ryujinx/Ryujinx.HLE/HOS/ProgramLoader.cs

312 lines
10 KiB
C#
Raw Normal View History

Add Profiled Persistent Translation Cache. (#769) * Delete DelegateTypes.cs * Delete DelegateCache.cs * Add files via upload * Update Horizon.cs * Update Program.cs * Update MainWindow.cs * Update Aot.cs * Update RelocEntry.cs * Update Translator.cs * Update MemoryManager.cs * Update InstEmitMemoryHelper.cs * Update Delegates.cs * Nit. * Nit. * Nit. * 10 fewer MSIL bytes for us * Add comment. Nits. * Update Translator.cs * Update Aot.cs * Nits. * Opt.. * Opt.. * Opt.. * Opt.. * Allow to change compression level. * Update MemoryManager.cs * Update Translator.cs * Manage corner cases during the save phase. Nits. * Update Aot.cs * Translator response tweak for Aot disabled. Nit. * Nit. * Nits. * Create DelegateHelpers.cs * Update Delegates.cs * Nit. * Nit. * Nits. * Fix due to #784. * Fixes due to #757 & #841. * Fix due to #846. * Fix due to #847. * Use MethodInfo for managed method calls. Use IR methods instead of managed methods about Max/Min (S/U). Follow-ups & Nits. * Add missing exception messages. Reintroduce slow path for Fmov_Vi. Implement slow path for Fmov_Si. * Switch to the new folder structure. Nits. * Impl. index-based relocation information. Impl. cache file version field. * Nit. * Address gdkchan comments. Mainly: - fixed cache file corruption issue on exit; - exposed a way to disable AOT on the GUI. * Address AcK77 comment. * Address Thealexbarney, jduncanator & emmauss comments. Header magic, CpuId (FI) & Aot -> Ptc. * Adaptation to the new application reloading system. Improvements to the call system of managed methods. Follow-ups. Nits. * Get the same boot times as on master when PTC is disabled. * Profiled Aot. * A32 support (#897). * #975 support (1 of 2). * #975 support (2 of 2). * Rebase fix & nits. * Some fixes and nits (still one bug left). * One fix & nits. * Tests fix (by gdk) & nits. * Support translations not only in high quality and rejit. Nits. * Added possibility to skip translations and continue execution, using `ESC` key. * Update SettingsWindow.cs * Update GLRenderer.cs * Update Ptc.cs * Disabled Profiled PTC by default as requested in the past by gdk. * Fix rejit bug. Increased number of parallel translations. Add stack unwinding stuffs support (1 of 2). Nits. * Add stack unwinding stuffs support (2 of 2). Tuned number of parallel translations. * Restored the ability to assemble jumps with 8-bit offset when Profiled PTC is disabled or during profiling. Modifications due to rebase. Nits. * Limited profiling of the functions to be translated to the addresses belonging to the range of static objects only. * Nits. * Nits. * Update Delegates.cs * Nit. * Update InstEmitSimdArithmetic.cs * Address riperiperi comments. * Fixed the issue of unjustifiably longer boot times at the second boot than at the first boot, measured at the same time or reference point and with the same number of translated functions. * Implemented a simple redundant load/save mechanism. Halved the value of Decoder.MaxInstsPerFunction more appropriate for the current performance of the Translator. Replaced by Logger.PrintError to Logger.PrintDebug in TexturePool.cs about the supposed invalid texture format to avoid the spawn of the log. Nits. * Nit. Improved Logger.PrintError in TexturePool.cs to avoid log spawn. Added missing code for FZ handling (in output) for fp max/min instructions (slow paths). * Add configuration migration for PTC Co-authored-by: Thog <me@thog.eu>
2020-06-16 14:28:02 -04:00
using ARMeilleure.Translation.PTC;
using Ryujinx.Common;
using Ryujinx.Common.Logging;
using Ryujinx.Cpu;
using Ryujinx.HLE.HOS.Kernel;
using Ryujinx.HLE.HOS.Kernel.Common;
using Ryujinx.HLE.HOS.Kernel.Memory;
using Ryujinx.HLE.HOS.Kernel.Process;
using Ryujinx.HLE.Loaders.Executables;
using Ryujinx.HLE.Loaders.Npdm;
namespace Ryujinx.HLE.HOS
{
