Hi Georges, On 2023/02/15 21:36, Aureau, Georges (Kernel Tools ERT) wrote: > Hello, > > For upstream crash, and this is more a general thought, what about ordering tests with most recent kernel changes being tested first: > > For example, in Lianbo's case, instead of: > - (STREQ(sp->name, "call_softirq") || STREQ(sp->name, "do_softirq_own_stack"))) > What about reversing the order of the tests: > + (STREQ(sp->name, " do_softirq_own_stack") || STREQ(sp->name, " call_softirq"))) > > As distros tend to move fast forward, ordering tests with most recent kernel changes being tested first would make sense. thank you for your comment, I really agree. Seeing the crash code, there are already many cases where a new test has been appended to the last else-if block, so I also tend to do the same thing for code consistency and etc... but I don't think it's very good especially in efficiency for recent kernels. On the other hand, there is often a case where adding a new test first will make the conditions complicated, e.g. when multiple symbols are related and different in the version of their introduction. But, generally I would like to place a new test first unless there are big cons of doing that. Thanks, Kazu -- Crash-utility mailing list Crash-utility@xxxxxxxxxx https://listman.redhat.com/mailman/listinfo/crash-utility Contribution Guidelines: https://github.com/crash-utility/crash/wiki
