SERVER-48291 Ensure runtime is dynamically linked in dynamic builds

Prior to this point, a dynamic build might have resulted in some runtime
libraries being statically linked into shared objects and executables in
cases where "shared" runtime libraries were actually linker scripts that
linked static versions. This was the case with the MongoDB toolchain and
some distro toolchains, including those installed as updated compiler
versions in RHEL.

The effect of having runtime libraries statically linked was that
symbols from those libraries would end up scattered over the compiled
objects, increasing object sizes and slowing down server startup. Now,
whenever a dynamic build is selected, the user can choose whether to
create "shim" runtime libraries that wrap the static ones.

The default behavior on Linux is that dynamic builds will detect whether
runtime libraries are linker scripts and create shim libraries if any are
found. On Windows, the default is to always use a dynamic runtime library
with dynamic builds. For other platforms, the prior behavior remains
unchanged.

1 files changed, 148 insertions, 1 deletions

diff --git a/src/SConscript b/src/SConscript
index fbf8e1fcecb..3c37fe2cd1c 100644
--- a/src/SConscript
+++ b/src/SConscript
@@ -3,21 +3,168 @@
 # This is the principle SConscript file, invoked by the SConstruct.  Its job is
 # to delegate to any and all per-module SConscript files.
 
+from functools import partial
+
+from site_scons.mongo import insort_wrapper
+
+import SCons
+
+Import('dynamicRT')
 Import('env')
 Import('get_option')
+Import('has_option')
 Import('module_sconscripts')
 
 
-env = env.Clone()
+def shim_hack(target, source, env, inject_target=None, disable_var=None):
+    # If we allowed conftests to become dependent, any TryLink
+    # that happened after we made the below modifications would
+    # cause the configure steps to try to compile tcmalloc and any
+    # of its dependencies. Oops!
+    if any('conftest' in str(t) for t in target):
+        return target, source
+
+    # It is possible that 'env' isn't a unique
+    # OverrideEnvironment, since if you didn't pass any kw args
+    # into your builder call, you just reuse the env you were
+    # called with. That could mean that we see the same
+    # environment here multiple times. But that is really OK,
+    # since the operation we are performing would be performed on
+    # all of them anyway. The flag serves as a way to disable the
+    # auto-injection for the handful of libraries where we must do
+    # so to avoid forming a cycle.
+    if not env.get(disable_var, False):
+        lds = env.get('LIBDEPS', [])
+        shim_target = f"$BUILD_DIR/{inject_target}"
+        if shim_target not in lds:
+            insort_wrapper(lds, shim_target)
+            env['LIBDEPS'] = lds
+
+    return target, source
+
+
+def add_shim_hack(env, hack_method):
+    for builder_name in ('Program', 'SharedLibrary', 'LoadableModule', 'StaticLibrary'):
+        builder = env['BUILDERS'][builder_name]
+        base_emitter = builder.emitter
+        builder.emitter = SCons.Builder.ListEmitter([hack_method, base_emitter])
+
+
+# Here, we "hoist" libgcc*.a symbols out of the toolchain and stuff them into
+# our own library so they don't get added piecemeal to every shared object and
+# executable in a build directly out of the toolchain.
+libcrtEnv = env.Clone(LIBS=[])
+libcxxEnv = env.Clone(LIBS=[])
+if dynamicRT == "force":
+
+    if env.ToolchainIs('gcc', 'clang'):
+
+        env.AppendUnique(
+            LINKFLAGS=[
+                '-nodefaultlibs',
+                '-l:libgcc_s.so'
+            ],
+        )
+
+        libcrtEnv.AppendUnique(
+            LINKFLAGS=[
+                "-Wl,-z,muldefs",
+                "-static-libgcc",
+                "-Wl,--push-state",
+                "-Wl,-Bstatic",
+                "-Wl,--no-warn-execstack",
+                "-Wl,--whole-archive",
+                "-lgcc",
+                "-lgcc_eh",
+                "-Wl,--no-whole-archive",
+                "-Wl,--pop-state",
+            ]
+        )
+
+        if has_option("libc++"):
+            cxx_lib = "c++"
+        else:
+            cxx_lib = "stdc++"
+
+        libcxxEnv.AppendUnique(
+            LINKFLAGS=[
+                "-Wl,-z,muldefs",
+                f"-static-lib{cxx_lib}",
+                "-Wl,--push-state",
+                "-Wl,-Bstatic",
+                "-Wl,--no-warn-execstack",
+                "-Wl,--whole-archive",
+                f"-l{cxx_lib}",
+                "-Wl,--no-whole-archive",
+                "-Wl,--pop-state",
+            ]
+        )
+
+libcrtEnv.ShimLibrary(
+    name="crt",
+    needs_link=(dynamicRT == "force" and env.ToolchainIs('gcc', 'clang')),
+    LIBDEPS_TAGS=[
+        # TODO: Remove all of these when SERVER-48291 is merged into stable build tools
+        # The shim allocator must be linked to every node, including what would
+        # be considered a leaf node to ensure the system allocator
+        # is not linked in before tcmalloc. This tag allows nodes tagged as
+        # leaf nodes to still get the correct allocator.
+        'lint-leaf-node-allowed-dep',
+        # This tag allows the allocator to be linked to nodes marked as not
+        # allowed to have public dependencies.
+        'lint-public-dep-allowed'
+    ],
+)
+
+libcxxEnv.ShimLibrary(
+    name="cxx",
+    needs_link=(dynamicRT == "force" and env.ToolchainIs('gcc', 'clang')),
+    LIBDEPS_TAGS=[
+        # TODO: Remove all of these when SERVER-48291 is merged into stable build tools
+        # The shim allocator must be linked to every node, including what would
+        # be considered a leaf node to ensure the system allocator
+        # is not linked in before tcmalloc. This tag allows nodes tagged as
+        # leaf nodes to still get the correct allocator.
+        'lint-leaf-node-allowed-dep',
+        # This tag allows the allocator to be linked to nodes marked as not
+        # allowed to have public dependencies.
+        'lint-public-dep-allowed'
+    ],
+)
+
+
 if get_option("build-tools") == "next":
     # Add any "global" dependencies here. This is where we make every build node
     # depend on a list of other build nodes, such as an allocator or libunwind
     # or libstdx or similar.
     env.AppendUnique(
         LIBDEPS_GLOBAL=[
+            '$BUILD_DIR/shim_crt' if dynamicRT == "force" else [],
+            '$BUILD_DIR/shim_cxx' if dynamicRT == "force" else [],
             '$BUILD_DIR/third_party/shim_allocator',
         ],
     )
+else:
+    add_shim_hack(
+        env,
+        partial(
+            shim_hack,
+            inject_target='third_party/shim_allocator',
+            disable_var='DISABLE_ALLOCATOR_SHIM_INJECTION'))
+    if dynamicRT == "force":
+        add_shim_hack(
+            env,
+            partial(
+                shim_hack,
+                inject_target='shim_cxx',
+                disable_var='DISABLE_CXX_SHIM_INJECTION'))
+        add_shim_hack(
+            env,
+            partial(
+                shim_hack,
+                inject_target='shim_crt',
+                disable_var='DISABLE_CRT_SHIM_INJECTION'))
+
 
 # NOTE: We must do third_party first as it adds methods to the environment
 # that we need in the mongo sconscript