without using the jobfile.
util/stats/db.py:
util/stats/profile.py:
Make it possible to send job as a string and to set the system
separately from the job.
--HG--
extra : convert_revision : 08aaebd3f9a1643bd41953b43f3b80dc97e6592f
cpu/simple/cpu.cc:
Properly set the Instruction Read bit in the Memory Request
--HG--
extra : convert_revision : e1a4756f32718fd8ef3ac3db16625bd6d8f07cc5
dev/pktfifo.hh:
we can't modify i because it's used further down to remove
the packet from the fifo. Instead, copy the iterator and
modify that to get the previous packet.
dev/sinic.cc:
- don't change the transmit state and kick the machine unless
we're at the head of the txList.
- add a couple of debugging statements to figure out how far
along we've gotten in processing a packet.
- assert that the current tx vnic has something to do when
we start processing the state machine.
--HG--
extra : convert_revision : 588fe2c7d810be0e3d8d39c5cc0ec8a72119517e
from #ifdef DEBUG to #ifndef NDEBUG
base/remote_gdb.cc:
make the remote debugger gdb stuff work in m5.opt
sim/system.cc:
sim/system.hh:
make the console panic break event happen in m5.opt
--HG--
extra : convert_revision : 044a9b7cdacb058112388a31315e45c5d8cf70fd
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
and to remove elements in the middle of the fifo. These elements
do not free space, they are just marked removed. Space is only
freed from the front of the fifo.
dev/etherpkt.cc:
serialize the current slack
dev/etherpkt.hh:
add "slack" to the ethernet packet. It is to be used by any fifo that
the packet is currently in to account for extra space that the packet
may be occupying due to the fifo organization.
--HG--
extra : convert_revision : 8e7c541ba316a9a76495c54cc5f707f8fc65b6d5
options, making existing options more visible and dealing with
holes in data better.
util/stats/barchart.py:
- move the options for BarChart to a base class ChartOptions so
they can be more easily set and copied.
- add an option to set the chart size (so you can adjust the aspect ratio)
- don't do the add_subplot thing, use add_axes directly so we can
affect the size of the figure itself to make room for the legend
- make the initial array bottom floating point so we don't lose precision
- add an option to set the limits on the y axis
- use a figure legend instead of an axes legend so we can put the legend
outside of the actual chart. Also add an option to set the fontsize of
the legend.
- initial hack at outputting csv files
util/stats/db.py:
don't print out an error when the run is missing from the database
just return None, the error will be print elsewhere.
util/stats/output.py:
- make StatOutput derive from ChartOptions so that it's easier to
set default chart options.
- make the various output functions (graph, display, etc.) take the
name of the data as a parameter instead of making it a parameter to
__init__. This allows me to create the StatOutput object with
generic parameters while still being able to specialize the name
after the fact
- add support for graph_group and graph_bars to be applied to multiple
configuration groups. This results in a cross product of the groups
to be generated and used.
- flush the html file output as we go so that we can load the file
while graphs are still being generated.
- make the proxy a parameter to the graph function so the proper system's
data can be graphed
- for any groups or bars that are completely missing, remove them from
the graph. This way, if we decide not to do a set of runs, there won't
be holes in the data.
- output eps and ps by default in addition to the png.
util/stats/profile.py:
- clean up the data structures that are used to store the function
profile information and try our best to avoid keeping extra data
around that isn't used.
- make get() return None if a job is missing so we know it was
missing rather than the all zeroes thing.
- make the function profile categorization stuff total up to 100%
- Fixup the x-axis and y-axis labels.
- fix the dot file output stuff.
util/stats/stats.py:
support the new options stuff for StatOutput
--HG--
extra : convert_revision : fae35df8c57a36257ea93bc3e0a0e617edc46bb7
on the test system.
add an option for pio_delay_write to run.py
util/stats/stats.py:
full0 -> run0 due to run.py change
sim_ticks doesn't make sense with tick = ps, so use
one of the cpu's numCycles paramter
--HG--
extra : convert_revision : db9dbe014549d823edc10395f5241db5e907df01
configs/boot/iscsi-client.rcS:
Clean up the newly updated iscsi-client script.
Reduce the writeback period so writes occur more frequently.
--HG--
extra : convert_revision : 21c84f781c6537b28c753291561d23c2d0144627
arch/alpha/alpha_linux_process.cc:
sim/syscall_emul.cc:
sim/syscall_emul.hh:
Apply patch for syscall emulation provided by Antti Miettinen (apm@brigitte.dna.fi).
