gem5  v21.1.0.2
Classes | Public Types | Public Member Functions | Public Attributes | Protected Attributes | Private Attributes | List of all members
gem5::Shader Class Reference

#include <shader.hh>

Inheritance diagram for gem5::Shader:
gem5::ClockedObject gem5::SimObject gem5::Clocked gem5::EventManager gem5::Serializable gem5::Drainable gem5::statistics::Group gem5::Named

Classes

struct  ShaderStats
 

Public Types

enum  hsail_mode_e { SIMT, VECTOR_SCALAR }
 
typedef ShaderParams Params
 
- Public Types inherited from gem5::ClockedObject
using Params = ClockedObjectParams
 Parameters of ClockedObject. More...
 
- Public Types inherited from gem5::SimObject
typedef SimObjectParams Params
 

Public Member Functions

GPUDispatcherdispatcher ()
 
void sampleLoad (const Tick accessTime)
 
void sampleStore (const Tick accessTime)
 
void sampleInstRoundTrip (std::vector< Tick > roundTripTime)
 
void sampleLineRoundTrip (const std::map< Addr, std::vector< Tick >> &roundTripTime)
 
const ApertureRegistergpuVmApe () const
 
const ApertureRegisterldsApe () const
 
const ApertureRegisterscratchApe () const
 
bool isGpuVmApe (Addr addr) const
 
bool isLdsApe (Addr addr) const
 
bool isScratchApe (Addr addr) const
 
Addr getScratchBase ()
 
Addr getHiddenPrivateBase ()
 
void initShHiddenPrivateBase (Addr queueBase, uint32_t offset)
 
 Shader (const Params &p)
 
 ~Shader ()
 
virtual void init ()
 init() is called after all C++ SimObjects have been created and all ports are connected. More...
 
void execScheduledAdds ()
 
void ScheduleAdd (int *val, Tick when, int x)
 
bool processTimingPacket (PacketPtr pkt)
 
void AccessMem (uint64_t address, void *ptr, uint32_t size, int cu_id, MemCmd cmd, bool suppress_func_errors)
 
void ReadMem (uint64_t address, void *ptr, uint32_t sz, int cu_id)
 
void ReadMem (uint64_t address, void *ptr, uint32_t sz, int cu_id, bool suppress_func_errors)
 
void WriteMem (uint64_t address, void *ptr, uint32_t sz, int cu_id)
 
void WriteMem (uint64_t address, void *ptr, uint32_t sz, int cu_id, bool suppress_func_errors)
 
void doFunctionalAccess (const RequestPtr &req, MemCmd cmd, void *data, bool suppress_func_errors, int cu_id)
 
void registerCU (int cu_id, ComputeUnit *compute_unit)
 
void prepareInvalidate (HSAQueueEntry *task)
 
void prepareFlush (GPUDynInstPtr gpuDynInst)
 dispatcher/shader arranges flush requests to the CUs More...
 
bool dispatchWorkgroups (HSAQueueEntry *task)
 
Addr mmap (int length)
 
void functionalTLBAccess (PacketPtr pkt, int cu_id, BaseMMU::Mode mode)
 
void updateContext (int cid)
 
void notifyCuSleep ()
 
void incVectorInstSrcOperand (int num_operands)
 
void incVectorInstDstOperand (int num_operands)
 
- Public Member Functions inherited from gem5::ClockedObject
 ClockedObject (const ClockedObjectParams &p)
 
void serialize (CheckpointOut &cp) const override
 Serialize an object. More...
 
void unserialize (CheckpointIn &cp) override
 Unserialize an object. More...
 
- Public Member Functions inherited from gem5::SimObject
const Paramsparams () const
 
 SimObject (const Params &p)
 
virtual ~SimObject ()
 
virtual void loadState (CheckpointIn &cp)
 loadState() is called on each SimObject when restoring from a checkpoint. More...
 
virtual void initState ()
 initState() is called on each SimObject when not restoring from a checkpoint. More...
 
virtual void regProbePoints ()
 Register probe points for this object. More...
 
virtual void regProbeListeners ()
 Register probe listeners for this object. More...
 
