Compare commits

...

38 Commits

Author SHA1 Message Date
Lioncash
f60d5aac3e gc_poller: Resolve compilation warnings on MSVC
We just need to make our intentional implicit truncations explicit.
2020-08-25 23:03:12 -04:00
bunnei
b8885aa03b Merge pull request #4582 from lioncash/xbyak
externals: Update Xbyak to 5.95
2020-08-25 22:02:20 -04:00
Lioncash
c024e5c69a externals: Update Xbyak to 5.95
5.95 contains a potentially backward-compatibility breaking change, so
we should be updating to this to ensure that our code remains
forward-compatible.
2020-08-25 18:16:52 -04:00
bunnei
4efaecf125 Merge pull request #4572 from lioncash/xbyak
externals: Update xbyak to v5.941
2020-08-25 11:24:32 -04:00
bunnei
902bf6d37d Merge pull request #4574 from lioncash/const-fn
memory_manager: Mark IsGranularRange() as a const member function
2020-08-25 11:24:13 -04:00
bunnei
2cd788b6ad Merge pull request #4565 from lioncash/memset
microprofile: Don't memset through std::atomic types
2020-08-25 11:23:16 -04:00
bunnei
dd828607e0 Merge pull request #4563 from lioncash/rcache
registered_cache: Make use of designated initializers
2020-08-25 10:07:33 -04:00
bunnei
b798813eee Merge pull request #4548 from lioncash/color
common/color: Migrate code over to the Common namespace
2020-08-25 10:07:13 -04:00
bunnei
bb752df736 Merge pull request #4542 from ReinUsesLisp/gpu-init-base
video_core: Initialize renderer with a GPU
2020-08-24 22:56:11 -04:00
bunnei
6fd3a1bf85 Merge pull request #4562 from lioncash/loop
cpu_manager: Make use of ranged for where applicable
2020-08-24 17:58:41 -04:00
Rodrigo Locatti
b72d2069ba Merge pull request #4573 from lioncash/label
gl_texture_cache: Take std::string by reference in DecorateViewName()
2020-08-24 01:40:22 -03:00
Lioncash
5bce81c3d6 memory_manager: Mark IsGranularRange() as a const member function
This doesn't modify internal member state, so it can be marked as const.
2020-08-24 00:37:57 -04:00
Lioncash
bae4e6c2f5 gl_texture_cache: Take std::string by reference in DecorateViewName()
LabelGLObject takes a string_view, so we don't need to make copies of
the std::string.
2020-08-23 23:36:33 -04:00
Rodrigo Locatti
fff2de8224 Merge pull request #4571 from lioncash/fence
video_core/fence_manager: Remove unnecessary includes
2020-08-24 00:32:29 -03:00
Lioncash
3121acdf4e externals: Update xbyak to v5.941
Keeps the tracked submodule up to date with the latest release.
2020-08-23 22:47:41 -04:00
Lioncash
f3bb52c0a9 video_core/fence_manager: Remove unnecessary includes
Avoids pulling in unnecessary things that can cause rebuilds when they
aren't required.
2020-08-23 21:44:50 -04:00
Lioncash
ec8d72a567 microprofile: Don't memset through std::atomic types
Two of the members of the MicroProfileThreadLog contains two std::atomic
instances. Given these aren't trivially-copyable types, we shouldn't be
memsetting the structure, given implementation details can contain other
members within it.

To avoid potential undefined behavior on platforms, we can use aggregate
initialization to zero out the members while still having well-defined
behavior.

While we're at it we can also silence some sign conversion warnings.
2020-08-23 21:19:01 -04:00
bunnei
ab350b8975 Merge pull request #4561 from lioncash/key-constexpr
key_manager: Make data arrays constexpr
2020-08-23 20:41:05 -04:00
bunnei
42a00877a2 Merge pull request #4549 from lioncash/files
vfs_real: Avoid redundant map lookups
2020-08-23 19:08:36 -04:00
bunnei
9619964e8c Merge pull request #4559 from lioncash/webresult
web_service: Move web_result.h into web_service
2020-08-23 17:54:43 -04:00
Lioncash
67a8e73101 registered_cache: Make use of ends_with for string suffix checking
Simplifies code.
2020-08-23 11:25:45 -04:00
Lioncash
f83c6e1e0c registered_cache: Make use of designated initializers
Removes the need for comments to indicate the fields being assigned.
2020-08-23 10:26:20 -04:00
bunnei
cfb7fd395c Merge pull request #4560 from lioncash/convert
core_timing: Resolve sign conversion warning
2020-08-23 02:11:14 -04:00
Lioncash
b62a8ca43e key_manager: Make data arrays constexpr
We can convert these maps into constexpr arrays to eliminate some
runtime static constructors.
2020-08-22 22:47:24 -04:00
Lioncash
544fbd5c84 cpu_manager: Make use of ranged for where applicable
We can simplify a few loops by making use of ranged for.
2020-08-22 22:43:19 -04:00
Lioncash
79504f1a39 core_timing: Remove unused header 2020-08-22 22:12:37 -04:00
Lioncash
25cd5d9dda core_timing: Move clock initializer into constructor initializer list
Same behavior, minus unnecessary zeroing out of the pointer.
2020-08-22 20:26:48 -04:00
Lioncash
2624b1eae6 core_timing: Resolve sign conversion warning
This constant is only ever assigned to downcount, which is a s64, not a
u64.
2020-08-22 20:26:45 -04:00
ReinUsesLisp
da53bcee60 video_core: Initialize renderer with a GPU
Add an extra step in GPU initialization to be able to initialize render
backends with a valid GPU instance.
2020-08-22 01:51:45 -03:00
bunnei
baff9ffcac Merge pull request #4521 from lioncash/optionalcache
gl_shader_disk_cache: Make use of std::nullopt where applicable
2020-08-21 23:56:55 -04:00
bunnei
66ac7cf730 Merge pull request #4541 from MerryMage/yolo
dynarmic: Add unsafe optimizations
2020-08-21 23:06:21 -04:00
bunnei
53fbf8e206 Merge pull request #4523 from lioncash/self-assign
macro-interpreter: Resolve -Wself-assign-field warning
2020-08-21 18:25:53 -04:00
Lioncash
551882e512 vfs_real: Resolve sign conversion warnings 2020-08-18 14:58:33 -04:00
Lioncash
ed9ae5a977 vfs_real: Avoid redundant map lookups
Avoids some trivially avoidable map lookups by keeping the result of
find operations around and querying them.
2020-08-18 14:58:29 -04:00
Lioncash
e864aa25d5 common/color: Migrate code over to the Common namespace
No external code makes use of this header, so we can freely change the
namespace.
2020-08-18 14:27:41 -04:00
MerryMage
836ec9176a dynarmic: Add unsafe optimizations 2020-08-16 14:15:39 +01:00
Lioncash
34ec64233a macro-interpreter: Resolve -Wself-assign-field warning
This was assigning the field to itself, which is a no-op. The size
doesn't change between its initial assignment and this one, so this is a
safe change to make.
2020-08-14 08:26:50 -04:00
Lioncash
c8135b3c18 gl_shader_disk_cache: Make use of std::nullopt where applicable
Allows the compiler to avoid unnecessarily zeroing out the internal
buffer of std::optional on some implementations.
2020-08-14 08:20:44 -04:00
50 changed files with 619 additions and 429 deletions

View File

@@ -152,9 +152,11 @@ typedef uint16_t MicroProfileGroupId;
#include <stdint.h>
#include <string.h>
#include <thread>
#include <mutex>
#include <algorithm>
#include <array>
#include <atomic>
#include <mutex>
#include <thread>
#ifndef MICROPROFILE_API
#define MICROPROFILE_API
@@ -605,28 +607,45 @@ struct MicroProfileFrameState
struct MicroProfileThreadLog
{
MicroProfileLogEntry Log[MICROPROFILE_BUFFER_SIZE];
std::array<MicroProfileLogEntry, MICROPROFILE_BUFFER_SIZE> Log{};
std::atomic<uint32_t> nPut;
std::atomic<uint32_t> nGet;
uint32_t nActive;
uint32_t nGpu;
ThreadIdType nThreadId;
std::atomic<uint32_t> nPut{0};
std::atomic<uint32_t> nGet{0};
uint32_t nActive = 0;
uint32_t nGpu = 0;
ThreadIdType nThreadId{};
uint32_t nStack[MICROPROFILE_STACK_MAX];
int64_t nChildTickStack[MICROPROFILE_STACK_MAX];
uint32_t nStackPos;
std::array<uint32_t, MICROPROFILE_STACK_MAX> nStack{};
std::array<int64_t, MICROPROFILE_STACK_MAX> nChildTickStack{};
uint32_t nStackPos = 0;
uint8_t nGroupStackPos[MICROPROFILE_MAX_GROUPS];
int64_t nGroupTicks[MICROPROFILE_MAX_GROUPS];
int64_t nAggregateGroupTicks[MICROPROFILE_MAX_GROUPS];
std::array<uint8_t, MICROPROFILE_MAX_GROUPS> nGroupStackPos{};
std::array<int64_t, MICROPROFILE_MAX_GROUPS> nGroupTicks{};
std::array<int64_t, MICROPROFILE_MAX_GROUPS> nAggregateGroupTicks{};
enum
{
THREAD_MAX_LEN = 64,
};
char ThreadName[64];
int nFreeListNext;
char ThreadName[64]{};
int nFreeListNext = 0;
void Reset() {
Log.fill({});
nPut = 0;
nGet = 0;
nActive = 0;
nGpu = 0;
nThreadId = {};
nStack.fill(0);
nChildTickStack.fill(0);
nStackPos = 0;
nGroupStackPos.fill(0);
nGroupTicks.fill(0);
nAggregateGroupTicks.fill(0);
std::fill(std::begin(ThreadName), std::end(ThreadName), '\0');
nFreeListNext = 0;
}
};
#if MICROPROFILE_GPU_TIMERS_D3D11
@@ -1151,6 +1170,7 @@ MicroProfileThreadLog* MicroProfileCreateThreadLog(const char* pName)
MP_ASSERT(pLog->nPut.load() == 0);
MP_ASSERT(pLog->nGet.load() == 0);
S.nFreeListHead = S.Pool[S.nFreeListHead]->nFreeListNext;
pLog->Reset();
}
else
{
@@ -1158,7 +1178,6 @@ MicroProfileThreadLog* MicroProfileCreateThreadLog(const char* pName)
S.nMemUsage += sizeof(MicroProfileThreadLog);
S.Pool[S.nNumLogs++] = pLog;
}
memset(pLog, 0, sizeof(*pLog));
int len = (int)strlen(pName);
int maxlen = sizeof(pLog->ThreadName)-1;
len = len < maxlen ? len : maxlen;
@@ -1206,8 +1225,8 @@ void MicroProfileOnThreadExit()
{
S.Frames[i].nLogStart[nLogIndex] = 0;
}
memset(pLog->nGroupStackPos, 0, sizeof(pLog->nGroupStackPos));
memset(pLog->nGroupTicks, 0, sizeof(pLog->nGroupTicks));
pLog->nGroupStackPos.fill(0);
pLog->nGroupTicks.fill(0);
}
}

