C-Box Performance Monitoring Events; An Overview; Acronyms Frequently Used In C-Box Events - Intel BX80571E7500 - Core 2 Duo 2.93 GHz Processor Programming Manual

I ® X ® P 7500 S
NTEL EON ROCESSOR
2.3.4

C-BOX Performance Monitoring Events

2.3.4.1

An Overview:

The performance monitoring events within the C-Box include all events internal to the LLC as well as
events which track ring related activity at the C-Box/Core ring stops. The only ring specific events that
are not tracked by the C-Box PMUs are those events that track ring activity at the S-Box ring stop (see
the S-Box chapter for details on those events).
C-Box performance monitoring events can be used to track LLC access rates, LLC hit/miss rates, LLC
eviction and fill rates, and to detect evidence of back pressure on the LLC pipelines. In addition, the C-
Box has performance monitoring events for tracking MESI state transitions that occur as a result of
data sharing across sockets in a multi-socket system. And finally, there are events in the C-Box for
tracking ring traffic at the C-Box/Core sink inject points.
Every event in the C-Box (with the exception of the P2C inject and *2P sink counts) are from the point
of view of the LLC and cannot be associated with any specific core since all cores in the socket send
their LLC transactions to all C-Boxes in the socket. The P2C inject and *2P sink counts serve as the
exception since those events are tracking ring activity at the cores' ring inject/sink points.
There are separate sets of counters for each C-Box instance. For any event, to get an aggregate count
of that event for the entire LLC, the counts across the C-Box instances must be added together. The
counts can be averaged across the C-Box instances to get a view of the typical count of an event from
the perspective of the individual C-Boxes. Individual per-C-Box deviations from the average can be
used to identify hot-spotting across the C-Boxes or other evidences of non-uniformity in LLC behavior
across the C-Boxes. Such hot-spotting should be rare, though a repetitive polling on a fixed physical
address is one obvious example of a case where an analysis of the deviations across the C-Boxes would
indicate hot-spotting.
2.3.4.2

Acronyms frequently used in C-Box Events:

The Rings:
AD (Address) Ring - Core Read/Write Requests and Intel QPI Snoops. Carries Intel QPI requests and
snoop responses from C to S-Box.
BL (Block or Data) Ring - Data == 2 transfers for 1 cache line
AK (Acknowledge) Ring - Acknowledges S-Box to C-Box and C-Box to Core. Carries snoop responses
from Core to C-Box.
IV (Invalidate) Ring - C-Box Snoop requests of core caches
Internal C-Box Queues:
IRQ - Ingress Request Queue on AD Ring. Associated with requests from core.
IPQ - Ingress Probe Queue on AD Ring. Associated with snoops from S-Box.
VIQ - Victim Queue internal to C-Box.
IDQ - Ingress Data Queue on BL Ring. For data from either Core or S-Box.
ICQ - S-Box Ingress Complete Queue on AK Ring
SRQ - Processor Snoop Response Queue on AK ring
IGQ - Ingress GO-pending (tracking GO's to core) Queue
MAF - Miss Address File. Intel QPI ordering buffer that also tracks local coherence.
U P G
ERIES NCORE ROGRAMMING UIDE
UNCORE PERFORMANCE MONITORING
2-16