| 2017-02-27 |
schulz | fixed typo
|
commit | commitdiff | tree |
| 2017-02-27 |
schulz | even more locking code...
|
commit | commitdiff | tree |
| 2017-02-27 |
schulz | further locking of exec lists added. More to come ...
|
commit | commitdiff | tree |
| 2017-02-27 |
schulz | lddemon uses execlock.resource now
|
commit | commitdiff | tree |
| 2017-02-27 |
schulz | execlock.resource fixed (proper table size when allocating...
|
commit | commitdiff | tree |
| 2017-02-27 |
schulz | Don't deliver extINT on other cores but BSP. This fixes...
|
commit | commitdiff | tree |
| 2017-02-26 |
schulz | Since a potentialy unsafe condition may occur (IRQ...
|
commit | commitdiff | tree |
| 2017-02-26 |
schulz | shut off annoying debug messages
|
commit | commitdiff | tree |
| 2017-02-26 |
schulz | first bunch of functions using new locking mechanism.
|
commit | commitdiff | tree |
| 2017-02-26 |
schulz | Exec locks provides means to protect exec's lists from...
|
commit | commitdiff | tree |
| 2017-02-26 |
schulz | when setting IDT gate for Heartbeat, actually enable...
|
commit | commitdiff | tree |
| 2017-02-26 |
schulz | support bounce buffer in ata.device in case of DMA...
|
commit | commitdiff | tree |
| 2017-02-26 |
schulz | fix typos
|
commit | commitdiff | tree |
| 2017-02-25 |
schulz | If se Enable() unconditionaly at the end of function...
|
commit | commitdiff | tree |
| 2017-02-24 |
schulz | beautify output on 64bit systems
|
commit | commitdiff | tree |
| 2017-02-24 |
schulz | Beautify output of C:TaskList in 64bit system
|
commit | commitdiff | tree |
| 2017-02-23 |
schulz | corrections for displaying large addresses on 64 bit...
|
commit | commitdiff | tree |
| 2017-02-23 |
schulz | increase width of address field in case of 64 bit systems
|
commit | commitdiff | tree |
| 2017-02-23 |
schulz | fixed rounding of 1/100 seconds calcul
|
commit | commitdiff | tree |
| 2017-02-23 |
schulz | add sorting by cpu time's microseconds, also display...
|
commit | commitdiff | tree |
| 2017-02-23 |
schulz | use %id types, since the data is stored in IPTR array...
|
commit | commitdiff | tree |
| 2017-02-23 |
schulz | support for IPI_CALL_HOOK added
|
commit | commitdiff | tree |
| 2017-02-23 |
schulz | IPI_CALL_HOOK code wip...
|
commit | commitdiff | tree |
| 2017-02-23 |
schulz | ignore %edx contents of rdmsr result.
|
commit | commitdiff | tree |
| 2017-02-23 |
schulz | Increase resolution of cpu meter in Sysmon, center...
|
commit | commitdiff | tree |
| 2017-02-23 |
schulz | Add CPU Frequency calculation for newer Core-i generations
|
commit | commitdiff | tree |
| 2017-02-22 |
schulz | forgotten freecpumask added
|
commit | commitdiff | tree |
| 2017-02-22 |
schulz | CPU usage is not a % number (0..100) anymore. Instead...
|
commit | commitdiff | tree |
| 2017-02-21 |
schulz | Since CPU mask is mask of real APIC IDs, do not convert...
|
commit | commitdiff | tree |
| 2017-02-21 |
schulz | enable MMU (share the same pages as bootstrap cpu)...
|
commit | commitdiff | tree |
| 2017-02-21 |
schulz | if AROSEXEC_SMP is defined, protect debug output by...
|
commit | commitdiff | tree |
| 2017-02-21 |
schulz | task register is set not according to APICID but rather...
|
commit | commitdiff | tree |
| 2017-02-21 |
schulz | moved allocation of IPI vectors to apic setup (so that...
|
commit | commitdiff | tree |
| 2017-02-21 |
schulz | when sending IPI, use APIC IDs and not ordinal cpu...