Add Profiled Persistent Translation Cache. (#769) * Delete DelegateTypes.cs * Delete DelegateCache.cs * Add files via upload * Update Horizon.cs * Update Program.cs * Update MainWindow.cs * Update Aot.cs * Update RelocEntry.cs * Update Translator.cs * Update MemoryManager.cs * Update InstEmitMemoryHelper.cs * Update Delegates.cs * Nit. * Nit. * Nit. * 10 fewer MSIL bytes for us * Add comment. Nits. * Update Translator.cs * Update Aot.cs * Nits. * Opt.. * Opt.. * Opt.. * Opt.. * Allow to change compression level. * Update MemoryManager.cs * Update Translator.cs * Manage corner cases during the save phase. Nits. * Update Aot.cs * Translator response tweak for Aot disabled. Nit. * Nit. * Nits. * Create DelegateHelpers.cs * Update Delegates.cs * Nit. * Nit. * Nits. * Fix due to #784. * Fixes due to #757 & #841. * Fix due to #846. * Fix due to #847. * Use MethodInfo for managed method calls. Use IR methods instead of managed methods about Max/Min (S/U). Follow-ups & Nits. * Add missing exception messages. Reintroduce slow path for Fmov_Vi. Implement slow path for Fmov_Si. * Switch to the new folder structure. Nits. * Impl. index-based relocation information. Impl. cache file version field. * Nit. * Address gdkchan comments. Mainly: - fixed cache file corruption issue on exit; - exposed a way to disable AOT on the GUI. * Address AcK77 comment. * Address Thealexbarney, jduncanator & emmauss comments. Header magic, CpuId (FI) & Aot -> Ptc. * Adaptation to the new application reloading system. Improvements to the call system of managed methods. Follow-ups. Nits. * Get the same boot times as on master when PTC is disabled. * Profiled Aot. * A32 support (#897). * #975 support (1 of 2). * #975 support (2 of 2). * Rebase fix & nits. * Some fixes and nits (still one bug left). * One fix & nits. * Tests fix (by gdk) & nits. * Support translations not only in high quality and rejit. Nits. * Added possibility to skip translations and continue execution, using `ESC` key. * Update SettingsWindow.cs * Update GLRenderer.cs * Update Ptc.cs * Disabled Profiled PTC by default as requested in the past by gdk. * Fix rejit bug. Increased number of parallel translations. Add stack unwinding stuffs support (1 of 2). Nits. * Add stack unwinding stuffs support (2 of 2). Tuned number of parallel translations. * Restored the ability to assemble jumps with 8-bit offset when Profiled PTC is disabled or during profiling. Modifications due to rebase. Nits. * Limited profiling of the functions to be translated to the addresses belonging to the range of static objects only. * Nits. * Nits. * Update Delegates.cs * Nit. * Update InstEmitSimdArithmetic.cs * Address riperiperi comments. * Fixed the issue of unjustifiably longer boot times at the second boot than at the first boot, measured at the same time or reference point and with the same number of translated functions. * Implemented a simple redundant load/save mechanism. Halved the value of Decoder.MaxInstsPerFunction more appropriate for the current performance of the Translator. Replaced by Logger.PrintError to Logger.PrintDebug in TexturePool.cs about the supposed invalid texture format to avoid the spawn of the log. Nits. * Nit. Improved Logger.PrintError in TexturePool.cs to avoid log spawn. Added missing code for FZ handling (in output) for fp max/min instructions (slow paths). * Add configuration migration for PTC Co-authored-by: Thog <me@thog.eu>
2020-06-16 14:28:02 -04:00
static class ProgramLoader
{
private const bool AslrEnabled = true;
private const int ArgsHeaderSize = 8;
private const int ArgsDataSize = 0x9000;
private const int ArgsTotalSize = ArgsHeaderSize + ArgsDataSize;
public static bool LoadKip(KernelContext context, KipExecutable kip)
{
int endOffset = kip.DataOffset + kip.Data.Length;
if (kip.BssSize != 0)
{
endOffset = kip.BssOffset + kip.BssSize;
}
int codeSize = BitUtils.AlignUp(kip.TextOffset + endOffset, KMemoryManager.PageSize);
int codePagesCount = codeSize / KMemoryManager.PageSize;
ulong codeBaseAddress = (kip.Header.Flags & 0x10) != 0 ? 0x8000000UL : 0x200000UL;
ulong codeAddress = codeBaseAddress + (ulong)kip.TextOffset;
int mmuFlags = 0;
if (AslrEnabled)
{
// TODO: Randomization.
