[fuzz] Superblock fuzz issues

Make idiomatic cmake install so third party cmake installers work right.

Fixes new fuzz issue

Credit to OSS-Fuzz

* Initializing unsigned value

* Initialilzing to 1 instead of 0 because its more conservative

* Unconditionoally setting to check first and then checking zero

* Moving bool to before block for c90

* Move check set before block

so that potential issue report do not confuse `dev` with latest release

Add Tool to Diagnose Whether Corrupt Blobs are Plausibly Bit-Flips

Check for fread failure

small refactoring : remove global variable g_displayOut

displaying into stderr or stdout is now an explicit operation,
no longer depending on some global state set elsewhere in the code.

fix recent issue combining -r with empty list of input files

This would resize the table of input filenames to zero,
delivering an empty table,
to which it was no longer possible to add stdin.

mingw cross compilation test

Fix MinGW cross

- only compiles zstd CLI
- enforces zero warning tolerance policy

note : mingw compilation of the library complains (warning) about -fPIC flag.
It mostly ignores it, so it's not a big issue.
Still, it would need to be fixed if library cross-compilation becomes part of CI.
At this stage, it's unclear if this is a scenario.

based on a script in RocksDB project.

also : pushed qemu aarch64 tests to master/cron,
since there are already hardware aarch64 tests on dev.

--filelist=FILE feature

On failure fread may return either a short read or 0. Need to use
ferror to detect error versus eof.

ppc cross-compilation seems broken outside of trusty
scan-build version changes with distrib version, resulting in a different list of false positives

since only UTIL_DISPLAY() depended on it.

from *.h to *.c
so that they don't get transitively included
into users of util.h.

trusty is out of long term support.
let's run (most) tests on "current" platform.

complaining about a `const` property on one side but not the other.

Fixes a fuzz issue where dictionary_round_trip failed because the compressor was generating corrupt files thanks to zero weights in the table.

* Only setting loaded dict huf table to valid on non-zero

* Adding hasNoZeroWeights test to fse tables

* Forbiding nbBits != 0 when weight == 0

* Reverting the last commit

* Setting table log to 0 when weight == 0

* Small (invalid) zero weight dict test

* Small (valid) zero weight dict test

* Initializing repeatMode vars to check before zero check

* Removing FSE changes to seperate pr

* Reverting accidentally changed file

* Negating bool, using unsigned, optimization nit

they are fast enough, and we want to catch such incompatibility early

also : joined c99 compatibility tests with gnu90 and c++ ones

as a guide when doing test selection

fix #1904