--HG--
extra : convert_revision : 37fbc78a927110b7798343afd2c5f37a269e42b4
output for sinic
dev/sinic.cc:
add the cpu number of the request to various panic and trace
output
--HG--
extra : convert_revision : e778a5c925e194652bec47af678488acf48c1ae0
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
pci device base class
dev/sinic.cc:
dev/sinic.hh:
use the new readBar/writeBar stuff that's in the pci device
base class
--HG--
extra : convert_revision : 8a0b2bde3cc13597785d6ea75d6e6811680bb01b
makes it easier to implement PCI device models.
dev/pcidev.cc:
default implementations for read/write and readBarX/writeBarX functions
--HG--
extra : convert_revision : bbe2e2a2a506e2dd94d98f8e0feaefef96380be9
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
python/m5/config.py:
Make Clock a NumericParamValue so you can do math with it
--HG--
extra : convert_revision : 7fa548d1a23c604a31d3ecae3853949b064a1830
util/stats/info.py:
If an operation results in a divide by zero, just return None
--HG--
extra : convert_revision : 19cb4319734a3a9cf02bb1966fed42eb0c8a8ade
base/loader/symtab.cc:
Add support for clearing out the symbol table
Add support for serializing the symbol table (clear on unserialize)
Don't allow empty symbols to be entered into the table
base/loader/symtab.hh:
Add support for clearing out the symbol table
Add support for serializing the symbol table
sim/system.cc:
Serialize the kernel, console, and palcode symbol tables so that
we can capture any dynamic symbols that are added and so that we
don't have to have the same kernel binary around to get the
symbols right
--HG--
extra : convert_revision : 779888c88aa530f3adcd37dc7600a335951d05f7
if we want something like allSymtab, we should create a symbol
table proxy class
--HG--
extra : convert_revision : 20ca551a693b0d6495c018cac8afd63af33f16da
better mysql error messages
base/stats/mysql.cc:
better placement of commit() calls to avoid failing transactions
due to what I think are timeouts.
print out the mysql error with every panic
--HG--
extra : convert_revision : bfc5ae172bcff733461adceffe2b381601839e82
profileNode if we got a new one
cpu/simple/cpu.cc:
Only update the exec context's profileNode if we got a new one
--HG--
extra : convert_revision : a16a7410070b0d811032dc4b86b1368df913b2be
are more efficient and reduce the number of new/delete calls
arch/alpha/stacktrace.cc:
- Change the StackTrace code so that the class can more easily be
cleaned out and reused to avoid extra allocations.
- Allow trace() to accept a static instruction pointer so it can
determine if the instruction is worth tracing. This is moved from
the CPU.
- provide constants for special meaning PCs (user, console, unknown),
instead of magic numbers
- switch to using kernelSymtab instead of allSymtab which will be
going away
- if the stack adjustment doesn't make any sense, exit and push
unknown so we don't get into an infinite loop or record garbage.
- check to see if we've made too many iterations through the stack
and panic to avoid an infinite loop
arch/alpha/stacktrace.hh:
- Change the StackTrace code so that the class can more easily be
cleaned out and reused to avoid extra allocations.
- Allow trace() to accept a static instruction pointer so it can
determine if the instruction is worth tracing. This is moved from
the CPU.
- provide constants for special meaning PCs (user, console, unknown),
instead of magic numbers
cpu/base.cc:
only clear the profile if we have one
include profile.hh here since base.hh doesn't do it anymore
cpu/base.hh:
no need to include cpu/profile.hh here
cpu/profile.cc:
use ProfileNode pointers instead of objects in the ChildList
Consume a vector of addresses since that's really all we
care about.
cpu/profile.hh:
Keep pointers to ProfileNodes to reduce the size of these structures
keep a StackTrace around so that we may reuse it.
provide consume functions that use the new StackTrace trace interface
one consume function is inline and tries to fastpath the no trace
condition, it calls the outlined consume function if a trace is generated.
cpu/simple/cpu.cc:
include cpu/profile.hh here since base.hh no longer does
use the new FunctionProfile::consume interface
(which contains the tracing functions)
--HG--
extra : convert_revision : 5a1d9265289a75f67a497b322926be1f8c2d8eb3
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
for retransmissions, out of order packets, lost packets, duplicate
ack, window full, etc. Easy way to see if you have a problem with a
run.
--HG--
extra : convert_revision : 95d8e8650b0fb3d120df107cd5281c56fefc3a1d
