rust/mk/tests.mk

333 lines
11 KiB
Makefile
Raw Normal View History

2011-05-01 22:18:52 +02:00
######################################################################
# Testing variables
######################################################################
ALL_TEST_INPUTS = $(wildcard $(S)src/test/*/*.rs \
$(S)src/test/*/*/*.rs \
$(S)src/test/*/*.rc)
2011-06-25 21:23:27 +02:00
RPASS_RC := $(wildcard $(S)src/test/run-pass/*.rc)
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
RPASS_RS := $(wildcard $(S)src/test/run-pass/*.rs)
2011-06-25 21:23:27 +02:00
RFAIL_RC := $(wildcard $(S)src/test/run-fail/*.rc)
RFAIL_RS := $(wildcard $(S)src/test/run-fail/*.rs)
CFAIL_RC := $(wildcard $(S)src/test/compile-fail/*.rc)
CFAIL_RS := $(wildcard $(S)src/test/compile-fail/*.rs)
BENCH_RS := $(wildcard $(S)src/test/bench/*.rs)
PRETTY_RS := $(wildcard $(S)src/test/pretty/*.rs)
2011-05-01 22:18:52 +02:00
# perf tests are the same as bench tests only they run under
# a performance monitor.
PERF_RS := $(wildcard $(S)src/test/bench/*.rs)
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
RPASS_TESTS := $(RPASS_RC) $(RPASS_RS)
RFAIL_TESTS := $(RFAIL_RC) $(RFAIL_RS)
CFAIL_TESTS := $(CFAIL_RC) $(CFAIL_RS)
BENCH_TESTS := $(BENCH_RS)
PERF_TESTS := $(PERF_RS)
PRETTY_TESTS := $(PRETTY_RS)
2011-05-01 22:18:52 +02:00
FT := run_pass_stage2
FT_LIB := $(call CFG_LIB_NAME,$(FT))
FT_DRIVER := $(FT)_driver
GENERATED += test/$(FT).rc test/$(FT_DRIVER).rs
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
# The arguments to all test runners
ifdef TESTNAME
TESTARGS += $(TESTNAME)
endif
ifdef CHECK_XFAILS
TESTARGS += --ignored
endif
# Arguments to the cfail/rfail/rpass/bench tests
ifdef CFG_VALGRIND
CTEST_RUNTOOL = --runtool "$(CFG_VALGRIND)"
endif
# Arguments to the perf tests
ifdef CFG_PERF_TOOL
CTEST_PERF_RUNTOOL = --runtool "$(CFG_PERF_TOOL)"
endif
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
CTEST_TESTARGS := $(TESTARGS)
2011-05-01 22:18:52 +02:00
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
ifdef VERBOSE
CTEST_TESTARGS += --verbose
endif
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
# The test runner that runs the cfail/rfail/rpass and bench tests
COMPILETEST_CRATE := $(S)src/test/compiletest/compiletest.rc
COMPILETEST_INPUTS := $(wildcard $(S)src/test/compiletest/*rs)
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
# The standard library test crate
STDTEST_CRATE := $(S)src/test/stdtest/stdtest.rc
STDTEST_INPUTS := $(wildcard $(S)src/test/stdtest/*rs)
# Run the compiletest runner itself under valgrind
ifdef CTEST_VALGRIND
CFG_RUN_CTEST=$(call CFG_RUN_TEST,$(2))
else
CFG_RUN_CTEST=$(call CFG_RUN,stage$(1)/lib,$(2))
endif
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
######################################################################
# Main test targets
######################################################################
check: tidy check-stage2 \
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
check-full: tidy check-stage1 check-stage2 check-stage3 \
2011-07-13 22:13:44 +02:00
check-fast: tidy \
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
check-stage2-rustc check-stage2-std \
test/$(FT_DRIVER).out \
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
tidy:
@$(call E, check: formatting)
