On Fri, 6 Aug 2021 at 15:43, Alexander Bulekov <alx...@bu.edu> wrote: > As a side note(not strictly related to this fix) should we continue > joining reproducer patches with the fixes? In order to test the > reproducer, you need to cleave the fix off the patch. At the same time > we don't want to mess up bisection, so does it make sense to have the > reproducer patch be separate but come last in the series?
My preference is for the test case as a separate patch, last in the series. For this kind of minor easy-to-review fix it matters less, but sometimes the right fix for a problem might be larger or more complicated, and then having the test case in the same patch makes that patch awkwardly large. Also the person able to review the code change and the person able to review the test case might not be the same... thanks -- PMM
