Quoting Brendan Higgins (2019-08-01 11:59:57)

On Thu, Aug 1, 2019 at 11:55 AM Brendan Higgins
[off-list ref] wrote:

quoted

On Fri, Jul 26, 2019 at 1:31 AM Petr Mladek [off-list ref] wrote:

quoted

To be honest I do not fully understand KUnit design. I am not
completely sure how the tested code is isolated from the running
system. Namely, I do not know if the tested code shares
the same locks with the system running the test.

No worries, I don't expect printk to be the hang up in those cases. It
sounds like KUnit has a long way to evolve before printk is going to
be a limitation.

So Stephen, what do you think?

Do you want me to go forward with the new kunit_assert API wrapping
the string_stream as I have it now? Would you prefer to punt this to a
later patch? Or would you prefer something else?

I like the struct based approach. If anything, it can be adjusted to
make the code throw some records into a spinlock later on and delay the
formatting of the assertion if need be. Can you resend with that
approach? I don't think I'll have any more comments after that.