$(Q)echo \
$(filter-out $(GENERATED) $(addprefix $(S)src/, $(GENERATED)) \
$(addprefix $(S)src/, $(RUSTLLVM_LIB_CS) $(RUSTLLVM_OBJS_CS) \
$(RUSTLLVM_HDR) $(PKG_3RDPARTY)) \
$(S)src/etc/%, $(PKG_FILES)) \
| xargs -n 10 python $(S)src/etc/tidy.py
2011-05-01 22:18:52 +02:00
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
# Cancel the implicit .out rule in GNU make
%.out: %
2011-05-01 22:18:52 +02:00
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
%.out: %.out.tmp
$(Q)mv $< $@
2011-05-01 22:18:52 +02:00
######################################################################
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
# Rules for the test runners
######################################################################
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
# StageN template: to stay consistent with stageN.mk, arge 2 is the
# stage being tested, arg 1 is stage N-1
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
define TEST_STAGEN
# FIXME: These rules are complicated by a scheme to produce .out files
# for each test, with the idea that if the targets produce actual
# output, subsequent "make check's" will just exit successfully
# without re-running the tests. Unfortunately this scheme is currently
# broken and the tests rerun with every invocation of "make check"
# anyway. Nobody seems to have noticed, so it would be simpler to just
# remove all the code here involved with producing .out files.
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
check-stage$(2): tidy \
check-stage$(2)-rustc \
check-stage$(2)-std \
check-stage$(2)-rpass \
check-stage$(2)-rfail \
check-stage$(2)-cfail \
check-stage$(2)-bench \
check-stage$(2)-pretty \
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
# Rules for the standard library test runner
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
check-stage$(2)-std: test/stdtest.stage$(2).out \
2011-05-01 22:18:52 +02:00
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
test/stdtest.stage$(2)$$(X): $$(STDTEST_CRATE) $$(STDTEST_INPUTS) \
$$(SREQ$(2))
@$$(call E, compile_and_link: $$@)
$$(STAGE$(2)) -o $$@ $$< --test
2011-06-25 21:23:27 +02:00
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
test/stdtest.stage$(2).out.tmp: test/stdtest.stage$(2)$$(X)
@$$(call E, run: $$<)
$$(Q)$$(call CFG_RUN_TEST,$$<) $$(TESTARGS)
$$(Q)touch $$@
2011-06-16 21:06:09 +02:00
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
# Rules for the rustc test runner
2011-06-16 21:06:09 +02:00
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
check-stage$(2)-rustc: test/rustctest.stage$(2).out \
2011-05-01 22:18:52 +02:00
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
test/rustctest.stage$(2)$$(X): $$(COMPILER_CRATE) $$(COMPILER_INPUTS) \
stage$(2)/$$(CFG_RUNTIME) \
$$(call CFG_STDLIB_DEFAULT,stage$(1),stage$(2)) \
stage$(2)/$$(CFG_RUSTLLVM) \
$$(SREQ$(1))
@$$(call E, compile_and_link: $$@)
$$(STAGE$(1)) -o $$@ $$< --test
2011-05-01 22:18:52 +02:00
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
test/rustctest.stage$(2).out.tmp: test/rustctest.stage$(2)$$(X)
@$$(call E, run: $$<)
$$(Q)$$(call CFG_RUN,stage$(2),$$(CFG_VALGRIND) $$<) \
$$(TESTARGS)
$$(Q)touch $$@
2011-05-01 22:18:52 +02:00
# Rules for the cfail/rfail/rpass/bench/perf test runner
2011-05-01 22:18:52 +02:00
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
check-stage$(2)-cfail: test/compile-fail.stage$(2).out \
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
check-stage$(2)-rfail: test/run-fail.stage$(2).out \
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
check-stage$(2)-rpass: test/run-pass.stage$(2).out \
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
check-stage$(2)-bench: test/bench.stage$(2).out \
