Windows warnings with gcc about %lld

[mattdowle]

Closes #4062

[codecov]

Codecov Report

Merging #4064 into master will increase coverage by <.01%.
The diff coverage is 100%.

@@            Coverage Diff            @@
##           master   #4064      +/-   ##
=========================================
+ Coverage    99.4%   99.4%   +<.01%     
=========================================
  Files          72      72              
  Lines       13674   13675       +1     
=========================================
+ Hits        13593   13594       +1     
  Misses         81      81

Impacted Files	Coverage Δ
src/freadR.c	100% <ø> (ø)	⬆️
src/chmatch.c	100% <ø> (ø)	⬆️
src/forder.c	100% <ø> (ø)	⬆️
src/frolladaptive.c	100% <100%> (ø)	⬆️
src/assign.c	100% <100%> (ø)	⬆️
src/rbindlist.c	100% <100%> (ø)	⬆️
src/fread.c	99.44% <100%> (ø)	⬆️
src/init.c	100% <100%> (ø)	⬆️
src/froll.c	100% <100%> (ø)	⬆️
src/fifelse.c	100% <100%> (ø)	⬆️
... and 4 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 1d33704...47c5ca8. Read the comment docs.

[mattdowle]

I submitted this to win-builder both R-devel and R-devel ATC. They came back passing fully OK, and I'd observed the same warnings before in a previous submission to win-builder so we know it isn't just OK for not doing the same thing as CRAN. Actually, when I click the links and look at the install logs, there are still warnings, but now a different one. I'll ask Uwe why the "Status: OK" doesn't reflect that there are actually still warnings, and I'll see how to fix the new warnings.

R-devel ATC
-------
Dear package maintainer,

this notification has been generated automatically.
Your package data.table_1.12.7.tar.gz has been built and checked for Windows.
Please check the log files and (if working) the binary package at:
https://win-builder.r-project.org/8l7esZ7J9J1Q
The files will be removed after roughly 72 hours.
Installation time in seconds: 102
Check time in seconds: 988
Status: OK
R Under development (unstable) (2019-11-14 r77415)

All the best,
Uwe Ligges
(CRAN maintainer of binary packages for Windows)

R-devel
-----------
Dear package maintainer,

this notification has been generated automatically.
Your package data.table_1.12.7.tar.gz has been built and checked for Windows.
Please check the log files and (if working) the binary package at:
https://win-builder.r-project.org/BjuunGK5u6H4
The files will be removed after roughly 72 hours.
Installation time in seconds: 103
Check time in seconds: 933
Status: OK
R Under development (unstable) (2019-11-20 r77445)

All the best,
Uwe Ligges
(CRAN maintainer of binary packages for Windows)

This confirms the warning on CRAN checks (r-devel-windows-ix86+x86_64-gcc8) will be resolved if this is submitted to CRAN. Since I was unable to reproduce any of the warnings locally even using gcc-8 with the extra compiler flags on (so it seems truly Windows only).

[mattdowle]

Email from From Matt to Uwe copied here for Jan and Michael and others, and the future.

---

Hi Uwe,
I can see new warnings in the install log here (see win-builder reply below) but the email says "Status: OK" and seems to have missed the warnings.
i.e. click the win-builder link below, click 00install.out and there are lots of compiler "warning: ISO C does not support the 'I64' ms_printf length modifier [-Wformat=]"
Why doesn't "Status: OK" show warning like it does on CRAN here: https://www.r-project.org/nosvn/R.check/r-devel-windows-ix86+x86_64-gcc8/data.table-00check.html
Also, it looks a bit odd to me that the -I path is to gcc 4.9.3 being used with rtools4 which is gcc-8. I don't know if that has anything to do with it.
Anyway, I'm investigating the new warning and I think there's a #define I can use to silence it.
Best, Matt