View File

@@ -169,14 +169,13 @@ MICROPROFILEUI_API void MicroProfileCustomGroup(const char* pCustomName, uint32_
MICROPROFILEUI_API void MicroProfileCustomGroupAddTimer(const char* pCustomName, const char* pGroup, const char* pTimer);
#ifdef MICROPROFILEUI_IMPL
#ifdef _WIN32
#define snprintf _snprintf
#endif
#include <inttypes.h>
#include <stdio.h>
#include <stdlib.h>
#include <stdarg.h>
#include <math.h>
#include <algorithm>
#include <array>
MICROPROFILE_DEFINE(g_MicroProfileDetailed, "MicroProfile", "Detailed View", 0x8888000);
MICROPROFILE_DEFINE(g_MicroProfileDrawGraph, "MicroProfile", "Draw Graph", 0xff44ee00);
@@ -227,10 +226,10 @@ struct SOptionDesc
uint8_t nIndex;
bool bSelected;
};
static uint32_t g_MicroProfileAggregatePresets[] = {0, 10, 20, 30, 60, 120};
static float g_MicroProfileReferenceTimePresets[] = {5.f, 10.f, 15.f,20.f, 33.33f, 66.66f, 100.f, 250.f, 500.f, 1000.f};
static uint32_t g_MicroProfileOpacityPresets[] = {0x40, 0x80, 0xc0, 0xff};
static const char* g_MicroProfilePresetNames[] =
static const std::array<uint32_t, 6> g_MicroProfileAggregatePresets{0, 10, 20, 30, 60, 120};
static const std::array<float, 10> g_MicroProfileReferenceTimePresets{5.f, 10.f, 15.f,20.f, 33.33f, 66.66f, 100.f, 250.f, 500.f, 1000.f};
static const std::array<uint32_t, 4> g_MicroProfileOpacityPresets{0x40, 0x80, 0xc0, 0xff};
static const std::array<const char*, 7> g_MicroProfilePresetNames
{
MICROPROFILE_DEFAULT_PRESET,
"Render",
@@ -243,8 +242,8 @@ static const char* g_MicroProfilePresetNames[] =
enum
{
MICROPROFILE_NUM_REFERENCE_PRESETS = sizeof(g_MicroProfileReferenceTimePresets)/sizeof(g_MicroProfileReferenceTimePresets[0]),
MICROPROFILE_NUM_OPACITY_PRESETS = sizeof(g_MicroProfileOpacityPresets)/sizeof(g_MicroProfileOpacityPresets[0]),
MICROPROFILE_NUM_REFERENCE_PRESETS = g_MicroProfileReferenceTimePresets.size(),
MICROPROFILE_NUM_OPACITY_PRESETS = g_MicroProfileOpacityPresets.size(),
#if MICROPROFILE_CONTEXT_SWITCH_TRACE
MICROPROFILE_OPTION_SIZE = MICROPROFILE_NUM_REFERENCE_PRESETS + MICROPROFILE_NUM_OPACITY_PRESETS * 2 + 2 + 7,
#else
@@ -326,9 +325,9 @@ struct MicroProfileUI
MicroProfileUI g_MicroProfileUI;
#define UI g_MicroProfileUI
static uint32_t g_nMicroProfileBackColors[2] = { 0x474747, 0x313131 };
static const std::array<uint32_t, 2> g_nMicroProfileBackColors{ 0x474747, 0x313131 };
#define MICROPROFILE_NUM_CONTEXT_SWITCH_COLORS 16
static uint32_t g_nMicroProfileContextSwitchThreadColors[MICROPROFILE_NUM_CONTEXT_SWITCH_COLORS] = //palette generated by http://tools.medialab.sciences-po.fr/iwanthue/index.php
static const std::array<uint32_t, MICROPROFILE_NUM_CONTEXT_SWITCH_COLORS> g_nMicroProfileContextSwitchThreadColors //palette generated by http://tools.medialab.sciences-po.fr/iwanthue/index.php
{
0x63607B,
0x755E2B,
@@ -356,7 +355,7 @@ void MicroProfileInitUI()
{
bInitialized = true;
memset(&g_MicroProfileUI, 0, sizeof(g_MicroProfileUI));
UI.nActiveMenu = (uint32_t)-1;
UI.nActiveMenu = UINT32_MAX;
UI.fDetailedOffsetTarget = UI.fDetailedOffset = 0.f;
UI.fDetailedRangeTarget = UI.fDetailedRange = 50.f;
@@ -368,7 +367,7 @@ void MicroProfileInitUI()
UI.nWidth = 100;
UI.nHeight = 100;
UI.nCustomActive = (uint32_t)-1;
UI.nCustomActive = UINT32_MAX;
UI.nCustomTimerCount = 0;
UI.nCustomCount = 0;
@@ -498,8 +497,8 @@ inline void MicroProfileDrawFloatWindow(uint32_t nX, uint32_t nY, const char** p
{
MicroProfileDrawBox(nX-MICROPROFILE_TEXT_WIDTH, nY, nX, nY + MICROPROFILE_TEXT_WIDTH, pColors[i]|0xff000000);
}
MicroProfileDrawText(nX + 1, nY + 1, (uint32_t)-1, ppStrings[i0], (uint32_t)strlen(ppStrings[i0]));
MicroProfileDrawText(nX + nWidth - nStringLengths[i0+1] * (MICROPROFILE_TEXT_WIDTH+1), nY + 1, (uint32_t)-1, ppStrings[i0+1], (uint32_t)strlen(ppStrings[i0+1]));
MicroProfileDrawText(nX + 1, nY + 1, UINT32_MAX, ppStrings[i0], (uint32_t)strlen(ppStrings[i0]));
MicroProfileDrawText(nX + nWidth - nStringLengths[i0+1] * (MICROPROFILE_TEXT_WIDTH+1), nY + 1, UINT32_MAX, ppStrings[i0+1], (uint32_t)strlen(ppStrings[i0+1]));
nY += (MICROPROFILE_TEXT_HEIGHT+1);
}
}
@@ -522,7 +521,7 @@ inline void MicroProfileDrawTextBox(uint32_t nX, uint32_t nY, const char** ppStr
MicroProfileDrawBox(nX, nY, nX + nWidth, nY + nHeight, 0xff000000);
for(uint32_t i = 0; i < nNumStrings; ++i)
{
MicroProfileDrawText(nX + 1, nY + 1, (uint32_t)-1, ppStrings[i], (uint32_t)strlen(ppStrings[i]));
MicroProfileDrawText(nX + 1, nY + 1, UINT32_MAX, ppStrings[i], (uint32_t)strlen(ppStrings[i]));
nY += (MICROPROFILE_TEXT_HEIGHT+1);
}
}
@@ -781,7 +780,7 @@ inline void MicroProfileDrawDetailedContextSwitchBars(uint32_t nY, uint32_t nThr
{
MicroProfile& S = *MicroProfileGet();
int64_t nTickIn = -1;
uint32_t nThreadBefore = -1;
uint32_t nThreadBefore = UINT32_MAX;
float fToMs = MicroProfileTickToMsMultiplier(MicroProfileTicksPerSecondCpu());
float fMsToScreen = UI.nWidth / UI.fDetailedRange;
float fMouseX = (float)UI.nMouseX;
@@ -949,10 +948,10 @@ inline void MicroProfileDrawDetailedBars(uint32_t nWidth, uint32_t nHeight, int
uint32_t nContextSwitchHoverThreadAfter = S.nContextSwitchHoverThreadAfter;
uint32_t nContextSwitchHoverThreadBefore = S.nContextSwitchHoverThreadBefore;
S.nContextSwitchHoverThread = S.nContextSwitchHoverThreadAfter = S.nContextSwitchHoverThreadBefore = -1;
S.nContextSwitchHoverThread = S.nContextSwitchHoverThreadAfter = S.nContextSwitchHoverThreadBefore = UINT32_MAX;
uint32_t nContextSwitchStart = -1;
uint32_t nContextSwitchEnd = -1;
uint32_t nContextSwitchStart = UINT32_MAX;
uint32_t nContextSwitchEnd = UINT32_MAX;
S.nContextSwitchHoverCpuNext = 0xff;
S.nContextSwitchHoverTickIn = -1;
S.nContextSwitchHoverTickOut = -1;
@@ -1005,9 +1004,10 @@ inline void MicroProfileDrawDetailedBars(uint32_t nWidth, uint32_t nHeight, int
}while(pFrameLogFirst != pFrameFirst);
if(nGet == (uint32_t)-1)
if (nGet == UINT32_MAX) {
continue;
MP_ASSERT(nGet != (uint32_t)-1);
}
MP_ASSERT(nGet != UINT32_MAX);
nPut = pFrameLogLast->nLogStart[i];
@@ -1023,9 +1023,9 @@ inline void MicroProfileDrawDetailedBars(uint32_t nWidth, uint32_t nHeight, int
int64_t nBaseTicks = bGpu ? nBaseTicksGpu : nBaseTicksCpu;
char ThreadName[MicroProfileThreadLog::THREAD_MAX_LEN + 16];
uint64_t nThreadId = pLog->nThreadId;
snprintf(ThreadName, sizeof(ThreadName)-1, "%04llx: %s", nThreadId, &pLog->ThreadName[0] );
snprintf(ThreadName, sizeof(ThreadName)-1, "%04" PRIx64 ": %s", nThreadId, &pLog->ThreadName[0] );
nY += 3;
uint32_t nThreadColor = -1;
uint32_t nThreadColor = UINT32_MAX;
if(pLog->nThreadId == nContextSwitchHoverThreadAfter || pLog->nThreadId == nContextSwitchHoverThreadBefore)
nThreadColor = UI.nHoverColorShared|0x906060;
MicroProfileDrawText(0, nY, nThreadColor, &ThreadName[0], (uint32_t)strlen(&ThreadName[0]));
@@ -1048,7 +1048,7 @@ inline void MicroProfileDrawDetailedBars(uint32_t nWidth, uint32_t nHeight, int
uint32_t nEnd = nRange[j][1];
for(uint32_t k = nStart; k < nEnd; ++k)
{
MicroProfileLogEntry* pEntry = pLog->Log + k;
MicroProfileLogEntry* pEntry = &pLog->Log[k];
int nType = MicroProfileLogType(*pEntry);
if(MP_LOG_ENTER == nType)
{
@@ -1066,7 +1066,7 @@ inline void MicroProfileDrawDetailedBars(uint32_t nWidth, uint32_t nHeight, int
continue;
}
MicroProfileLogEntry* pEntryEnter = pLog->Log + nStack[nStackPos-1];
MicroProfileLogEntry* pEntryEnter = &pLog->Log[nStack[nStackPos-1]];
if(MicroProfileLogTimerIndex(*pEntryEnter) != MicroProfileLogTimerIndex(*pEntry))
{
//uprintf("mismatch %llx %llx\n", pEntryEnter->nToken, pEntry->nToken);
@@ -1126,7 +1126,7 @@ inline void MicroProfileDrawDetailedBars(uint32_t nWidth, uint32_t nHeight, int
uint32_t nIntegerWidth = (uint32_t)(fXEnd - fXStart);
if(nIntegerWidth)
{
if(bHover && UI.nActiveMenu == -1)
if(bHover && UI.nActiveMenu == UINT32_MAX)
{
nHoverToken = MicroProfileLogTimerIndex(*pEntry);
#if MICROPROFILE_DEBUG
@@ -1146,7 +1146,7 @@ inline void MicroProfileDrawDetailedBars(uint32_t nWidth, uint32_t nHeight, int
int nCharacters = (nTextWidth - 2*MICROPROFILE_TEXT_WIDTH) / MICROPROFILE_TEXT_WIDTH;
if(nCharacters>0)
{
MicroProfileDrawText(fXStartText+1, fYStart+1, -1, S.TimerInfo[nTimerIndex].pName, MicroProfileMin<uint32_t>(S.TimerInfo[nTimerIndex].nNameLen, nCharacters));
MicroProfileDrawText(fXStartText + 1, fYStart + 1, UINT32_MAX, S.TimerInfo[nTimerIndex].pName, MicroProfileMin<uint32_t>(S.TimerInfo[nTimerIndex].nNameLen, nCharacters));
}
}
#endif
@@ -1158,7 +1158,7 @@ inline void MicroProfileDrawDetailedBars(uint32_t nWidth, uint32_t nHeight, int
int nLineX = (int)floor(fXAvg+0.5f);
if(nLineX != (int)nLinesDrawn[nStackPos])
{
if(bHover && UI.nActiveMenu == -1)
if(bHover && UI.nActiveMenu == UINT32_MAX)
{
nHoverToken = (uint32_t)MicroProfileLogTimerIndex(*pEntry);
nHoverTime = MicroProfileLogTickDifference(nTickStart, nTickEnd);
@@ -1235,9 +1235,9 @@ inline void MicroProfileDrawDetailedBars(uint32_t nWidth, uint32_t nHeight, int
// nThreadId is 32-bit on Windows
int nStrLen = snprintf(ThreadName, sizeof(ThreadName)-1, "%04x: %s%s", nThreadId, cLocal, i < nNumThreadsBase ? &S.Pool[i]->ThreadName[0] : MICROPROFILE_THREAD_NAME_FROM_ID(nThreadId) );
#else
int nStrLen = snprintf(ThreadName, sizeof(ThreadName)-1, "%04llx: %s%s", nThreadId, cLocal, i < nNumThreadsBase ? &S.Pool[i]->ThreadName[0] : MICROPROFILE_THREAD_NAME_FROM_ID(nThreadId) );
int nStrLen = snprintf(ThreadName, sizeof(ThreadName)-1, "%04" PRIx64 ": %s%s", nThreadId, cLocal, i < nNumThreadsBase ? &S.Pool[i]->ThreadName[0] : MICROPROFILE_THREAD_NAME_FROM_ID(nThreadId) );
#endif
uint32_t nThreadColor = -1;
uint32_t nThreadColor = UINT32_MAX;
if(nThreadId == nContextSwitchHoverThreadAfter || nThreadId == nContextSwitchHoverThreadBefore)
nThreadColor = UI.nHoverColorShared|0x906060;
MicroProfileDrawDetailedContextSwitchBars(nY+2, nThreadId, nContextSwitchStart, nContextSwitchEnd, nBaseTicksCpu, nBaseY);
@@ -1249,9 +1249,6 @@ inline void MicroProfileDrawDetailedBars(uint32_t nWidth, uint32_t nHeight, int
S.nContextSwitchHoverCpu = S.nContextSwitchHoverCpuNext;
UI.pDisplayMouseOver = pMouseOverNext;
if(!S.nRunning)
@@ -1286,10 +1283,10 @@ inline void MicroProfileDrawDetailedBars(uint32_t nWidth, uint32_t nHeight, int
float fStartTextWidth = (float)((1+MICROPROFILE_TEXT_WIDTH) * nLenStart);
float fStartTextX = fXStart - fStartTextWidth - 2;
MicroProfileDrawBox(fStartTextX, nBaseY, fStartTextX + fStartTextWidth + 2, MICROPROFILE_TEXT_HEIGHT + 2 + nBaseY, 0x33000000, MicroProfileBoxTypeFlat);
MicroProfileDrawText(fStartTextX+1, nBaseY, (uint32_t)-1, sBuffer, nLenStart);
MicroProfileDrawText(fStartTextX+1, nBaseY, UINT32_MAX, sBuffer, nLenStart);
uint32_t nLenEnd = snprintf(sBuffer, sizeof(sBuffer)-1, "%.2fms", fMsEnd);
MicroProfileDrawBox(fXEnd+1, nBaseY, fXEnd+1+(1+MICROPROFILE_TEXT_WIDTH) * nLenEnd + 3, MICROPROFILE_TEXT_HEIGHT + 2 + nBaseY, 0x33000000, MicroProfileBoxTypeFlat);
MicroProfileDrawText(fXEnd+2, nBaseY+1, (uint32_t)-1, sBuffer, nLenEnd);
MicroProfileDrawText(fXEnd+2, nBaseY+1, UINT32_MAX, sBuffer, nLenEnd);
if(UI.nMouseRight)
{
@@ -1316,10 +1313,10 @@ inline void MicroProfileDrawDetailedBars(uint32_t nWidth, uint32_t nHeight, int
float fStartTextWidth = (float)((1+MICROPROFILE_TEXT_WIDTH) * nLenStart);
float fStartTextX = fXStart - fStartTextWidth - 2;
MicroProfileDrawBox(fStartTextX, nBaseY, fStartTextX + fStartTextWidth + 2, MICROPROFILE_TEXT_HEIGHT + 2 + nBaseY, 0x33000000, MicroProfileBoxTypeFlat);
MicroProfileDrawText(fStartTextX+1, nBaseY, (uint32_t)-1, sBuffer, nLenStart);