check-stage$(2)-perf: test/perf.stage$(2).out \
check-stage$(2)-pretty-rpass: test/pretty-rpass.stage$(2).out \
check-stage$(2)-pretty-rfail: test/pretty-rfail.stage$(2).out \
check-stage$(2)-pretty-bench: test/pretty-bench.stage$(2).out \
check-stage$(2)-pretty-pretty: test/pretty-pretty.stage$(2).out \
check-stage$(2)-pretty: check-stage$(2)-pretty-rpass \
check-stage$(2)-pretty-rfail \
check-stage$(2)-pretty-bench \
check-stage$(2)-pretty-pretty \
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
CTEST_COMMON_ARGS$(2) := --compile-lib-path stage$(2) \
--run-lib-path stage$(2)/lib \
--rustc-path stage$(2)/rustc$$(X) \
--stage-id stage$(2) \
--rustcflags "$$(CFG_RUSTC_FLAGS)" \
$$(CTEST_TESTARGS) \
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
CFAIL_ARGS$(2) := $$(CTEST_COMMON_ARGS$(2)) \
--src-base $$(S)src/test/compile-fail/ \
--build-base test/compile-fail/ \
--mode compile-fail \
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
RFAIL_ARGS$(2) := $$(CTEST_COMMON_ARGS$(2)) \
--src-base $$(S)src/test/run-fail/ \
--build-base test/run-fail/ \
--mode run-fail \
$$(CTEST_RUNTOOL) \
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
RPASS_ARGS$(2) := $$(CTEST_COMMON_ARGS$(2)) \
--src-base $(S)src/test/run-pass/ \
--build-base test/run-pass/ \
--mode run-pass \
$$(CTEST_RUNTOOL) \
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
BENCH_ARGS$(2) := $$(CTEST_COMMON_ARGS$(2)) \
--src-base $(S)src/test/bench/ \
--build-base test/bench/ \
--mode run-pass \
$$(CTEST_RUNTOOL) \
PERF_ARGS$(2) := $$(CTEST_COMMON_ARGS$(2)) \
--src-base $(S)src/test/bench/ \
--build-base test/perf/ \
--mode run-pass \
$$(CTEST_PERF_RUNTOOL) \
PRETTY_RPASS_ARGS$(2) := $$(CTEST_COMMON_ARGS$(2)) \
--src-base $$(S)src/test/run-pass/ \
--build-base test/run-pass/ \
--mode pretty \
PRETTY_RFAIL_ARGS$(2) := $$(CTEST_COMMON_ARGS$(2)) \
--src-base $$(S)src/test/run-fail/ \
--build-base test/run-fail/ \
--mode pretty \
PRETTY_BENCH_ARGS$(2) := $$(CTEST_COMMON_ARGS$(2)) \
--src-base $$(S)src/test/bench/ \
--build-base test/bench/ \
--mode pretty \
PRETTY_PRETTY_ARGS$(2) := $$(CTEST_COMMON_ARGS$(2)) \
--src-base $$(S)src/test/pretty/ \
--build-base test/pretty/ \
--mode pretty \
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
test/compiletest.stage$(2)$$(X): $$(COMPILETEST_CRATE) \
$$(COMPILETEST_INPUTS) \
$$(SREQ$(2))
@$$(call E, compile_and_link: $$@)
$$(STAGE$(2)) -o $$@ $$<
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
test/compile-fail.stage$(2).out.tmp: test/compiletest.stage$(2)$$(X) \
$$(CFAIL_TESTS)
@$$(call E, run: $$<)
$$(Q)$$(call CFG_RUN_CTEST,$(2),$$<) $$(CFAIL_ARGS$(2))
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
$$(Q)touch $$@
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
test/run-fail.stage$(2).out.tmp: test/compiletest.stage$(2)$$(X) \
$$(RFAIL_TESTS)
@$$(call E, run: $$<)
$$(Q)$$(call CFG_RUN_CTEST,$(2),$$<) $$(RFAIL_ARGS$(2))
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
$$(Q)touch $$@
test/run-pass.stage$(2).out.tmp: test/compiletest.stage$(2)$$(X) \
$$(RPASS_TESTS)
@$$(call E, run: $$<)
$$(Q)$$(call CFG_RUN_CTEST,$(2),$$<) $$(RPASS_ARGS$(2))
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
$$(Q)touch $$@
test/bench.stage$(2).out.tmp: test/compiletest.stage$(2)$$(X) \
$$(BENCH_TESTS)
@$$(call E, run: $$<)
$$(Q)$$(call CFG_RUN_CTEST,$(2),$$<) $$(BENCH_ARGS$(2))
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
$$(Q)touch $$@
test/perf.stage$(2).out.tmp: test/compiletest.stage$(2)$$(X) \
$$(BENCH_TESTS)
@$$(call E, perf: $$<)
$$(Q)$$(call CFG_RUN_CTEST,$(2),$$<) $$(PERF_ARGS$(2))
$$(Q)touch $$@
test/pretty-rpass.stage$(2).out.tmp: test/compiletest.stage$(2)$$(X) \
$$(RPASS_TESTS)
@$$(call E, run: $$<)
