gcc -O0  -Wall -g -o segment_map -pie -fpic  segment_map.c  -lpthread
setcap cap_sys_resource=ep mov_code_segment
mmap.c is a microbenchmark written to determine mmap, mprotect latencies

1. Process pinned to a random core
2. Anonymous region of specified size (4KB, 400KB, 4MB, 400MB, 4GB, 40GB) mmap-ed 
3. Every page in region untouched/read/written
4. Mprotect called to update region permissions to none

Things to watch out for-
Transparent Huge Pages - enable/disable using /sys/kernel/mm/transparent_hugepage/enabled
Kernel same-page merging?
Clear Page cache- sync; echo 1 | sudo tee /proc/sys/vm/drop_caches   
Turn off hardware prefetechers using msr tool
https://software.intel.com/en-us/articles/disclosure-of-hw-prefetcher-control-on-some-intel-processors
0x1a4
sudo rdmsr 0x1a4
Value of 0 means all h/w prefetchers enabled

Setting the bits to 1 will disable prefetch
sudo wrmsr 0x1a4 0xf

To verify use perf-
(from man perf-list)
If the Intel docs for a QM720 Core i7 describe an event as:
Event  Umask  Event Mask
Num.   Value  Mnemonic    Description                        Comment
A8H      01H  LSD.UOPS    Counts the number of micro-ops     Use cmask=1 and
delivered by loop stream detector  invert to count
cycles
raw encoding of 0x1A8 can be used
perf stat -e r1a8 -a sleep 1
perf record -e r1a8 ...

Frequency Scaling - ondemand frequency scaling means variable performance so use performance governor
https://wiki.archlinux.org/index.php/CPU_frequency_scaling
cpupower frequency-set -g governor


If using Ftrace, follow the following steps
echo 0 | sudo tee /sys/kernel/debug/tracing/tracing_on
echo function_graph | sudo tee /sys/kernel/debug/tracing/current_tracer


use filefrag to find how many extents in a file
filefrag <file>