mmuFlags |= 0x20;
}
if ((kip.Header.Flags & 0x10) != 0)
{
mmuFlags |= (int)AddressSpaceType.Addr39Bits << 1;
}
if ((kip.Header.Flags & 0x08) != 0)
{
mmuFlags |= 1;
}
ProcessCreationInfo creationInfo = new ProcessCreationInfo(
kip.Header.Name,
kip.Header.ProcessCategory,
kip.Header.TitleId,
codeAddress,
codePagesCount,
mmuFlags,
0,
0);
MemoryRegion memoryRegion = (kip.Header.Flags & 0x20) != 0
? MemoryRegion.Service
: MemoryRegion.Application;
KMemoryRegionManager region = context.MemoryRegions[(int)memoryRegion];
KernelResult result = region.AllocatePages((ulong)codePagesCount, false, out KPageList pageList);
if (result != KernelResult.Success)
{
Logger.PrintError(LogClass.Loader, $"Process initialization returned error \"{result}\".");
return false;
}
KProcess process = new KProcess(context);
result = process.InitializeKip(
creationInfo,
kip.Capabilities,
pageList,
context.ResourceLimit,
memoryRegion);
if (result != KernelResult.Success)
{
Logger.PrintError(LogClass.Loader, $"Process initialization returned error \"{result}\".");
return false;
}
result = LoadIntoMemory(process, kip, codeBaseAddress);
if (result != KernelResult.Success)
{
Logger.PrintError(LogClass.Loader, $"Process initialization returned error \"{result}\".");
return false;
}
process.DefaultCpuCore = kip.Header.DefaultCore;
result = process.Start(kip.Header.MainThreadPriority, (ulong)kip.Header.Sections[1].Attribute);
if (result != KernelResult.Success)
{
Logger.PrintError(LogClass.Loader, $"Process start returned error \"{result}\".");
return false;
}
context.Processes.TryAdd(process.Pid, process);
return true;
}
public static bool LoadNsos(
KernelContext context,
Npdm metaData,
byte[] arguments = null,
params IExecutable[] executables)
{
ulong argsStart = 0;
int argsSize = 0;
ulong codeStart = metaData.Is64Bit ? 0x8000000UL : 0x200000UL;
int codeSize = 0;
ulong[] nsoBase = new ulong[executables.Length];
for (int index = 0; index < executables.Length; index++)
{
IExecutable staticObject = executables[index];
int textEnd = staticObject.TextOffset + staticObject.Text.Length;
int roEnd = staticObject.RoOffset + staticObject.Ro.Length;
int dataEnd = staticObject.DataOffset + staticObject.Data.Length + staticObject.BssSize;
int nsoSize = textEnd;
if ((uint)nsoSize < (uint)roEnd)
{
nsoSize = roEnd;
}
if ((uint)nsoSize < (uint)dataEnd)
{
nsoSize = dataEnd;
}
nsoSize = BitUtils.AlignUp(nsoSize, KMemoryManager.PageSize);
nsoBase[index] = codeStart + (ulong)codeSize;
codeSize += nsoSize;
if (arguments != null && argsSize == 0)
{
argsStart = (ulong)codeSize;
argsSize = BitUtils.AlignDown(arguments.Length * 2 + ArgsTotalSize - 1, KMemoryManager.PageSize);
codeSize += argsSize;
}
}
Add Profiled Persistent Translation Cache. (#769) * Delete DelegateTypes.cs * Delete DelegateCache.cs * Add files via upload * Update Horizon.cs * Update Program.cs * Update MainWindow.cs * Update Aot.cs * Update RelocEntry.cs * Update Translator.cs * Update MemoryManager.cs * Update InstEmitMemoryHelper.cs * Update Delegates.cs * Nit. * Nit. * Nit. * 10 fewer MSIL bytes for us * Add comment. Nits. * Update Translator.cs * Update Aot.cs * Nits. * Opt.. * Opt.. * Opt.. * Opt.. * Allow to change compression level. * Update MemoryManager.cs * Update Translator.cs * Manage corner cases during the save