Commit graph

3821 commits

Author SHA1 Message Date
Gabe Black
32a927b85f Only bother with the device SConscript if you're in FULL_SYSTEM
--HG--
extra : convert_revision : ac52f548afb98dd0437e7d7c2600ff9b8ebfd1fa
2006-11-06 19:55:42 -05:00
Gabe Black
02abca6b9e Merge zizzer.eecs.umich.edu:/bk/newmem/
into  zeep.eecs.umich.edu:/home/gblack/m5/newmemmemops

src/SConscript:
    SCCS merged

--HG--
extra : convert_revision : f130c8a2d33f58d857e5d5a02bb9698c1bceb23b
2006-11-06 19:52:32 -05:00
Gabe Black
dd14c86ec8 Moved the tsunami devices into the dev/alpha directory. Other devices "generic" devices are dependent on some of those files. That will either need to change, or most likely those devices will have to be considered architecture dependent.
--HG--
rename : src/dev/tsunami.cc => src/dev/alpha/tsunami.cc
rename : src/dev/tsunami.hh => src/dev/alpha/tsunami.hh
rename : src/dev/tsunami_cchip.cc => src/dev/alpha/tsunami_cchip.cc
rename : src/dev/tsunami_cchip.hh => src/dev/alpha/tsunami_cchip.hh
rename : src/dev/tsunami_io.cc => src/dev/alpha/tsunami_io.cc
rename : src/dev/tsunami_io.hh => src/dev/alpha/tsunami_io.hh
rename : src/dev/tsunami_pchip.cc => src/dev/alpha/tsunami_pchip.cc
rename : src/dev/tsunami_pchip.hh => src/dev/alpha/tsunami_pchip.hh
rename : src/dev/tsunamireg.h => src/dev/alpha/tsunamireg.h
extra : convert_revision : ffbb6fd93341d2623a6932bf096019b8976da694
2006-11-06 19:45:00 -05:00
Gabe Black
5b152b970b Got rid of stray alpha include
--HG--
extra : convert_revision : eddd64dd9291d6656821fe6387aeab2f9ddbaf58
2006-11-06 19:10:13 -05:00
Gabe Black
b04a2653f9 Got rid of obsolete ivlb and ivle psuedo instructions.
--HG--
extra : convert_revision : c3c2dd5a6e7181ad94194146d7fa2b33b21074fb
2006-11-06 19:09:23 -05:00
Gabe Black
ef1a92eb9b Stub for SPARC interrupt handling object.
--HG--
extra : convert_revision : 7257e3387c01e84e5a1018a9cdcc09a79edfa934
2006-11-06 18:30:28 -05:00
Gabe Black
85a6079db7 Remote GDB support has been changed to use inheritance. Alpha should work, but isn't tested. Other architectures will not.
--HG--
extra : convert_revision : fc7e1e73e2f3b1a4ab9905a1eb98c5f07c6c8707
2006-11-06 18:29:58 -05:00
Gabe Black
e39de58d21 Took the Alpha prefix off of AlphaArguments, and made sure it was being used from TheISA:: rather than AlphaISA::
--HG--
extra : convert_revision : 17c143d3cbc2f58a7a9d01366a8f649810ff7f33
2006-11-06 18:28:10 -05:00
Gabe Black
1ffff78ca9 Created seperate SConscript for the dev directory. Made subdirectories for Alpha and SPARC and put SConscripts in them.
--HG--
rename : src/base/kgdb.h => src/arch/alpha/kgdb.h
rename : src/dev/alpha_access.h => src/dev/alpha/access.h
rename : src/dev/alpha_console.cc => src/dev/alpha/console.cc
rename : src/dev/alpha_console.hh => src/dev/alpha/console.hh
extra : convert_revision : a7dd466308cb83edc40528689aacb72413089cdf
2006-11-06 18:26:11 -05:00
Ali Saidi
bf3223d7ce delete pcifake, tsunamifake. Combine BadAddr/IsaFake into one
src/SConscript:
    remove pcifake and tsunami fake from sconscript
src/dev/isa_fake.cc:
src/dev/isa_fake.hh:
    combine badaddr and isa fake into one
src/python/m5/objects/Pci.py:
    remove pcifake
src/python/m5/objects/Tsunami.py:
    make badaddr derive from isafake

