| 2017-04-04 |
schulz | Don't check for maximal depth, rather control stack...
|
commit | commitdiff | tree |
| 2017-04-04 |
schulz | Don't try to optimize too much, or else the code will...
|
commit | commitdiff | tree |
| 2017-04-01 |
schulz | re-read the mmap pointer since it was changed previously...
|
commit | commitdiff | tree |
| 2017-04-01 |
schulz | local APIC may throw error if we mask unconfigured...
|
commit | commitdiff | tree |
| 2017-04-01 |
schulz | make krnLeaveSupervisorRing macro 64-bit safe...
|
commit | commitdiff | tree |
| 2017-04-01 |
schulz | add some IPTR casting in order to avoid sign extensions...
|
commit | commitdiff | tree |
| 2017-04-01 |
schulz | Add at least minimal protection against comprised memory...
|
commit | commitdiff | tree |
| 2017-03-30 |
schulz | Big change - moved local APIC interrupts to absolute...
|
commit | commitdiff | tree |
| 2017-03-26 |
schulz | clean the code a little by reducing amout of places...
|
commit | commitdiff | tree |
| 2017-03-26 |
schulz | in case of APIC_ERROR_IRQ display a message with error...
|
commit | commitdiff | tree |
| 2017-03-23 |
schulz | move DEFINESET() to another file, otherwise gcc4.6...
|
commit | commitdiff | tree |
| 2017-03-23 |
schulz | less debug
|
commit | commitdiff | tree |
| 2017-03-23 |
schulz | make sure the bouncing buffer has a length
|
commit | commitdiff | tree |
| 2017-03-21 |
schulz | adjust debug output a little (less talking form uhci)
|
commit | commitdiff | tree |
| 2017-03-21 |
schulz | adjust alignment in EhciQH
|
commit | commitdiff | tree |
| 2017-03-21 |
schulz | name the INT lines of PCI properly
|
commit | commitdiff | tree |
| 2017-03-21 |
schulz | when calculating modulo we calculate the IRQLines from...
|
commit | commitdiff | tree |
| 2017-03-21 |
schulz | Add support for IRQ routing tables in PCI.
|
commit | commitdiff | tree |
| 2017-03-21 |
schulz | allow setting INTLine in PCIDevice
|
commit | commitdiff | tree |
| 2017-03-21 |
schulz | provide PCIDriver_IRQRoutingTable attribute
|
commit | commitdiff | tree |
| 2017-03-21 |
schulz | PCI routing table entries unified und moved to hidds...
|
commit | commitdiff | tree |
| 2017-03-21 |
schulz | create IRQ routing table in PCI Driver class. Use it...
|
commit | commitdiff | tree |
| 2017-03-20 |
schulz | wip means to provide IRQ routing table.
|
commit | commitdiff | tree |
| 2017-03-20 |
schulz | shut down tons of debug...
|
commit | commitdiff | tree |
| 2017-03-20 |
schulz | nothing valuable yet, just start playing with finding...
|
commit | commitdiff | tree |
| 2017-03-20 |
schulz | quieten some debug from poseidon, leave only pciusb...
|
commit | commitdiff | tree |
| 2017-03-20 |
schulz | handle bouncing buffer in case of control transfers...
|
commit | commitdiff | tree |
| 2017-03-20 |
schulz | Aww. errornously commited my coding mistake. reverted
|
commit | commitdiff | tree |
| 2017-03-20 |
schulz | in CTRL transfers SetupData packet is always UHDIR_OUT...
|
commit | commitdiff | tree |
| 2017-03-20 |
schulz | Adjust layout of structures on 64bit systems
|
commit | commitdiff | tree |
| 2017-03-17 |
schulz | further 64bit wip changes...
|
commit | commitdiff | tree |
| 2017-03-17 |
schulz | Do some pointer checking before using them. Protects...
|
commit | commitdiff | tree |
| 2017-03-17 |
schulz | Measure and display rendering time
|
commit | commitdiff | tree |
| 2017-03-17 |
schulz | Don'T use KrnAllocCPUMask() in Signal, since Signal...
|
commit | commitdiff | tree |
| 2017-03-15 |
schulz | align the ipi hooks to 128 byte boundary
|
commit | commitdiff | tree |
| 2017-03-15 |
schulz | Do not signal GUI task to redraw *whole* screen when...
|
commit | commitdiff | tree |
| 2017-03-15 |
schulz | Put new IPIs at the *end* of the queue, requeue used...
|
commit | commitdiff | tree |
| 2017-03-15 |
schulz | Pass IPI_CALL_HOOK parameters in IPIHook, use asynchronous...
|
commit | commitdiff | tree |
| 2017-03-15 |
schulz | Allow certain amount of parameters (5 at the moment...
|
commit | commitdiff | tree |
| 2017-03-14 |
schulz | send MSG_STATS only *if* we are not dying currently
|
commit | commitdiff | tree |
| 2017-03-14 |
schulz | if IPI_CALL_HOOK was sent, don't try to set the signal...
|
commit | commitdiff | tree |
| 2017-03-14 |
schulz | fix those badly used macros :)
|
commit | commitdiff | tree |
| 2017-03-14 |
schulz | don't enforce c mode anymore, compile using c++ linker
|
commit | commitdiff | tree |
| 2017-03-14 |
schulz | also support systems with stack growing upwards :)
|
commit | commitdiff | tree |
| 2017-03-14 |
schulz | don't ask for stack pointer explicitly, use AROS_GET_SP...
