Fix for test_pty_no_color when cache is cleared. NFC
#17412
Conversation
sbc100
force-pushed
the
test_pty_no_color
branch
2 times, most recently
from
5a999b0 to
d4b3f8d
Compare
tests/test_other.py
Outdated
| }) | ||
| @no_windows('ptys and select are not available on windows') | ||
| def test_pty_no_color(self, flag): | ||
| # Ensure that the sysroot is already constructed so that emcc doesn't | ||
| # output any log information in color. | ||
| self.emcc(test_file('hello_world.c'), ['-c']) |
There was a problem hiding this comment.
why is this needed? shouldn't the flag affect colors in the sysroot construction logging?
There was a problem hiding this comment.
No the -fno-diagnostics-color currently only effects clang... I guess we could change it to also effect python's logging module, but it doesn't do that today.
I guess I could file a bug? But its seems like maybe its not worth it?
There was a problem hiding this comment.
(BTW we don't use python's logging mechaism for emcc error or warnings... it really only used for debugging and for niche cases like this where we report building of internal libraries).
So normal errors and warnings that emcc generates are not effected by this.
There was a problem hiding this comment.
I see, makes sense. lgtm with an explanation of that in the test code.
sbc100
force-pushed
the
test_pty_no_color
branch
3 times, most recently
from
e0339f3 to
9a8abad
Compare
|
Looks like the fix for this depends on #17434 |
Withouth this change the color message that emscripten generated when it sets up the sysroot will cause the test to fail.
sbc100
force-pushed
the
test_pty_no_color
branch
from
9a8abad to
c8f78e5
Compare
Withouth this change the color message that emscripten generated when
it sets up the sysroot will cause the test to fail.