Messages in this thread | | | Date | Mon, 29 Apr 2013 18:42:25 -0400 | From | KOSAKI Motohiro <> | Subject | Re: [PATCH v2 1/3] process cputimer is moving faster than its corresponding clock |
| |
>> tick lost doesn't occur an issue. because glibc only test posix conformance and >> posix allow inacculacy. In other words, timer must not run faster than real clock. >> but lost and makes slower are accepted in the test. >> > What is lost isn't cputimer tick. They are accounted > account_group_exec_runtime(). What is lost it is what is added to > curr->sum_exec_runtime. Thus making the thread group clock running > slower than the cputimer. > > Please spend some time reading the code and less time writing e-mails. > > Read the code of release_task() and where it is called in do_exit(). > > Once it is done, it should be clear to you.
After several grepping, I may got your point. Guys, sig->sum_sched_runtime is gathered in __exit_signal, not release_task() nor do_exit().
So, I'm now convinced your patch is correct. Even though the comments is terribuly misleading.
| |