ProbeManagergetProbeManager ()
 Get the probe manager for this object. More...
 
virtual PortgetPort (const std::string &if_name, PortID idx=InvalidPortID)
 Get a port with a given name and index. More...
 
virtual void startup ()
 startup() is the final initialization call before simulation. More...
 
DrainState drain () override
 Provide a default implementation of the drain interface for objects that don't need draining. More...
 
virtual void memWriteback ()
 Write back dirty buffers to memory using functional writes. More...
 
virtual void memInvalidate ()
 Invalidate the contents of memory buffers. More...
 
void serialize (CheckpointOut &cp) const override
 Serialize an object. More...
 
void unserialize (CheckpointIn &cp) override
 Unserialize an object. More...
 
- Public Member Functions inherited from gem5::EventManager
EventQueueeventQueue () const
 
void schedule (Event &event, Tick when)
 
void deschedule (Event &event)
 
void reschedule (Event &event, Tick when, bool always=false)
 
void schedule (Event *event, Tick when)
 
void deschedule (Event *event)
 
void reschedule (Event *event, Tick when, bool always=false)
 
void wakeupEventQueue (Tick when=(Tick) -1)
 This function is not needed by the usual gem5 event loop but may be necessary in derived EventQueues which host gem5 on other schedulers. More...
 
void setCurTick (Tick newVal)
 
 EventManager (EventManager &em)
 Event manger manages events in the event queue. More...
 
 EventManager (EventManager *em)
 
 EventManager (EventQueue *eq)
 
- Public Member Functions inherited from gem5::Serializable
 Serializable ()
 
virtual ~Serializable ()
 
void serializeSection (CheckpointOut &cp, const char *name) const
 Serialize an object into a new section. More...
 
void serializeSection (CheckpointOut &cp, const std::string &name) const
 
void unserializeSection (CheckpointIn &cp, const char *name)
 Unserialize an a child object. More...
 
void unserializeSection (CheckpointIn &cp, const std::string &name)
 
- Public Member Functions inherited from gem5::Drainable
DrainState drainState () const
 Return the current drain state of an object. More...
 
virtual void notifyFork ()
 Notify a child process of a fork. More...
 
- Public Member Functions inherited from gem5::statistics::Group
 Group (Group *parent, const char *name=nullptr)
 Construct a new statistics group. More...
 
virtual ~Group ()
 
virtual void regStats ()
 Callback to set stat parameters. More...
 
virtual void resetStats ()
 Callback to reset stats. More...
 
virtual void preDumpStats ()
 Callback before stats are dumped. More...
 
void addStat (statistics::Info *info)
 Register a stat with this group. More...
 
const std::map< std::string, Group * > & getStatGroups () const
 Get all child groups associated with this object. More...
 
const std::vector< Info * > & getStats () const
 Get all stats associated with this object. More...
 
void addStatGroup (const char *name, Group *block)
 Add a stat block as a child of this block. More...
 
const InforesolveStat (std::string name) const
 Resolve a stat by its name within this group. More...
 
void mergeStatGroup (Group *block)
 Merge the contents (stats & children) of a block to this block. More...
 
 Group ()=delete
 
 Group (const Group &)=delete
 
Groupoperator= (const Group &)=delete
 
- Public Member Functions inherited from gem5::Named
 Named (const std::string &name_)
 
virtual ~Named ()=default
 
virtual std::string name () const
 
- Public Member Functions inherited from gem5::Clocked
void updateClockPeriod ()
 Update the tick to the current tick. More...
 
Tick clockEdge (Cycles cycles=Cycles(0)) const
 Determine the tick when a cycle begins, by default the current one, but the argument also enables the caller to determine a future cycle. More...
 
Cycles curCycle () const
 Determine the current cycle, corresponding to a tick aligned to a clock edge. More...
 
