Skip to content

Latest commit

 

History

History
58 lines (40 loc) · 2.5 KB

README.md

File metadata and controls

58 lines (40 loc) · 2.5 KB
Raw

Course OS Testing framework

This directory is a custom-build testing framework by Jonathan Dönszelmann and Victor Roest. To run tests, run

make test

If you are sitting here, thinking: the way this test framework works is one enormous hack: yes it is. Unfortunately that's how it is in operating systems. The testing framework depends on the kernel booting, because how else could we run any code? Testing complex interactions in the kernel is very hard as test may use resources of the kernel itself. My advice: deal with it.

Creating tests.

To create a new test, you can use the TEST_CREATE macro. It takes two arguments, a test name (which must be globally unique) and a block of code. An example test would be

#include <test.h>

TEST_CREATE(test_one_plus_one, {
    ASSERT_EQ(1 + 1, 2);
})

Tests created like this will be automatically detected and compiled into a file src/test/test.c which will then be compiled by gcc into the kernel. Tests can be created in any file in any subdirectory of the src file. Tests will be randomized in order, and you cannot depend on the order in which tests are executed. This has been done deliberately since we have had many problems with accidentally depending on this.

When the MEM_DEBUG definition is given in the makefile (default for tests), the testing framework will record the allocator's number of bytes allocated before the test. If this number is not equal after the test, the test will fail. When running with ENABLE_TESTS on, WARN macros will report file and line number, and DATA_ABORT handlers will panic and fail the test.

Assert macros

There are a number of assertion macros included in test.h.

They are listed here:

Macro name usage
PASS() Will immediately return and pass the test
FAIL() Will immediately return and fail the test
ASSERT(expr) Will fail the test if expr evaluates to zero/false
ASSERT_EQ(l, r) Will fail the test if l != r
ASSERT_GT(l, r) Will fail the test if !(l >= r)
ASSERT_GTEQ(l, r) Will fail the test if !(l > r)
ASSERT_GT(l, r) Will fail the test if !(l >= r)
ASSERT_LT(l, r) Will fail the test if !(l < r)
ASSERT_LTEQ(l, r) Will fail the test if !(l <= r)
ASSERT_NEQ(l, r) Will fail the test if l == r
ASSERT_NOT_NULL(expr) Will fail the test if expr evaluates to NULL
ASSERT_NULL(expr) Will fail the test if expr doesn't evaluates to NULL