update to fmt 10.1.1, fast_double_parser 0.7.0

[jameslamb]

Contributes to #3763.
Contributes to #5691.

Proposes updated the projects vendored in here as git submodules to their latest versions:

• boostorg/compute: no change (already on v1.83.0, September 2021)
• fmtlib/fmt: from v8.1.1 (January 2022) to v10.1.1 (August 2023)
• lemire/fast_double_parser: from v0.3.0 (November 2020) to v0.7.0 (December 2022)
• libeigen/eigen: no change (already on v3.4.0, August 2021)

How I did that

# fast_double_parser
cd ./external_libs/fast_double_parser
git fetch origin v0.7.0
git checkout v0.7.0
cd ../../
git add ./external_libs/fast_double_parser

# fmt
cd ./external_libs/fmt
git fetch origin 10.1.1
git checkout 10.1.1
cd ../../
git add ./external_libs/fmt

Benefits of this change

Pulls in bugfixes, security patches, performance improvements.

I'm hoping it'll help with R 4.3 support via changes like this:

• regression in 'constexpr' expansion of 'fmt::v8::basic_format_string...' fmtlib/fmt#2898
• GCC 13 reports OOB write fmtlib/fmt#3533

[jameslamb]

This is used to suppress the following warning that shows up only with MinGW (gcc 10):

[ 60%] Building CXX object CMakeFiles/lightgbm_objs.dir/src/network/linkers_mpi.cpp.obj

d:\a\lightgbm\lightgbm\lightgbm.rcheck\00_pkg_src\lightgbm\src\external_libs\fmt\include\fmt\format.h: In function 'void fmt::v10::detail::format_hexfloat(Float, int, fmt::v10::detail::float_specs, fmt::v10::detail::buffer<char>&) [with Float = double; typename std::enable_if<(! std::integral_constant<bool, (std::numeric_limits<_Tp>::digits == 106)>::value), int>::type <anonymous> = 0]':
d:\a\lightgbm\lightgbm\lightgbm.rcheck\00_pkg_src\lightgbm\src\external_libs\fmt\include\fmt\format.h:1370:8: note: at offset -2 to object 'buffer' with size 10 declared here
 1370 |   Char buffer[digits10<UInt>() + 1] = {};
      |        ^~~~~~

(example build link)

Per fmtlib/fmt#1810 (comment), this was a false positive and a bug in gcc 10 (https://gcc.gnu.org/bugzilla/show_bug.cgi?id=95353). Seems that maybe it wasn't backported in MinGW as well?

I saw the same error in only the MinGW + R 4.3 + Rtools43 build on #6061.

[jameslamb]

/gha run r-valgrind

Workflow R valgrind tests has been triggered! 🚀
https://github.com/microsoft/LightGBM/actions/runs/6068766481

Status: success ✔️.

[jameslamb]

@guolinke @shiyu1994 could I get help with a review on this?

In addition to the bugfixes and performance improvements in these new versions of these libraries, I think this PR will help with fully testing the R package against R 4.3 and with efforts like #5981 to make the project compatible with C++20.

[jameslamb]

Thanks so much @guolinke !

[github-actions]

This pull request has been automatically locked since there has not been any recent activity since it was closed. To start a new related discussion, open a new issue at https://github.com/microsoft/LightGBM/issues including a reference to this.