MicroProfileDrawText(fStartTextX+1, nBaseY, UINT32_MAX, sBuffer, nLenStart);
uint32_t nLenEnd = snprintf(sBuffer, sizeof(sBuffer)-1, "%.2fms", fMsEnd);
MicroProfileDrawBox(fXEnd+1, nBaseY, fXEnd+1+(1+MICROPROFILE_TEXT_WIDTH) * nLenEnd + 3, MICROPROFILE_TEXT_HEIGHT + 2 + nBaseY, 0x33000000, MicroProfileBoxTypeFlat);
MicroProfileDrawText(fXEnd+2, nBaseY+1, (uint32_t)-1, sBuffer, nLenEnd);
MicroProfileDrawText(fXEnd+2, nBaseY+1, UINT32_MAX, sBuffer, nLenEnd);
}
}
}
@@ -1365,7 +1362,7 @@ inline void MicroProfileDrawDetailedFrameHistory(uint32_t nWidth, uint32_t nHeig
fBaseX = fXStart;
uint32_t nColor = MICROPROFILE_FRAME_HISTORY_COLOR_CPU;
if(nIndex == nSelectedFrame)
nColor = (uint32_t)-1;
nColor = UINT32_MAX;
MicroProfileDrawBox(fXStart, nBaseY + fScale * nBarHeight, fXEnd, nBaseY+MICROPROFILE_FRAME_HISTORY_HEIGHT, nColor, MicroProfileBoxTypeBar);
if(pNext->nFrameStartCpu > nCpuStart)
{
@@ -1387,7 +1384,7 @@ inline void MicroProfileDrawDetailedView(uint32_t nWidth, uint32_t nHeight)
uint32_t nBaseY = MICROPROFILE_TEXT_HEIGHT + 1;
int nSelectedFrame = -1;
if(UI.nMouseY > nBaseY && UI.nMouseY <= nBaseY + MICROPROFILE_FRAME_HISTORY_HEIGHT && UI.nActiveMenu == -1)
if(UI.nMouseY > nBaseY && UI.nMouseY <= nBaseY + MICROPROFILE_FRAME_HISTORY_HEIGHT && UI.nActiveMenu == UINT32_MAX)
{
nSelectedFrame = ((MICROPROFILE_NUM_FRAMES) * (UI.nWidth-UI.nMouseX) / UI.nWidth);
@@ -1425,7 +1422,7 @@ inline void MicroProfileDrawHeader(int32_t nX, uint32_t nWidth, const char* pNam
if(pName)
{
MicroProfileDrawBox(nX-8, MICROPROFILE_TEXT_HEIGHT + 2, nX + nWidth+5, MICROPROFILE_TEXT_HEIGHT + 2 + (MICROPROFILE_TEXT_HEIGHT+1), 0xff000000|g_nMicroProfileBackColors[1]);
MicroProfileDrawText(nX, MICROPROFILE_TEXT_HEIGHT + 2, (uint32_t)-1, pName, (uint32_t)strlen(pName));
MicroProfileDrawText(nX, MICROPROFILE_TEXT_HEIGHT + 2, UINT32_MAX, pName, (uint32_t)strlen(pName));
}
}
@@ -1440,7 +1437,7 @@ inline void MicroProfileLoopActiveGroupsDraw(int32_t nX, int32_t nY, const char*
uint32_t nCount = 0;
for(uint32_t j = 0; j < MICROPROFILE_MAX_GROUPS; ++j)
{
uint64_t nMask = 1ll << j;
uint64_t nMask = 1ULL << j;
if(nMask & nGroup)
{
nY += MICROPROFILE_TEXT_HEIGHT + 1;
@@ -1521,7 +1518,7 @@ inline void MicroProfileCalcTimers(float* pTimers, float* pAverage, float* pMax,
}
}
}
nMask <<= 1ll;
nMask <<= 1;
}
}
@@ -1543,7 +1540,7 @@ inline void MicroProfileDrawBarArrayCallback(uint32_t nTimer, uint32_t nIdx, uin
snprintf(sBuffer, SBUF_MAX-1, "%5.2f", pTimers[nIdx]);
if (!pTimers2)
MicroProfileDrawBox(nX + nTextWidth, nY, nX + nTextWidth + fWidth * pTimers[nIdx+1], nY + nHeight, UI.nOpacityForeground|S.TimerInfo[nTimer].nColor, MicroProfileBoxTypeBar);
MicroProfileDrawText(nX, nY, (uint32_t)-1, sBuffer, (uint32_t)strlen(sBuffer));
MicroProfileDrawText(nX, nY, UINT32_MAX, sBuffer, (uint32_t)strlen(sBuffer));
}
@@ -1564,7 +1561,7 @@ inline void MicroProfileDrawBarCallCountCallback(uint32_t nTimer, uint32_t nIdx,
MicroProfile& S = *MicroProfileGet();
char sBuffer[SBUF_MAX];
int nLen = snprintf(sBuffer, SBUF_MAX-1, "%5d", S.Frame[nTimer].nCount);//fix
MicroProfileDrawText(nX, nY, (uint32_t)-1, sBuffer, nLen);
MicroProfileDrawText(nX, nY, UINT32_MAX, sBuffer, nLen);
}
inline uint32_t MicroProfileDrawBarCallCount(int32_t nX, int32_t nY, const char* pName)
@@ -1588,7 +1585,7 @@ inline void MicroProfileDrawBarMetaAverageCallback(uint32_t nTimer, uint32_t nId
float fRcpFrames = pArgs->fRcpFrames;
char sBuffer[SBUF_MAX];
int nLen = snprintf(sBuffer, SBUF_MAX-1, "%5.2f", pCounters[nTimer] * fRcpFrames);
MicroProfileDrawText(nX - nLen * (MICROPROFILE_TEXT_WIDTH+1), nY, (uint32_t)-1, sBuffer, nLen);
MicroProfileDrawText(nX - nLen * (MICROPROFILE_TEXT_WIDTH+1), nY, UINT32_MAX, sBuffer, nLen);
}
inline uint32_t MicroProfileDrawBarMetaAverage(int32_t nX, int32_t nY, uint64_t* pCounters, const char* pName, uint32_t nTotalHeight)
@@ -1609,8 +1606,8 @@ inline void MicroProfileDrawBarMetaCountCallback(uint32_t nTimer, uint32_t nIdx,
{
uint64_t* pCounters = (uint64_t*)pExtra;
char sBuffer[SBUF_MAX];
int nLen = snprintf(sBuffer, SBUF_MAX-1, "%5llu", pCounters[nTimer]);
MicroProfileDrawText(nX - nLen * (MICROPROFILE_TEXT_WIDTH+1), nY, (uint32_t)-1, sBuffer, nLen);
int nLen = snprintf(sBuffer, SBUF_MAX-1, "%5" PRIu64, pCounters[nTimer]);
MicroProfileDrawText(nX - nLen * (MICROPROFILE_TEXT_WIDTH+1), nY, UINT32_MAX, sBuffer, nLen);
}
inline uint32_t MicroProfileDrawBarMetaCount(int32_t nX, int32_t nY, uint64_t* pCounters, const char* pName, uint32_t nTotalHeight)
@@ -1667,7 +1664,7 @@ bool MicroProfileDrawGraph(uint32_t nScreenWidth, uint32_t nScreenHeight)
if(bMouseOver)
{
float fXAvg = fMouseXPrc * MICROPROFILE_GRAPH_WIDTH + nX;
MicroProfileDrawLineVertical(fXAvg, nY, nY + MICROPROFILE_GRAPH_HEIGHT, (uint32_t)-1);
MicroProfileDrawLineVertical(fXAvg, nY, nY + MICROPROFILE_GRAPH_HEIGHT, UINT32_MAX);
}
@@ -1706,7 +1703,7 @@ bool MicroProfileDrawGraph(uint32_t nScreenWidth, uint32_t nScreenHeight)
char buf[32];
int nLen = snprintf(buf, sizeof(buf)-1, "%5.2fms", S.fReferenceTime);
MicroProfileDrawText(nX+1, fY1 - (2+MICROPROFILE_TEXT_HEIGHT), (uint32_t)-1, buf, nLen);
MicroProfileDrawText(nX+1, fY1 - (2+MICROPROFILE_TEXT_HEIGHT), UINT32_MAX, buf, nLen);
}
@@ -1782,7 +1779,7 @@ void MicroProfileDumpTimers()
for(uint32_t j = 0; j < MICROPROFILE_MAX_GROUPS; ++j)
{
uint64_t nMask = 1ll << j;
uint64_t nMask = 1ULL << j;
if(nMask & nActiveGroup)
{
MICROPROFILE_PRINTF("%s\n", S.GroupInfo[j].pName);
@@ -1823,7 +1820,7 @@ inline void MicroProfileDrawBarView(uint32_t nScreenWidth, uint32_t nScreenHeigh
uint32_t nNumGroups = 0;
for(uint32_t j = 0; j < MICROPROFILE_MAX_GROUPS; ++j)
{
if(nActiveGroup & (1ll << j))
if(nActiveGroup & (1ULL << j))
{
nNumTimers += S.GroupInfo[j].nNumTimers;
nNumGroups += 1;
@@ -1878,7 +1875,7 @@ inline void MicroProfileDrawBarView(uint32_t nScreenWidth, uint32_t nScreenHeigh
for(uint32_t i = 0; i < nNumTimers+nNumGroups+1; ++i)
{
uint32_t nY0 = nY + i * (nHeight + 1);
bool bInside = (UI.nActiveMenu == -1) && ((UI.nMouseY >= nY0) && (UI.nMouseY < (nY0 + nHeight + 1)));
bool bInside = (UI.nActiveMenu == UINT32_MAX) && ((UI.nMouseY >= nY0) && (UI.nMouseY < (nY0 + nHeight + 1)));
MicroProfileDrawBox(nX, nY0, nWidth+nX, nY0 + (nHeight+1)+1, UI.nOpacityBackground | (g_nMicroProfileBackColors[nColorIndex++ & 1] + ((bInside) ? 0x002c2c2c : 0)));
}
nX += 10;
@@ -1927,22 +1924,22 @@ inline void MicroProfileDrawBarView(uint32_t nScreenWidth, uint32_t nScreenHeigh
nY = nHeight + 3 - UI.nOffsetY;
for(uint32_t i = 0; i < nNumTimers+nNumGroups+1; ++i)
{
uint32_t nY0 = nY + i * (nHeight + 1);
bool bInside = (UI.nActiveMenu == -1) && ((UI.nMouseY >= nY0) && (UI.nMouseY < (nY0 + nHeight + 1)));
const uint32_t nY0 = nY + i * (nHeight + 1);
const bool bInside = (UI.nActiveMenu == UINT32_MAX) && ((UI.nMouseY >= nY0) && (UI.nMouseY < (nY0 + nHeight + 1)));
MicroProfileDrawBox(nX, nY0, nTimerWidth, nY0 + (nHeight+1)+1, 0xff0000000 | (g_nMicroProfileBackColors[nColorIndex++ & 1] + ((bInside) ? 0x002c2c2c : 0)));
}
nX += MicroProfileDrawBarLegend(nX, nY, nTotalHeight, nTimerWidth-5) + 1;
for(uint32_t j = 0; j < MICROPROFILE_MAX_GROUPS; ++j)
{
if(nActiveGroup & (1ll << j))
if(nActiveGroup & (1ULL << j))
{
MicroProfileDrawText(nX, nY + (1+nHeight) * nLegendOffset, (uint32_t)-1, S.GroupInfo[j].pName, S.GroupInfo[j].nNameLen);
MicroProfileDrawText(nX, nY + (1+nHeight) * nLegendOffset, UINT32_MAX, S.GroupInfo[j].pName, S.GroupInfo[j].nNameLen);
nLegendOffset += S.GroupInfo[j].nNumTimers+1;
}
}
MicroProfileDrawHeader(nX, nTimerWidth-5, "Group");
MicroProfileDrawTextRight(nTimerWidth-3, MICROPROFILE_TEXT_HEIGHT + 2, (uint32_t)-1, "Timer", 5);
MicroProfileDrawTextRight(nTimerWidth-3, MICROPROFILE_TEXT_HEIGHT + 2, UINT32_MAX, "Timer", 5);
MicroProfileDrawLineVertical(nTimerWidth, 0, nTotalHeight+nY, UI.nOpacityBackground|g_nMicroProfileBackColors[0]|g_nMicroProfileBackColors[1]);
MicroProfileDrawLineHorizontal(0, nWidth, 2*MICROPROFILE_TEXT_HEIGHT + 3, UI.nOpacityBackground|g_nMicroProfileBackColors[0]|g_nMicroProfileBackColors[1]);
}
@@ -2003,7 +2000,7 @@ inline const char* MicroProfileUIMenuGroups(int nIndex, bool* bSelected)
}
else
{
*bSelected = 0 != (S.nActiveGroupWanted & (1ll << Item.nIndex));
*bSelected = 0 != (S.nActiveGroupWanted & (1ULL << Item.nIndex));
snprintf(buffer, sizeof(buffer)-1, " %s", Item.pName);
}
return buffer;
@@ -2015,16 +2012,18 @@ inline const char* MicroProfileUIMenuGroups(int nIndex, bool* bSelected)
inline const char* MicroProfileUIMenuAggregate(int nIndex, bool* bSelected)
{
MicroProfile& S = *MicroProfileGet();
if(nIndex < sizeof(g_MicroProfileAggregatePresets)/sizeof(g_MicroProfileAggregatePresets[0]))
if(static_cast<uint32_t>(nIndex) < g_MicroProfileAggregatePresets.size())
{
int val = g_MicroProfileAggregatePresets[nIndex];
*bSelected = (int)S.nAggregateFlip == val;
if(0 == val)
uint32_t val = g_MicroProfileAggregatePresets[nIndex];
*bSelected = S.nAggregateFlip == val;
if (0 == val)
{
return "Infinite";
}
else
{
static char buf[128];
snprintf(buf, sizeof(buf)-1, "%7d", val);
snprintf(buf, sizeof(buf)-1, "%7u", val);
return buf;
}
}
@@ -2098,11 +2097,13 @@ inline const char* MicroProfileUIMenuPreset(int nIndex, bool* bSelected)
{
static char buf[128];
*bSelected = false;
int nNumPresets = sizeof(g_MicroProfilePresetNames) / sizeof(g_MicroProfilePresetNames[0]);
int nNumPresets = static_cast<int>(g_MicroProfilePresetNames.size());
int nIndexSave = nIndex - nNumPresets - 1;
if(nIndex == nNumPresets)
if (nIndex == nNumPresets)
{
return "--";
else if(nIndexSave >=0 && nIndexSave <nNumPresets)
}
else if(nIndexSave >=0 && nIndexSave < nNumPresets)
{
snprintf(buf, sizeof(buf)-1, "Save '%s'", g_MicroProfilePresetNames[nIndexSave]);
return buf;
@@ -2120,13 +2121,13 @@ inline const char* MicroProfileUIMenuPreset(int nIndex, bool* bSelected)
inline const char* MicroProfileUIMenuCustom(int nIndex, bool* bSelected)
{
if((uint32_t)-1 == UI.nCustomActive)
if(UINT32_MAX == UI.nCustomActive)
{
*bSelected = nIndex == 0;
}
else
{
*bSelected = nIndex-2 == UI.nCustomActive;
*bSelected = nIndex-2 == static_cast<int>(UI.nCustomActive);
}
switch(nIndex)
{
@@ -2202,7 +2203,7 @@ inline void MicroProfileUIClickGroups(int nIndex)
else
{
MP_ASSERT(Item.nIndex < S.nGroupCount);
S.nActiveGroupWanted ^= (1ll << Item.nIndex);
S.nActiveGroupWanted ^= (1ULL << Item.nIndex);
}
}
}
@@ -2273,7 +2274,7 @@ inline void MicroProfileUIClickOptions(int nIndex)
inline void MicroProfileUIClickPreset(int nIndex)
{
int nNumPresets = sizeof(g_MicroProfilePresetNames) / sizeof(g_MicroProfilePresetNames[0]);
int nNumPresets = static_cast<int>(g_MicroProfilePresetNames.size());
int nIndexSave = nIndex - nNumPresets - 1;
if(nIndexSave >= 0 && nIndexSave < nNumPresets)
{
@@ -2310,7 +2311,7 @@ inline void MicroProfileDrawMenu(uint32_t nWidth, uint32_t nHeight)
uint32_t nX = 0;
uint32_t nY = 0;
bool bMouseOver = UI.nMouseY < MICROPROFILE_TEXT_HEIGHT + 1;
#define SBUF_SIZE 256
char buffer[256];
MicroProfileDrawBox(nX, nY, nX + nWidth, nY + (MICROPROFILE_TEXT_HEIGHT+1)+1, 0xff000000|g_nMicroProfileBackColors[1]);
@@ -2321,7 +2322,7 @@ inline void MicroProfileDrawMenu(uint32_t nWidth, uint32_t nHeight)
uint32_t nNumMenuItems = 0;
int nLen = snprintf(buffer, 127, "MicroProfile");
MicroProfileDrawText(nX, nY, (uint32_t)-1, buffer, nLen);
MicroProfileDrawText(nX, nY, UINT32_MAX, buffer, nLen);
