Substructure of signal_struct and which varies for uniprocessor versus This is included in the structure "thread_group_cputime," which is a new Scheduler times and use these in appropriate places: We define a new structure, "task_cputime," to contain user, system and Running processors.) To do this, at each tick we now update fields in Or multiprocessor and, in the latter case, depending upon the number of One machine and another depending upon whether the kernel is built as single. Run in constant time for a particular machine. This rework corrects the implementation of run_posix_cpu_timers() to make it Very much to Roland McGrath, without whom this patch wouldn't exist in this This patch fixes bug 9906, "Weird hang with NPTL and SIGPROF." And thanks Which point things degrade rather quickly. That the performance of the old implementation of run_posix_cpu_timers() wasĪt least O(n*3) (where "n" is the number of threads in a process) or worse.Įverything is fine with an increasing number of threads until the time takenįor that routine to run becomes the same as or greater than the tick time, at The problem we ran into, and the reason for this rework, has to do with usingĪ profiling timer in a process with a large number of threads. With the help of Roland McGrath, the owner and original writer of this code. ITIMER_PROF, ITIMER_VIRT timers and rlimit handling. This patch reworks the handling of POSIX CPU timers, including the 3:59 ` Roland McGrath 0 siblings, 2 replies 11+ messages in thread LKML Archive on help / color / mirror / Atom feed * Fix itimer/many thread hang.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |