TLB TranslationState: this currently is a somewhat bastardization of the usage of SenderState, whereby the receiver of a packet is not usually supposed to need to look at the contents of the senderState, you're really only supposed to look at what you pushed on, pop it off, and send it back. More...

#include <gpu_tlb.hh>

Inheritance diagram for X86ISA::GpuTLB::TranslationState:

Public Member Functions
	TranslationState (Mode tlb_mode, ThreadContext _tc, bool _prefetch=false, Packet::SenderState _saved=nullptr)

Public Member Functions inherited from Packet::SenderState
	SenderState ()

virtual	~SenderState ()

Public Attributes
Mode	tlbMode

ThreadContext *	tc

GpuTlbEntry *	tlbEntry

bool	prefetch

uint64_t	issueTime

std::vector< SlavePort * >	ports

std::vector< int >	reqCnt

int	hitLevel

Packet::SenderState *	saved

Public Attributes inherited from Packet::SenderState
SenderState *	predecessor

Detailed Description

TLB TranslationState: this currently is a somewhat bastardization of the usage of SenderState, whereby the receiver of a packet is not usually supposed to need to look at the contents of the senderState, you're really only supposed to look at what you pushed on, pop it off, and send it back.

However, since there is state that we want to pass to the TLBs using the send/recv Timing/Functional/etc. APIs, which don't allow for new arguments, we need a common TLB senderState to pass between TLBs, both "forwards" and "backwards."

So, basically, the rule is that any packet received by a TLB port (cpuside OR memside) must be safely castable to a TranslationState.

Definition at line 343 of file gpu_tlb.hh.

Constructor & Destructor Documentation

X86ISA::GpuTLB::TranslationState::TranslationState	(	Mode	tlb_mode,
		ThreadContext *	_tc,
		bool	_prefetch = `false`,
		Packet::SenderState *	_saved = `nullptr`
	)