Speed socket

[thiagoftsm]

Summary

After a talk with product team and to have a more clear picture about what we are going to deliver, we are creating this PR that will reduce I/O events in kernel ring.

Test Plan

1. Get binaries according your LIBC from this link and extract them inside a directory, for example: ../artifacts. You can also get everything for glibc here.

2. Extract them running:

   $ for i in `ls *.zip`; do unzip $i; rm .gitkeep ; rm $i; done
   $ for i in `ls *.xz`; do tar -xf $i; rm $i* ; done

3. Compile branch an run the following tests:

   # make clean; make tester
   # for i in `seq 0 3`; do ./kernel/legacy_test --netdata-path ../artifacts --content --iteration 1 --pid $i --log-path file_pid$i.txt; done

4. Every test should ends with Success, unless you do not have a specific target (function) available.

Additional information

This PR was tested on:

Linux Distribution	Environment	Kernel Version	Real Parent	Parent	All PIDs	Without PIDs
Slackware current	Bare metal	6.1.46	slackware_6_1_pid0.txt	slackware_6_1_pid1.txt	slackware_6_1_pid2.txt	slackware_6_1_pid3.txt
Arch Linux	Libvirt	6.4.11-arch2-1	arch_6_4_pid0.txt	arch_6_4_pid1.txt	arch_6_4_pid2.txt	arch_6_4_pid3.txt
Slackware current	Qemu	4.14.290	slackware_4_14_pid0.txt	slackware_4_14_pid1.txt	slackware_4_14_pid2.txt	slackware_4_14_pid3.txt
CentOS 7.9	Libvirt	3.10	centos_3_10_pid0.txt	centos_3_10_pid1.txt	centos_3_10_pid2.txt	centos_3_10_pid3.txt