OpenE2K/rust
2
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
a6a9e09f98
rust/mk/rt.mk

373 lines
13 KiB
Makefile
Raw Normal View History

Refactor the build system for easily adding crates Before this patch, if you wanted to add a crate to the build system you had to change about 100 lines across 8 separate makefiles. This is highly error prone and opaque to all but a few. This refactoring is targeted at consolidating this effort so adding a new crate adds one line in one file in a way that everyone can understand it.
2014-01-25 04:27:22 +01:00
# Copyright 2014 The Rust Project Developers. See the COPYRIGHT
# file at the top-level directory of this distribution and at
# http://rust-lang.org/COPYRIGHT.
virtualize the runtime per target triple.
2011-11-02 00:50:47 +01:00
#
Refactor the build system for easily adding crates Before this patch, if you wanted to add a crate to the build system you had to change about 100 lines across 8 separate makefiles. This is highly error prone and opaque to all but a few. This refactoring is targeted at consolidating this effort so adding a new crate adds one line in one file in a way that everyone can understand it.
2014-01-25 04:27:22 +01:00
# Licensed under the Apache License, Version 2.0 <LICENSE-APACHE or
# http://www.apache.org/licenses/LICENSE-2.0> or the MIT license
# <LICENSE-MIT or http://opensource.org/licenses/MIT>, at your
# option. This file may not be copied, modified, or distributed
# except according to those terms.
################################################################################
# Native libraries built as part of the rust build process
virtualize the runtime per target triple.
2011-11-02 00:50:47 +01:00
#
Refactor the build system for easily adding crates Before this patch, if you wanted to add a crate to the build system you had to change about 100 lines across 8 separate makefiles. This is highly error prone and opaque to all but a few. This refactoring is targeted at consolidating this effort so adding a new crate adds one line in one file in a way that everyone can understand it.
2014-01-25 04:27:22 +01:00
# This portion of the rust build system is meant to keep track of native
# dependencies and how to build them. It is currently required that all native
# dependencies are built as static libraries, as slinging around dynamic
# libraries isn't exactly the most fun thing to do.
virtualize the runtime per target triple.
2011-11-02 00:50:47 +01:00
#
Refactor the build system for easily adding crates Before this patch, if you wanted to add a crate to the build system you had to change about 100 lines across 8 separate makefiles. This is highly error prone and opaque to all but a few. This refactoring is targeted at consolidating this effort so adding a new crate adds one line in one file in a way that everyone can understand it.
2014-01-25 04:27:22 +01:00
# This section should need minimal modification to add new libraries. The
# relevant variables are:
virtualize the runtime per target triple.
2011-11-02 00:50:47 +01:00
#
Refactor the build system for easily adding crates Before this patch, if you wanted to add a crate to the build system you had to change about 100 lines across 8 separate makefiles. This is highly error prone and opaque to all but a few. This refactoring is targeted at consolidating this effort so adding a new crate adds one line in one file in a way that everyone can understand it.
2014-01-25 04:27:22 +01:00
# NATIVE_LIBS
# This is a list of all native libraries which are built as part of the
# build process. It will build all libraries into RT_OUTPUT_DIR with the
# appropriate name of static library as dictated by the target platform
#
# NATIVE_DEPS_<lib>
# This is a list of files relative to the src/rt directory which are
# needed to build the native library. Each file will be compiled to an
# object file, and then all the object files will be assembled into an
# archive (static library). The list contains files of any extension
#
# If adding a new library, you should update the NATIVE_LIBS list, and then list
# the required files below it. The list of required files is a list of files
# that's per-target so you're allowed to conditionally add files based on the
# target.
################################################################################
std: Extract librustrt out of libstd As part of the libstd facade efforts, this commit extracts the runtime interface out of the standard library into a standalone crate, librustrt. This crate will provide the following services: * Definition of the rtio interface * Definition of the Runtime interface * Implementation of the Task structure * Implementation of task-local-data * Implementation of task failure via unwinding via libunwind * Implementation of runtime initialization and shutdown * Implementation of thread-local-storage for the local rust Task Notably, this crate avoids the following services: * Thread creation and destruction. The crate does not require the knowledge of an OS threading system, and as a result it seemed best to leave out the `rt::thread` module from librustrt. The librustrt module does depend on mutexes, however. * Implementation of backtraces. There is no inherent requirement for the runtime to be able to generate backtraces. As will be discussed later, this functionality continues to live in libstd rather than librustrt. As usual, a number of architectural changes were required to make this crate possible. Users of "stable" functionality will not be impacted by this change, but users of the `std::rt` module will likely note the changes. A list of architectural changes made is: * The stdout/stderr handles no longer live directly inside of the `Task` structure. This is a consequence of librustrt not knowing about `std::io`. These two handles are now stored inside of task-local-data. The handles were originally stored inside of the `Task` for perf reasons, and TLD is not currently as fast as it could be. For comparison, 100k prints goes from 59ms to 68ms (a 15% slowdown). This appeared to me to be an acceptable perf loss for the successful extraction of a librustrt crate. * The `rtio` module was forced to duplicate more functionality of `std::io`. As the module no longer depends on `std::io`, `rtio` now defines structures such as socket addresses, addrinfo fiddly bits, etc. The primary change made was that `rtio` now defines its own `IoError` type. This type is distinct from `std::io::IoError` in that it does not have an enum for what error occurred, but rather a platform-specific error code. The native and green libraries will be updated in later commits for this change, and the bulk of this effort was put behind updating the two libraries for this change (with `rtio`). * Printing a message on task failure (along with the backtrace) continues to live in libstd, not in librustrt. This is a consequence of the above decision to move the stdout/stderr handles to TLD rather than inside the `Task` itself. The unwinding API now supports registration of global callback functions which will be invoked when a task fails, allowing for libstd to register a function to print a message and a backtrace. The API for registering a callback is experimental and unsafe, as the ramifications of running code on unwinding is pretty hairy. * The `std::unstable::mutex` module has moved to `std::rt::mutex`. * The `std::unstable::sync` module has been moved to `std::rt::exclusive` and the type has been rewritten to not internally have an Arc and to have an RAII guard structure when locking. Old code should stop using `Exclusive` in favor of the primitives in `libsync`, but if necessary, old code should port to `Arc<Exclusive<T>>`. * The local heap has been stripped down to have fewer debugging options. None of these were tested, and none of these have been used in a very long time. [breaking-change]
2014-06-04 04:11:49 +02:00
NATIVE_LIBS := rust_builtin hoedown uv_support morestack miniz context_switch \
rustrt_native rust_test_helpers
Refactor the build system for easily adding crates Before this patch, if you wanted to add a crate to the build system you had to change about 100 lines across 8 separate makefiles. This is highly error prone and opaque to all but a few. This refactoring is targeted at consolidating this effort so adding a new crate adds one line in one file in a way that everyone can understand it.