Tick nextCycle () const
 Based on the clock of the object, determine the start tick of the first cycle that is at least one cycle in the future. More...
 
uint64_t frequency () const
 
Tick clockPeriod () const
 
double voltage () const
 
Cycles ticksToCycles (Tick t) const
 
Tick cyclesToTicks (Cycles c) const
 

Public Attributes

SimpleThreadcpuThread
 
ThreadContextgpuTc
 
BaseCPUcpuPointer
 
EventFunctionWrapper tickEvent
 
bool timingSim
 
hsail_mode_e hsail_mode
 
int impl_kern_launch_acq
 
int impl_kern_end_rel
 
int coissue_return
 
int trace_vgpr_all
 
int n_cu
 
int n_wf
 
int globalMemSize
 
int nextSchedCu
 
uint32_t sa_n
 
std::vector< int * > sa_val
 
std::vector< uint64_t > sa_when
 
std::vector< int32_t > sa_x
 
std::vector< ComputeUnit * > cuList
 
GPUCommandProcessorgpuCmdProc
 
GPUDispatcher_dispatcher
 
int64_t max_valu_insts
 
int64_t total_valu_insts
 
- Public Attributes inherited from gem5::ClockedObject
PowerStatepowerState
 

Protected Attributes

gem5::Shader::ShaderStats stats
 
- Protected Attributes inherited from gem5::SimObject
const SimObjectParams & _params
 Cached copy of the object parameters. More...
 
- Protected Attributes inherited from gem5::EventManager
EventQueueeventq
 A pointer to this object's event queue. More...
 

Private Attributes

ApertureRegister _gpuVmApe
 
ApertureRegister _ldsApe
 
ApertureRegister _scratchApe
 
Addr shHiddenPrivateBaseVmid
 
int _activeCus
 
Tick _lastInactiveTick
 

Additional Inherited Members

- Static Public Member Functions inherited from gem5::SimObject
static void serializeAll (const std::string &cpt_dir)
 Create a checkpoint by serializing all SimObjects in the system. More...
 
static SimObjectfind (const char *name)
 Find the SimObject with the given name and return a pointer to it. More...
 
static void setSimObjectResolver (SimObjectResolver *resolver)
 There is a single object name resolver, and it is only set when simulation is restoring from checkpoints. More...
 
static SimObjectResolvergetSimObjectResolver ()
 There is a single object name resolver, and it is only set when simulation is restoring from checkpoints. More...
 
- Static Public Member Functions inherited from gem5::Serializable
static const std::string & currentSection ()
 Gets the fully-qualified name of the active section. More...
 
static void generateCheckpointOut (const std::string &cpt_dir, std::ofstream &outstream)
 Generate a checkpoint file so that the serialization can be routed to it. More...
 
- Protected Member Functions inherited from gem5::Drainable
 Drainable ()
 
virtual ~Drainable ()
 
virtual void drainResume ()
 Resume execution after a successful drain. More...
 
void signalDrainDone () const
 Signal that an object is drained. More...
 
- Protected Member Functions inherited from gem5::Clocked
 Clocked (ClockDomain &clk_domain)
 Create a clocked object and set the clock domain based on the parameters. More...
 
 Clocked (Clocked &)=delete
 
Clockedoperator= (Clocked &)=delete
 
virtual ~Clocked ()
 Virtual destructor due to inheritance. More...
 
void resetClock () const
 Reset the object's clock using the current global tick value. More...
 
virtual void clockPeriodUpdated ()
 A hook subclasses can implement so they can do any extra work that's needed when the clock rate is changed. More...
 

Detailed Description

Definition at line 84 of file shader.hh.

Member Typedef Documentation

◆ Params

typedef ShaderParams gem5::Shader::Params

Definition at line 99 of file shader.hh.

Member Enumeration Documentation

◆ hsail_mode_e

Enumerator
SIMT 
VECTOR_SCALAR 

Definition at line 100 of file shader.hh.

Constructor & Destructor Documentation

◆ Shader()

gem5::Shader::Shader ( const Params p)

Definition at line 58 of file shader.cc.

References execScheduledAdds().