--HG--
extra : convert_revision : 91470db60aa1de6b85827304e27bd3414cc9d8d1
2006-11-06 16:24:25 -05:00
Ali Saidi
84e07f3a51 Merge zizzer:/bk/newmem
into  zeep.pool:/z/saidi/work/m5.newmem

--HG--
extra : convert_revision : f77e5cf8cc5b99960d28e1cc109d140f1013cfca
2006-11-06 14:14:49 -05:00
Ali Saidi
0de8850136 small fixes for solaris
--HG--
extra : convert_revision : 3546b2cecf7e7e8e62295abc1ed08b3b6d2b0a8b
2006-11-06 14:14:18 -05:00
Kevin Lim
2506c15620 Merge ktlim@zizzer:/bk/newmem
into  zamp.eecs.umich.edu:/z/ktlim2/clean/newmem

--HG--
extra : convert_revision : d6bb87586cf7ee63ca32e36944c3755fae0b55d0
2006-11-06 13:27:57 -05:00
Kevin Lim
652281a61c Clean up clock phase drift code a bit.
src/cpu/base.cc:
    Move clock phase drift code to the base CPU so that any CPU model can use it.
src/cpu/base.hh:
    Added two functions to help get the next cycle the CPU should be scheduled.
src/cpu/simple/atomic.cc:
src/cpu/simple/timing.cc:
    Use the function now in BaseCPU.

--HG--
extra : convert_revision : 444494b66ffc85fc473c23f57683c5f9458ad80c
2006-11-06 13:27:45 -05:00
Ali Saidi
430622c173 replace NULL with 0.... Why isn't NULL defined by default on Mac OS X I don't know
--HG--
extra : convert_revision : b60403445bd4e855732fd4e6753068abd90ecc9d
2006-11-06 10:15:27 -05:00
Kevin Lim
257e09d626 Update refs.
--HG--
extra : convert_revision : 61d298fb0d9a66a76209a6bfcdb7c14f2efca947
2006-11-05 20:42:05 -05:00
Kevin Lim
067c9c5531 Initialize pointer to NULL.
src/cpu/o3/lsq_unit_impl.hh:
    Be sure to initialize pointer to NULL.

--HG--
extra : convert_revision : 917d5119e4bd8eae10959ed07069d8c694315c7a
2006-11-05 20:29:38 -05:00
Ali Saidi
de90ae4825 Merge zizzer:/bk/newmem
into  zeep.pool:/z/saidi/work/m5.newmem

--HG--
extra : convert_revision : d7133e32cfca9f15869ee9ab7a93e3470e7d9038
2006-11-04 21:41:10 -05:00
Ali Saidi
21cf4a46b9 fixes so that M5 will compile under solaris
SConstruct:
    Add check to see if we need to include libsocket
src/arch/sparc/floatregfile.cc:
src/arch/sparc/intregfile.cc:
    use memset rather than bzero and include the appropriate headerfile
src/base/pollevent.cc:
    If we're compling under solaris we need sys/file.h
src/base/random.cc:
src/base/random.hh:
    solaris doesn't have random(), so use rint with the correct rounding mode
    if we're compiling on solaris
src/base/stats/flags.hh:
    u_int32_t??
src/base/time.hh:
    grab the timersub() define from freebsd since it doesn't exist in solaris
src/cpu/inst_seq.hh:
    we don't need to include stdint here
src/sim/byteswap.hh:
    the method to detect endianness on Solaris is a little more complex...