|
commit | commitdiff | tree |
| 2017-02-21 |
schulz | enable spuriouis interrupt in IDT table and report...
|
commit | commitdiff | tree |
| 2017-02-21 |
schulz | corrected wait loop in smp_Wake
|
commit | commitdiff | tree |
| 2017-02-21 |
schulz | silent debug output of CPU usage
|
commit | commitdiff | tree |
| 2017-02-21 |
schulz | GetCpuInfo calls kernel resource (KrnGetSystemAttr...
|
commit | commitdiff | tree |
| 2017-02-21 |
schulz | add new system attributes - cpu and system load
|
commit | commitdiff | tree |
| 2017-02-21 |
schulz | match prototype of rdmsr used here with asm/cpu one
|
commit | commitdiff | tree |
| 2017-02-21 |
schulz | calculate CPU usage
|
commit | commitdiff | tree |
| 2017-02-21 |
schulz | use timestamp counter to calculate CPU usage per core
|
commit | commitdiff | tree |
| 2017-02-21 |
schulz | Calibrate TSC too, while calibrating LocalAPIC timer
|
commit | commitdiff | tree |
| 2017-02-21 |
schulz | fixed rdtsc for x86_64
|
commit | commitdiff | tree |
| 2017-02-21 |
schulz | not yet working CPU load calculation code deactivated
|
commit | commitdiff | tree |
| 2017-02-21 |
schulz | Do not set LAPIC timer in periodic mode, use OneShoot...
|
commit | commitdiff | tree |
| 2017-02-21 |
schulz | Add new fields for CPU calculation into CPUData
|
commit | commitdiff | tree |
| 2017-02-21 |
schulz | fixed the "bit test and..." instructions
|
commit | commitdiff | tree |
| 2017-02-21 |
schulz | use ChangePMState system call for sleeping. The Syscall...
|
commit | commitdiff | tree |
| 2017-02-21 |
schulz | Move spinlock protection of debug output into krnBug...
|
commit | commitdiff | tree |
| 2017-02-20 |
schulz | handle rare cases, where cpu time is negative (omitted...
|
commit | commitdiff | tree |
| 2017-02-20 |
schulz | do not set timer CCR register, it's read only :)
|
commit | commitdiff | tree |
| 2017-02-20 |
schulz | update task time calculations... something is still...
|
commit | commitdiff | tree |
| 2017-02-20 |
schulz | configure frequency of LAPIC timer correctly
|
commit | commitdiff | tree |
| 2017-02-20 |
schulz | Every heartbeat update the LAPIC Tick, used to calcualte...
|
commit | commitdiff | tree |
| 2017-02-19 |
schulz | Combine Forbid() with spinlock. The former protects...
|
commit | commitdiff | tree |
| 2017-02-19 |
schulz | task's priority is signed byte...
|
commit | commitdiff | tree |
| 2017-02-19 |
schulz | currected cpu time calculation
|
commit | commitdiff | tree |
| 2017-02-19 |
schulz | When sending Signal() to task on another CPU, send...
|
commit | commitdiff | tree |
| 2017-02-19 |
schulz | fix in setting IDT gates
|
commit | commitdiff | tree |
| 2017-02-19 |
schulz | less debug from IPI, enable reschedule IPI only if...
|
commit | commitdiff | tree |
| 2017-02-19 |
schulz | Handle IPIs first...
|
commit | commitdiff | tree |
| 2017-02-19 |
schulz | fix copyright character
|
commit | commitdiff | tree |
| 2017-02-19 |
schulz | enable IPI vectors on secondary CPUs
|
commit | commitdiff | tree |
| 2017-02-19 |
schulz | protect displayalert by semaphore on smp systems
|
commit | commitdiff | tree |
| 2017-02-18 |
schulz | all-pc version of KrnScheduleCPU
|
commit | commitdiff | tree |
| 2017-02-18 |
schulz | KrnScheduleCPU function to trigger scheduler on CPU...
