This is a list of PPC 7450's performance counter event types. Please see MPC7450 RISC Microprocessor Family Reference Manual, Chapter 11: Performance Monitor. Downloadable from freescale.com
Name | Description | Counters usable | Unit mask options |
CYCLES | Processor cycles | 0, 1, 2, 3 | |
COMPLETED_INSNS | Completed Instructions | 0, 1, 2, 3 | |
TBL_BIT_TRANSTNS | TBL Bit Transitions | 0, 1, 2, 3 | |
DISPATCHED_INSNS | Dispatched Instructions | 0, 1, 2, 3 | |
PROC_PERFMON_EXC | Process Performance Monitor Exception | 0, 1, 2, 3 | |
VPU_INSNS | VPU Instructions Completed | 0, 1, 3 | |
VFPU_INSNS | VFPU Instructions Completed | 0, 1, 3 | |
VIU1_INSNS | VIU1 Instructions Completed | 0, 1, 3 | |
VIU2_INSNS | VIU2 Instructions Completed | 0, 1, 3 | |
VPU_CYCLES | Cycles a VPU Instruction | 0, 1 | |
VFPU_CYCLES | Cycles a VFPU Instruction | 0, 1 | |
VIU1_CYCLES | Cycles a VIU1 Instruction | 0, 1 | |
VIU2_CYCLES | Cycles a VIU2 Instruction | 0, 1 | |
DTLB_MISSES | DTLB misses | 2 | |
STORE_INSNS | Store Instructions | 0, 1 | |
L1_ICACHE_MISSES | L1 Instruction Cache Misses | 0, 1 | |
L1_DATA_SNOOPS | L1 Data Snoops | 0, 1 | |
UNRESOLVED_BRANCHES | Unresolved Branches | 0, 1 | |
MISPREDICTED_BRANCHES | Mispredicted branches | 3 | |
FOLDED_BRANCHES | Folded branches | 3 | |
BR_LN_STACK_MIS | Branch Link Stack Mispredicted | 2 | |
ITLB_TABLE_CYCLES | ITLM Hardware Table Search Cycles | 0 | |
L1_ICACHE_ACCESSES | L1 Instruction Cache Accesses | 0 | |
INSN_BP_MATCHES | Instruction Breakpoint Matches | 0 | |
L1_DSNOOP_HITS | L1 data snoop hits | 0 | |
WRITETHRU_STORES | Write-through stores | 0 | |
CACHEINH_STORES | Cache-inhibited stores | 0 | |
L1_DLOAD_HIT | L1 data load hit | 0 | |
L1_DTOUCH_HIT | L1 data touch hit | 0 | |
L1_DSTORE_HIT | L1 data store hit | 0 | |
L1_DATA_HITS | L1 data total hits | 0 | |
ALTIVEC_LD_INSNS_COMPLETED | Altivec load instructions completed | 0 | |
FP_STORE_INSNS_COMPLETED_LSU | Floating point store instructions completed in LSU | 0 | |
FP_LOAD_INSNS_COMPLETED_LSU | Floating point load instructions completed in LSU | 0 | |
FP_LDSINGLE_INSNS_COMPLETED_LSU | Floating point load single instructions completed in LSU | 0 | |
FP_DENORMALIZED_RESULT | Floating point denormalized result | 0 |
Speed, it seems to me, provides the one genuinely modern pleasure.- Aldous Huxley