From b64d38562079a7d4720c15c9f6309912f4090795 Mon Sep 17 00:00:00 2001 From: Tor Aamodt Date: Fri, 8 Oct 2010 13:05:25 -0800 Subject: 1. modify shader_core_ctx::execute_pipe() to model instruction throughput correctly (off-by-one error) 2. adding code to dump_pipeline to display reason for memory stage stalling 3. removing dead code resulting from prior changes correlation vs. GT200 is 0.95 (need to add back modeling of memory writebacks, shared memory latency) [git-p4: depot-paths = "//depot/gpgpu_sim_research/fermi/distribution/": change = 7831] --- src/gpgpu-sim/shader.cc | 26 ++++++++++++++++++-------- 1 file changed, 18 insertions(+), 8 deletions(-) (limited to 'src/gpgpu-sim/shader.cc') diff --git a/src/gpgpu-sim/shader.cc b/src/gpgpu-sim/shader.cc index 54e4ffa..1d51004 100644 --- a/src/gpgpu-sim/shader.cc +++ b/src/gpgpu-sim/shader.cc @@ -402,6 +402,7 @@ shader_core_ctx::shader_core_ctx( class gpgpu_sim *gpu, for (unsigned i = 0; i < config->max_warps_per_shader; ++i) m_pdom_warp[i] = new pdom_warp_ctx_t(i,this); m_shader_memory_new_instruction_processed = false; + m_mem_rc = NO_RC_FAIL, // Initialize scoreboard m_scoreboard = new Scoreboard(m_sid, m_config->max_warps_per_shader); @@ -419,11 +420,6 @@ shader_core_ctx::shader_core_ctx( class gpgpu_sim *gpu, m_memory_queue.texture.reserve(warp_size); m_memory_queue.local_global.reserve(warp_size); - // writeback - m_pl_tid = (int*) malloc(sizeof(int)*warp_size); - m_mshr_lat_info = (insn_latency_info*) malloc(sizeof(insn_latency_info) * warp_size); - m_pl_lat_info = (insn_latency_info*) malloc(sizeof(insn_latency_info) * warp_size); - // fetch m_last_warp_fetched = 0; m_last_warp_issued = 0; @@ -840,7 +836,7 @@ void shader_core_ctx::execute_pipe( unsigned pipeline, unsigned next_stage ) { if( !m_pipeline_reg[next_stage]->empty() ) return; - if( m_pipeline_reg[pipeline]->cycles ) { + if( m_pipeline_reg[pipeline]->cycles > 1 ) { m_pipeline_reg[pipeline]->cycles--; return; } @@ -1417,7 +1413,7 @@ void shader_core_ctx::memory() done &= memory_constant_cycle(rc_fail, type); done &= memory_texture_cycle(rc_fail, type); done &= memory_cycle(rc_fail, type); - + m_mem_rc = rc_fail; if (!done) { // log stall types and return assert(rc_fail != NO_RC_FAIL); m_stats->gpu_stall_shd_mem++; @@ -1425,7 +1421,7 @@ void shader_core_ctx::memory() return; } if( not m_pipeline_reg[MM_WB]->empty() ) - return; // writeback stalled + return; // writeback stalled move_warp(m_pipeline_reg[MM_WB],m_pipeline_reg[EX_MM]); } @@ -1631,6 +1627,20 @@ void shader_core_ctx::display_pipeline(FILE *fout, int print_mem, int mask ) } fprintf(fout, "EX/MEM = "); print_stage(EX_MM, fout, print_mem, mask); + if( m_mem_rc != NO_RC_FAIL ) { + fprintf(fout, "EX/MEM (stall condition: "); + switch ( m_mem_rc ) { + case BK_CONF: fprintf(fout,"BK_CONF"); break; + case MSHR_RC_FAIL: fprintf(fout,"MSHR_RC_FAIL"); break; + case ICNT_RC_FAIL: fprintf(fout,"ICNT_RC_FAIL"); break; + case COAL_STALL: fprintf(fout,"COAL_STALL"); break; + case WB_ICNT_RC_FAIL: fprintf(fout,"WB_ICNT_RC_FAIL"); break; + case WB_CACHE_RSRV_FAIL: fprintf(fout,"WB_CACHE_RSRV_FAIL"); break; + case N_MEM_STAGE_STALL_TYPE: fprintf(fout,"N_MEM_STAGE_STALL_TYPE"); break; + default: abort(); + } + fprintf(fout,")\n"); + } fprintf(fout, "MEM/WB = "); print_stage(MM_WB, fout, print_mem, mask); fprintf(fout, "\n"); -- cgit v1.3