欢迎访问ic37.com |
会员登录 免费注册
发布采购

RM7965A-900UI 参数 Datasheet PDF下载

RM7965A-900UI图片预览
型号: RM7965A-900UI
PDF下载: 下载PDF文件 查看货源
内容描述: [Microprocessor, 64-Bit, 900MHz, CMOS, PBGA256, 27 X 27 MM, 1.62 MM HEIGHT, MO-192BAL-2, CSBGA-256]
分类和应用: 时钟外围集成电路
文件页数/大小: 74 页 / 508 K
品牌: PMC [ PMC-SIERRA, INC ]
 浏览型号RM7965A-900UI的Datasheet PDF文件第42页浏览型号RM7965A-900UI的Datasheet PDF文件第43页浏览型号RM7965A-900UI的Datasheet PDF文件第44页浏览型号RM7965A-900UI的Datasheet PDF文件第45页浏览型号RM7965A-900UI的Datasheet PDF文件第47页浏览型号RM7965A-900UI的Datasheet PDF文件第48页浏览型号RM7965A-900UI的Datasheet PDF文件第49页浏览型号RM7965A-900UI的Datasheet PDF文件第50页  
RM7965A-900UI 900 MHz 64-bit Microprocessor Data Sheet  
7.4 Performance Counters  
The RM7965A supports two CP0 performance-counter registers with the PerfCount and  
PerfControl registers. The PerfCount register is a 64-bit register divided into two independent  
32-bit counters, PerfCounter0, PerfCounter1. The counters can be written to by software to  
initialize event monitoring, and they generate a performance-counter interrupt when the most  
significant bit in either counter (bit 63 in Counter 2, and bit 31 in Counter 1) is set.  
The PerfControl register is a 32-bit register containing two 5-bit fields used to select one of  
twenty-four event types counted by each counter, as well as a handful of bits which control the  
overall counting function. Note that only one event type can be counted per counter at a time,  
and that counting can occur for user code, kernel code or both. The event types and control bits  
are listed in Table 18.  
Table 18 Performance Counter Control  
PerfControl  
Field  
Description  
4:0  
Event Type  
00:  
01:  
02:  
03:  
04:  
05:  
06:  
07:  
08:  
09:  
0A:  
0B:  
0C:  
0D:  
0E:  
0F:  
10:  
11:  
12:  
13:  
14:  
15:  
Clock cycles  
Total instructions issued (Integer and Floating Point)  
Floating-point instructions issued (any COP1 or COP3).  
Integer instructions issued (no COP1 or COP3).  
Load instructions issued  
Store instructions issued  
Dual issued instruction pairs  
Branch mispredictions  
External Cache Misses  
Stall cycles  
Secondary cache misses  
Instruction cache misses  
Data cache misses  
Data TLB misses  
Instruction TLB misses  
Joint TLB instruction misses  
Joint TLB data misses  
Branches taken  
Branches issued  
Secondary cache writebacks  
Data cache writebacks  
Data cache miss stall cycles (A stall occurs when the data cache is processing  
two misses and a third miss occurs).  
16:  
17:  
18:  
19:  
1A:  
1B:  
1C:  
1D:  
1E:  
Cache misses (all caches).  
FP possible exception cycles  
Slip Cycles due to multiplier busy  
Coprocessor 0 slip cycles  
Slip cycles due to pending non-blocking loads  
Stall cycles due to full Write buffer  
Stall cycles due to Cache instruction  
Unused  
Stall cycles due to pending non-blocking loads - stall start of exception  
Proprietary and Confidential to PMC-Sierra, Inc., and for its customers’ internal use.  
Document No.: PMC-2100294, Issue 2  
46  
 
 
 复制成功!