Use cut instead of $CUT

[RadekCap]

Add check of CUT.

CUT is needed for a proper setting of LIB_DIR in CYGWIN.

[sophia-guo]

cut is only use once for CYGWIN check UNAME_OS := $(shell $(UNAME) -s | $(CUT) -f1 -d_). May just use cut instead.

[judovana]

@sophia-guo can you please elaborate more? Do yo recomend to use cut instead $(CUT) ? Whats the benefit? The cut will still be missing from path. And it will enad as command not found...

I'm missing some detection of the cut. Where do you expect it to be set? Can you point us?

[judovana]

@RadekCap maybe you can add fallback?

if $(CUT) is missing, maybe yo may check path for cut it, and if it is there, then use that, if not then indeed die...

[sophia-guo]

I mean use cut directly there is no need to define CUT as it is used only once and presumably that will be true for a long time.

@RadekCap and I have talked about it in a meeting and @RadekCap will update the PR accordingly.

[RadekCap]

@sophia-guo updated as requested.

I've added a debug message for LIB_DIR as well.

[judovana]

I think the verification that the cut is on path should be done. Also i think that calling cut without $(CUT) is bad code.

[sophia-guo]

I think the cut command comes pre-installed in most Linux distributions. Though agreed @judovana it doesn't hurt to verify.

[judovana]

-UNAME_OS := $(shell $(UNAME) -s | $(CUT) -f1 -d_)
+UNAME_OS := $(shell $(UNAME) -s | cut -f1 -d_)

Just to summ this up - we agreed it should be checked - but the patch was approvred without the check.
In addition we caused (microscopical) regression, that if I have CUT declared, it will be ignored.

Am I right or am reading it wrong?

I know the proper gues/set/check if is nto exactly plesant, but Radek's original commit was very close to it. Also I agree it is super minor thing, but I feel very very uncomfortable about it:( Do we agree on finish it properly (@RadekCap please?), or do you really want to set this precedent?