◆ ~Shader()

gem5::Shader::~Shader ( )

Definition at line 148 of file shader.cc.

References cuList, gem5::ArmISA::j, and n_cu.

Member Function Documentation

◆ AccessMem()

void gem5::Shader::AccessMem ( uint64_t  address,
void *  ptr,
uint32_t  size,
int  cu_id,
MemCmd  cmd,
bool  suppress_func_errors 
)

Definition at line 378 of file shader.cc.

References cuList, doFunctionalAccess(), and gem5::ChunkGenerator::done().

Referenced by ReadMem(), and WriteMem().

◆ dispatcher()

GPUDispatcher & gem5::Shader::dispatcher ( )

◆ dispatchWorkgroups()

bool gem5::Shader::dispatchWorkgroups ( HSAQueueEntry task)

◆ doFunctionalAccess()

void gem5::Shader::doFunctionalAccess ( const RequestPtr req,
MemCmd  cmd,
void *  data,
bool  suppress_func_errors,
int  cu_id 
)

◆ execScheduledAdds()

void gem5::Shader::execScheduledAdds ( )

◆ functionalTLBAccess()

void gem5::Shader::functionalTLBAccess ( PacketPtr  pkt,
int  cu_id,
BaseMMU::Mode  mode 
)

Definition at line 429 of file shader.cc.

References cuList, gpuTc, gem5::ArmISA::mode, and gem5::Packet::senderState.

Referenced by doFunctionalAccess().

◆ getHiddenPrivateBase()

Addr gem5::Shader::getHiddenPrivateBase ( )
inline

Definition at line 163 of file shader.hh.

References shHiddenPrivateBaseVmid.

Referenced by gem5::GPUDynInst::resolveFlatSegment().

◆ getScratchBase()

Addr gem5::Shader::getScratchBase ( )
inline

Definition at line 157 of file shader.hh.

References _scratchApe, and gem5::ApertureRegister::base.

Referenced by gem5::GPUDynInst::resolveFlatSegment().

◆ gpuVmApe()

const ApertureRegister& gem5::Shader::gpuVmApe ( ) const
inline

Definition at line 114 of file shader.hh.

References _gpuVmApe.

◆ incVectorInstDstOperand()

void gem5::Shader::incVectorInstDstOperand ( int  num_operands)
inline

Definition at line 281 of file shader.hh.

References stats, and gem5::Shader::ShaderStats::vectorInstDstOperand.

Referenced by gem5::Wavefront::exec().

◆ incVectorInstSrcOperand()

void gem5::Shader::incVectorInstSrcOperand ( int  num_operands)
inline

Definition at line 275 of file shader.hh.

References stats, and gem5::Shader::ShaderStats::vectorInstSrcOperand.

Referenced by gem5::Wavefront::exec().

◆ init()

void gem5::Shader::init ( )
virtual

init() is called after all C++ SimObjects have been created and all ports are connected.

Initializations that are independent of unserialization but rely on a fully instantiated and connected SimObject graph should be done here.

Reimplemented from gem5::SimObject.

Definition at line 140 of file shader.cc.

References cpuPointer, gem5::BaseCPU::getContext(), and gpuTc.

◆ initShHiddenPrivateBase()

void gem5::Shader::initShHiddenPrivateBase ( Addr  queueBase,
uint32_t  offset 
)
inline

Definition at line 169 of file shader.hh.

References gem5::ArmISA::offset, panic_if, and shHiddenPrivateBaseVmid.

Referenced by gem5::Wavefront::initRegState().

◆ isGpuVmApe()

bool gem5::Shader::isGpuVmApe ( Addr  addr) const
inline

◆ isLdsApe()

bool gem5::Shader::isLdsApe ( Addr  addr) const
inline

◆ isScratchApe()

bool gem5::Shader::isScratchApe ( Addr  addr) const
inline

◆ ldsApe()

const ApertureRegister& gem5::Shader::ldsApe ( ) const
inline

Definition at line 120 of file shader.hh.

References _ldsApe.