2014-01-25 04:27:22 +01:00
# $(1) is the target triple
define NATIVE_LIBRARIES
rustdoc: Migrate from sundown to hoedown This primary fix brought on by this upgrade is the proper matching of the ``` and ~~~ doc blocks. This also moves hoedown to a git submodule rather than a bundled repository. Additionally, hoedown is stricter about code blocks, so this ended up fixing a lot of invalid code blocks (ending with " ```" instead of "```", or ending with "~~~~" instead of "~~~"). Closes #12776
2014-05-03 02:56:35 +02:00
NATIVE_DEPS_hoedown_$(1) := hoedown/src/autolink.c \
hoedown/src/buffer.c \
hoedown/src/document.c \
hoedown/src/escape.c \
hoedown/src/html.c \
hoedown/src/html_blocks.c \
hoedown/src/html_smartypants.c \
hoedown/src/stack.c \
hoedown/src/version.c
Refactor the build system for easily adding crates Before this patch, if you wanted to add a crate to the build system you had to change about 100 lines across 8 separate makefiles. This is highly error prone and opaque to all but a few. This refactoring is targeted at consolidating this effort so adding a new crate adds one line in one file in a way that everyone can understand it.
2014-01-25 04:27:22 +01:00
NATIVE_DEPS_uv_support_$(1) := rust_uv.c
Move extra::flate to libflate This is hopefully the beginning of the long-awaited dissolution of libextra. Using the newly created build infrastructure for building libraries, I decided to move the first module out of libextra. While not being a particularly meaty module in and of itself, the flate module is required by rustc and additionally has a native C dependency. I was able to very easily split out the C dependency from rustrt, update librustc, and magically everything gets installed to the right locations and built automatically. This is meant to be a proof-of-concept commit to how easy it is to remove modules from libextra now. I didn't put any effort into modernizing the interface of libflate or updating it other than to remove the one glob import it had.
2014-01-25 06:00:31 +01:00
NATIVE_DEPS_miniz_$(1) = miniz.c
std: Extract librustrt out of libstd As part of the libstd facade efforts, this commit extracts the runtime interface out of the standard library into a standalone crate, librustrt. This crate will provide the following services: * Definition of the rtio interface * Definition of the Runtime interface * Implementation of the Task structure * Implementation of task-local-data * Implementation of task failure via unwinding via libunwind * Implementation of runtime initialization and shutdown * Implementation of thread-local-storage for the local rust Task Notably, this crate avoids the following services: * Thread creation and destruction. The crate does not require the knowledge of an OS threading system, and as a result it seemed best to leave out the `rt::thread` module from librustrt. The librustrt module does depend on mutexes, however. * Implementation of backtraces. There is no inherent requirement for the runtime to be able to generate backtraces. As will be discussed later, this functionality continues to live in libstd rather than librustrt. As usual, a number of architectural changes were required to make this crate possible. Users of "stable" functionality will not be impacted by this change, but users of the `std::rt` module will likely note the changes. A list of architectural changes made is: * The stdout/stderr handles no longer live directly inside of the `Task` structure. This is a consequence of librustrt not knowing about `std::io`. These two handles are now stored inside of task-local-data. The handles were originally stored inside of the `Task` for perf reasons, and TLD is not currently as fast as it could be. For comparison, 100k prints goes from 59ms to 68ms (a 15% slowdown). This appeared to me to be an acceptable perf loss for the successful extraction of a librustrt crate. * The `rtio` module was forced to duplicate more functionality of `std::io`. As the module no longer depends on `std::io`, `rtio` now defines structures such as socket addresses, addrinfo fiddly bits, etc. The primary change made was that `rtio` now defines its own `IoError` type. This type is distinct from `std::io::IoError` in that it does not have an enum for what error occurred, but rather a platform-specific error code. The native and green libraries will be updated in later commits for this change, and the bulk of this effort was put behind updating the two libraries for this change (with `rtio`). * Printing a message on task failure (along with the backtrace) continues to live in libstd, not in librustrt. This is a consequence of the above decision to move the stdout/stderr handles to TLD rather than inside the `Task` itself. The unwinding API now supports registration of global callback functions which will be invoked when a task fails, allowing for libstd to register a function to print a message and a backtrace. The API for registering a callback is experimental and unsafe, as the ramifications of running code on unwinding is pretty hairy. * The `std::unstable::mutex` module has moved to `std::rt::mutex`. * The `std::unstable::sync` module has been moved to `std::rt::exclusive` and the type has been rewritten to not internally have an Arc and to have an RAII guard structure when locking. Old code should stop using `Exclusive` in favor of the primitives in `libsync`, but if necessary, old code should port to `Arc<Exclusive<T>>`. * The local heap has been stripped down to have fewer debugging options. None of these were tested, and none of these have been used in a very long time. [breaking-change]
2014-06-04 04:11:49 +02:00
NATIVE_DEPS_rust_builtin_$(1) := rust_builtin.c \
rust_android_dummy.c
NATIVE_DEPS_rustrt_native_$(1) := \
Refactor the build system for easily adding crates Before this patch, if you wanted to add a crate to the build system you had to change about 100 lines across 8 separate makefiles. This is highly error prone and opaque to all but a few. This refactoring is targeted at consolidating this effort so adding a new crate adds one line in one file in a way that everyone can understand it.