nX += (sizeof("MicroProfile")+2) * (MICROPROFILE_TEXT_WIDTH+1);
pMenuText[nNumMenuItems++] = "Mode";
pMenuText[nNumMenuItems++] = "Groups";
@@ -2409,7 +2410,7 @@ inline void MicroProfileDrawMenu(uint32_t nWidth, uint32_t nHeight)
};
uint32_t nSelectMenu = (uint32_t)-1;
uint32_t nSelectMenu = UINT32_MAX;
for(uint32_t i = 0; i < nNumMenuItems; ++i)
{
nMenuX[i] = nX;
@@ -2419,17 +2420,17 @@ inline void MicroProfileDrawMenu(uint32_t nWidth, uint32_t nHeight)
{
MicroProfileDrawBox(nX-1, nY, nX + nLen * (MICROPROFILE_TEXT_WIDTH+1), nY +(MICROPROFILE_TEXT_HEIGHT+1)+1, 0xff888888);
nSelectMenu = i;
if((UI.nMouseLeft || UI.nMouseRight) && i == (int)nPauseIndex)
if((UI.nMouseLeft || UI.nMouseRight) && i == (uint32_t)nPauseIndex)
{
S.nToggleRunning = 1;
}
}
MicroProfileDrawText(nX, nY, (uint32_t)-1, pMenuText[i], (uint32_t)strlen(pMenuText[i]));
MicroProfileDrawText(nX, nY, UINT32_MAX, pMenuText[i], (uint32_t)strlen(pMenuText[i]));
nX += (nLen+1) * (MICROPROFILE_TEXT_WIDTH+1);
}
uint32_t nMenu = nSelectMenu != (uint32_t)-1 ? nSelectMenu : UI.nActiveMenu;
uint32_t nMenu = nSelectMenu != UINT32_MAX ? nSelectMenu : UI.nActiveMenu;
UI.nActiveMenu = nMenu;
if((uint32_t)-1 != nMenu)
if(UINT32_MAX != nMenu)
{
nX = nMenuX[nMenu];
nY += MICROPROFILE_TEXT_HEIGHT+1;
@@ -2450,9 +2451,9 @@ inline void MicroProfileDrawMenu(uint32_t nWidth, uint32_t nHeight)
{
UI.nActiveMenu = nMenu;
}
else if(nSelectMenu == (uint32_t)-1)
else if(nSelectMenu == UINT32_MAX)
{
UI.nActiveMenu = (uint32_t)-1;
UI.nActiveMenu = UINT32_MAX;
}
MicroProfileDrawBox(nX, nY, nX + nWidth, nY + nHeight, 0xff000000|g_nMicroProfileBackColors[1]);
for(int i = 0; i < nNumLines; ++i)
@@ -2461,7 +2462,6 @@ inline void MicroProfileDrawMenu(uint32_t nWidth, uint32_t nHeight)
const char* pString = CB(i, &bSelected);
if(UI.nMouseY >= nY && UI.nMouseY < nY + MICROPROFILE_TEXT_HEIGHT + 1)
{
bMouseOver = true;
if(UI.nMouseLeft || UI.nMouseRight)
{
CBClick[nMenu](i);
@@ -2469,7 +2469,7 @@ inline void MicroProfileDrawMenu(uint32_t nWidth, uint32_t nHeight)
MicroProfileDrawBox(nX, nY, nX + nWidth, nY + MICROPROFILE_TEXT_HEIGHT + 1, 0xff888888);
}
int nLen = snprintf(buffer, SBUF_SIZE-1, "%c %s", bSelected ? '*' : ' ' ,pString);
MicroProfileDrawText(nX, nY, (uint32_t)-1, buffer, nLen);
MicroProfileDrawText(nX, nY, UINT32_MAX, buffer, nLen);
nY += MICROPROFILE_TEXT_HEIGHT+1;
}
}
@@ -2484,7 +2484,7 @@ inline void MicroProfileDrawMenu(uint32_t nWidth, uint32_t nHeight)
float fMaxMs = fToMs * S.nFlipMaxDisplay;
int nLen = snprintf(FrameTimeMessage, sizeof(FrameTimeMessage)-1, "Time[%6.2f] Avg[%6.2f] Max[%6.2f]", fMs, fAverageMs, fMaxMs);
pMenuText[nNumMenuItems++] = &FrameTimeMessage[0];
MicroProfileDrawText(nWidth - nLen * (MICROPROFILE_TEXT_WIDTH+1), 0, -1, FrameTimeMessage, nLen);
MicroProfileDrawText(nWidth - nLen * (MICROPROFILE_TEXT_WIDTH+1), 0, UINT32_MAX, FrameTimeMessage, nLen);
}
}
@@ -2538,7 +2538,7 @@ inline void MicroProfileMoveGraph()
inline void MicroProfileDrawCustom(uint32_t nWidth, uint32_t nHeight)
{
if((uint32_t)-1 != UI.nCustomActive)
if(UINT32_MAX != UI.nCustomActive)
{
MicroProfile& S = *MicroProfileGet();
MP_ASSERT(UI.nCustomActive < MICROPROFILE_CUSTOM_MAX);
@@ -2571,8 +2571,8 @@ inline void MicroProfileDrawCustom(uint32_t nWidth, uint32_t nHeight)
pColors[i] = S.TimerInfo[nTimerIndex].nColor;
}
MicroProfileDrawText(MICROPROFILE_CUSTOM_PADDING + 3*MICROPROFILE_TEXT_WIDTH, nOffsetY, (uint32_t)-1, "Avg", sizeof("Avg")-1);
MicroProfileDrawText(MICROPROFILE_CUSTOM_PADDING + 13*MICROPROFILE_TEXT_WIDTH, nOffsetY, (uint32_t)-1, "Max", sizeof("Max")-1);
MicroProfileDrawText(MICROPROFILE_CUSTOM_PADDING + 3*MICROPROFILE_TEXT_WIDTH, nOffsetY, UINT32_MAX, "Avg", sizeof("Avg")-1);
MicroProfileDrawText(MICROPROFILE_CUSTOM_PADDING + 13*MICROPROFILE_TEXT_WIDTH, nOffsetY, UINT32_MAX, "Max", sizeof("Max")-1);
for(uint32_t i = 0; i < nCount; ++i)
{
nOffsetY += (1+MICROPROFILE_TEXT_HEIGHT);
@@ -2582,10 +2582,10 @@ inline void MicroProfileDrawCustom(uint32_t nWidth, uint32_t nHeight)
int nSize;
uint32_t nOffsetX = MICROPROFILE_CUSTOM_PADDING;
nSize = snprintf(Buffer, sizeof(Buffer)-1, "%6.2f", pTimeAvg[i]);
MicroProfileDrawText(nOffsetX, nOffsetY, (uint32_t)-1, Buffer, nSize);
MicroProfileDrawText(nOffsetX, nOffsetY, UINT32_MAX, Buffer, nSize);
nOffsetX += (nSize+2) * (MICROPROFILE_TEXT_WIDTH+1);
nSize = snprintf(Buffer, sizeof(Buffer)-1, "%6.2f", pTimeMax[i]);
MicroProfileDrawText(nOffsetX, nOffsetY, (uint32_t)-1, Buffer, nSize);
MicroProfileDrawText(nOffsetX, nOffsetY, UINT32_MAX, Buffer, nSize);
nOffsetX += (nSize+2) * (MICROPROFILE_TEXT_WIDTH+1);
nSize = snprintf(Buffer, sizeof(Buffer)-1, "%s:%s", S.GroupInfo[nGroupIndex].pName, pTimerInfo->pName);
MicroProfileDrawText(nOffsetX, nOffsetY, pTimerInfo->nColor, Buffer, nSize);
@@ -2599,9 +2599,9 @@ inline void MicroProfileDrawCustom(uint32_t nWidth, uint32_t nHeight)
nOffsetY = nOffsetYBase;
float* pMs = pCustom->nFlags & MICROPROFILE_CUSTOM_BAR_SOURCE_MAX ? pTimeMax : pTimeAvg;
const char* pString = pCustom->nFlags & MICROPROFILE_CUSTOM_BAR_SOURCE_MAX ? "Max" : "Avg";
MicroProfileDrawText(nMaxOffsetX, nOffsetY, (uint32_t)-1, pString, static_cast<uint32_t>(strlen(pString)));
MicroProfileDrawText(nMaxOffsetX, nOffsetY, UINT32_MAX, pString, static_cast<uint32_t>(strlen(pString)));
int nSize = snprintf(Buffer, sizeof(Buffer)-1, "%6.2fms", fReference);
MicroProfileDrawText(nReducedWidth - (1+nSize) * (MICROPROFILE_TEXT_WIDTH+1), nOffsetY, (uint32_t)-1, Buffer, nSize);
MicroProfileDrawText(nReducedWidth - (1+nSize) * (MICROPROFILE_TEXT_WIDTH+1), nOffsetY, UINT32_MAX, Buffer, nSize);
for(uint32_t i = 0; i < nCount; ++i)
{
nOffsetY += (1+MICROPROFILE_TEXT_HEIGHT);
@@ -2613,9 +2613,9 @@ inline void MicroProfileDrawCustom(uint32_t nWidth, uint32_t nHeight)
{
nOffsetY += 2*(1+MICROPROFILE_TEXT_HEIGHT);
const char* pString = pCustom->nFlags & MICROPROFILE_CUSTOM_STACK_SOURCE_MAX ? "Max" : "Avg";
MicroProfileDrawText(MICROPROFILE_CUSTOM_PADDING, nOffsetY, (uint32_t)-1, pString, static_cast<uint32_t>(strlen(pString)));
MicroProfileDrawText(MICROPROFILE_CUSTOM_PADDING, nOffsetY, UINT32_MAX, pString, static_cast<uint32_t>(strlen(pString)));
int nSize = snprintf(Buffer, sizeof(Buffer)-1, "%6.2fms", fReference);
MicroProfileDrawText(nReducedWidth - (1+nSize) * (MICROPROFILE_TEXT_WIDTH+1), nOffsetY, (uint32_t)-1, Buffer, nSize);
MicroProfileDrawText(nReducedWidth - (1+nSize) * (MICROPROFILE_TEXT_WIDTH+1), nOffsetY, UINT32_MAX, Buffer, nSize);
nOffsetY += (1+MICROPROFILE_TEXT_HEIGHT);
float fPosX = MICROPROFILE_CUSTOM_PADDING;
float* pMs = pCustom->nFlags & MICROPROFILE_CUSTOM_STACK_SOURCE_MAX ? pTimeMax : pTimeAvg;
@@ -2668,7 +2668,7 @@ inline void MicroProfileDraw(uint32_t nWidth, uint32_t nHeight)
UI.nHoverTime = 0;
UI.nHoverFrame = -1;
if(S.nDisplay != MP_DRAW_DETAILED)
S.nContextSwitchHoverThread = S.nContextSwitchHoverThreadAfter = S.nContextSwitchHoverThreadBefore = -1;
S.nContextSwitchHoverThread = S.nContextSwitchHoverThreadAfter = S.nContextSwitchHoverThreadBefore = UINT32_MAX;
MicroProfileMoveGraph();
@@ -2798,13 +2798,13 @@ inline void MicroProfileDraw(uint32_t nWidth, uint32_t nHeight)
if(UI.nActiveMenu == -1 && !bMouseOverGraph)
if(UI.nActiveMenu == UINT32_MAX && !bMouseOverGraph)
{
if(UI.nHoverToken != MICROPROFILE_INVALID_TOKEN)
{
MicroProfileDrawFloatTooltip(UI.nMouseX, UI.nMouseY, UI.nHoverToken, UI.nHoverTime);
}
else if(S.nContextSwitchHoverThreadAfter != -1 && S.nContextSwitchHoverThreadBefore != -1)
else if(S.nContextSwitchHoverThreadAfter != UINT32_MAX && S.nContextSwitchHoverThreadBefore != UINT32_MAX)
{
float fToMs = MicroProfileTickToMsMultiplier(MicroProfileTicksPerSecondCpu());
MicroProfileStringArray ToolTip;
@@ -2820,7 +2820,7 @@ inline void MicroProfileDraw(uint32_t nWidth, uint32_t nHeight)
MicroProfileStringArrayFormat(&ToolTip, "%6.2fms", fToMs * nDifference );
MicroProfileStringArrayAddLiteral(&ToolTip, "CPU");
MicroProfileStringArrayFormat(&ToolTip, "%d", S.nContextSwitchHoverCpu);
MicroProfileDrawFloatWindow(UI.nMouseX, UI.nMouseY+20, &ToolTip.ppStrings[0], ToolTip.nNumStrings, -1);
MicroProfileDrawFloatWindow(UI.nMouseX, UI.nMouseY+20, &ToolTip.ppStrings[0], ToolTip.nNumStrings, UINT32_MAX);
}
@@ -2858,7 +2858,7 @@ inline void MicroProfileDraw(uint32_t nWidth, uint32_t nHeight)
}
}
#endif
MicroProfileDrawFloatWindow(UI.nMouseX, UI.nMouseY+20, &ToolTip.ppStrings[0], ToolTip.nNumStrings, -1);
MicroProfileDrawFloatWindow(UI.nMouseX, UI.nMouseY+20, &ToolTip.ppStrings[0], ToolTip.nNumStrings, UINT32_MAX);
}
if(UI.nMouseLeft)
{
@@ -2883,7 +2883,7 @@ inline void MicroProfileDraw(uint32_t nWidth, uint32_t nHeight)
#endif
m.unlock();
}
else if(UI.nCustomActive != (uint32_t)-1)
else if(UI.nCustomActive != UINT32_MAX)
{
std::recursive_mutex& m = MicroProfileGetMutex();
m.lock();
@@ -3179,7 +3179,7 @@ void MicroProfileLoadPreset(const char* pSuffix)
{
if(0 == MP_STRCASECMP(pGroupName, S.GroupInfo[j].pName))
{
S.nActiveGroupWanted |= (1ll << j);
S.nActiveGroupWanted |= (1ULL << j);
}
}
}
@@ -3212,7 +3212,7 @@ void MicroProfileLoadPreset(const char* pSuffix)
uint64_t nGroupIndex = S.TimerInfo[j].nGroupIndex;
if(0 == MP_STRCASECMP(pGraphName, S.TimerInfo[j].pName) && 0 == MP_STRCASECMP(pGraphGroupName, S.GroupInfo[nGroupIndex].pName))
{
MicroProfileToken nToken = MicroProfileMakeToken(1ll << nGroupIndex, (uint16_t)j);
MicroProfileToken nToken = MicroProfileMakeToken(1ULL << nGroupIndex, (uint16_t)j);
S.Graph[i].nToken = nToken; // note: group index is stored here but is checked without in MicroProfileToggleGraph()!
S.TimerInfo[j].bGraph = true;
if(nToken != nPrevToken)
@@ -3235,7 +3235,7 @@ inline uint32_t MicroProfileCustomGroupFind(const char* pCustomName)
return i;
}
}
return (uint32_t)-1;
return UINT32_MAX;
}
inline uint32_t MicroProfileCustomGroup(const char* pCustomName)
@@ -3251,7 +3251,7 @@ inline uint32_t MicroProfileCustomGroup(const char* pCustomName)
uint32_t nIndex = UI.nCustomCount;
UI.nCustomCount++;
memset(&UI.Custom[nIndex], 0, sizeof(UI.Custom[nIndex]));
uint32_t nLen = (uint32_t)strlen(pCustomName);
size_t nLen = strlen(pCustomName);
if(nLen > MICROPROFILE_NAME_MAX_LEN-1)
nLen = MICROPROFILE_NAME_MAX_LEN-1;
memcpy(&UI.Custom[nIndex].pName[0], pCustomName, nLen);
@@ -3309,7 +3309,7 @@ inline void MicroProfileCustomGroupEnable(uint32_t nIndex)
void MicroProfileCustomGroupToggle(const char* pCustomName)
{
uint32_t nIndex = MicroProfileCustomGroupFind(pCustomName);
if(nIndex == (uint32_t)-1 || nIndex == UI.nCustomActive)
if(nIndex == UINT32_MAX || nIndex == UI.nCustomActive)
{
MicroProfileCustomGroupDisable();
}
@@ -3328,13 +3328,13 @@ void MicroProfileCustomGroupDisable()
{
MicroProfile& S = *MicroProfileGet();
S.nForceGroupUI = 0;
UI.nCustomActive = (uint32_t)-1;
UI.nCustomActive = UINT32_MAX;
}
void MicroProfileCustomGroupAddTimer(const char* pCustomName, const char* pGroup, const char* pTimer)
{
uint32_t nIndex = MicroProfileCustomGroupFind(pCustomName);
if((uint32_t)-1 == nIndex)
if(UINT32_MAX == nIndex)
{
return;
}
@@ -3344,7 +3344,7 @@ void MicroProfileCustomGroupAddTimer(const char* pCustomName, const char* pGroup
MP_ASSERT(nToken != MICROPROFILE_INVALID_TOKEN); //Timer must be registered first.
UI.Custom[nIndex].pTimers[nTimerIndex] = nToken;
uint16_t nGroup = MicroProfileGetGroupIndex(nToken);
UI.Custom[nIndex].nGroupMask |= (1ll << nGroup);
UI.Custom[nIndex].nGroupMask |= (1ULL << nGroup);
UI.Custom[nIndex].nNumTimers++;
}

