From dc93f319051a9a9936a02cd9c1f7843a382a2da0 Mon Sep 17 00:00:00 2001 From: Tor Aamodt Date: Thu, 21 Oct 2010 07:16:49 -0800 Subject: 1. rewriting memory access generation code (from scratch), why not... passing CUDA 3.1 and ptxplus correlation, but correlation still bad (0.62)... after debugging 1 to get it working with ptxplus, problem is very clear: shared and constant cache accesses not occuring for operations that combine these with ALU operations. TODO: have a "read-operands" stage, which somehow combines operand collector register reading with shared and const memory accesses... [git-p4: depot-paths = "//depot/gpgpu_sim_research/fermi/distribution/": change = 7895] --- src/gpgpu-sim/shader.h | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-) (limited to 'src/gpgpu-sim/shader.h') diff --git a/src/gpgpu-sim/shader.h b/src/gpgpu-sim/shader.h index b9e0bda..278fafc 100644 --- a/src/gpgpu-sim/shader.h +++ b/src/gpgpu-sim/shader.h @@ -105,7 +105,6 @@ #define WRITE_PACKET_SIZE 8 #define WRITE_MASK_SIZE 8 -#define NO_PARTIAL_WRITE (partial_write_mask_t()) //Set a hard limit of 32 CTAs per shader [cuda only has 8] #define MAX_CTA_PER_SHADER 32 @@ -863,15 +862,13 @@ public: void print(FILE *fout) const; private: - void generate_mem_accesses(warp_inst_t &pipe_reg); - bool shared_cycle( warp_inst_t &inst, mem_stage_stall_type &rc_fail, mem_stage_access_type &fail_type); bool constant_cycle( warp_inst_t &inst, mem_stage_stall_type &rc_fail, mem_stage_access_type &fail_type); bool texture_cycle( warp_inst_t &inst, mem_stage_stall_type &rc_fail, mem_stage_access_type &fail_type); bool memory_cycle( warp_inst_t &inst, mem_stage_stall_type &rc_fail, mem_stage_access_type &fail_type); mem_stage_stall_type process_memory_access_queue( cache_t *cache, warp_inst_t &inst ); - mem_fetch *create_data_mem_fetch(warp_inst_t &inst, mem_access_t &access); + mem_fetch *create_data_mem_fetch(const warp_inst_t &inst, const mem_access_t &access); const memory_config *m_memory_config; class shader_memory_interface *m_icnt; -- cgit v1.3