2014-01-25 04:27:22 +01:00
rust_try.ll \
arch/$$(HOST_$(1))/record_sp.S
mk: Move rust_test_helpers out of libstd There's no need to distribute these ABI helpers for tests with the standard rust distribution they're only needed for our tests. Closes #2665
2014-06-05 20:08:43 +02:00
NATIVE_DEPS_rust_test_helpers_$(1) := rust_test_helpers.c
Refactor the build system for easily adding crates Before this patch, if you wanted to add a crate to the build system you had to change about 100 lines across 8 separate makefiles. This is highly error prone and opaque to all but a few. This refactoring is targeted at consolidating this effort so adding a new crate adds one line in one file in a way that everyone can understand it.
2014-01-25 04:27:22 +01:00
NATIVE_DEPS_morestack_$(1) := arch/$$(HOST_$(1))/morestack.S
Remove two allocations from spawning a green task Two unfortunate allocations were wrapping a proc() in a proc() with GreenTask::build_start_wrapper, and then boxing this proc in a ~proc() inside of Context::new(). Both of these allocations were a direct result from two conditions: 1. The Context::new() function has a nice api of taking a procedure argument to start up a new context with. This inherently required an allocation by build_start_wrapper because extra code needed to be run around the edges of a user-provided proc() for a new task. 2. The initial bootstrap code only understood how to pass one argument to the next function. By modifying the assembly and entry points to understand more than one argument, more information is passed through in registers instead of allocating a pointer-sized context. This is sadly where I end up throwing mips under a bus because I have no idea what's going on in the mips context switching code and don't know how to modify it. Closes #7767 cc #11389
2014-02-11 01:13:50 +01:00
NATIVE_DEPS_context_switch_$(1) := \
arch/$$(HOST_$(1))/_context.S
Refactor the build system for easily adding crates Before this patch, if you wanted to add a crate to the build system you had to change about 100 lines across 8 separate makefiles. This is highly error prone and opaque to all but a few. This refactoring is targeted at consolidating this effort so adding a new crate adds one line in one file in a way that everyone can understand it.
2014-01-25 04:27:22 +01:00
################################################################################
# You shouldn't find it that necessary to edit anything below this line.
################################################################################
# While we're defining the native libraries for each target, we define some
# common rules used to build files for various targets.
virtualize the runtime per target triple.
2011-11-02 00:50:47 +01:00
Refactor the build system for easily adding crates Before this patch, if you wanted to add a crate to the build system you had to change about 100 lines across 8 separate makefiles. This is highly error prone and opaque to all but a few. This refactoring is targeted at consolidating this effort so adding a new crate adds one line in one file in a way that everyone can understand it.
2014-01-25 04:27:22 +01:00
RT_OUTPUT_DIR_$(1) := $(1)/rt
Remove VPATH usage in Makefiles This is just messing up all the doc dependencies because of such similar directory names and file names. Closes #11422
2014-01-28 23:14:51 +01:00
$$(RT_OUTPUT_DIR_$(1))/%.o: $(S)src/rt/%.ll $$(MKFILE_DEPS) \
$$(LLVM_CONFIG_$$(CFG_BUILD))
Refactor the build system for easily adding crates Before this patch, if you wanted to add a crate to the build system you had to change about 100 lines across 8 separate makefiles. This is highly error prone and opaque to all but a few. This refactoring is targeted at consolidating this effort so adding a new crate adds one line in one file in a way that everyone can understand it.
2014-01-25 04:27:22 +01:00
@mkdir -p $$(@D)
@$$(call E, compile: $$@)
$$(Q)$$(LLC_$$(CFG_BUILD)) $$(CFG_LLC_FLAGS_$(1)) \
-filetype=obj -mtriple=$(1) -relocation-model=pic -o $$@ $$<
Remove VPATH usage in Makefiles This is just messing up all the doc dependencies because of such similar directory names and file names. Closes #11422
2014-01-28 23:14:51 +01:00
$$(RT_OUTPUT_DIR_$(1))/%.o: $(S)src/rt/%.c $$(MKFILE_DEPS)
Refactor the build system for easily adding crates Before this patch, if you wanted to add a crate to the build system you had to change about 100 lines across 8 separate makefiles. This is highly error prone and opaque to all but a few. This refactoring is targeted at consolidating this effort so adding a new crate adds one line in one file in a way that everyone can understand it.
2014-01-25 04:27:22 +01:00
@mkdir -p $$(@D)
@$$(call E, compile: $$@)
$$(Q)$$(call CFG_COMPILE_C_$(1), $$@, \
rustdoc: Migrate from sundown to hoedown This primary fix brought on by this upgrade is the proper matching of the ``` and ~~~ doc blocks. This also moves hoedown to a git submodule rather than a bundled repository. Additionally, hoedown is stricter about code blocks, so this ended up fixing a lot of invalid code blocks (ending with " ```" instead of "```", or ending with "~~~~" instead of "~~~"). Closes #12776
2014-05-03 02:56:35 +02:00
-I $$(S)src/rt/hoedown/src \
Refactor the build system for easily adding crates Before this patch, if you wanted to add a crate to the build system you had to change about 100 lines across 8 separate makefiles. This is highly error prone and opaque to all but a few. This refactoring is targeted at consolidating this effort so adding a new crate adds one line in one file in a way that everyone can understand it.
2014-01-25 04:27:22 +01:00
-I $$(S)src/libuv/include -I $$(S)src/rt \
$$(RUNTIME_CFLAGS_$(1))) $$<
Remove VPATH usage in Makefiles This is just messing up all the doc dependencies because of such similar directory names and file names. Closes #11422
2014-01-28 23:14:51 +01:00
$$(RT_OUTPUT_DIR_$(1))/%.o: $(S)src/rt/%.S $$(MKFILE_DEPS) \
$$(LLVM_CONFIG_$$(CFG_BUILD))
Refactor the build system for easily adding crates Before this patch, if you wanted to add a crate to the build system you had to change about 100 lines across 8 separate makefiles. This is highly error prone and opaque to all but a few. This refactoring is targeted at consolidating this effort so adding a new crate adds one line in one file in a way that everyone can understand it.