$$(Q)$$(call CFG_RUN_CTEST,$(2),$$<) $$(PRETTY_RPASS_ARGS$(2))
$$(Q)touch $$@
test/pretty-rfail.stage$(2).out.tmp: test/compiletest.stage$(2)$$(X) \
$$(RFAIL_TESTS)
@$$(call E, run: $$<)
$$(Q)$$(call CFG_RUN_CTEST,$(2),$$<) $$(PRETTY_RFAIL_ARGS$(2))
$$(Q)touch $$@
test/pretty-bench.stage$(2).out.tmp: test/compiletest.stage$(2)$$(X) \
$$(BENCH_TESTS)
@$$(call E, run: $$<)
$$(Q)$$(call CFG_RUN_CTEST,$(2),$$<) $$(PRETTY_BENCH_ARGS$(2))
$$(Q)touch $$@
test/pretty-pretty.stage$(2).out.tmp: test/compiletest.stage$(2)$$(X) \
$$(PRETTY_TESTS)
@$$(call E, run: $$<)
$$(Q)$$(call CFG_RUN_CTEST,$(2),$$<) $$(PRETTY_PRETTY_ARGS$(2))
$$(Q)touch $$@
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
endef
# Instantiate the template for stage 0, 1, 2, 3
$(eval $(call TEST_STAGEN,0,0))
The Big Test Suite Overhaul This replaces the make-based test runner with a set of Rust-based test runners. I believe that all existing functionality has been preserved. The primary objective is to dogfood the Rust test framework. A few main things happen here: 1) The run-pass/lib-* tests are all moved into src/test/stdtest. This is a standalone test crate intended for all standard library tests. It compiles to build/test/stdtest.stageN. 2) rustc now compiles into yet another build artifact, this one a test runner that runs any tests contained directly in the rustc crate. This allows much more fine-grained unit testing of the compiler. It compiles to build/test/rustctest.stageN. 3) There is a new custom test runner crate at src/test/compiletest that reproduces all the functionality for running the compile-fail, run-fail, run-pass and bench tests while integrating with Rust's test framework. It compiles to build/test/compiletest.stageN. 4) The build rules have been completely changed to use the new test runners, while also being less redundant, following the example of the recent stageN.mk rewrite. It adds two new features to the cfail/rfail/rpass/bench tests: 1) Tests can specify multiple 'error-pattern' directives which must be satisfied in order. 2) Tests can specify a 'compile-flags' directive which will make the test runner provide additional command line arguments to rustc. There are some downsides, the primary being that Rust has to be functioning pretty well just to run _any_ tests, which I imagine will be the source of some frustration when the entire test suite breaks. Will also cause some headaches during porting. Not having individual make rules, each rpass, etc test no longer remembers between runs whether it completed successfully. As a result, it's not possible to incrementally fix multiple tests by just running 'make check', fixing a test, and repeating without re-running all the tests contained in the test runner. Instead you can filter just the tests you want to run by using the TESTNAME environment variable. This also dispenses with the ability to run stage0 tests, but they tended to be broken more often than not anyway.
2011-07-13 04:01:09 +02:00
$(eval $(call TEST_STAGEN,0,1))
$(eval $(call TEST_STAGEN,1,2))
$(eval $(call TEST_STAGEN,2,3))
######################################################################
# Fast-test rules
######################################################################
test/$(FT).rc test/$(FT_DRIVER).rs: $(TEST_RPASS_SOURCES_STAGE2) \
$(S)src/etc/combine-tests.py
@$(call E, check: building combined stage2 test runner)
$(Q)$(S)src/etc/combine-tests.py
stage2/lib/$(FT_LIB): test/$(FT).rc $(SREQ2)
@$(call E, compile_and_link: $@)
$(STAGE2) --lib -o $@ $<
test/$(FT_DRIVER)$(X): test/$(FT_DRIVER).rs stage2/lib/$(FT_LIB) $(SREQ2)
@$(call E, compile_and_link: $@)
$(STAGE2) -o $@ $<
test/$(FT_DRIVER).out: test/$(FT_DRIVER)$(X) $(SREQ2)
$(Q)$(call CFG_RUN_TEST, $<)