On Thu, Jul 27, 2023, Sean Christopherson wrote: > On Thu, Jul 27, 2023, JinrongLiang wrote: > > > > 在 2023/7/27 06:41, Sean Christopherson 写道: > > > Side topic, if anyone lurking out there wants an easy (but tedious and boring) > > > starter project, we should convert all tests to the newfangled formatting and > > > drop GUEST_ASSERT_N entirely. Once all tests are converted, GUEST_ASSERT_FMT() > > > and REPORT_GUEST_ASSERT_FMT can drop the "FMT" postfix. > > > > I'd be happy to get the job done. > > > > However, before I proceed, could you please provide a more detailed example > > or further guidance on the desired formatting and the specific changes you > > would like to see? > > Hrm, scratch that request. I was thinking we could convert tests one-by-one, but > that won't work well because to do a one-by-one conversions, tests that use > GUEST_ASSERT_EQ() would need to first convert to e.g. GUEST_ASSERT_EQ_FMT() and > then convert back, which would be a silly amount of churn just to a void a single > selftests-wide patch. > > It probably makes sense to just convert everything as part of this series. There > are quite a few asserts that need a message, but not *that* many. Aha! And there's already a "tree"-wide patch in this area to rename ASSERT_EQ() to TEST_ASSERT_EQ()[*]. I'll include that in v4 as well, and then piggyback on it to implement the new and improved GUEST_ASSERT_EQ(). [*] https://lore.kernel.org/all/20230712075910.22480-2-thuth@xxxxxxxxxx