2014-01-25 04:27:22 +01:00
@mkdir -p $$(@D)
@$$(call E, compile: $$@)
$$(Q)$$(call CFG_ASSEMBLE_$(1),$$@,$$<)
endef
$(foreach target,$(CFG_TARGET),$(eval $(call NATIVE_LIBRARIES,$(target))))
# A macro for devining how to build third party libraries listed above (based
# on their dependencies).
#
# $(1) is the target
# $(2) is the lib name
define THIRD_PARTY_LIB
OBJS_$(2)_$(1) := $$(NATIVE_DEPS_$(2)_$(1):%=$$(RT_OUTPUT_DIR_$(1))/%)
OBJS_$(2)_$(1) := $$(OBJS_$(2)_$(1):.c=.o)
OBJS_$(2)_$(1) := $$(OBJS_$(2)_$(1):.cpp=.o)
OBJS_$(2)_$(1) := $$(OBJS_$(2)_$(1):.ll=.o)
OBJS_$(2)_$(1) := $$(OBJS_$(2)_$(1):.S=.o)
NATIVE_$(2)_$(1) := $$(call CFG_STATIC_LIB_NAME_$(1),$(2))
$$(RT_OUTPUT_DIR_$(1))/$$(NATIVE_$(2)_$(1)): $$(OBJS_$(2)_$(1))
@$$(call E, link: $$@)
$$(Q)$$(AR_$(1)) rcs $$@ $$^
endef
$(foreach target,$(CFG_TARGET), \
$(eval $(call RUNTIME_RULES,$(target))))
$(foreach lib,$(NATIVE_LIBS), \
$(foreach target,$(CFG_TARGET), \
$(eval $(call THIRD_PARTY_LIB,$(target),$(lib)))))
Disable some advanced (post glibc-2.3) libuv features when building snaps.
2012-03-27 00:54:22 +02:00
Refactor the build system for easily adding crates Before this patch, if you wanted to add a crate to the build system you had to change about 100 lines across 8 separate makefiles. This is highly error prone and opaque to all but a few. This refactoring is targeted at consolidating this effort so adding a new crate adds one line in one file in a way that everyone can understand it.
2014-01-25 04:27:22 +01:00
################################################################################
# Building third-party targets with external build systems
#
add back jemalloc to the tree This adds a `std::rt::heap` module with a nice allocator API. It's a step towards fixing #13094 and is a starting point for working on a generic allocator trait. The revision used for the jemalloc submodule is the stable 3.6.0 release. Closes #11807
2014-04-25 08:19:34 +02:00
# This location is meant for dependencies which have external build systems. It
# is still assumed that the output of each of these steps is a static library
# in the correct location.
################################################################################
################################################################################
# libuv
Refactor the build system for easily adding crates Before this patch, if you wanted to add a crate to the build system you had to change about 100 lines across 8 separate makefiles. This is highly error prone and opaque to all but a few. This refactoring is targeted at consolidating this effort so adding a new crate adds one line in one file in a way that everyone can understand it.
2014-01-25 04:27:22 +01:00
################################################################################
Disable some advanced (post glibc-2.3) libuv features when building snaps.
2012-03-27 00:54:22 +02:00
Fix compilation issues with libuv on OSX It turns out that gyp (libuv's new build system) wants x64 for a 64-bit x86 architecture and ia32 for a 32-bit architecture, so this performs the relevant mapping and then invokes libuv's configure script with the appropriate target architecture. This can be verified by running make with VERBOSE=1 and seeing that beforehand on a 64-bit build libuv was passed "-arch i386" and now it's passed "-arch x86_64" Closes #8826
2013-08-29 02:01:44 +02:00
define DEF_LIBUV_ARCH_VAR
LIBUV_ARCH_$(1) = $$(subst i386,ia32,$$(subst x86_64,x64,$$(HOST_$(1))))
endef
support for GNU configure syntax
2013-10-21 11:18:21 +02:00
$(foreach t,$(CFG_TARGET),$(eval $(call DEF_LIBUV_ARCH_VAR,$(t))))
Fix compilation issues with libuv on OSX It turns out that gyp (libuv's new build system) wants x64 for a 64-bit x86 architecture and ia32 for a 32-bit architecture, so this performs the relevant mapping and then invokes libuv's configure script with the appropriate target architecture. This can be verified by running make with VERBOSE=1 and seeing that beforehand on a 64-bit build libuv was passed "-arch i386" and now it's passed "-arch x86_64" Closes #8826
2013-08-29 02:01:44 +02:00
Build libuv/jemalloc only once (not per stage) Reorganize the makefiles to stop building these once per stage because there's no need to do this at all. Closes #7002
2013-10-08 18:20:17 +02:00
ifdef CFG_ENABLE_FAST_MAKE
LIBUV_DEPS := $(S)/.gitmodules
else
LIBUV_DEPS := $(wildcard \
$(S)src/libuv/* \
$(S)src/libuv/*/* \
$(S)src/libuv/*/*/* \
$(S)src/libuv/*/*/*/*)
endif
LIBUV_NO_LOAD = run-benchmarks.target.mk run-tests.target.mk \
uv_dtrace_header.target.mk uv_dtrace_provider.target.mk
export PYTHONPATH := $(PYTHONPATH):$(S)src/gyp/pylib
define DEF_THIRD_PARTY_TARGETS
# $(1) is the target triple
ifeq ($$(CFG_WINDOWSY_$(1)), 1)
LIBUV_OSTYPE_$(1) := win
add back jemalloc to the tree This adds a `std::rt::heap` module with a nice allocator API. It's a step towards fixing #13094 and is a starting point for working on a generic allocator trait. The revision used for the jemalloc submodule is the stable 3.6.0 release. Closes #11807