View File

@@ -10,7 +10,7 @@
#include "common/swap.h"
#include "common/vector_math.h"
namespace Color {
namespace Common::Color {
/// Convert a 1-bit color component to 8 bit
[[nodiscard]] constexpr u8 Convert1To8(u8 value) {
@@ -268,4 +268,4 @@ inline void EncodeX24S8(u8 stencil, u8* bytes) {
bytes[3] = stencil;
}
} // namespace Color
} // namespace Common::Color

View File

@@ -143,7 +143,7 @@ std::shared_ptr<Dynarmic::A32::Jit> ARM_Dynarmic_32::MakeJit(Common::PageTable&
config.wall_clock_cntpct = uses_wall_clock;
// Safe optimizations
if (Settings::values.cpu_accuracy != Settings::CPUAccuracy::Accurate) {
if (Settings::values.cpu_accuracy == Settings::CPUAccuracy::DebugMode) {
if (!Settings::values.cpuopt_page_tables) {
config.page_table = nullptr;
}
@@ -170,6 +170,17 @@ std::shared_ptr<Dynarmic::A32::Jit> ARM_Dynarmic_32::MakeJit(Common::PageTable&
}
}
// Unsafe optimizations
if (Settings::values.cpu_accuracy == Settings::CPUAccuracy::Unsafe) {
config.unsafe_optimizations = true;
if (Settings::values.cpuopt_unsafe_unfuse_fma) {
config.optimizations |= Dynarmic::OptimizationFlag::Unsafe_UnfuseFMA;
}
if (Settings::values.cpuopt_unsafe_reduce_fp_error) {
config.optimizations |= Dynarmic::OptimizationFlag::Unsafe_ReducedErrorFP;
}
}
return std::make_unique<Dynarmic::A32::Jit>(config);
}

View File

@@ -195,7 +195,7 @@ std::shared_ptr<Dynarmic::A64::Jit> ARM_Dynarmic_64::MakeJit(Common::PageTable&
config.wall_clock_cntpct = uses_wall_clock;
// Safe optimizations
if (Settings::values.cpu_accuracy != Settings::CPUAccuracy::Accurate) {
if (Settings::values.cpu_accuracy == Settings::CPUAccuracy::DebugMode) {
if (!Settings::values.cpuopt_page_tables) {
config.page_table = nullptr;
}
@@ -222,6 +222,17 @@ std::shared_ptr<Dynarmic::A64::Jit> ARM_Dynarmic_64::MakeJit(Common::PageTable&
}
}
// Unsafe optimizations
if (Settings::values.cpu_accuracy == Settings::CPUAccuracy::Unsafe) {
config.unsafe_optimizations = true;
if (Settings::values.cpuopt_unsafe_unfuse_fma) {
config.optimizations |= Dynarmic::OptimizationFlag::Unsafe_UnfuseFMA;
}
if (Settings::values.cpuopt_unsafe_reduce_fp_error) {
config.optimizations |= Dynarmic::OptimizationFlag::Unsafe_ReducedErrorFP;
}
}
return std::make_shared<Dynarmic::A64::Jit>(config);
}

View File

@@ -7,14 +7,14 @@
#include <string>
#include <tuple>
#include "common/assert.h"
#include "common/microprofile.h"
#include "core/core_timing.h"
#include "core/core_timing_util.h"
#include "core/hardware_properties.h"
namespace Core::Timing {
constexpr u64 MAX_SLICE_LENGTH = 4000;
constexpr s64 MAX_SLICE_LENGTH = 4000;
std::shared_ptr<EventType> CreateEvent(std::string name, TimedCallback&& callback) {
return std::make_shared<EventType>(std::move(callback), std::move(name));
@@ -37,10 +37,8 @@ struct CoreTiming::Event {
}
};
CoreTiming::CoreTiming() {
clock =
Common::CreateBestMatchingClock(Core::Hardware::BASE_CLOCK_RATE, Core::Hardware::CNTFREQ);
}
CoreTiming::CoreTiming()
: clock{Common::CreateBestMatchingClock(Hardware::BASE_CLOCK_RATE, Hardware::CNTFREQ)} {}
CoreTiming::~CoreTiming() = default;
@@ -136,7 +134,7 @@ void CoreTiming::UnscheduleEvent(const std::shared_ptr<EventType>& event_type,
void CoreTiming::AddTicks(u64 ticks) {
this->ticks += ticks;
downcount -= ticks;
downcount -= static_cast<s64>(ticks);
}
void CoreTiming::Idle() {

View File

@@ -8,6 +8,7 @@
#include <limits>
#include "common/logging/log.h"
#include "common/uint128.h"
#include "core/hardware_properties.h"
namespace Core::Timing {

View File

@@ -6,7 +6,6 @@
#include <chrono>
#include "common/common_types.h"
#include "core/hardware_properties.h"
namespace Core::Timing {

View File

@@ -41,9 +41,9 @@ void CpuManager::Shutdown() {
running_mode = false;
Pause(false);
if (is_multicore) {
for (std::size_t core = 0; core < Core::Hardware::NUM_CPU_CORES; core++) {
core_data[core].host_thread->join();
core_data[core].host_thread.reset();
for (auto& data : core_data) {
data.host_thread->join();
data.host_thread.reset();
}
} else {
core_data[0].host_thread->join();
@@ -166,25 +166,23 @@ void CpuManager::MultiCorePause(bool paused) {
bool all_not_barrier = false;
while (!all_not_barrier) {
all_not_barrier = true;
for (std::size_t core = 0; core < Core::Hardware::NUM_CPU_CORES; core++) {
all_not_barrier &=
!core_data[core].is_running.load() && core_data[core].initialized.load();
for (const auto& data : core_data) {
all_not_barrier &= !data.is_running.load() && data.initialized.load();
}
}
for (std::size_t core = 0; core < Core::Hardware::NUM_CPU_CORES; core++) {
core_data[core].enter_barrier->Set();
for (auto& data : core_data) {
data.enter_barrier->Set();
}
if (paused_state.load()) {
bool all_barrier = false;
while (!all_barrier) {
all_barrier = true;
for (std::size_t core = 0; core < Core::Hardware::NUM_CPU_CORES; core++) {
all_barrier &=
core_data[core].is_paused.load() && core_data[core].initialized.load();
for (const auto& data : core_data) {
all_barrier &= data.is_paused.load() && data.initialized.load();
}
}
for (std::size_t core = 0; core < Core::Hardware::NUM_CPU_CORES; core++) {
core_data[core].exit_barrier->Set();
for (auto& data : core_data) {
data.exit_barrier->Set();
}
}
} else {
@@ -192,9 +190,8 @@ void CpuManager::MultiCorePause(bool paused) {
bool all_barrier = false;
while (!all_barrier) {
all_barrier = true;
for (std::size_t core = 0; core < Core::Hardware::NUM_CPU_CORES; core++) {
all_barrier &=
core_data[core].is_paused.load() && core_data[core].initialized.load();
for (const auto& data : core_data) {
all_barrier &= data.is_paused.load() && data.initialized.load();
}
}
/// Don't release the barrier