|
commit | commitdiff | tree |
| 2017-03-12 |
schulz | RAYDEPTH (maximal recursion depth for raytracing) is...
|
commit | commitdiff | tree |
| 2017-03-12 |
schulz | enforce C linker (we are using *no* features which...
|
commit | commitdiff | tree |
| 2017-03-12 |
schulz | add AROS_STACKSIZE to worker stack size
|
commit | commitdiff | tree |
| 2017-03-12 |
schulz | break iteration loop of raytracer unconditionaly if...
|
commit | commitdiff | tree |
| 2017-03-12 |
schulz | Initialize variable
|
commit | commitdiff | tree |
| 2017-03-12 |
schulz | explicit mode is commandline option now
|
commit | commitdiff | tree |
| 2017-03-12 |
schulz | Explicit mode (from smallpt-Authors website) added...
|
commit | commitdiff | tree |
| 2017-03-11 |
schulz | if width/height not given, use screen dimensions...
|
commit | commitdiff | tree |
| 2017-03-11 |
schulz | add width and height options to define render window...
|
commit | commitdiff | tree |
| 2017-03-11 |
schulz | extend the rendering scene a little
|
commit | commitdiff | tree |
| 2017-03-11 |
schulz | add status updates in titlebar :)
|
commit | commitdiff | tree |
| 2017-03-11 |
schulz | small fixups towars clean shutdown of app
|
commit | commitdiff | tree |
| 2017-03-11 |
schulz | simple renderer using SMP
|
commit | commitdiff | tree |
| 2017-03-11 |
schulz | use comparison, not assignment...
|
commit | commitdiff | tree |
| 2017-03-10 |
schulz | After moving Disable() in front of lock function I...
|
commit | commitdiff | tree |
| 2017-03-04 |
schulz | Locking of fragile objects reworked...
|
commit | commitdiff | tree |
| 2017-03-04 |
schulz | use atomic macros to increase DispCount and IdleCount
|
commit | commitdiff | tree |
| 2017-03-04 |
schulz | removed bluish tint in drawn fractal
|
commit | commitdiff | tree |
| 2017-03-04 |
schulz | that change was not necessary. core_ExitInterrupt does...
|
commit | commitdiff | tree |
| 2017-03-03 |
schulz | priorities changed again. Workers are at Pri 0, master...
|
commit | commitdiff | tree |
| 2017-03-03 |
schulz | added OVERSAMPLE switch, added limits on user-defined...
|
commit | commitdiff | tree |
| 2017-03-03 |
schulz | use store fence before sending an IPI? Is it necessary...
|
commit | commitdiff | tree |
| 2017-03-03 |
schulz | enable building of BUDDHA-style fractal generator ...
|
commit | commitdiff | tree |
| 2017-03-03 |
schulz | use spinlock to protect entire trajectory drawing,...
|
commit | commitdiff | tree |
| 2017-03-03 |
schulz | smp-test reworked a little. Now rendering works as...
|
commit | commitdiff | tree |
| 2017-03-02 |
schulz | increase priority of worker and use signals to synchronize...
|
commit | commitdiff | tree |
| 2017-03-02 |
schulz | IPI_CALL_HOOK reworked a little and Signal() improved.
|
commit | commitdiff | tree |
| 2017-03-02 |
schulz | When Enable() is called in supervisor, don't call Sti...
|
commit | commitdiff | tree |
| 2017-03-02 |
schulz | WiP, subject of change. Use IPI_CALL_HOOK If Signal...
|
commit | commitdiff | tree |
| 2017-03-02 |
schulz | fixed busyloop waiting for initialization of worker...
|
commit | commitdiff | tree |
| 2017-03-01 |
schulz | lock the task lists one after another and not all at...
|
commit | commitdiff | tree |
| 2017-02-28 |
schulz | fixed the calculation of per-task CPU usage
|
commit | commitdiff | tree |
| 2017-02-28 |
schulz | Provide per-task CPU usage calculation. On cores above...
|
commit | commitdiff | tree |
| 2017-02-28 |
schulz | even more protected accessed to Execs Lists
|
commit | commitdiff | tree |
| 2017-02-28 |
schulz | LeakWatch and devicetest not protected yet, the rest...
|
commit | commitdiff | tree |
| 2017-02-28 |
schulz | Rev, Lib and ResList use proper lockign of SysBase...
|
commit | commitdiff | tree |
| 2017-02-28 |
schulz | if ExecLockBase is null call single-cpu locking function
|
commit | commitdiff | tree |
| 2017-02-28 |
schulz | use AROSPLATFORM_SMP and check if ExecLockBase is not...
|
commit | commitdiff | tree |
| 2017-02-28 |
schulz | C:ResList uses proper locking of the list
|
commit | commitdiff | tree |
| 2017-02-28 |
schulz | aww, fixed header and function names...
|
commit | commitdiff | tree |
| 2017-02-28 |
schulz | fixed old bug in showlist (it was Forbid()+Enable(...
|
commit | commitdiff | tree |
| 2017-02-28 |
schulz | lower idle task priority to -128
|
commit | commitdiff | tree |
| 2017-02-28 |
schulz | use proper locking of exec's task lists
|
commit | commitdiff | tree |
| 2017-02-28 |
schulz | fixed duplicate macro parameter error
|
commit | commitdiff | tree |
| 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 |
