Hi Tom,
On Fri, 1 Nov 2024 at 22:46, Tom Rini trini@konsulko.com wrote:
On Mon, Oct 28, 2024 at 01:41:06PM +0100, Simon Glass wrote:
This test doesn't belong at the top level. Move it into the common/ directory, to match its implementation.
This also fixes the build rule for the bootm test, which was depending on bloblist by mistake.
Signed-off-by: Simon Glass sjg@chromium.org
test/Makefile | 3 --- test/common/Makefile | 3 +++ test/{ => common}/bloblist.c | 0 3 files changed, 3 insertions(+), 3 deletions(-) rename test/{ => common}/bloblist.c (100%)
On rpi_3_32b (and presumably others): test/py/tests/test_ut.py ss....F
=================================== FAILURES =================================== _______________________ test_ut[ut_bootm_bootm_test_nop] _______________________ test/py/tests/test_ut.py:592: in test_ut assert output.endswith('Failures: 0') E AssertionError: assert False E + where False = <built-in method endswith of str object at 0x2bfcb00>('Failures: 0') E + where <built-in method endswith of str object at 0x2bfcb00> = 'ut - unit tests\r\n\r\nUsage:\r\nut [-r] [-f] [<suite>] - run unit tests\r\n -r<runs> Number of times to run each...elated commands\r\nprint - printing things to the console\r\nsetexpr - setexpr command\r\nunicode - Unicode functions'.endswith ----------------------------- Captured stdout call ----------------------------- U-Boot> ut bootm bootm_test_nop ut - unit tests
Usage: ut [-r] [-f] [<suite>] - run unit tests -r<runs> Number of times to run each test -f Force 'manual' tests to run as well <suite> Test suite to run, or all
Options for <suite>: all - execute all enabled tests info - show info about tests bdinfo - bdinfo command bootstd - standard boot implementation cmd - test various commands env - environment fdt - fdt command hush - Test hush behavior lib - library functions log - logging functions mem - memory-related commands print - printing things to the console setexpr - setexpr command unicode - Unicode functions U-Boot> =========================== short test summary info ============================
Oh, that explains why the test depended on bloblist. I thought it was a mistake but it was just a way to make it only run on sandbox. I will resend the series, leaving the bloblist suite in place.
I'm looking forward to getting my lab running...
Regards, Simon