View File

@@ -36,6 +36,7 @@
#include "core/settings.h"
namespace Core::Crypto {
namespace {
constexpr u64 CURRENT_CRYPTO_REVISION = 0x5;
constexpr u64 FULL_TICKET_SIZE = 0x400;
@@ -49,7 +50,72 @@ constexpr std::array eticket_source_hashes{
};
// clang-format on
const std::map<std::pair<S128KeyType, u64>, std::string> KEYS_VARIABLE_LENGTH{
constexpr std::array<std::pair<std::string_view, KeyIndex<S128KeyType>>, 30> s128_file_id{{
{"eticket_rsa_kek", {S128KeyType::ETicketRSAKek, 0, 0}},
{"eticket_rsa_kek_source",
{S128KeyType::Source, static_cast<u64>(SourceKeyType::ETicketKek), 0}},
{"eticket_rsa_kekek_source",
{S128KeyType::Source, static_cast<u64>(SourceKeyType::ETicketKekek), 0}},
{"rsa_kek_mask_0", {S128KeyType::RSAKek, static_cast<u64>(RSAKekType::Mask0), 0}},
{"rsa_kek_seed_3", {S128KeyType::RSAKek, static_cast<u64>(RSAKekType::Seed3), 0}},
{"rsa_oaep_kek_generation_source",
{S128KeyType::Source, static_cast<u64>(SourceKeyType::RSAOaepKekGeneration), 0}},
{"sd_card_kek_source", {S128KeyType::Source, static_cast<u64>(SourceKeyType::SDKek), 0}},
{"aes_kek_generation_source",
{S128KeyType::Source, static_cast<u64>(SourceKeyType::AESKekGeneration), 0}},
{"aes_key_generation_source",
{S128KeyType::Source, static_cast<u64>(SourceKeyType::AESKeyGeneration), 0}},
{"package2_key_source", {S128KeyType::Source, static_cast<u64>(SourceKeyType::Package2), 0}},
{"master_key_source", {S128KeyType::Source, static_cast<u64>(SourceKeyType::Master), 0}},
{"header_kek_source", {S128KeyType::Source, static_cast<u64>(SourceKeyType::HeaderKek), 0}},
{"key_area_key_application_source",
{S128KeyType::Source, static_cast<u64>(SourceKeyType::KeyAreaKey),
static_cast<u64>(KeyAreaKeyType::Application)}},
{"key_area_key_ocean_source",
{S128KeyType::Source, static_cast<u64>(SourceKeyType::KeyAreaKey),
static_cast<u64>(KeyAreaKeyType::Ocean)}},
{"key_area_key_system_source",
{S128KeyType::Source, static_cast<u64>(SourceKeyType::KeyAreaKey),
static_cast<u64>(KeyAreaKeyType::System)}},
{"titlekek_source", {S128KeyType::Source, static_cast<u64>(SourceKeyType::Titlekek), 0}},
{"keyblob_mac_key_source",
{S128KeyType::Source, static_cast<u64>(SourceKeyType::KeyblobMAC), 0}},
{"tsec_key", {S128KeyType::TSEC, 0, 0}},
{"secure_boot_key", {S128KeyType::SecureBoot, 0, 0}},
{"sd_seed", {S128KeyType::SDSeed, 0, 0}},
{"bis_key_0_crypt", {S128KeyType::BIS, 0, static_cast<u64>(BISKeyType::Crypto)}},
{"bis_key_0_tweak", {S128KeyType::BIS, 0, static_cast<u64>(BISKeyType::Tweak)}},
{"bis_key_1_crypt", {S128KeyType::BIS, 1, static_cast<u64>(BISKeyType::Crypto)}},
{"bis_key_1_tweak", {S128KeyType::BIS, 1, static_cast<u64>(BISKeyType::Tweak)}},
{"bis_key_2_crypt", {S128KeyType::BIS, 2, static_cast<u64>(BISKeyType::Crypto)}},
{"bis_key_2_tweak", {S128KeyType::BIS, 2, static_cast<u64>(BISKeyType::Tweak)}},
{"bis_key_3_crypt", {S128KeyType::BIS, 3, static_cast<u64>(BISKeyType::Crypto)}},
{"bis_key_3_tweak", {S128KeyType::BIS, 3, static_cast<u64>(BISKeyType::Tweak)}},
{"header_kek", {S128KeyType::HeaderKek, 0, 0}},
{"sd_card_kek", {S128KeyType::SDKek, 0, 0}},
}};
auto Find128ByName(std::string_view name) {
return std::find_if(s128_file_id.begin(), s128_file_id.end(),
[&name](const auto& pair) { return pair.first == name; });
}
constexpr std::array<std::pair<std::string_view, KeyIndex<S256KeyType>>, 6> s256_file_id{{
{"header_key", {S256KeyType::Header, 0, 0}},
{"sd_card_save_key_source", {S256KeyType::SDKeySource, static_cast<u64>(SDKeyType::Save), 0}},
{"sd_card_nca_key_source", {S256KeyType::SDKeySource, static_cast<u64>(SDKeyType::NCA), 0}},
{"header_key_source", {S256KeyType::HeaderSource, 0, 0}},
{"sd_card_save_key", {S256KeyType::SDKey, static_cast<u64>(SDKeyType::Save), 0}},
{"sd_card_nca_key", {S256KeyType::SDKey, static_cast<u64>(SDKeyType::NCA), 0}},
}};
auto Find256ByName(std::string_view name) {
return std::find_if(s256_file_id.begin(), s256_file_id.end(),
[&name](const auto& pair) { return pair.first == name; });
}
using KeyArray = std::array<std::pair<std::pair<S128KeyType, u64>, std::string_view>, 7>;
constexpr KeyArray KEYS_VARIABLE_LENGTH{{
{{S128KeyType::Master, 0}, "master_key_"},
{{S128KeyType::Package1, 0}, "package1_key_"},
{{S128KeyType::Package2, 0}, "package2_key_"},
@@ -57,14 +123,13 @@ const std::map<std::pair<S128KeyType, u64>, std::string> KEYS_VARIABLE_LENGTH{
{{S128KeyType::Source, static_cast<u64>(SourceKeyType::Keyblob)}, "keyblob_key_source_"},
{{S128KeyType::Keyblob, 0}, "keyblob_key_"},
{{S128KeyType::KeyblobMAC, 0}, "keyblob_mac_key_"},
};
}};
namespace {
template <std::size_t Size>
bool IsAllZeroArray(const std::array<u8, Size>& array) {
return std::all_of(array.begin(), array.end(), [](const auto& elem) { return elem == 0; });
}
} // namespace
} // Anonymous namespace
u64 GetSignatureTypeDataSize(SignatureType type) {
switch (type) {
@@ -564,13 +629,13 @@ void KeyManager::LoadFromFile(const std::string& filename, bool is_title_keys) {
s128_keys[{S128KeyType::Titlekey, rights_id[1], rights_id[0]}] = key;
} else {
out[0] = Common::ToLower(out[0]);
if (s128_file_id.find(out[0]) != s128_file_id.end()) {
const auto index = s128_file_id.at(out[0]);
Key128 key = Common::HexStringToArray<16>(out[1]);
if (const auto iter128 = Find128ByName(out[0]); iter128 != s128_file_id.end()) {
const auto& index = iter128->second;
const Key128 key = Common::HexStringToArray<16>(out[1]);
s128_keys[{index.type, index.field1, index.field2}] = key;
} else if (s256_file_id.find(out[0]) != s256_file_id.end()) {
const auto index = s256_file_id.at(out[0]);
Key256 key = Common::HexStringToArray<32>(out[1]);
} else if (const auto iter256 = Find256ByName(out[0]); iter256 != s256_file_id.end()) {
const auto& index = iter256->second;
const Key256 key = Common::HexStringToArray<32>(out[1]);
s256_keys[{index.type, index.field1, index.field2}] = key;
} else if (out[0].compare(0, 8, "keyblob_") == 0 &&
out[0].compare(0, 9, "keyblob_k") != 0) {
@@ -742,8 +807,7 @@ void KeyManager::SetKey(S128KeyType id, Key128 key, u64 field1, u64 field2) {
}
const auto iter2 = std::find_if(
s128_file_id.begin(), s128_file_id.end(),
[&id, &field1, &field2](const std::pair<std::string, KeyIndex<S128KeyType>> elem) {
s128_file_id.begin(), s128_file_id.end(), [&id, &field1, &field2](const auto& elem) {
return std::tie(elem.second.type, elem.second.field1, elem.second.field2) ==
std::tie(id, field1, field2);
});
@@ -753,9 +817,11 @@ void KeyManager::SetKey(S128KeyType id, Key128 key, u64 field1, u64 field2) {
// Variable cases
if (id == S128KeyType::KeyArea) {
static constexpr std::array<const char*, 3> kak_names = {"key_area_key_application_{:02X}",
"key_area_key_ocean_{:02X}",
"key_area_key_system_{:02X}"};
static constexpr std::array<const char*, 3> kak_names = {
"key_area_key_application_{:02X}",
"key_area_key_ocean_{:02X}",
"key_area_key_system_{:02X}",
};
WriteKeyToFile(category, fmt::format(kak_names.at(field2), field1), key);
} else if (id == S128KeyType::Master) {
WriteKeyToFile(category, fmt::format("master_key_{:02X}", field1), key);
@@ -781,8 +847,7 @@ void KeyManager::SetKey(S256KeyType id, Key256 key, u64 field1, u64 field2) {
return;
}
const auto iter = std::find_if(
s256_file_id.begin(), s256_file_id.end(),
[&id, &field1, &field2](const std::pair<std::string, KeyIndex<S256KeyType>> elem) {
s256_file_id.begin(), s256_file_id.end(), [&id, &field1, &field2](const auto& elem) {
return std::tie(elem.second.type, elem.second.field1, elem.second.field2) ==
std::tie(id, field1, field2);
});
@@ -1245,58 +1310,4 @@ bool KeyManager::AddTicketPersonalized(Ticket raw) {
SetKey(S128KeyType::Titlekey, key, rights_id[1], rights_id[0]);
return true;
}
const boost::container::flat_map<std::string, KeyIndex<S128KeyType>> KeyManager::s128_file_id = {
{"eticket_rsa_kek", {S128KeyType::ETicketRSAKek, 0, 0}},
{"eticket_rsa_kek_source",
{S128KeyType::Source, static_cast<u64>(SourceKeyType::ETicketKek), 0}},
{"eticket_rsa_kekek_source",
{S128KeyType::Source, static_cast<u64>(SourceKeyType::ETicketKekek), 0}},
{"rsa_kek_mask_0", {S128KeyType::RSAKek, static_cast<u64>(RSAKekType::Mask0), 0}},
{"rsa_kek_seed_3", {S128KeyType::RSAKek, static_cast<u64>(RSAKekType::Seed3), 0}},
{"rsa_oaep_kek_generation_source",
{S128KeyType::Source, static_cast<u64>(SourceKeyType::RSAOaepKekGeneration), 0}},
{"sd_card_kek_source", {S128KeyType::Source, static_cast<u64>(SourceKeyType::SDKek), 0}},
{"aes_kek_generation_source",
{S128KeyType::Source, static_cast<u64>(SourceKeyType::AESKekGeneration), 0}},
{"aes_key_generation_source",
{S128KeyType::Source, static_cast<u64>(SourceKeyType::AESKeyGeneration), 0}},
{"package2_key_source", {S128KeyType::Source, static_cast<u64>(SourceKeyType::Package2), 0}},
{"master_key_source", {S128KeyType::Source, static_cast<u64>(SourceKeyType::Master), 0}},
{"header_kek_source", {S128KeyType::Source, static_cast<u64>(SourceKeyType::HeaderKek), 0}},
{"key_area_key_application_source",
{S128KeyType::Source, static_cast<u64>(SourceKeyType::KeyAreaKey),
static_cast<u64>(KeyAreaKeyType::Application)}},
{"key_area_key_ocean_source",
{S128KeyType::Source, static_cast<u64>(SourceKeyType::KeyAreaKey),
static_cast<u64>(KeyAreaKeyType::Ocean)}},
{"key_area_key_system_source",
{S128KeyType::Source, static_cast<u64>(SourceKeyType::KeyAreaKey),
static_cast<u64>(KeyAreaKeyType::System)}},
{"titlekek_source", {S128KeyType::Source, static_cast<u64>(SourceKeyType::Titlekek), 0}},
{"keyblob_mac_key_source",
{S128KeyType::Source, static_cast<u64>(SourceKeyType::KeyblobMAC), 0}},
{"tsec_key", {S128KeyType::TSEC, 0, 0}},
{"secure_boot_key", {S128KeyType::SecureBoot, 0, 0}},
{"sd_seed", {S128KeyType::SDSeed, 0, 0}},
{"bis_key_0_crypt", {S128KeyType::BIS, 0, static_cast<u64>(BISKeyType::Crypto)}},
{"bis_key_0_tweak", {S128KeyType::BIS, 0, static_cast<u64>(BISKeyType::Tweak)}},
{"bis_key_1_crypt", {S128KeyType::BIS, 1, static_cast<u64>(BISKeyType::Crypto)}},
{"bis_key_1_tweak", {S128KeyType::BIS, 1, static_cast<u64>(BISKeyType::Tweak)}},
{"bis_key_2_crypt", {S128KeyType::BIS, 2, static_cast<u64>(BISKeyType::Crypto)}},
{"bis_key_2_tweak", {S128KeyType::BIS, 2, static_cast<u64>(BISKeyType::Tweak)}},
{"bis_key_3_crypt", {S128KeyType::BIS, 3, static_cast<u64>(BISKeyType::Crypto)}},
{"bis_key_3_tweak", {S128KeyType::BIS, 3, static_cast<u64>(BISKeyType::Tweak)}},
{"header_kek", {S128KeyType::HeaderKek, 0, 0}},
{"sd_card_kek", {S128KeyType::SDKek, 0, 0}},
};
const boost::container::flat_map<std::string, KeyIndex<S256KeyType>> KeyManager::s256_file_id = {
{"header_key", {S256KeyType::Header, 0, 0}},
{"sd_card_save_key_source", {S256KeyType::SDKeySource, static_cast<u64>(SDKeyType::Save), 0}},
{"sd_card_nca_key_source", {S256KeyType::SDKeySource, static_cast<u64>(SDKeyType::NCA), 0}},
{"header_key_source", {S256KeyType::HeaderSource, 0, 0}},
{"sd_card_save_key", {S256KeyType::SDKey, static_cast<u64>(SDKeyType::Save), 0}},
{"sd_card_nca_key", {S256KeyType::SDKey, static_cast<u64>(SDKeyType::NCA), 0}},
};
} // namespace Core::Crypto

View File

@@ -10,7 +10,6 @@
#include <string>
#include <variant>
#include <boost/container/flat_map.hpp>
#include <fmt/format.h>
#include "common/common_funcs.h"
#include "common/common_types.h"
@@ -293,9 +292,6 @@ private:
void SetKeyWrapped(S128KeyType id, Key128 key, u64 field1 = 0, u64 field2 = 0);
void SetKeyWrapped(S256KeyType id, Key256 key, u64 field1 = 0, u64 field2 = 0);
static const boost::container::flat_map<std::string, KeyIndex<S128KeyType>> s128_file_id;
static const boost::container::flat_map<std::string, KeyIndex<S256KeyType>> s256_file_id;
};
Key128 GenerateKeyEncryptionKey(Key128 source, Key128 master, Key128 kek_seed, Key128 key_seed);