phase. Nits. * Update Aot.cs * Translator response tweak for Aot disabled. Nit. * Nit. * Nits. * Create DelegateHelpers.cs * Update Delegates.cs * Nit. * Nit. * Nits. * Fix due to #784. * Fixes due to #757 & #841. * Fix due to #846. * Fix due to #847. * Use MethodInfo for managed method calls. Use IR methods instead of managed methods about Max/Min (S/U). Follow-ups & Nits. * Add missing exception messages. Reintroduce slow path for Fmov_Vi. Implement slow path for Fmov_Si. * Switch to the new folder structure. Nits. * Impl. index-based relocation information. Impl. cache file version field. * Nit. * Address gdkchan comments. Mainly: - fixed cache file corruption issue on exit; - exposed a way to disable AOT on the GUI. * Address AcK77 comment. * Address Thealexbarney, jduncanator & emmauss comments. Header magic, CpuId (FI) & Aot -> Ptc. * Adaptation to the new application reloading system. Improvements to the call system of managed methods. Follow-ups. Nits. * Get the same boot times as on master when PTC is disabled. * Profiled Aot. * A32 support (#897). * #975 support (1 of 2). * #975 support (2 of 2). * Rebase fix & nits. * Some fixes and nits (still one bug left). * One fix & nits. * Tests fix (by gdk) & nits. * Support translations not only in high quality and rejit. Nits. * Added possibility to skip translations and continue execution, using `ESC` key. * Update SettingsWindow.cs * Update GLRenderer.cs * Update Ptc.cs * Disabled Profiled PTC by default as requested in the past by gdk. * Fix rejit bug. Increased number of parallel translations. Add stack unwinding stuffs support (1 of 2). Nits. * Add stack unwinding stuffs support (2 of 2). Tuned number of parallel translations. * Restored the ability to assemble jumps with 8-bit offset when Profiled PTC is disabled or during profiling. Modifications due to rebase. Nits. * Limited profiling of the functions to be translated to the addresses belonging to the range of static objects only. * Nits. * Nits. * Update Delegates.cs * Nit. * Update InstEmitSimdArithmetic.cs * Address riperiperi comments. * Fixed the issue of unjustifiably longer boot times at the second boot than at the first boot, measured at the same time or reference point and with the same number of translated functions. * Implemented a simple redundant load/save mechanism. Halved the value of Decoder.MaxInstsPerFunction more appropriate for the current performance of the Translator. Replaced by Logger.PrintError to Logger.PrintDebug in TexturePool.cs about the supposed invalid texture format to avoid the spawn of the log. Nits. * Nit. Improved Logger.PrintError in TexturePool.cs to avoid log spawn. Added missing code for FZ handling (in output) for fp max/min instructions (slow paths). * Add configuration migration for PTC Co-authored-by: Thog <me@thog.eu>
2020-06-16 14:28:02 -04:00
PtcProfiler.StaticCodeStart = codeStart;
PtcProfiler.StaticCodeSize = codeSize;
int codePagesCount = codeSize / KMemoryManager.PageSize;
int personalMmHeapPagesCount = metaData.PersonalMmHeapSize / KMemoryManager.PageSize;
ProcessCreationInfo creationInfo = new ProcessCreationInfo(
metaData.TitleName,
metaData.ProcessCategory,
metaData.Aci0.TitleId,
codeStart,
codePagesCount,
metaData.MmuFlags,
0,
personalMmHeapPagesCount);
KernelResult result;