2014-04-25 08:19:34 +02:00
# This isn't necessarily a desired option, but it's harmless and works around
# what appears to be a mingw-w64 bug.
#
# https://sourceforge.net/p/mingw-w64/bugs/395/
JEMALLOC_ARGS_$(1) := --enable-lazy-lock
Build libuv/jemalloc only once (not per stage) Reorganize the makefiles to stop building these once per stage because there's no need to do this at all. Closes #7002
2013-10-08 18:20:17 +02:00
else ifeq ($(OSTYPE_$(1)), apple-darwin)
LIBUV_OSTYPE_$(1) := mac
else ifeq ($(OSTYPE_$(1)), unknown-freebsd)
LIBUV_OSTYPE_$(1) := freebsd
else ifeq ($(OSTYPE_$(1)), linux-androideabi)
LIBUV_OSTYPE_$(1) := android
LIBUV_ARGS_$(1) := PLATFORM=android host=android OS=linux
add back jemalloc to the tree This adds a `std::rt::heap` module with a nice allocator API. It's a step towards fixing #13094 and is a starting point for working on a generic allocator trait. The revision used for the jemalloc submodule is the stable 3.6.0 release. Closes #11807
2014-04-25 08:19:34 +02:00
JEMALLOC_ARGS_$(1) := --disable-tls
support a fast-make mode that avoids globbing into llvm and libuv
2012-02-10 21:07:01 +01:00
else
Build libuv/jemalloc only once (not per stage) Reorganize the makefiles to stop building these once per stage because there's no need to do this at all. Closes #7002
2013-10-08 18:20:17 +02:00
LIBUV_OSTYPE_$(1) := linux
support a fast-make mode that avoids globbing into llvm and libuv
2012-02-10 21:07:01 +01:00
endif
Build libuv/jemalloc only once (not per stage) Reorganize the makefiles to stop building these once per stage because there's no need to do this at all. Closes #7002
2013-10-08 18:20:17 +02:00
LIBUV_NAME_$(1) := $$(call CFG_STATIC_LIB_NAME_$(1),uv)
uv: Suppress a warning by using an absolute path Turns out libuv's build system doesn't like us telling them that the build directory is a relative location, as it always spits out a warning about a circular dependency being dropped. By using an absolute path, turns out the warnings isn't spit out, who knew? Closes #11067
2013-12-19 08:44:10 +01:00
LIBUV_DIR_$(1) := $$(RT_OUTPUT_DIR_$(1))/libuv
Refactor the build system for easily adding crates Before this patch, if you wanted to add a crate to the build system you had to change about 100 lines across 8 separate makefiles. This is highly error prone and opaque to all but a few. This refactoring is targeted at consolidating this effort so adding a new crate adds one line in one file in a way that everyone can understand it.
2014-01-25 04:27:22 +01:00
LIBUV_LIB_$(1) := $$(RT_OUTPUT_DIR_$(1))/$$(LIBUV_NAME_$(1))
Upgrade libuv to the current master (again) This is a reopening of the libuv-upgrade part of #8645. Hopefully this won't cause random segfaults all over the place. The windows regression in testing should also be fixed (it shouldn't build the whole compiler twice). A notable difference from before is that gyp is now a git submodule instead of always git-cloned at make time. This allows bundling for releases more easily. Closes #8850
2013-08-19 02:11:45 +02:00
Build libuv/jemalloc only once (not per stage) Reorganize the makefiles to stop building these once per stage because there's no need to do this at all. Closes #7002
2013-10-08 18:20:17 +02:00
LIBUV_MAKEFILE_$(1) := $$(CFG_BUILD_DIR)$$(RT_OUTPUT_DIR_$(1))/libuv/Makefile
Upgrade libuv to the current master (again) This is a reopening of the libuv-upgrade part of #8645. Hopefully this won't cause random segfaults all over the place. The windows regression in testing should also be fixed (it shouldn't build the whole compiler twice). A notable difference from before is that gyp is now a git submodule instead of always git-cloned at make time. This allows bundling for releases more easily. Closes #8850
2013-08-19 02:11:45 +02:00
uv: Suppress a warning by using an absolute path Turns out libuv's build system doesn't like us telling them that the build directory is a relative location, as it always spits out a warning about a circular dependency being dropped. By using an absolute path, turns out the warnings isn't spit out, who knew? Closes #11067
2013-12-19 08:44:10 +01:00
LIBUV_STAMP_$(1) = $$(LIBUV_DIR_$(1))/libuv-auto-clean-stamp
$$(LIBUV_STAMP_$(1)): $(S)src/rt/libuv-auto-clean-trigger
$$(Q)rm -rf $$(LIBUV_DIR_$(1))
$$(Q)mkdir -p $$(@D)
touch $$@
Remove sjlj stuff from rust_upcall and don't pass -Werror to libuv.
2013-11-23 03:21:37 +01:00
# libuv triggers a few warnings on some platforms
LIBUV_CFLAGS_$(1) := $(subst -Werror,,$(CFG_GCCISH_CFLAGS_$(1)))
uv: Suppress a warning by using an absolute path Turns out libuv's build system doesn't like us telling them that the build directory is a relative location, as it always spits out a warning about a circular dependency being dropped. By using an absolute path, turns out the warnings isn't spit out, who knew? Closes #11067
2013-12-19 08:44:10 +01:00
$$(LIBUV_MAKEFILE_$(1)): $$(LIBUV_DEPS) $$(MKFILE_DEPS) $$(LIBUV_STAMP_$(1))
Upgrade libuv to the current master (again) This is a reopening of the libuv-upgrade part of #8645. Hopefully this won't cause random segfaults all over the place. The windows regression in testing should also be fixed (it shouldn't build the whole compiler twice). A notable difference from before is that gyp is now a git submodule instead of always git-cloned at make time. This allows bundling for releases more easily. Closes #8850
2013-08-19 02:11:45 +02:00
(cd $(S)src/libuv/ && \
Update to the latest libuv At this time, also point the libuv submodule to the official repo instead of my own off to the side. cc #10246 Closes #10329
2013-11-08 00:26:47 +01:00
$$(CFG_PYTHON) ./gyp_uv.py -f make -Dtarget_arch=$$(LIBUV_ARCH_$(1)) \
Build libuv/jemalloc only once (not per stage) Reorganize the makefiles to stop building these once per stage because there's no need to do this at all. Closes #7002
2013-10-08 18:20:17 +02:00
-D ninja \
-DOS=$$(LIBUV_OSTYPE_$(1)) \
Upgrade libuv to the current master (again) This is a reopening of the libuv-upgrade part of #8645. Hopefully this won't cause random segfaults all over the place. The windows regression in testing should also be fixed (it shouldn't build the whole compiler twice). A notable difference from before is that gyp is now a git submodule instead of always git-cloned at make time. This allows bundling for releases more easily. Closes #8850