View File

@@ -257,8 +257,7 @@ std::vector<NcaID> PlaceholderCache::List() const {
for (const auto& sdir : dir->GetSubdirectories()) {
for (const auto& file : sdir->GetFiles()) {
const auto name = file->GetName();
if (name.length() == 36 && name[32] == '.' && name[33] == 'n' && name[34] == 'c' &&
name[35] == 'a') {
if (name.length() == 36 && name.ends_with(".nca")) {
out.push_back(Common::HexStringToArray<0x10>(name.substr(0, 32)));
}
}
@@ -621,25 +620,25 @@ InstallResult RegisteredCache::InstallEntry(const NSP& nsp, bool overwrite_if_ex
InstallResult RegisteredCache::InstallEntry(const NCA& nca, TitleType type,
bool overwrite_if_exists, const VfsCopyFunction& copy) {
CNMTHeader header{
nca.GetTitleId(), // Title ID
0, // Ignore/Default title version
type, // Type
{}, // Padding
0x10, // Default table offset
1, // 1 Content Entry
0, // No Meta Entries
{}, // Padding
{}, // Reserved 1
0, // Is committed
0, // Required download system version
{}, // Reserved 2
const CNMTHeader header{
.title_id = nca.GetTitleId(),
.title_version = 0,
.type = type,
.reserved = {},
.table_offset = 0x10,
.number_content_entries = 1,
.number_meta_entries = 0,
.attributes = 0,
.reserved2 = {},
.is_committed = 0,
.required_download_system_version = 0,
.reserved3 = {},
};
OptionalHeader opt_header{0, 0};
const OptionalHeader opt_header{0, 0};
ContentRecord c_rec{{}, {}, {}, GetCRTypeFromNCAType(nca.GetType()), {}};
const auto& data = nca.GetBaseFile()->ReadBytes(0x100000);
mbedtls_sha256_ret(data.data(), data.size(), c_rec.hash.data(), 0);
memcpy(&c_rec.nca_id, &c_rec.hash, 16);
std::memcpy(&c_rec.nca_id, &c_rec.hash, 16);
const CNMT new_cnmt(header, opt_header, {c_rec}, {});
if (!RawInstallYuzuMeta(new_cnmt)) {
return InstallResult::ErrorMetaFailed;

View File

@@ -72,8 +72,10 @@ VfsEntryType RealVfsFilesystem::GetEntryType(std::string_view path_) const {
VirtualFile RealVfsFilesystem::OpenFile(std::string_view path_, Mode perms) {
const auto path = FS::SanitizePath(path_, FS::DirectorySeparator::PlatformDefault);
if (cache.find(path) != cache.end()) {
auto weak = cache[path];
if (const auto weak_iter = cache.find(path); weak_iter != cache.cend()) {
const auto& weak = weak_iter->second;
if (!weak.expired()) {
return std::shared_ptr<RealVfsFile>(new RealVfsFile(*this, weak.lock(), path, perms));
}
@@ -84,7 +86,7 @@ VirtualFile RealVfsFilesystem::OpenFile(std::string_view path_, Mode perms) {
}
auto backing = std::make_shared<FS::IOFile>(path, ModeFlagsToString(perms).c_str());
cache[path] = backing;
cache.insert_or_assign(path, backing);
// Cannot use make_shared as RealVfsFile constructor is private
return std::shared_ptr<RealVfsFile>(new RealVfsFile(*this, backing, path, perms));
@@ -116,11 +118,12 @@ VirtualFile RealVfsFilesystem::CopyFile(std::string_view old_path_, std::string_
VirtualFile RealVfsFilesystem::MoveFile(std::string_view old_path_, std::string_view new_path_) {
const auto old_path = FS::SanitizePath(old_path_, FS::DirectorySeparator::PlatformDefault);
const auto new_path = FS::SanitizePath(new_path_, FS::DirectorySeparator::PlatformDefault);
const auto cached_file_iter = cache.find(old_path);
if (cache.find(old_path) != cache.end()) {
auto file = cache[old_path].lock();
if (cached_file_iter != cache.cend()) {
auto file = cached_file_iter->second.lock();
if (!cache[old_path].expired()) {
if (!cached_file_iter->second.expired()) {
file->Close();
}
@@ -131,7 +134,7 @@ VirtualFile RealVfsFilesystem::MoveFile(std::string_view old_path_, std::string_
cache.erase(old_path);
file->Open(new_path, "r+b");
cache[new_path] = file;
cache.insert_or_assign(new_path, std::move(file));
} else {
UNREACHABLE();
return nullptr;
@@ -142,12 +145,15 @@ VirtualFile RealVfsFilesystem::MoveFile(std::string_view old_path_, std::string_
bool RealVfsFilesystem::DeleteFile(std::string_view path_) {
const auto path = FS::SanitizePath(path_, FS::DirectorySeparator::PlatformDefault);
if (cache.find(path) != cache.end()) {
if (!cache[path].expired()) {
cache[path].lock()->Close();
const auto cached_iter = cache.find(path);
if (cached_iter != cache.cend()) {
if (!cached_iter->second.expired()) {
cached_iter->second.lock()->Close();
}
cache.erase(path);
}
return FS::Delete(path);
}
@@ -192,21 +198,25 @@ VirtualDir RealVfsFilesystem::MoveDirectory(std::string_view old_path_,
}
for (auto& kv : cache) {
// Path in cache starts with old_path
if (kv.first.rfind(old_path, 0) == 0) {
const auto file_old_path =
FS::SanitizePath(kv.first, FS::DirectorySeparator::PlatformDefault);
const auto file_new_path =
FS::SanitizePath(new_path + DIR_SEP + kv.first.substr(old_path.size()),
FS::DirectorySeparator::PlatformDefault);
auto cached = cache[file_old_path];
if (!cached.expired()) {
auto file = cached.lock();
file->Open(file_new_path, "r+b");
cache.erase(file_old_path);
cache[file_new_path] = file;
}
// If the path in the cache doesn't start with old_path, then bail on this file.
if (kv.first.rfind(old_path, 0) != 0) {
continue;
}
const auto file_old_path =
FS::SanitizePath(kv.first, FS::DirectorySeparator::PlatformDefault);
auto file_new_path = FS::SanitizePath(new_path + DIR_SEP + kv.first.substr(old_path.size()),
FS::DirectorySeparator::PlatformDefault);
const auto& cached = cache[file_old_path];
if (cached.expired()) {
continue;
}
auto file = cached.lock();
file->Open(file_new_path, "r+b");
cache.erase(file_old_path);
cache.insert_or_assign(std::move(file_new_path), std::move(file));
}
return OpenDirectory(new_path, Mode::ReadWrite);
@@ -214,15 +224,21 @@ VirtualDir RealVfsFilesystem::MoveDirectory(std::string_view old_path_,
bool RealVfsFilesystem::DeleteDirectory(std::string_view path_) {
const auto path = FS::SanitizePath(path_, FS::DirectorySeparator::PlatformDefault);
for (auto& kv : cache) {
// Path in cache starts with old_path
if (kv.first.rfind(path, 0) == 0) {
if (!cache[kv.first].expired()) {
cache[kv.first].lock()->Close();
}
cache.erase(kv.first);
// If the path in the cache doesn't start with path, then bail on this file.
if (kv.first.rfind(path, 0) != 0) {
continue;
}
const auto& entry = cache[kv.first];
if (!entry.expired()) {
entry.lock()->Close();
}
cache.erase(kv.first);
}
return FS::DeleteDirRecursively(path);
}
@@ -260,14 +276,14 @@ bool RealVfsFile::IsReadable() const {
}
std::size_t RealVfsFile::Read(u8* data, std::size_t length, std::size_t offset) const {
if (!backing->Seek(offset, SEEK_SET)) {
if (!backing->Seek(static_cast<s64>(offset), SEEK_SET)) {
return 0;
}
return backing->ReadBytes(data, length);
}
std::size_t RealVfsFile::Write(const u8* data, std::size_t length, std::size_t offset) {
if (!backing->Seek(offset, SEEK_SET)) {
if (!backing->Seek(static_cast<s64>(offset), SEEK_SET)) {
return 0;
}
return backing->WriteBytes(data, length);

View File

@@ -359,7 +359,8 @@ enum class GPUAccuracy : u32 {
enum class CPUAccuracy {
Accurate = 0,
DebugMode = 1,
Unsafe = 1,
DebugMode = 2,
};
extern bool configuring_global;
@@ -419,6 +420,9 @@ struct Values {
bool cpuopt_misc_ir;
bool cpuopt_reduce_misalign_checks;
bool cpuopt_unsafe_unfuse_fma;
bool cpuopt_unsafe_reduce_fp_error;
// Renderer
Setting<RendererBackend> renderer_backend;
bool renderer_debug;

View File

@@ -38,7 +38,8 @@ public:
explicit GCAxisButton(int port_, int axis_, float threshold_, bool trigger_if_greater_,
GCAdapter::Adapter* adapter)
: port(port_), axis(axis_), threshold(threshold_), trigger_if_greater(trigger_if_greater_),
gcadapter(adapter), origin_value(adapter->GetOriginValue(port_, axis_)) {}
gcadapter(adapter),
origin_value(static_cast<float>(adapter->GetOriginValue(port_, axis_))) {}
bool GetStatus() const override {
if (gcadapter->DeviceConnected(port)) {
@@ -151,8 +152,9 @@ public:
GCAnalog(int port_, int axis_x_, int axis_y_, float deadzone_, GCAdapter::Adapter* adapter,
float range_)
: port(port_), axis_x(axis_x_), axis_y(axis_y_), deadzone(deadzone_), gcadapter(adapter),
origin_value_x(adapter->GetOriginValue(port_, axis_x_)),
origin_value_y(adapter->GetOriginValue(port_, axis_y_)), range(range_) {}
origin_value_x(static_cast<float>(adapter->GetOriginValue(port_, axis_x_))),
origin_value_y(static_cast<float>(adapter->GetOriginValue(port_, axis_y_))),
range(range_) {}
float GetAxis(int axis) const {
if (gcadapter->DeviceConnected(port)) {

View File

@@ -10,7 +10,13 @@
namespace Tegra::Engines {
Fermi2D::Fermi2D(VideoCore::RasterizerInterface& rasterizer) : rasterizer{rasterizer} {}
Fermi2D::Fermi2D() = default;
Fermi2D::~Fermi2D() = default;
void Fermi2D::BindRasterizer(VideoCore::RasterizerInterface& rasterizer_) {
rasterizer = &rasterizer_;
}
void Fermi2D::CallMethod(u32 method, u32 method_argument, bool is_last_call) {
ASSERT_MSG(method < Regs::NUM_REGS,
@@ -87,7 +93,7 @@ void Fermi2D::HandleSurfaceCopy() {
copy_config.src_rect = src_rect;
copy_config.dst_rect = dst_rect;
if (!rasterizer.AccelerateSurfaceCopy(regs.src, regs.dst, copy_config)) {
if (!rasterizer->AccelerateSurfaceCopy(regs.src, regs.dst, copy_config)) {
UNIMPLEMENTED();
}
}

View File

@@ -34,8 +34,11 @@ namespace Tegra::Engines {
class Fermi2D final : public EngineInterface {
public:
explicit Fermi2D(VideoCore::RasterizerInterface& rasterizer);
~Fermi2D() = default;
explicit Fermi2D();
~Fermi2D();
/// Binds a rasterizer to this engine.
void BindRasterizer(VideoCore::RasterizerInterface& rasterizer);
/// Write the value to the register identified by method.
void CallMethod(u32 method, u32 method_argument, bool is_last_call) override;
@@ -149,7 +152,7 @@ public:
};
private:
VideoCore::RasterizerInterface& rasterizer;
VideoCore::RasterizerInterface* rasterizer;
/// Performs the copy from the source surface to the destination surface as configured in the
/// registers.