--HG--
extra : convert_revision : 6b7db0e900e7bccfc250d65c125065f27280dda1
2006-11-04 21:41:01 -05:00
Gabe Black
601822c6b5 Make things compile in SE again.
--HG--
extra : convert_revision : cf7faf5001b31d61c61ddce2386d61c919075800
2006-11-03 14:42:12 -05:00
Gabe Black
8778d85b2d Use a PowerOnReset to initialize the cpu.
--HG--
extra : convert_revision : 9e65af095c37c7c67db377424d2d4363fa8065f9
2006-11-03 14:41:27 -05:00
Gabe Black
6ad386f1a8 Calling syscalls from within the trap instruction's invoke method won't work because apparently you need an xc for that and not a tc. Cleaned up the TrapInstruction fault in light of this.
--HG--
extra : convert_revision : 1805c9244cfd62d0ee7862d8fd7c9983e00c5747
2006-11-03 14:40:35 -05:00
Gabe Black
4a5cb3f425 The tc needs to be protected instead of private so that the CpuEventWrapper can access it.
--HG--
extra : convert_revision : bd836d63ac3630b20dda552e7b289730f3c114ef
2006-11-03 11:05:56 -05:00
Gabe Black
29a79acb7c Gutted out the old Alpha stuff.
--HG--
extra : convert_revision : 6767dc1305a58e3e7eb0ee909d54768e51744927
2006-11-03 11:05:13 -05:00
Gabe Black
3f4b098985 Added a stub initCPU function. This would be a good place to force in a PowerOnReset fault to kick start the CPU.
--HG--
extra : convert_revision : 79e1fa2ef40e326682069639e260db255fd29d93
2006-11-03 11:04:10 -05:00
Gabe Black
6b701a6d25 Compilation fixes.
--HG--
extra : convert_revision : 44d67a3bb95f875f17586499aa4a04268aa2fd46
2006-11-03 11:03:03 -05:00
Gabe Black
70c0ba22c4 Added this constant to get compilation to work. The value is bogus since I don't know what it actually represents.
--HG--
extra : convert_revision : ab579c1275bfcfb7ffe21633bd8c5b9bea24015e
2006-11-03 10:59:24 -05:00
Gabe Black
ab651344dd Add the syscall number as the second parameter for the trap fault. This could be improved and syscalls could be called from the trap's invoke method.
--HG--
extra : convert_revision : 127a3673a076110fb3605c0fbc93e8d7e9fec84b
2006-11-03 10:56:47 -05:00
Gabe Black
e6fed44625 Add an invoke function for PowerOnReset
--HG--
extra : convert_revision : a1cdd35c74f6e85f42a04061b466ec7617da8ac2
2006-11-03 10:55:29 -05:00
Gabe Black
694323b7c4 Move around misc reg code
src/arch/sparc/faults.cc:
    Moved some code here from miscregfile.cc
src/arch/sparc/miscregfile.cc:
    Moved code from here to faults.cc, and merged (read|set)MiscRegWithEffect and it's FS version from ua2005.cc
src/arch/sparc/miscregfile.hh:
    readFSRegWithEffect is no longer a seperate function, and is instead done in the main readRegWith Effect.

--HG--
extra : convert_revision : 0b45f0f78e83929b32ddd2f443c8b1dbf9bc04fb
2006-11-03 10:54:34 -05:00
Gabe Black
7c5a859243 removed ua2005.cc since it's been obsorbed into the miscregfile, and added system.cc
--HG--
extra : convert_revision : 2a124adcefe0d15860632a05e8788d3fd34008c2
2006-11-03 10:52:08 -05:00
Gabe Black
118b9dc1f9 Got rid of "inPalMode". Some places are still effectively checking if they are in PAL mode, however.
--HG--
extra : convert_revision : b52d9642efc474eaf97437fa2df879efefa0062b
2006-11-03 04:25:33 -05:00
Gabe Black
c8fc116c76 Add a new file which describes an ISA's interrupt handling mechanism. It records when interrupts are requested, and returns an interrupt to execute if the
--HG--
extra : convert_revision : c535000a6a170caefd441687b60f940513d29739
2006-11-03 02:25:39 -05:00
Gabe Black
fa91832900 Fixed a comment
--HG--
extra : convert_revision : bebc701508e1d38ee74a07377c634d5e46e89abe
2006-11-03 01:15:31 -05:00
Kevin Lim
e71ccde663 Merge ktlim@zizzer:/bk/newmem
into  zamp.eecs.umich.edu:/z/ktlim2/clean/newmem-busfix

--HG--
extra : convert_revision : a9a41e2c292bd95aa148e1cf4d9a77c0622a462b
2006-11-02 15:20:47 -05:00
Kevin Lim
45363ea658 Have bus use the BadAddress device to handle bad addresses. The O3 CPU should be able to boot into Linux with caches on after this change.
src/mem/bus.cc:
src/mem/bus.hh:
    Bus now will be setup with a default responder, unless the user overrides it.  This default responder should return BadAddress if no matching port is found.
src/python/m5/objects/Bus.py:
    Bus now has a default responder for FS mode if the user doesn't override it.  It returns BadAddress if no matching port is found.
src/python/m5/objects/Tsunami.py:
    Add bad address device.  Also record when the user has specified their own default responder.

