mirror of
https://github.com/ryujinx-mirror/ryujinx.git
synced 2024-11-25 03:25:13 +00:00
0c87bf9ea4
* Refactor CPU interface * Use IExecutionContext interface on SVC handler, change how CPU interrupts invokes the handlers * Make CpuEngine take a ITickSource rather than returning one The previous implementation had the scenario where the CPU engine had to implement the tick source in mind, like for example, when we have a hypervisor and the game can read CNTPCT on the host directly. However given that we need to do conversion due to different frequencies anyway, it's not worth it. It's better to just let the user pass the tick source and redirect any reads to CNTPCT to the user tick source * XML docs for the public interfaces * PPTC invalidation due to NativeInterface function name changes * Fix build of the CPU tests * PR feedback
39 lines
1.9 KiB
C#
39 lines
1.9 KiB
C#
namespace Ryujinx.Cpu
|
|
{
|
|
/// <summary>
|
|
/// CPU context interface.
|
|
/// </summary>
|
|
public interface ICpuContext
|
|
{
|
|
/// <summary>
|
|
/// Creates a new execution context that will store thread CPU register state when executing guest code.
|
|
/// </summary>
|
|
/// <param name="exceptionCallbacks">Optional functions to be called when the CPU receives an interrupt</param>
|
|
/// <returns>Execution context</returns>
|
|
IExecutionContext CreateExecutionContext(ExceptionCallbacks exceptionCallbacks);
|
|
|
|
/// <summary>
|
|
/// Starts executing code at a specified entry point address.
|
|
/// </summary>
|
|
/// <remarks>
|
|
/// This function only returns when the execution is stopped, by calling <see cref="IExecutionContext.StopRunning"/>.
|
|
/// </remarks>
|
|
/// <param name="context">Execution context to be used for this run</param>
|
|
/// <param name="address">Entry point address</param>
|
|
void Execute(IExecutionContext context, ulong address);
|
|
|
|
/// <summary>
|
|
/// Invalidates the instruction cache for a given memory region.
|
|
/// </summary>
|
|
/// <remarks>
|
|
/// This should be called if code is modified to make the CPU emulator aware of the modifications,
|
|
/// otherwise it might run stale code which will lead to errors and crashes.
|
|
/// Calling this function is not necessary if the code memory was modified by guest code,
|
|
/// as the expectation is that it will do it on its own using the appropriate cache invalidation instructions,
|
|
/// except on Arm32 where those instructions can't be used in unprivileged mode.
|
|
/// </remarks>
|
|
/// <param name="address">Address of the region to be invalidated</param>
|
|
/// <param name="size">Size of the region to be invalidated</param>
|
|
void InvalidateCacheRegion(ulong address, ulong size);
|
|
}
|
|
}
|