/freebsd/freebsd_events.cc.
arch/alpha/isa_traits.hh:
    Added constant for argument register 2 as it is needed by FreebsdSystem::doCalibrateClocks().
cpu/exec_context.hh:
cpu/o3/alpha_cpu.hh:
    Replaced htoa()s with gtoh() and htog().
cpu/o3/fetch_impl.hh:
cpu/simple/cpu.cc:
    Replaced htoa() with gtoh().
dev/disk_image.cc:
    Replaced htoa()s with letoh()s.
dev/ide_ctrl.cc:
    Got rid of magic numbers.
    Added IdeChannel and IdeRegType type names where necessary.
dev/ide_ctrl.hh:
    Got rid of unnecessary macros.
    Changed RegType_t to IdeRegType.
    Changed bmi_regs to allow accessing registers by name instead of just by array index.
    Added IdeChannel enum type to use in place of bool variables which were used to specify IDE channel.
dev/ide_disk.cc:
    Rewrote IdeDisk::read and IdeDisk::write functions to specify registers by name instead of indexing through an array.
dev/ide_disk.hh:
    Updated command register struct.
dev/ns_gige.cc:
dev/ns_gige.hh:
    Made ReadConfig and WriteConfig begin with a lower-case letter.
    writeConfig() now takes a pointer to data as a parameter instead of a copy of data.
dev/pciconfigall.cc:
    writeConfig() now takes a pointer to data as a parameter instead of a copy of data.
dev/pcidev.cc:
    Cleaned up readConfig() and writeConfig() functions.
dev/pcidev.hh:
    Added macros to make code that works with the BARs (base adress registers) more readable. writeConfig() now takes a pointer to data.
dev/pcireg.h:
    Changed PCIConfig struct to make accessing elements more straight forward. Removed type 1 (for PCI-to-PCI bridges) PCI configuration space struct since it is not used.
dev/rtcreg.h:
    Added macros for bit fields in RTC status registers A & B.
dev/sinic.cc:
    Function name change: WriteConfig --> writeConfig.
    writeConfig() now takes a pointer to data instead of a copy of data.
    The accessing of elements of PCIConfig structure is updated.
dev/sinic.hh:
    Function name change: WriteConfig --> writeConfig.
    writeConfig() now takes a pointer to data instead of a copy of data.
dev/tsunami_io.cc:
    Added implementation of new RTC and PIT classes.
dev/tsunami_io.hh:
    Added classes for RTC and PIT modules.
dev/tsunamireg.h:
    Added macros for DMA ports used by Tsunami-Tru64.
dev/uart8250.cc:
    Got rid of a magic number.
    Transmit (Tx) interrupts should clear upon a read of the Interrupt ID register.
dev/uart8250.hh:
    Added comments and macros dealing with the UART Interrupt ID register.
kern/linux/linux_system.cc:
    Replaced htoa() with htog().
python/m5/objects/Pci.py:
    PciFake is a python class for Pci Devices that do nothing.
python/m5/objects/Tsunami.py:
    TsunamiFake was renamed as IsaFake.
sim/system.cc:
    Replaced htoa()s with htog()s.
dev/isa_fake.cc:
    New BitKeeper file ``dev/isa_fake.cc''
    TsunamiFake was renamed as IsaFake.
dev/isa_fake.hh:
    New BitKeeper file ``dev/isa_fake.hh''
    TsunmaiFake was renamed as IsaFake.
dev/pitreg.h:
    New BitKeeper file ``dev/pitreg.h''
    Useful macros for working with PIT (Periodic Interval Timer) registers.

--HG--
extra : convert_revision : 33f3a8a1034af4f6c71b32dd743e371c8613e780
2005-08-15 16:59:58 -04:00
Nathan Binkert
38da461fd7 Fix NextEthernetAddr
python/m5/config.py:
    NextEthernetAddr shouldnt' be a Singleton since we want __init__ to be
    called more than once.
    Make the EthernetAddr class a "proxy" so that unproxy will
    be called and NextEthernetAddr will generally work correctly.