2013-08-19 02:11:45 +02:00
-Goutput_dir=$$(@D) --generator-output $$(@D))
uv: Suppress a warning by using an absolute path Turns out libuv's build system doesn't like us telling them that the build directory is a relative location, as it always spits out a warning about a circular dependency being dropped. By using an absolute path, turns out the warnings isn't spit out, who knew? Closes #11067
2013-12-19 08:44:10 +01:00
touch $$@
Upgrade libuv to the current master (again) This is a reopening of the libuv-upgrade part of #8645. Hopefully this won't cause random segfaults all over the place. The windows regression in testing should also be fixed (it shouldn't build the whole compiler twice). A notable difference from before is that gyp is now a git submodule instead of always git-cloned at make time. This allows bundling for releases more easily. Closes #8850
2013-08-19 02:11:45 +02:00
Build libuv/jemalloc only once (not per stage) Reorganize the makefiles to stop building these once per stage because there's no need to do this at all. Closes #7002
2013-10-08 18:20:17 +02:00
# Windows has a completely different build system for libuv because of mingw. In
# theory when we support msvc then we should be using gyp's msvc output instead
# of mingw's makefile for windows
mk: cleanup - lib and executable suffix handling
2013-03-02 13:25:12 +01:00
ifdef CFG_WINDOWSY_$(1)
uv: Suppress a warning by using an absolute path Turns out libuv's build system doesn't like us telling them that the build directory is a relative location, as it always spits out a warning about a circular dependency being dropped. By using an absolute path, turns out the warnings isn't spit out, who knew? Closes #11067
2013-12-19 08:44:10 +01:00
$$(LIBUV_LIB_$(1)): $$(LIBUV_DEPS) $$(MKFILE_DEPS)
Upgrade libuv to the current master (again) This is a reopening of the libuv-upgrade part of #8645. Hopefully this won't cause random segfaults all over the place. The windows regression in testing should also be fixed (it shouldn't build the whole compiler twice). A notable difference from before is that gyp is now a git submodule instead of always git-cloned at make time. This allows bundling for releases more easily. Closes #8850
2013-08-19 02:11:45 +02:00
$$(Q)$$(MAKE) -C $$(S)src/libuv -f Makefile.mingw \
Remove sjlj stuff from rust_upcall and don't pass -Werror to libuv.
2013-11-23 03:21:37 +01:00
LDFLAGS="$$(CFG_GCCISH_LINK_FLAGS_$(1))" \
CC="$$(CC_$(1)) $$(LIBUV_CFLAGS_$(1)) $$(SNAP_DEFINES)" \
CXX="$$(CXX_$(1))" \
Upgrade libuv to the current master (again) This is a reopening of the libuv-upgrade part of #8645. Hopefully this won't cause random segfaults all over the place. The windows regression in testing should also be fixed (it shouldn't build the whole compiler twice). A notable difference from before is that gyp is now a git submodule instead of always git-cloned at make time. This allows bundling for releases more easily. Closes #8850
2013-08-19 02:11:45 +02:00
AR="$$(AR_$(1))" \
build: ifdef for mingw/non-mingw builds
2013-02-02 21:14:04 +01:00
V=$$(VERBOSE)
Upgrade libuv to the current master (again) This is a reopening of the libuv-upgrade part of #8645. Hopefully this won't cause random segfaults all over the place. The windows regression in testing should also be fixed (it shouldn't build the whole compiler twice). A notable difference from before is that gyp is now a git submodule instead of always git-cloned at make time. This allows bundling for releases more easily. Closes #8850
2013-08-19 02:11:45 +02:00
$$(Q)cp $$(S)src/libuv/libuv.a $$@
mk: rewrite make files
2013-02-27 06:53:35 +01:00
else
uv: Suppress a warning by using an absolute path Turns out libuv's build system doesn't like us telling them that the build directory is a relative location, as it always spits out a warning about a circular dependency being dropped. By using an absolute path, turns out the warnings isn't spit out, who knew? Closes #11067
2013-12-19 08:44:10 +01:00
$$(LIBUV_LIB_$(1)): $$(LIBUV_DIR_$(1))/Release/libuv.a $$(MKFILE_DEPS)
Refactor the build system for easily adding crates Before this patch, if you wanted to add a crate to the build system you had to change about 100 lines across 8 separate makefiles. This is highly error prone and opaque to all but a few. This refactoring is targeted at consolidating this effort so adding a new crate adds one line in one file in a way that everyone can understand it.
2014-01-25 04:27:22 +01:00
$$(Q)cp $$< $$@
uv: Suppress a warning by using an absolute path Turns out libuv's build system doesn't like us telling them that the build directory is a relative location, as it always spits out a warning about a circular dependency being dropped. By using an absolute path, turns out the warnings isn't spit out, who knew? Closes #11067
2013-12-19 08:44:10 +01:00
$$(LIBUV_DIR_$(1))/Release/libuv.a: $$(LIBUV_DEPS) $$(LIBUV_MAKEFILE_$(1)) \
$$(MKFILE_DEPS)
$$(Q)$$(MAKE) -C $$(LIBUV_DIR_$(1)) \
Remove sjlj stuff from rust_upcall and don't pass -Werror to libuv.
2013-11-23 03:21:37 +01:00
CFLAGS="$$(LIBUV_CFLAGS_$(1)) $$(SNAP_DEFINES)" \
mk: Get rid of redundant LIBUV_FLAGS.
2013-11-20 12:00:49 +01:00
LDFLAGS="$$(CFG_GCCISH_LINK_FLAGS_$(1))" \
mk: mips toolchain config
2013-03-19 06:32:03 +01:00
CC="$$(CC_$(1))" \
CXX="$$(CXX_$(1))" \
AR="$$(AR_$(1))" \
Build libuv/jemalloc only once (not per stage) Reorganize the makefiles to stop building these once per stage because there's no need to do this at all. Closes #7002
2013-10-08 18:20:17 +02:00
$$(LIBUV_ARGS_$(1)) \
Upgrade libuv to the current master (again) This is a reopening of the libuv-upgrade part of #8645. Hopefully this won't cause random segfaults all over the place. The windows regression in testing should also be fixed (it shouldn't build the whole compiler twice). A notable difference from before is that gyp is now a git submodule instead of always git-cloned at make time. This allows bundling for releases more easily. Closes #8850
2013-08-19 02:11:45 +02:00
BUILDTYPE=Release \
NO_LOAD="$$(LIBUV_NO_LOAD)" \
build: tweak rt.mk, as per graydon.. CFLAGS had to stay
2013-02-02 16:40:40 +01:00
V=$$(VERBOSE)
mk: Always touch libuv.a libuv.a always looks out of date to the makefile, causing make to always descend into the libuv makefile, even when there's nothing to build.
