Commit graph

8517 commits

Author SHA1 Message Date
Korey Sewell
078f914e69 inorder: SE mode TLB faults
handle them like we do in FS mode, by blocking the TLB until the fault
is handled by the fault->invoke()
2011-06-19 21:43:42 -04:00
Korey Sewell
3cb23bd3a2 inorder:tracing: fix fault tracing bug 2011-06-19 21:43:42 -04:00
Korey Sewell
fe3a2aa4a3 inorder: se compile fixes 2011-06-19 21:43:42 -04:00
Korey Sewell
e572c01120 inorder: add necessary debug flag header files 2011-06-19 21:43:41 -04:00
Korey Sewell
59686795e3 mips: mark unaligned access flag as true 2011-06-19 21:43:41 -04:00
Korey Sewell
91a88ae8ce inorder: clear fetchbuffer on traps
implement clearfetchbufferfunction
extend predecoder to use multiple threads and clear those on trap
2011-06-19 21:43:41 -04:00
Korey Sewell
2dae0e8735 inorder: use separate float-reg bits function in dyninst
this will make sure we get the correct view of a FP register
2011-06-19 21:43:41 -04:00
Korey Sewell
8c0def8d03 inorder: use trapPending flag to manage traps 2011-06-19 21:43:41 -04:00
Korey Sewell
5ef0b7a9db inorder/dtb: make sure DTB translate correct address
The DTB expects the correct PC in the ThreadContext
but how if the memory accesses are speculative? Shouldn't
we send along the requestor's PC to the translate functions?
2011-06-19 21:43:41 -04:00
Korey Sewell
716e447da8 inorder: handle serializing instructions
including IPR accesses and store-conditionals. These class of instructions will not
execute correctly in a superscalar machine
2011-06-19 21:43:41 -04:00
Korey Sewell
83a0fd24f7 alpha: fix warn_once for prefetches 2011-06-19 21:43:40 -04:00
Korey Sewell
33a6020f2e alpha: naming for dtb faults
Just "dfault" gets confusing while debugging. Why not
differentiate whether it's an access violation or page
fault
2011-06-19 21:43:40 -04:00
Korey Sewell
561c33f082 inorder: dont handle multiple faults on same cycle
if a faulting instruction reaches an execution unit,
then ignore it and pass it through the pipeline.

Once we recognize the fault in the graduation unit,
dont allow a second fault to creep in on the same cycle.
2011-06-19 21:43:40 -04:00
Korey Sewell
c4deabfb97 inorder: register ports for FS mode
handle "snoop" port registration as well as functional
port setup for FS mode
2011-06-19 21:43:40 -04:00
Korey Sewell
f1c3691356 inorder: check for interrupts each tick
use a dummy instruction to facilitate the squash after
the interrupts trap
2011-06-19 21:43:40 -04:00
Korey Sewell
0bfdf342da inorder: explicit fault check
Before graduating an instruction, explicitly check fault
by making the fault check it's own separate command
that can be put on an instruction schedule.
2011-06-19 21:43:40 -04:00
Korey Sewell
5f608dd2e9 inorder: squash and trap behind a tlb fault 2011-06-19 21:43:39 -04:00
Korey Sewell
e0e387c2a9 inorder: stall stores on store conditionals & compare/swaps 2011-06-19 21:43:39 -04:00
Korey Sewell
f268d7d004 alpha: make hwrei a control inst
this always changes the PC and is basically an impromptu branch instruction. why
not speculate on this instead of always be forced to mispredict/squash after the
hwrei gets resolved?