Referenced by gem5::GPUDynInst::resolveFlatSegment().

◆ mmap()

Addr gem5::Shader::mmap ( int  length)

◆ notifyCuSleep()

void gem5::Shader::notifyCuSleep ( )

◆ prepareFlush()

void gem5::Shader::prepareFlush ( GPUDynInstPtr  gpuDynInst)

◆ prepareInvalidate()

void gem5::Shader::prepareInvalidate ( HSAQueueEntry task)

◆ processTimingPacket()

bool gem5::Shader::processTimingPacket ( PacketPtr  pkt)

◆ ReadMem() [1/2]

void gem5::Shader::ReadMem ( uint64_t  address,
void *  ptr,
uint32_t  sz,
int  cu_id 
)

Definition at line 396 of file shader.cc.

References AccessMem(), and gem5::MemCmd::ReadReq.

◆ ReadMem() [2/2]

void gem5::Shader::ReadMem ( uint64_t  address,
void *  ptr,
uint32_t  sz,
int  cu_id,
bool  suppress_func_errors 
)

Definition at line 402 of file shader.cc.

References AccessMem(), and gem5::MemCmd::ReadReq.

◆ registerCU()

void gem5::Shader::registerCU ( int  cu_id,
ComputeUnit compute_unit 
)
inline

Definition at line 260 of file shader.hh.

References cuList.

◆ sampleInstRoundTrip()

void gem5::Shader::sampleInstRoundTrip ( std::vector< Tick roundTripTime)

◆ sampleLineRoundTrip()

void gem5::Shader::sampleLineRoundTrip ( const std::map< Addr, std::vector< Tick >> &  roundTripTime)

◆ sampleLoad()

void gem5::Shader::sampleLoad ( const Tick  accessTime)

◆ sampleStore()

void gem5::Shader::sampleStore ( const Tick  accessTime)

◆ ScheduleAdd()

void gem5::Shader::ScheduleAdd ( int *  val,
Tick  when,
int  x 
)

◆ scratchApe()

const ApertureRegister& gem5::Shader::scratchApe ( ) const
inline

Definition at line 126 of file shader.hh.

References _scratchApe.

◆ updateContext()

void gem5::Shader::updateContext ( int  cid)

Definition at line 155 of file shader.cc.

References cpuPointer, gem5::BaseCPU::getContext(), and gpuTc.

Referenced by gem5::GPUDispatcher::exec().

◆ WriteMem() [1/2]

void gem5::Shader::WriteMem ( uint64_t  address,
void *  ptr,
uint32_t  sz,
int  cu_id 
)

Definition at line 410 of file shader.cc.

References AccessMem(), and gem5::MemCmd::WriteReq.

◆ WriteMem() [2/2]

void gem5::Shader::WriteMem ( uint64_t  address,
void *  ptr,
uint32_t  sz,
int  cu_id,
bool  suppress_func_errors 
)

Definition at line 416 of file shader.cc.

References AccessMem(), and gem5::MemCmd::WriteReq.

Member Data Documentation

◆ _activeCus

int gem5::Shader::_activeCus
private

Definition at line 93 of file shader.hh.

Referenced by dispatchWorkgroups(), and notifyCuSleep().

◆ _dispatcher

GPUDispatcher& gem5::Shader::_dispatcher

Definition at line 227 of file shader.hh.

Referenced by dispatcher(), prepareFlush(), and prepareInvalidate().

◆ _gpuVmApe

ApertureRegister gem5::Shader::_gpuVmApe
private

Definition at line 87 of file shader.hh.

Referenced by gpuVmApe(), and isGpuVmApe().

◆ _lastInactiveTick

Tick gem5::Shader::_lastInactiveTick
private

Definition at line 96 of file shader.hh.

Referenced by dispatchWorkgroups(), and notifyCuSleep().

◆ _ldsApe

ApertureRegister gem5::Shader::_ldsApe
private

Definition at line 88 of file shader.hh.

Referenced by isLdsApe(), and ldsApe().