2014-03-27 22:29:07 +01:00
$$(Q)touch $$@
Statically link libuv to librustuv Similarly to the previous commit, libuv is only used by this library, so there's no need for it to be linked into librustrt and available to all crates by default.
2013-10-31 07:57:27 +01:00
Refactor the build system for easily adding crates Before this patch, if you wanted to add a crate to the build system you had to change about 100 lines across 8 separate makefiles. This is highly error prone and opaque to all but a few. This refactoring is targeted at consolidating this effort so adding a new crate adds one line in one file in a way that everyone can understand it.
2014-01-25 04:27:22 +01:00
endif
Statically link sundown to librustdoc Closes #10103
2013-10-31 07:22:04 +01:00
add back jemalloc to the tree This adds a `std::rt::heap` module with a nice allocator API. It's a step towards fixing #13094 and is a starting point for working on a generic allocator trait. The revision used for the jemalloc submodule is the stable 3.6.0 release. Closes #11807
2014-04-25 08:19:34 +02:00
################################################################################
# jemalloc
################################################################################
ifdef CFG_ENABLE_FAST_MAKE
JEMALLOC_DEPS := $(S)/.gitmodules
else
JEMALLOC_DEPS := $(wildcard \
$(S)src/jemalloc/* \
$(S)src/jemalloc/*/* \
$(S)src/jemalloc/*/*/* \
$(S)src/jemalloc/*/*/*/*)
endif
JEMALLOC_NAME_$(1) := $$(call CFG_STATIC_LIB_NAME_$(1),jemalloc)
ifeq ($$(CFG_WINDOWSY_$(1)),1)
JEMALLOC_REAL_NAME_$(1) := $$(call CFG_STATIC_LIB_NAME_$(1),jemalloc_s)
else
JEMALLOC_REAL_NAME_$(1) := $$(call CFG_STATIC_LIB_NAME_$(1),jemalloc_pic)
endif
JEMALLOC_LIB_$(1) := $$(RT_OUTPUT_DIR_$(1))/$$(JEMALLOC_NAME_$(1))
JEMALLOC_BUILD_DIR_$(1) := $$(RT_OUTPUT_DIR_$(1))/jemalloc
$$(JEMALLOC_LIB_$(1)): $$(JEMALLOC_DEPS) $$(MKFILE_DEPS)
@$$(call E, make: jemalloc)
cd "$$(JEMALLOC_BUILD_DIR_$(1))"; "$(S)src/jemalloc/configure" \
$$(JEMALLOC_ARGS_$(1)) --enable-cc-silence --with-jemalloc-prefix=je_ \
--disable-experimental --build=$(CFG_BUILD) --host=$(1) \
CC="$$(CC_$(1))" \
AR="$$(AR_$(1))" \
RANLIB="$$(AR_$(1)) s" \
make sure jemalloc valgrind support is enabled This requires pointing it at the valgrind headers we carry in-tree.
2014-05-12 01:31:23 +02:00
CPPFLAGS="-I $(S)src/rt/" \
mk: Don't build jemalloc with -g3 By default, jemalloc is building itself with -g3 if the local compiler supports it. It looks like this is generating a good deal of debug info that windows isn't optimizing out (on the order of 18MB). Windows gcc/ld is also not optimizing this data away, causing hello world to be 18MB in size. There's no current real need for debugging jemalloc to a great extent, so this commit manually passes -g1 to override -g3 which jemalloc is using. This is confirmed to drop the size of executables on windows back to a more reasonable size (2.0MB, as they were before). Closes #14144
2014-05-16 00:43:40 +02:00
EXTRA_CFLAGS="$$(CFG_CFLAGS_$(1)) -g1"
add back jemalloc to the tree This adds a `std::rt::heap` module with a nice allocator API. It's a step towards fixing #13094 and is a starting point for working on a generic allocator trait. The revision used for the jemalloc submodule is the stable 3.6.0 release. Closes #11807
2014-04-25 08:19:34 +02:00
$$(Q)$$(MAKE) -C "$$(JEMALLOC_BUILD_DIR_$(1))" build_lib_static
$$(Q)cp $$(JEMALLOC_BUILD_DIR_$(1))/lib/$$(JEMALLOC_REAL_NAME_$(1)) $$(JEMALLOC_LIB_$(1))
Build compiler-rt and link it to all crates, similarly to morestack.
2014-02-04 07:54:09 +01:00
################################################################################
# compiler-rt
################################################################################
ifdef CFG_ENABLE_FAST_MAKE
COMPRT_DEPS := $(S)/.gitmodules
else
COMPRT_DEPS := $(wildcard \
$(S)src/compiler-rt/* \
$(S)src/compiler-rt/*/* \
$(S)src/compiler-rt/*/*/* \
$(S)src/compiler-rt/*/*/*/*)
endif
COMPRT_NAME_$(1) := $$(call CFG_STATIC_LIB_NAME_$(1),compiler-rt)
COMPRT_LIB_$(1) := $$(RT_OUTPUT_DIR_$(1))/$$(COMPRT_NAME_$(1))
COMPRT_BUILD_DIR_$(1) := $$(RT_OUTPUT_DIR_$(1))/compiler-rt
Add MKFILE_DEPS to compiler-rt target Currently when you run `make -jN` it's likely that you'll remove compiler-rt and then it won't get cp'd back into the right place. I believe the reason for this is that the compiler-rt library target never got updated so make decided it never needed to copy the files back into place. The files were all there at the beginning of `make`, but then we may clean out the stage0 versions if we unzip the snapshot again.
2014-02-14 17:13:19 +01:00
$$(COMPRT_LIB_$(1)): $$(COMPRT_DEPS) $$(MKFILE_DEPS)
Build compiler-rt and link it to all crates, similarly to morestack.