The InOrder model needs this marked as "isControl" so it knows to update the PC
after the ALU executes it. If this isnt marked as control, then it's going to
force the model to check the PC of every instruction at commit (what O3 does?),
and that would be a wasteful check for a very high percentage of instructions.
2011-06-19 21:43:39 -04:00
Korey Sewell
e8b7df072b inorder: make InOrder CPU FS compilable/visible
make syscall a SE mode only functionality
copy over basic FS functions (hwrei) to make FS compile
2011-06-19 21:43:39 -04:00
Korey Sewell
d71b95d84d inorder: remove memdep tracking for default pipeline
speculative load/store pipelines can reenable this
2011-06-19 21:43:39 -04:00
Korey Sewell
b72bdcf4f8 inorder: fetchBuffer tracking
calculate blocks in use for the fetch buffer to figure out how many total blocks
are pending
2011-06-19 21:43:39 -04:00
Korey Sewell
4d4c7d79d0 inorder: redefine DynInst FP result type
Sharing the FP value w/the integer values was giving inconsistent results esp. when
their is a 32-bit integer register matched w/a 64-bit float value
2011-06-19 21:43:38 -04:00
Korey Sewell
db8b1e4b78 inorder: treat SE mode syscalls as a trapping instruction
define a syscallContext to schedule the syscall and then use syscall() to actually perform the action
2011-06-19 21:43:38 -04:00
Korey Sewell
c95fe261ab inorder: bug in mdu
segfault was caused by squashed multiply thats in the process of an event.
use isProcessing flag to handle this and cleanup the MDU code
2011-06-19 21:43:38 -04:00
Korey Sewell
4c979f9325 inorder: optionally track faulting instructions 2011-06-19 21:43:38 -04:00
Korey Sewell
22ba1718c4 inorder: cleanup events in resource pool
remove events in the resource pool that can be called from the CPU event, since the CPU
event is scheduled at the same time at the resource pool event.
----
Also, match the resPool event function names to the cpu event function names
----
2011-06-19 21:43:38 -04:00
Korey Sewell
e8082a28c8 inorder: don't stall after stores
once a ST is sent off, it's OK to keep processing, however it's a little more
complicated to handle the packet acknowledging the store is completed
2011-06-19 21:43:38 -04:00
Korey Sewell
379c23199e inorder: don't stall after stores
once a ST is sent off, it's OK to keep processing, however it's a little more
complicated to handle the packet acknowledging the store is completed
2011-06-19 21:43:37 -04:00
Korey Sewell
4c9ad53cc5 inorder: remove decode squash
also, cleanup comments for gem5.fast compilation
2011-06-19 21:43:37 -04:00
Korey Sewell
a444133e73 inorder: support for compare and swap insts
dont treat read() and write() fields as mut. exclusive
2011-06-19 21:43:37 -04:00
Korey Sewell
89d0f95bf0 inorder: branch predictor update
only update BTB on a taken branch and update branch predictor w/pcstate from instruction
---
only pay attention to branch predictor updates if the the inst. is in fact a branch
2011-06-19 21:43:37 -04:00
Korey Sewell
479195d4cf inorder: priority for grad/squash events
define separate priority resource pool squash and graduate events
2011-06-19 21:43:37 -04:00
Korey Sewell
71018f5e8b inorder: remove stalls on trap squash 2011-06-19 21:43:37 -04:00
Korey Sewell
34b2500f09 inorder: no dep. tracking for zero reg
this causes forwarding a bad value register value
2011-06-19 21:43:37 -04:00
Korey Sewell
d02fa0f6b6 imported patch recoverPCfromTrap 2011-06-19 21:43:37 -04:00
Korey Sewell
264e8178ff imported patch squash_from_next_stage 2011-06-19 21:43:36 -04:00
Korey Sewell
f0f33ae2b9 inorder: add flatDestReg member to dyninst
use it in reg. dep. tracking
2011-06-19 21:43:36 -04:00
Korey Sewell
555bd4d842 inorder: update event priorities
dont use offset to calculate this but rather an enum
that can be updated
2011-06-19 21:43:36 -04:00
Korey Sewell
7dea79535c inorder: implement trap handling 2011-06-19 21:43:36 -04:00
Korey Sewell
061b369d28 inorder: cleanup intercomm. structs/squash info 2011-06-19 21:43:35 -04:00
Korey Sewell
b195da9345 inorder: use setupSquash for misspeculation
implement a clean interface to handle branch misprediction and eventually all pipeline
flushing
2011-06-19 21:43:35 -04:00
Korey Sewell
d5d4e47f76 sparc: init. cache state in TLB
valgrind complains and its a potential source of instability, so go ahead
and set it to 0 to start
2011-06-19 21:43:35 -04:00
Korey Sewell
73cfab8b23 inorder: DynInst handling of stores for big-endian ISAs
The DynInst was not performing the host-to-guest translation
which ended up breaking stores for SPARC
2011-06-19 21:43:35 -04:00
Korey Sewell
4f34bc8b7b inorder: make marking of dest. regs an explicit request
formerly, this was implicit when you accessed the execution unit
or the use-def unit but it's better that this just be something
that a user can specify.
2011-06-19 21:43:35 -04:00
Korey Sewell
946b0ed4f4 inorder: simplify handling of split accesses 2011-06-19 21:43:35 -04:00
Korey Sewell
1a6d25dc47 inorder: addtl functionaly for inst. skeds
add find and end functions for inst. schedules
that can search by stage number
2011-06-19 21:43:35 -04:00
Korey Sewell
8b54858831 inorder: register file stats
keep stats for int/float reg file usage instead
of aggregating across reg file types
2011-06-19 21:43:34 -04:00
Korey Sewell
085f30ff9c inorder: scheduling for nonspec insts
make handling of speculative and nonspeculative insts
more explicit
2011-06-19 21:43:34 -04:00
Korey Sewell
3c417ea23a inorder: find register dependencies "lazily"
Architectures like SPARC need to read the window pointer
in order to figure out it's register dependence. However,
this may not get updated until after an instruction gets
executed, so now we lazily detect the register dependence
in the EXE stage (execution unit or use_def). This
makes sure we get the mapping after the most current change.
2011-06-19 21:43:34 -04:00