|
commit | commitdiff | tree |
| 2017-02-18 |
schulz | if IPI irq is send, pass it to IPI handler. Also added...
|
commit | commitdiff | tree |
| 2017-02-18 |
schulz | add IPI IRQ boundaries
|
commit | commitdiff | tree |
| 2017-02-18 |
schulz | store iet_CpuTime in timespec to get nanosecond resolution...
|
commit | commitdiff | tree |
| 2017-02-18 |
schulz | use HLT instruction to sleep in idle task.
|
commit | commitdiff | tree |
| 2017-02-18 |
schulz | added spinlock for debug output. It shouldn't be here...
|
commit | commitdiff | tree |
| 2017-02-17 |
schulz | correct copyright symbol
|
commit | commitdiff | tree |
| 2017-02-17 |
schulz | properly calculate CPU time spent in each task
|
commit | commitdiff | tree |
| 2017-02-17 |
schulz | after calculating cpu frequency reset local apic to...
|
commit | commitdiff | tree |
| 2017-02-17 |
schulz | use correct order of hours, minutes and seconds in...
|
commit | commitdiff | tree |
| 2017-02-17 |
schulz | make sure owner set before comparing priorities...
|
commit | commitdiff | tree |
| 2017-02-17 |
schulz | aww, temporary variable to store current state of lock...
|
commit | commitdiff | tree |
| 2017-02-17 |
schulz | If using some fixed signal instead of the one allocated...
|
commit | commitdiff | tree |
| 2017-02-17 |
schulz | Don't reallocate new signal if we used CreateMsgPort...
|
commit | commitdiff | tree |
| 2017-02-17 |
schulz | free signal of msgport (we're using SIGB_INTUITION...
|
commit | commitdiff | tree |
| 2017-02-17 |
schulz | use spinlock's owner. Reduce priority of spinning task...
|
commit | commitdiff | tree |
| 2017-02-17 |
schulz | Initialize spinlocks owner
|
commit | commitdiff | tree |
| 2017-02-17 |
schulz | add owner (struct Task *) to spinlock
|
commit | commitdiff | tree |
| 2017-02-17 |
schulz | clear MsgPort structure before using, in case it has...
|
commit | commitdiff | tree |
| 2017-02-17 |
schulz | If requested, do not provide prototypes of spinlock...
|
commit | commitdiff | tree |
| 2017-02-14 |
schulz | Cleaned up spinlock stuff. Moved cmpxchg and bt.. constructs...
|
commit | commitdiff | tree |
| 2017-02-14 |
schulz | temporary gag, use spinlocks to protec memory allocations...
|
commit | commitdiff | tree |
| 2017-02-14 |
schulz | disable sleeping in idle task until we know the rest...
|
commit | commitdiff | tree |
| 2017-02-14 |
schulz | thread safe locking of debug output
|
commit | commitdiff | tree |
| 2017-02-14 |
schulz | safedebug is volatile (can be changed by any cpu core)
|
commit | commitdiff | tree |
| 2017-02-14 |
schulz | pit_wait should work more stable on fast machines now...
|
commit | commitdiff | tree |
| 2017-02-14 |
schulz | KrnSpinTryLock and KrnSpinUnlock reimplemented with...
|
commit | commitdiff | tree |
| 2017-02-14 |
schulz | lock is 32bits long. Using type "long" would promote...
|
commit | commitdiff | tree |
| 2017-02-13 |
schulz | atomic shared lock implemented
|
commit | commitdiff | tree |
| 2017-02-13 |
schulz | added new field in spinlock - lock for updating readcount
|
commit | commitdiff | tree |
| 2017-02-13 |
schulz | fixed stupid typos...
|
commit | commitdiff | tree |
| 2017-02-13 |
schulz | write mode spinlock implemented using cmpxchg instruction
|
commit | commitdiff | tree |
| 2017-02-13 |
schulz | added missing include
|
commit | commitdiff | tree |
| 2017-02-13 |
schulz | added functions for primitive atomic spinlock (just...
|
commit | commitdiff | tree |