--HG--
extra : convert_revision : 59070477ae313ee711b2d59baa2369c9a91c5b85
2006-11-02 15:20:37 -05:00
Kevin Lim
c3485a6548 Implement device that will return BadAddress.
--HG--
extra : convert_revision : d833c20f691e01c84a0678f19f7d83f3ee50c0c1
2006-11-02 15:18:35 -05:00
Kevin Lim
8d53f298a6 Caches return a new functional port whenever asked for one.
src/mem/cache/base_cache.cc:
    Have caches return a new functional port whenever asked for them.  I'm pretty sure this is desired behavior.  Ron can correct me if it's not.

--HG--
extra : convert_revision : e1fadf895a7d714968128ff900d10e86fde53387
2006-11-02 15:17:45 -05:00
Kevin Lim
dd5e2cd959 More proper handling of the ports.
src/cpu/simple_thread.cc:
    Fix up port handling to share code.
src/cpu/thread_state.cc:
    Separate code off into a function.
src/cpu/thread_state.hh:
    Make a separate function that will get the CPU's memory's functional port.

--HG--
extra : convert_revision : 96a9bb3c5e4b9ba5511678c0fd17f0017c8cd312
2006-11-02 14:58:31 -05:00
Kevin Lim
64f8cd12c6 Remove function that should have been deleted.
src/cpu/simple_thread.cc:
    This function should have been deleted from an earlier push.
src/cpu/simple_thread.hh:
    Delete this function; it's now in thread_state.hh/.cc.

--HG--
extra : convert_revision : f78dcf9c2b388418030d48d0ea4911c8b8b1f5ff
2006-11-02 13:12:36 -05:00
Kevin Lim
ccaf80cc46 Use ISA specific makeExtMI.
src/arch/alpha/utility.hh:
    For now makeExtMI will be specific to the ISA.

--HG--
extra : convert_revision : 89959c6499efcc3df9301ad8ea039580764a1496
2006-11-02 13:11:38 -05:00
Lisa Hsu
683d8f0831 Merge zizzer:/bk/newmem
into  zed.eecs.umich.edu:/z/hsul/work/m5/newmem

--HG--
extra : convert_revision : a30e2da1f0a272b8c867c0e7a3491118be92bc5e
2006-11-01 19:25:20 -05:00
Lisa Hsu
74ff45d353 factor some more commone code and enable going from checkpoint into arbitrary CPU with or without caches.
configs/common/Simulation.py:
    enable going from checkpoint into arbitrary CPU with or without caches.

--HG--
extra : convert_revision : 02e7ff8982fdb3a08bc609f89bd58df5b3a581b2
2006-11-01 19:25:09 -05:00
Gabe Black
b565660c42 Merge zizzer.eecs.umich.edu:/bk/newmem/
into  zeep.eecs.umich.edu:/home/gblack/m5/newmemmemops

--HG--
extra : convert_revision : c2f7398a0d14dd11108579bb243ada7420285a22
2006-11-01 19:00:59 -05:00
Gabe Black
8dbab9f701 Added code to handle draining.
--HG--
extra : convert_revision : 3861f553bde5865cd21a8a58a4c410896726f0a3
2006-11-01 19:00:49 -05:00
Gabe Black
6f78d49410 Fix a range check on the ipr_index.
--HG--
extra : convert_revision : 84e25abd4bb2de0c877c883804d39feb019c7030
2006-11-01 18:46:18 -05:00
Gabe Black
2b11b47357 Adjustments for the AlphaTLB changing to AlphaISA::TLB and changing register file functions to not take faults
--HG--
extra : convert_revision : 1cef0734462ee2e4db12482462c2ab3c134d3675
2006-11-01 16:44:45 -05:00
Lisa Hsu
7665be4f70 make it so that you can do a standard switch without the caches option. this will have only the o3 cpu have a cache, rather than timing (warmup) + o3 have cache.
--HG--
extra : convert_revision : d733de7ebb362bbd7376a0235ee7f117df2d6d37
2006-11-01 11:49:39 -05:00
Lisa Hsu
9ef8bf74c7 change name of 2nd switch_cpu so that ckpt recovery with multiple cpus doens't get confused.
--HG--
extra : convert_revision : 16c710c4196c520d03c1993a26f38cf1f04ab637
2006-11-01 11:40:49 -05:00
Gabe Black
f3ba6d20f6 Arg!
--HG--
extra : convert_revision : 8328d002780c0291e7eb264076a62084de88b7a5
2006-10-31 18:59:50 -05:00