2014-02-04 07:54:09 +01:00
@$$(call E, make: compiler-rt)
$$(Q)$$(MAKE) -C "$(S)src/compiler-rt" \
ProjSrcRoot="$(S)src/compiler-rt" \
ProjObjRoot="$$(abspath $$(COMPRT_BUILD_DIR_$(1)))" \
CC="$$(CC_$(1))" \
AR="$$(AR_$(1))" \
RANLIB="$$(AR_$(1)) s" \
CFLAGS="$$(CFG_GCCISH_CFLAGS_$(1))" \
TargetTriple=$(1) \
Upgrade compiler-rt
2014-04-22 03:43:19 +02:00
triple-builtins
$$(Q)cp $$(COMPRT_BUILD_DIR_$(1))/triple/builtins/libcompiler_rt.a $$(COMPRT_LIB_$(1))
Build compiler-rt and link it to all crates, similarly to morestack.
2014-02-04 07:54:09 +01:00
Add basic backtrace functionality Whenever a failure happens, if a program is run with `RUST_LOG=std::rt::backtrace` a backtrace will be printed to the task's stderr handle. Stack traces are uncondtionally printed on double-failure and rtabort!(). This ended up having a nontrivial implementation, and here's some highlights of it: * We're bundling libbacktrace for everything but OSX and Windows * We use libgcc_s and its libunwind apis to get a backtrace of instruction pointers * On OSX we use dladdr() to go from an instruction pointer to a symbol * On unix that isn't OSX, we use libbacktrace to get symbols * Windows, as usual, has an entirely separate implementation Lots more fun details and comments can be found in the source itself. Closes #10128
2014-02-06 00:19:40 +01:00
################################################################################
# libbacktrace
#
# We use libbacktrace on linux to get symbols in backtraces, but only on linux.
# Elsewhere we use other system utilities, so this library is only built on
# linux.
################################################################################
BACKTRACE_NAME_$(1) := $$(call CFG_STATIC_LIB_NAME_$(1),backtrace)
BACKTRACE_LIB_$(1) := $$(RT_OUTPUT_DIR_$(1))/$$(BACKTRACE_NAME_$(1))
BACKTRACE_BUILD_DIR_$(1) := $$(RT_OUTPUT_DIR_$(1))/libbacktrace
ifeq ($$(findstring darwin,$$(OSTYPE_$(1))),darwin)
# We don't use this on platforms that aren't linux-based, so just make the file
# available, the compilation of libstd won't actually build it.
$$(BACKTRACE_LIB_$(1)):
touch $$@
else
ifeq ($$(CFG_WINDOWSY_$(1)),1)
$$(BACKTRACE_LIB_$(1)):
touch $$@
else
ifdef CFG_ENABLE_FAST_MAKE
BACKTRACE_DEPS := $(S)/.gitmodules
else
BACKTRACE_DEPS := $(wildcard $(S)src/libbacktrace/*)
endif
# We need to export CFLAGS because otherwise it doesn't pick up cross compile
# builds. If libbacktrace doesn't realize this, it will attempt to read 64-bit
# elf headers when compiled for a 32-bit system, yielding blank backtraces.
#
# This also removes the -Werror flag specifically to prevent errors during
# configuration.
#
# Down below you'll also see echos into the config.h generated by the
# ./configure script. This is done to force libbacktrace to *not* use the
# atomic/sync functionality because it pulls in unnecessary dependencies and we
# never use it anyway.
$$(BACKTRACE_BUILD_DIR_$(1))/Makefile: \
export CFLAGS:=$$(CFG_GCCISH_CFLAGS_$(1):-Werror=) \
-fno-stack-protector
$$(BACKTRACE_BUILD_DIR_$(1))/Makefile: export CC:=$$(CC_$(1))
$$(BACKTRACE_BUILD_DIR_$(1))/Makefile: export AR:=$$(AR_$(1))
$$(BACKTRACE_BUILD_DIR_$(1))/Makefile: export RANLIB:=$$(AR_$(1)) s
$$(BACKTRACE_BUILD_DIR_$(1))/Makefile: $$(BACKTRACE_DEPS) $$(MKFILE_DEPS)
$$(Q)rm -rf $$(BACKTRACE_BUILD_DIR_$(1))
$$(Q)mkdir -p $$(BACKTRACE_BUILD_DIR_$(1))
$$(Q)(cd $$(BACKTRACE_BUILD_DIR_$(1)) && \
$(S)src/libbacktrace/configure --target=$(1) --host=$(CFG_BUILD))
$$(Q)echo '#undef HAVE_ATOMIC_FUNCTIONS' >> \
$$(BACKTRACE_BUILD_DIR_$(1))/config.h
$$(Q)echo '#undef HAVE_SYNC_FUNCTIONS' >> \
$$(BACKTRACE_BUILD_DIR_$(1))/config.h
$$(BACKTRACE_LIB_$(1)): $$(BACKTRACE_BUILD_DIR_$(1))/Makefile $$(MKFILE_DEPS)
@$$(call E, make: libbacktrace)
$$(Q)$$(MAKE) -C $$(BACKTRACE_BUILD_DIR_$(1)) \
INCDIR=$(S)src/libbacktrace
$$(Q)cp $$(BACKTRACE_BUILD_DIR_$(1))/.libs/libbacktrace.a $$@
endif # endif for windowsy
endif # endif for darwin
virtualize the runtime per target triple.
2011-11-02 00:50:47 +01:00
endef
Build libuv/jemalloc only once (not per stage) Reorganize the makefiles to stop building these once per stage because there's no need to do this at all. Closes #7002
2013-10-08 18:20:17 +02:00
# Instantiate template for all stages/targets
support for GNU configure syntax
2013-10-21 11:18:21 +02:00
$(foreach target,$(CFG_TARGET), \
Build libuv/jemalloc only once (not per stage) Reorganize the makefiles to stop building these once per stage because there's no need to do this at all. Closes #7002
2013-10-08 18:20:17 +02:00
$(eval $(call DEF_THIRD_PARTY_TARGETS,$(target))))
Reference in New Issue Copy Permalink