KResourceLimit resourceLimit = new KResourceLimit(context);
long applicationRgSize = (long)context.MemoryRegions[(int)MemoryRegion.Application].Size;
result = resourceLimit.SetLimitValue(LimitableResource.Memory, applicationRgSize);
result |= resourceLimit.SetLimitValue(LimitableResource.Thread, 608);
result |= resourceLimit.SetLimitValue(LimitableResource.Event, 700);
result |= resourceLimit.SetLimitValue(LimitableResource.TransferMemory, 128);
result |= resourceLimit.SetLimitValue(LimitableResource.Session, 894);
if (result != KernelResult.Success)
{
Logger.PrintError(LogClass.Loader, $"Process initialization failed setting resource limit values.");
return false;
}
KProcess process = new KProcess(context);
MemoryRegion memoryRegion = (MemoryRegion)((metaData.Acid.Flags >> 2) & 0xf);
if (memoryRegion > MemoryRegion.NvServices)
{
Logger.PrintError(LogClass.Loader, $"Process initialization failed due to invalid ACID flags.");
return false;
}
result = process.Initialize(
creationInfo,
metaData.Aci0.KernelAccessControl.Capabilities,
resourceLimit,
memoryRegion);
if (result != KernelResult.Success)
{
Logger.PrintError(LogClass.Loader, $"Process initialization returned error \"{result}\".");
return false;
}
for (int index = 0; index < executables.Length; index++)
{
Logger.PrintInfo(LogClass.Loader, $"Loading image {index} at 0x{nsoBase[index]:x16}...");
result = LoadIntoMemory(process, executables[index], nsoBase[index]);
if (result != KernelResult.Success)
{
Logger.PrintError(LogClass.Loader, $"Process initialization returned error \"{result}\".");
return false;
}
}
process.DefaultCpuCore = metaData.DefaultCpuId;
result = process.Start(metaData.MainThreadPriority, (ulong)metaData.MainThreadStackSize);
if (result != KernelResult.Success)
{
Logger.PrintError(LogClass.Loader, $"Process start returned error \"{result}\".");
return false;
}
context.Processes.TryAdd(process.Pid, process);
return true;
}
private static KernelResult LoadIntoMemory(KProcess process, IExecutable image, ulong baseAddress)
{
ulong textStart = baseAddress + (ulong)image.TextOffset;
ulong roStart = baseAddress + (ulong)image.RoOffset;
ulong dataStart = baseAddress + (ulong)image.DataOffset;
ulong bssStart = baseAddress + (ulong)image.BssOffset;
ulong end = dataStart + (ulong)image.Data.Length;
if (image.BssSize != 0)
{
end = bssStart + (ulong)image.BssSize;
}
process.CpuMemory.Write(textStart, image.Text);
process.CpuMemory.Write(roStart, image.Ro);
process.CpuMemory.Write(dataStart, image.Data);
MemoryHelper.FillWithZeros(process.CpuMemory, (long)bssStart, image.BssSize);
KernelResult SetProcessMemoryPermission(ulong address, ulong size, MemoryPermission permission)
{
if (size == 0)
{
return KernelResult.Success;
}
size = BitUtils.AlignUp(size, KMemoryManager.PageSize);
return process.MemoryManager.SetProcessMemoryPermission(address, size, permission);
}
KernelResult result = SetProcessMemoryPermission(textStart, (ulong)image.Text.Length, MemoryPermission.ReadAndExecute);
if (result != KernelResult.Success)
{
return result;
}
result = SetProcessMemoryPermission(roStart, (ulong)image.Ro.Length, MemoryPermission.Read);
if (result != KernelResult.Success)
{
return result;
}
return SetProcessMemoryPermission(dataStart, end - dataStart, MemoryPermission.ReadAndWrite);
}
}
}