https://easyperf.net/blog/2018/06/01/PMU-counters-and-profiling-basics
In a really simplified view our processor looks like this:
There is a clock generator that sends pulses to every piece of the system to make everything moving to the next stage. This is called a cycle. If we add just a little bit of silicon and connect it to the pulse generator we can count a number of cycles, yay!
If we don’t set the scaling governor policy to be performance kernel can decide that it’s better to save power and throttle. Setting scaling_governor to ‘performance’ helps to avoid sub-nominal clocking. Here is the documentation about Linux CPU frequency governors.
Here is how we can set it for all the cores:
|
|
https://easyperf.net/blog/2019/08/02/Perf-measurement-environment-on-Linux
https://www.kernel.org/doc/Documentation/cpu-freq/governors.txt
pool1-n104-vpod1-wpool1-n23:~/pmu-tools # x86info -c
x86info vVERSION
Found 80 identical CPUsMP Configuration Table Header MISSING!
Extended Family: 0 Extended Model: 5 Family: 6 Model: 85 Stepping: 7
Type: 0 (Original OEM)
CPU Model (x86info's best guess): Unknown model.
Processor name string (BIOS programmed): Intel(R) Xeon(R) Gold 6230N CPU @ 2.30GHz
Cache info
L1 Data Cache: 32KB, 8-way associative, 64 byte line size
L1 Instruction Cache: 32KB, 8-way associative, 64 byte line size
L2 Unified Cache: 1024KB, 16-way associative, 64 byte line size
L3 Unified Cache: 28160KB, 11-way associative, 64 byte line size
TLB info
Instruction TLB: 2M/4M pages, fully associative, 8 entries
Instruction TLB: 4K pages, 8-way associative, 64 entries
Data TLB: 1GB pages, 4-way set associative, 4 entries
Data TLB: 4KB pages, 4-way associative, 64 entries
Shared L2 TLB: 4KB/2MB pages, 6-way associative, 1536 entries
64 byte prefetching.
Total processor threads: 80
This system has 2 20-core processors with hyper-threading (2 threads per core) running at an estimated 2.30GHz
Huge pages part 5: A deeper look at TLBs and costs
TLB and Java
A translation lookaside buffer (TLB) is a memory cache that is used to reduce the time taken to access a user memory location. It is a part of the chip’s memory-management unit (MMU). The TLB stores the recent translations of virtual memory to physical memory and can be called an address-translation cache. A TLB may reside between the CPU and the CPU cache, between CPU cache and the main memory or between the different levels of the multi-level cache
Intel Turbo Boost is a feature that automatically raises CPU operating frequency when demanding tasks are running. It can be permanently disabled in BIOS. Check FAQ for more information. To disable turbo in Linux do:
# Intel
echo 1 > /sys/devices/system/cpu/intel_pstate/no_turbo
# AMD
echo 0 > /sys/devices/system/cpu/cpufreq/boost
Also you might want to take a look at how it’s done in uarch-bench.
Example (single-threaded workload running on Intel® Core™ i5-8259U):
# TurboBoost enabled
$ cat /sys/devices/system/cpu/intel_pstate/no_turbo
0
$ perf stat -e task-clock,cycles -- ./a.out
11984.691958 task-clock (msec) # 1.000 CPUs utilized
32,427,294,227 cycles # 2.706 GHz
11.989164338 seconds time elapsed
# TurboBoost disabled
$ echo 1 | sudo tee /sys/devices/system/cpu/intel_pstate/no_turbo
1
$ perf stat -e task-clock,cycles -- ./a.out
13055.200832 task-clock (msec) # 0.993 CPUs utilized
29,946,969,255 cycles # 2.294 GHz
13.142983989 seconds time elapsed
You can see the average frequency is much higher when TurboBoost is on.
https://accedian.com/blog/close-tcp-sessions-diagnose-disconnections/
Figure 1 – Simplified TCP closing with FIN.
The standard way to close TCP sessions is to send a FIN packet, then wait for a FIN response from the other party.
This works fine in a perfect world. However, what happens when one part of the conversation is broken? That’s why the Reset (RST) packet exists.
https://accedian.com/blog/diagnose-tcp-connection-setup-issues/
A TCP connection, also called 3-way Handshake is achieved with SYN, SYN+ACK and ACK packets. From this handshake, we can extract a performance metric called Connection Time (CT), which summarizes how fast session a can be set up between a client and a server over a network. For more details, see this excellent article on Wikipedia.