◆ _scratchApe

ApertureRegister gem5::Shader::_scratchApe
private

Definition at line 89 of file shader.hh.

Referenced by getScratchBase(), isScratchApe(), and scratchApe().

◆ coissue_return

int gem5::Shader::coissue_return

◆ cpuPointer

BaseCPU* gem5::Shader::cpuPointer

Definition at line 111 of file shader.hh.

Referenced by init(), and updateContext().

◆ cpuThread

SimpleThread* gem5::Shader::cpuThread

Definition at line 109 of file shader.hh.

◆ cuList

std::vector<ComputeUnit*> gem5::Shader::cuList

◆ globalMemSize

int gem5::Shader::globalMemSize

Definition at line 208 of file shader.hh.

Referenced by gem5::GlobalMemPipeline::init().

◆ gpuCmdProc

GPUCommandProcessor& gem5::Shader::gpuCmdProc

Definition at line 226 of file shader.hh.

Referenced by gem5::ComputeUnit::sendRequest().

◆ gpuTc

ThreadContext* gem5::Shader::gpuTc

◆ hsail_mode

hsail_mode_e gem5::Shader::hsail_mode

Definition at line 192 of file shader.hh.

Referenced by gem5::Wavefront::exec().

◆ impl_kern_end_rel

int gem5::Shader::impl_kern_end_rel

◆ impl_kern_launch_acq

int gem5::Shader::impl_kern_launch_acq

Definition at line 195 of file shader.hh.

Referenced by gem5::GPUDispatcher::exec().

◆ max_valu_insts

int64_t gem5::Shader::max_valu_insts

Definition at line 229 of file shader.hh.

Referenced by gem5::ComputeUnit::updateInstStats().

◆ n_cu

int gem5::Shader::n_cu

Definition at line 203 of file shader.hh.

Referenced by dispatchWorkgroups(), prepareInvalidate(), and ~Shader().

◆ n_wf

int gem5::Shader::n_wf

◆ nextSchedCu

int gem5::Shader::nextSchedCu

Definition at line 211 of file shader.hh.

Referenced by dispatchWorkgroups().

◆ sa_n

uint32_t gem5::Shader::sa_n

Definition at line 214 of file shader.hh.

Referenced by execScheduledAdds(), and ScheduleAdd().

◆ sa_val

std::vector<int*> gem5::Shader::sa_val

Definition at line 217 of file shader.hh.

Referenced by execScheduledAdds(), and ScheduleAdd().

◆ sa_when

std::vector<uint64_t> gem5::Shader::sa_when

Definition at line 219 of file shader.hh.

Referenced by execScheduledAdds(), and ScheduleAdd().

◆ sa_x

std::vector<int32_t> gem5::Shader::sa_x

Definition at line 221 of file shader.hh.

Referenced by execScheduledAdds(), and ScheduleAdd().

◆ shHiddenPrivateBaseVmid

Addr gem5::Shader::shHiddenPrivateBaseVmid
private

Definition at line 90 of file shader.hh.

Referenced by getHiddenPrivateBase(), and initShHiddenPrivateBase().

◆ stats

gem5::Shader::ShaderStats gem5::Shader::stats
protected

◆ tickEvent

EventFunctionWrapper gem5::Shader::tickEvent

Definition at line 188 of file shader.hh.

Referenced by dispatchWorkgroups(), execScheduledAdds(), and ScheduleAdd().

◆ timingSim

bool gem5::Shader::timingSim

Definition at line 191 of file shader.hh.

Referenced by gem5::FetchUnit::init(), and gem5::ComputeUnit::sendRequest().

◆ total_valu_insts

int64_t gem5::Shader::total_valu_insts

Definition at line 230 of file shader.hh.

Referenced by gem5::ComputeUnit::updateInstStats().

◆ trace_vgpr_all

int gem5::Shader::trace_vgpr_all

Definition at line 201 of file shader.hh.


The documentation for this class was generated from the following files:

Generated on Tue Sep 21 2021 12:28:09 for gem5 by doxygen 1.8.17