--HG--
extra : convert_revision : c89bf268e805e202ae71030fcea4833867c7e477
2005-08-15 16:12:19 -04:00
Steve Reinhardt
1e2c16c912 Fix for passing functional memory param to timing mem.
python/m5/config.py:
    Fix error message.

--HG--
extra : convert_revision : 4e57f7bdd4ea7dfdd3e88c60080f993997b0bda2
2005-07-18 19:58:43 -04:00
Nathan Binkert
8a0bc84022 Allow CPUs to specify their own CPU ids.
Make the AlphaConsole calculate the number of CPUs instead
of passing that in as a parameter.

cpu/base.cc:
    pass the desired cpu_id into registerExecContext, offsetting it
    by the thread number.  a cpu_id of -1 means that it should be
    generated for you.
cpu/base.hh:
    Take the cpu_id as a parameter
cpu/o3/alpha_cpu_builder.cc:
cpu/simple/cpu.cc:
    Accept the cpu_id as a parameter
    while we're here, let's remove the multiplier since it is
    not used.
dev/alpha_console.cc:
    don't take the number of CPUs as a parameter.  Calculate it from
    the system based on the number of CPUs that have been registered.
    move init() code to startup() to ensure that all CPUs are registerd.
dev/alpha_console.hh:
python/m5/objects/AlphaConsole.py:
    don't take the number of CPUs as a parameter.
    move init() code to startup() to ensure that all CPUs are registerd.
python/m5/objects/BaseCPU.py:
    take the cpu_id as a parameter.  Default it to -1 which means
    that it will be generated.
sim/system.cc:
    allow the registerExecContext functioin to take a desired
    cpu_id as a parameter.  Check to ensure that the id isn't
    already used.  Accept -1 as a request to have an id assigned.
sim/system.hh:
    keep track of the number of registered exec contexts.
    provide a function for accessing the number of exec contexts
    that checks to ensure that they are all registered correctly.

--HG--
extra : convert_revision : 8e12f96ff8a49fa16cdbbdb4c05c651376c35788
2005-06-29 01:20:41 -04:00
Nathan Binkert
c4029ecb30 Implement a state machine clock that acutally limits how fast
the nsgige state machine can run. The frequency is of the actual
state transitions, and not the rate of what underlying
instructions might run at.

dev/ns_gige.cc:
    Implement a state machine clock that acutally limits how fast
    the state machine can run.  After each state transition, a
    variable is kept to hold the next state transition until the
    next clock.  The frequency is of the actual state transitions,
    and not the rate of what underlying instructions might run at.
dev/ns_gige.hh:
    Add back the rxKickEvent and txKickEvent events.
python/m5/objects/Ethernet.py:
    Default the state machine clock to '0ns' so the default
    behaviour doesn't change when we actually implement the
    state machine clock.

--HG--
extra : convert_revision : 2db1943dee4e91ea75aaee6a91e88f27f01a09dd
2005-06-27 17:02:40 -04:00
Ali Saidi
cad549d7aa Added Float class
Fixed printing so the tokenizer in m5 doesn't get confused
Expanded NullSimObject so it could be used as an element in a VectorParam

--HG--
extra : convert_revision : 661b1916967d663ab7aee891f15f7ca190deeba6
2005-06-23 01:07:04 -07:00
Nathan Binkert
91f5736fd3 Move max_time and progress_interval parameters to the Root
object and get rid of the ParamContext that each used to have.

python/m5/objects/Root.py:
    Add max_time and progress_interval to the Root object
sim/root.cc:
    Add max_time and progress_interval to the Root object.  These
    parameters used to be in their own contexts in sim_events.cc
sim/sim_events.cc:
    Get rid of the ParamContext for max cycles and the progress
    event.  Move the functionality to the Root object
sim/sim_events.hh:
    Move ProgressEvent declaration to the header so that it can
    be used in other files.

--HG--
extra : convert_revision : ff664b806855e8eb9201b8a25392aa53204464f0
2005-06-22 09:59:13 -04:00
Nathan Binkert
ebb35e57ce BaseSystem was renamed to System
--HG--
extra : convert_revision : 74e03fe9447d9d2be59e675b034dc6df0afcde51
2005-06-09 15:09:35 -04:00