Support boost 1.69

[seelabs]

No description provided.

[mellery451]

I think our local FindBoost.cmake should be updated with latest: https://github.com/Kitware/CMake/blob/master/Modules/FindBoost.cmake

[ripplelabs-jenkins]

Jenkins Build Summary

Built from this commit

Built at 20190130 - 16:33:38

Test Results

Build Type	Log	Result	Status
msvc.Debug	logfile	1142 cases, 0 failed, t: 637s	PASS ✅
gcc.Release -Dassert=ON, MANUAL_TESTS=true	logfile	915 cases, 0 failed, t: 6m12s	PASS ✅
docs, TARGET=docs	logfile	1 cases, 0 failed, t: 0m2s	PASS ✅
gcc.Debug -Dcoverage=ON, TARGET=coverage_report, SKIP_TESTS=true	logfile	1145 cases, 0 failed, t: 17m15s	PASS ✅
clang.Debug	logfile	1145 cases, 0 failed, t: 3m12s	PASS ✅
msvc.Debug, NINJA_BUILD=true	logfile	1142 cases, 0 failed, t: 656s	PASS ✅
rpm	logfile	1145 cases, 0 failed, t: n/a	PASS ✅
gcc.Debug	logfile	1145 cases, 0 failed, t: 3m28s	PASS ✅
clang.Debug -Dunity=OFF	logfile	1145 cases, 0 failed, t: 10m53s	PASS ✅
clang.Release -Dassert=ON	logfile	1145 cases, 0 failed, t: 5m43s	PASS ✅
gcc.Debug -Dunity=OFF	logfile	1145 cases, 0 failed, t: 10m20s	PASS ✅
msvc.Debug -Dunity=OFF	logfile	1142 cases, 0 failed, t: 1287s	PASS ✅
gcc.Debug -Dstatic=OFF	logfile	1145 cases, 0 failed, t: 3m28s	PASS ✅
gcc.Release -Dassert=ON	logfile	1145 cases, 0 failed, t: 6m8s	PASS ✅
gcc.Debug -Dstatic=OFF -DBUILD_SHARED_LIBS=ON	logfile	1145 cases, 0 failed, t: 3m28s	PASS ✅
gcc.Debug, NINJA_BUILD=true	logfile	1145 cases, 0 failed, t: 3m2s	PASS ✅
msvc.Release	logfile	1142 cases, 0 failed, t: 418s	PASS ✅
clang.Debug -Dunity=OFF -Dsan=address, PARALLEL_TESTS=false, DEBUGGER=false	logfile	1145 cases, 0 failed, t: 4m49s	PASS ✅
clang.Debug -Dunity=OFF -Dsan=undefined, PARALLEL_TESTS=false	logfile	1145 cases, 0 failed, t: 7m57s	PASS ✅

[HowardHinnant]

It isn't looking good on macOS:

In file included from /Users/howardhinnant/Development/rippled/src/ripple/app/main/Main.cpp:49:
In file included from /Users/howardhinnant/Development/boost_1_69_0/boost/process.hpp:24:
In file included from /Users/howardhinnant/Development/boost_1_69_0/boost/process/async_system.hpp:22:
In file included from /Users/howardhinnant/Development/boost_1_69_0/boost/process/child.hpp:21:
In file included from /Users/howardhinnant/Development/boost_1_69_0/boost/process/detail/child_decl.hpp:30:
/Users/howardhinnant/Development/boost_1_69_0/boost/process/detail/posix/wait_for_exit.hpp:60:7: error: expected unqualified-id
    ::sigemptyset(&sigset);
      ^
/usr/include/signal.h:125:26: note: expanded from macro 'sigemptyset'
#define sigemptyset(set)        (*(set) = 0, 0)
                                ^

I've got a comment in cpplang.slack.com #boost about this. Waiting for responses.

[seelabs]

Pushed a commit to update FindBoost.cmake to the latest, as suggested by @mellery451

[HowardHinnant]

Despite the macOS failure for boost 1.69, this PR tests out fine with boost 1.68 and 1.67.

[miguelportilla]

👍 with boost 1.69.0 on Windows. I did see the following Cmake warning:

CMake Warning (dev) at C:\ripple\rippled\CMakeLists.txt:599 (find_package):
Policy CMP0074 is not set: find_package uses <PackageName>_ROOT variables.
Run "cmake --help-policy CMP0074" for policy details.  Use the cmake_policy
command to set the policy and suppress this warning.

Line 599 in CMakeLists.txt find_package (OpenSSL ${_ssl_min_ver} REQUIRED)

Policy: https://cmake.org/cmake/help/git-stage/policy/CMP0074.html

[mellery451]

@miguelportilla I would expect that warning for cmake 3.12.4 or greater, but I don't think it has anything to do with this PR. Maybe you recently upgraded your cmake version (or visual studio if it includes CMake)? We should go ahead and explicitly set policy CMP0074 in our cmake, but I don't think it needs to be in this PR.

[MarkusTeufelberger]

Anything missing from this PR?

Arch Linux ships boost 1.69 by now (https://www.archlinux.org/packages/extra/x86_64/boost/) and this would be nice to have.

[seelabs]

@MarkusTeufelberger There is an issue with the mac build that we don't have a good solution for so we decided not to merge this yet. I wasn't aware any distros were shipping with boost 1.69, so it seemed like a good idea to delay this.

However, this change should be backward's compatable with older boosts, so it should be OK to merge. Since arch ships with boost 1.69, I'll rebase/retest this and see if it can be marked passed. Thanks for the info about arch.

[seelabs]

Rebased, squashed and retested locally on boost 1.68 and 1.69. Unless there are objections from the team or CI issues I'll mark this passed when CI finishes.

[HowardHinnant]

Still 👍 from me.

[miguelportilla]

Latest beast changes LGTM. On Windows, all targets build and tests pass. 👍

[MarkusTeufelberger]

Fedora 30 will ship with Boost 1.69 too: https://apps.fedoraproject.org/packages/boost

[seelabs]

Thanks for the info @MarkusTeufelberger. BTW, this PR will go into the first 1.3.0 beta;