--- Log opened Fri Jul 07 00:00:03 2017 | ||
stekern | shorne: I don't quite understand the master slave logic, is cpu0 always the master (e.g. boot cpu is master the other slaves)? | 01:32 |
---|---|---|
stekern | if so, should the master sync really be in _cpu_up()? | 01:36 |
stekern | afaict, that's being called for every cpu that is being brought up | 01:36 |
shorne | stekern: well cpu up is called buy cpu[0] to kick cpu[n] | 04:52 |
shorne | (as I understand it) | 04:52 |
shorne | so, after boot_secondary() is called, the master (cpu[0]) calls sync_count_master(cpu). | 04:53 |
shorne | To wait for the secondary cpu (cpu[n]) to call syn_count_clast(cpu) with same value for `cpu`. Then they will sync clocks | 04:54 |
shorne | then master (cpu[0]) will return form sync_count_master(), and continue booting other cpus | 04:55 |
shorne | and secondary cpu/slave (cpu[n]) will finish its startup and go to idle task | 04:56 |
stekern | ok, that may be the case | 05:44 |
--- Log closed Sat Jul 08 00:00:04 2017 |
Generated by irclog2html.py 2.15.2 by Marius Gedminas - find it at mg.pov.lt!