recovery code as well as the architecturally visible instruction. You can eliminate idle loops effects by using the command-line option --exclude-idle True (which is the default). The effects of failed speculative operations and TLB misses cannot be directly eliminated, but you can get an estimate of the impact of events from the cspec, dspec, and tlb event sets. You can use the cpi event set to obtain the fraction of all instructions retired that have an architecturally visible result, except for predicated off branches, which are counted as useful instructions (non-taken branch) by the Itanium 2 PMU.

Metrics Available from this Measurement

The following metrics are available from this event set. These descriptions do not take into account any command-line options you might use.

The metrics are:

•Cycles

This is the number of CPU cycles that occurred during the sampling interval.

•Total Instr Retired

This is the total number of instructions retired during the sampling interval. It includes predicated off and nop instructions.

•Chka Retired

This is the total number advanced check loads (chk.a) and check loads (ld.c) that were retired during the sample interval.

•Chka Failed

This is the total number advanced check loads (chk.a) and check loads (ld.c) that failed during the sample interval.

•ALAT Access

This is the number of times the ALAT is accessed during the sample interval. In effect, this is the count of all instructions that access the ALAT. Instructions that access the ALAT include ld.a, ld.sa, ldf.a, ldf.sa, and ld.c.nc.

•ALAT Misses

This is the number of misses that incurred by instructions accessing the ALAT during the sample interval. Instructions that access the ALAT include ld.a, ld.sa, ldf.a, ldf.sa, and ld.c.nc.

•Data Speculation:

◦Total Specs Per Sec

This is the total number of data speculation events per second.

◦Failed Specs Per Sec

This is the number of data speculation fail events per second.

◦Total Specs Per Kinst

This is the total number of data speculation events per 1000 retired instructions. The instruction count includes predicated off and nop instructions.

◦Failed Specs Per Kinst

This is the number of data speculation fail events per 1000 retired instructions. The instruction count includes predicated off and nop instructions.

◦%Miss

This is the percentage of failed check advance loads (chk.a) and check loads (ld.c) out of the total number of retired check advance loads and check loads. Ideally, this

dspec Event Set 229