aboutsummaryrefslogtreecommitdiff
path: root/arch/riscv64/kernel
diff options
context:
space:
mode:
authorKimplul <kimi.h.kuparinen@gmail.com>2024-07-17 22:02:47 +0300
committerKimplul <kimi.h.kuparinen@gmail.com>2024-07-17 22:02:47 +0300
commit1a0d61b33c2365cf413bf584289058d0d9097377 (patch)
tree13ed35e7cc0c4360a429eab137caddcf85844807 /arch/riscv64/kernel
parent83f8e82c421fa9013ac813b1560abf1c749822ca (diff)
downloadkmi-1a0d61b33c2365cf413bf584289058d0d9097377.tar.gz
kmi-1a0d61b33c2365cf413bf584289058d0d9097377.zip
make tests a bit more sensible
+ At least by my standards. Some amount of scripting going on, but should still be fine. Effectively, source.mk contains information to build the test, check.mk checks the generated log from running the test. As a results, check.mk should output OK into reports/$TEST/OK if everything went well, and anything else otherwise. The user can then look at reports/$TEST/log to see what went wrong. I expect to add some more features, such as starting QEMU with gdb and running a single test a bit easier than right now (same limitation as in the top level makefile, i.e. you have to specify make -f scripts/makefile $TEST from within `tests` and you must have ran `make check` at that point so that `tests.mk` is fully generated. Not huge issues, but slightly annoying, somewhat unsure how to work around them properly but we'll see.
Diffstat (limited to 'arch/riscv64/kernel')
0 files changed, 0 insertions, 0 deletions