View File

@@ -16,14 +16,15 @@
namespace Tegra::Engines {
KeplerCompute::KeplerCompute(Core::System& system, VideoCore::RasterizerInterface& rasterizer,
MemoryManager& memory_manager)
: system{system}, rasterizer{rasterizer}, memory_manager{memory_manager}, upload_state{
memory_manager,
regs.upload} {}
KeplerCompute::KeplerCompute(Core::System& system_, MemoryManager& memory_manager_)
: system{system_}, memory_manager{memory_manager_}, upload_state{memory_manager, regs.upload} {}
KeplerCompute::~KeplerCompute() = default;
void KeplerCompute::BindRasterizer(VideoCore::RasterizerInterface& rasterizer_) {
rasterizer = &rasterizer_;
}
void KeplerCompute::CallMethod(u32 method, u32 method_argument, bool is_last_call) {
ASSERT_MSG(method < Regs::NUM_REGS,
"Invalid KeplerCompute register, increase the size of the Regs structure");
@@ -104,11 +105,11 @@ SamplerDescriptor KeplerCompute::AccessSampler(u32 handle) const {
}
VideoCore::GuestDriverProfile& KeplerCompute::AccessGuestDriverProfile() {
return rasterizer.AccessGuestDriverProfile();
return rasterizer->AccessGuestDriverProfile();
}
const VideoCore::GuestDriverProfile& KeplerCompute::AccessGuestDriverProfile() const {
return rasterizer.AccessGuestDriverProfile();
return rasterizer->AccessGuestDriverProfile();
}
void KeplerCompute::ProcessLaunch() {
@@ -119,7 +120,7 @@ void KeplerCompute::ProcessLaunch() {
const GPUVAddr code_addr = regs.code_loc.Address() + launch_description.program_start;
LOG_TRACE(HW_GPU, "Compute invocation launched at address 0x{:016x}", code_addr);
rasterizer.DispatchCompute(code_addr);
rasterizer->DispatchCompute(code_addr);
}
Texture::TICEntry KeplerCompute::GetTICEntry(u32 tic_index) const {

View File

@@ -42,10 +42,12 @@ namespace Tegra::Engines {
class KeplerCompute final : public ConstBufferEngineInterface, public EngineInterface {
public:
explicit KeplerCompute(Core::System& system, VideoCore::RasterizerInterface& rasterizer,
MemoryManager& memory_manager);
explicit KeplerCompute(Core::System& system, MemoryManager& memory_manager);
~KeplerCompute();
/// Binds a rasterizer to this engine.
void BindRasterizer(VideoCore::RasterizerInterface& rasterizer);
static constexpr std::size_t NumConstBuffers = 8;
struct Regs {
@@ -230,11 +232,6 @@ public:
const VideoCore::GuestDriverProfile& AccessGuestDriverProfile() const override;
private:
Core::System& system;
VideoCore::RasterizerInterface& rasterizer;
MemoryManager& memory_manager;
Upload::State upload_state;
void ProcessLaunch();
/// Retrieves information about a specific TIC entry from the TIC buffer.
@@ -242,6 +239,11 @@ private:
/// Retrieves information about a specific TSC entry from the TSC buffer.
Texture::TSCEntry GetTSCEntry(u32 tsc_index) const;
Core::System& system;
MemoryManager& memory_manager;
VideoCore::RasterizerInterface* rasterizer = nullptr;
Upload::State upload_state;
};
#define ASSERT_REG_POSITION(field_name, position) \

View File

@@ -22,14 +22,19 @@ using VideoCore::QueryType;
/// First register id that is actually a Macro call.
constexpr u32 MacroRegistersStart = 0xE00;
Maxwell3D::Maxwell3D(Core::System& system, VideoCore::RasterizerInterface& rasterizer,
MemoryManager& memory_manager)
: system{system}, rasterizer{rasterizer}, memory_manager{memory_manager},
macro_engine{GetMacroEngine(*this)}, upload_state{memory_manager, regs.upload} {
Maxwell3D::Maxwell3D(Core::System& system_, MemoryManager& memory_manager_)
: system{system_}, memory_manager{memory_manager_}, macro_engine{GetMacroEngine(*this)},
upload_state{memory_manager, regs.upload} {
dirty.flags.flip();
InitializeRegisterDefaults();
}
Maxwell3D::~Maxwell3D() = default;
void Maxwell3D::BindRasterizer(VideoCore::RasterizerInterface& rasterizer_) {
rasterizer = &rasterizer_;
}
void Maxwell3D::InitializeRegisterDefaults() {
// Initializes registers to their default values - what games expect them to be at boot. This is
// for certain registers that may not be explicitly set by games.
@@ -192,7 +197,7 @@ void Maxwell3D::CallMethod(u32 method, u32 method_argument, bool is_last_call) {
switch (method) {
case MAXWELL3D_REG_INDEX(wait_for_idle): {
rasterizer.WaitForIdle();
rasterizer->WaitForIdle();
break;
}
case MAXWELL3D_REG_INDEX(shadow_ram_control): {
@@ -402,7 +407,7 @@ void Maxwell3D::FlushMMEInlineDraw() {
const bool is_indexed = mme_draw.current_mode == MMEDrawMode::Indexed;
if (ShouldExecute()) {
rasterizer.Draw(is_indexed, true);
rasterizer->Draw(is_indexed, true);
}
// TODO(bunnei): Below, we reset vertex count so that we can use these registers to determine if
@@ -465,7 +470,7 @@ void Maxwell3D::ProcessQueryGet() {
switch (regs.query.query_get.operation) {
case Regs::QueryOperation::Release:
if (regs.query.query_get.fence == 1) {
rasterizer.SignalSemaphore(regs.query.QueryAddress(), regs.query.query_sequence);
rasterizer->SignalSemaphore(regs.query.QueryAddress(), regs.query.query_sequence);
} else {
StampQueryResult(regs.query.query_sequence, regs.query.query_get.short_query == 0);
}
@@ -533,7 +538,7 @@ void Maxwell3D::ProcessQueryCondition() {
void Maxwell3D::ProcessCounterReset() {
switch (regs.counter_reset) {
case Regs::CounterReset::SampleCnt:
rasterizer.ResetCounter(QueryType::SamplesPassed);
rasterizer->ResetCounter(QueryType::SamplesPassed);
break;
default:
LOG_DEBUG(Render_OpenGL, "Unimplemented counter reset={}",
@@ -547,7 +552,7 @@ void Maxwell3D::ProcessSyncPoint() {
const u32 increment = regs.sync_info.increment.Value();
[[maybe_unused]] const u32 cache_flush = regs.sync_info.unknown.Value();
if (increment) {
rasterizer.SignalSyncPoint(sync_point);
rasterizer->SignalSyncPoint(sync_point);
}
}
@@ -570,7 +575,7 @@ void Maxwell3D::DrawArrays() {
const bool is_indexed{regs.index_array.count && !regs.vertex_buffer.count};
if (ShouldExecute()) {
rasterizer.Draw(is_indexed, false);
rasterizer->Draw(is_indexed, false);
}
// TODO(bunnei): Below, we reset vertex count so that we can use these registers to determine if
@@ -590,8 +595,8 @@ std::optional<u64> Maxwell3D::GetQueryResult() {
return 0;
case Regs::QuerySelect::SamplesPassed:
// Deferred.
rasterizer.Query(regs.query.QueryAddress(), VideoCore::QueryType::SamplesPassed,
system.GPU().GetTicks());
rasterizer->Query(regs.query.QueryAddress(), VideoCore::QueryType::SamplesPassed,
system.GPU().GetTicks());
return {};
default:
LOG_DEBUG(HW_GPU, "Unimplemented query select type {}",
@@ -718,7 +723,7 @@ void Maxwell3D::ProcessClearBuffers() {
regs.clear_buffers.R == regs.clear_buffers.B &&
regs.clear_buffers.R == regs.clear_buffers.A);
rasterizer.Clear();
rasterizer->Clear();
}
u32 Maxwell3D::AccessConstBuffer32(ShaderType stage, u64 const_buffer, u64 offset) const {
@@ -752,11 +757,11 @@ SamplerDescriptor Maxwell3D::AccessSampler(u32 handle) const {
}
VideoCore::GuestDriverProfile& Maxwell3D::AccessGuestDriverProfile() {
return rasterizer.AccessGuestDriverProfile();
return rasterizer->AccessGuestDriverProfile();
}
const VideoCore::GuestDriverProfile& Maxwell3D::AccessGuestDriverProfile() const {
return rasterizer.AccessGuestDriverProfile();
return rasterizer->AccessGuestDriverProfile();
}
} // namespace Tegra::Engines

View File

@@ -51,9 +51,11 @@ namespace Tegra::Engines {
class Maxwell3D final : public ConstBufferEngineInterface, public EngineInterface {
public:
explicit Maxwell3D(Core::System& system, VideoCore::RasterizerInterface& rasterizer,
MemoryManager& memory_manager);
~Maxwell3D() = default;
explicit Maxwell3D(Core::System& system, MemoryManager& memory_manager);
~Maxwell3D();
/// Binds a rasterizer to this engine.
void BindRasterizer(VideoCore::RasterizerInterface& rasterizer);
/// Register structure of the Maxwell3D engine.
/// TODO(Subv): This structure will need to be made bigger as more registers are discovered.
@@ -1418,12 +1420,12 @@ public:
return execute_on;
}
VideoCore::RasterizerInterface& GetRasterizer() {
return rasterizer;
VideoCore::RasterizerInterface& Rasterizer() {
return *rasterizer;
}
const VideoCore::RasterizerInterface& GetRasterizer() const {
return rasterizer;
const VideoCore::RasterizerInterface& Rasterizer() const {
return *rasterizer;
}
/// Notify a memory write has happened.
@@ -1460,11 +1462,10 @@ private:
void InitializeRegisterDefaults();
Core::System& system;
VideoCore::RasterizerInterface& rasterizer;
MemoryManager& memory_manager;
VideoCore::RasterizerInterface* rasterizer = nullptr;
/// Start offsets of each macro in macro_memory
std::array<u32, 0x80> macro_positions = {};

View File

@@ -5,15 +5,10 @@
#pragma once
#include <algorithm>
#include <array>
#include <memory>
#include <queue>
#include "common/assert.h"
#include "common/common_types.h"
#include "core/core.h"
#include "core/memory.h"
#include "core/settings.h"
#include "video_core/gpu.h"
#include "video_core/memory_manager.h"
#include "video_core/rasterizer_interface.h"

View File

@@ -27,21 +27,28 @@ namespace Tegra {
MICROPROFILE_DEFINE(GPU_wait, "GPU", "Wait for the GPU", MP_RGB(128, 128, 192));
GPU::GPU(Core::System& system, std::unique_ptr<VideoCore::RendererBase>&& renderer_, bool is_async)
: system{system}, renderer{std::move(renderer_)}, is_async{is_async} {
auto& rasterizer{renderer->Rasterizer()};
memory_manager = std::make_unique<Tegra::MemoryManager>(system, rasterizer);
dma_pusher = std::make_unique<Tegra::DmaPusher>(system, *this);
maxwell_3d = std::make_unique<Engines::Maxwell3D>(system, rasterizer, *memory_manager);
fermi_2d = std::make_unique<Engines::Fermi2D>(rasterizer);
kepler_compute = std::make_unique<Engines::KeplerCompute>(system, rasterizer, *memory_manager);
maxwell_dma = std::make_unique<Engines::MaxwellDMA>(system, *memory_manager);
kepler_memory = std::make_unique<Engines::KeplerMemory>(system, *memory_manager);
shader_notify = std::make_unique<VideoCore::ShaderNotify>();
}
GPU::GPU(Core::System& system_, bool is_async_)
: system{system_}, dma_pusher{std::make_unique<Tegra::DmaPusher>(system, *this)},
memory_manager{std::make_unique<Tegra::MemoryManager>(system)},
maxwell_3d{std::make_unique<Engines::Maxwell3D>(system, *memory_manager)},
fermi_2d{std::make_unique<Engines::Fermi2D>()},
kepler_compute{std::make_unique<Engines::KeplerCompute>(system, *memory_manager)},
maxwell_dma{std::make_unique<Engines::MaxwellDMA>(system, *memory_manager)},
kepler_memory{std::make_unique<Engines::KeplerMemory>(system, *memory_manager)},
shader_notify{std::make_unique<VideoCore::ShaderNotify>()}, is_async{is_async_} {}
GPU::~GPU() = default;
void GPU::BindRenderer(std::unique_ptr<VideoCore::RendererBase> renderer_) {
renderer = std::move(renderer_);
VideoCore::RasterizerInterface& rasterizer = renderer->Rasterizer();
memory_manager->BindRasterizer(rasterizer);
maxwell_3d->BindRasterizer(rasterizer);
fermi_2d->BindRasterizer(rasterizer);
kepler_compute->BindRasterizer(rasterizer);
}
Engines::Maxwell3D& GPU::Maxwell3D() {
return *maxwell_3d;
}

View File

@@ -142,11 +142,6 @@ class MemoryManager;
class GPU {
public:
explicit GPU(Core::System& system, std::unique_ptr<VideoCore::RendererBase>&& renderer,
bool is_async);
virtual ~GPU();
struct MethodCall {
u32 method{};
u32 argument{};
@@ -162,6 +157,12 @@ public:
method_count(method_count) {}
};
explicit GPU(Core::System& system, bool is_async);
virtual ~GPU();
/// Binds a renderer to the GPU.
void BindRenderer(std::unique_ptr<VideoCore::RendererBase> renderer);
/// Calls a GPU method.
void CallMethod(const MethodCall& method_call);
@@ -345,8 +346,8 @@ private:
bool ExecuteMethodOnEngine(u32 method);
protected:
std::unique_ptr<Tegra::DmaPusher> dma_pusher;
Core::System& system;
std::unique_ptr<Tegra::DmaPusher> dma_pusher;
std::unique_ptr<VideoCore::RendererBase> renderer;
private:

View File

@@ -10,16 +10,14 @@
namespace VideoCommon {
GPUAsynch::GPUAsynch(Core::System& system, std::unique_ptr<VideoCore::RendererBase>&& renderer_,
std::unique_ptr<Core::Frontend::GraphicsContext>&& context)
: GPU(system, std::move(renderer_), true), gpu_thread{system},
cpu_context(renderer->GetRenderWindow().CreateSharedContext()),
gpu_context(std::move(context)) {}
GPUAsynch::GPUAsynch(Core::System& system) : GPU{system, true}, gpu_thread{system} {}
GPUAsynch::~GPUAsynch() = default;
void GPUAsynch::Start() {
gpu_thread.StartThread(*renderer, *gpu_context, *dma_pusher);
gpu_thread.StartThread(*renderer, renderer->Context(), *dma_pusher);
cpu_context = renderer->GetRenderWindow().CreateSharedContext();
cpu_context->MakeCurrent();
}
void GPUAsynch::ObtainContext() {

View File

@@ -20,8 +20,7 @@ namespace VideoCommon {
/// Implementation of GPU interface that runs the GPU asynchronously
class GPUAsynch final : public Tegra::GPU {
public:
explicit GPUAsynch(Core::System& system, std::unique_ptr<VideoCore::RendererBase>&& renderer,
std::unique_ptr<Core::Frontend::GraphicsContext>&& context);
explicit GPUAsynch(Core::System& system);
~GPUAsynch() override;
void Start() override;
@@ -42,7 +41,6 @@ protected:
private:
GPUThread::ThreadManager gpu_thread;
std::unique_ptr<Core::Frontend::GraphicsContext> cpu_context;
std::unique_ptr<Core::Frontend::GraphicsContext> gpu_context;
};
} // namespace VideoCommon

View File

@@ -7,20 +7,18 @@
namespace VideoCommon {
GPUSynch::GPUSynch(Core::System& system, std::unique_ptr<VideoCore::RendererBase>&& renderer,
std::unique_ptr<Core::Frontend::GraphicsContext>&& context)
: GPU(system, std::move(renderer), false), context{std::move(context)} {}
GPUSynch::GPUSynch(Core::System& system) : GPU{system, false} {}
GPUSynch::~GPUSynch() = default;
void GPUSynch::Start() {}
void GPUSynch::ObtainContext() {
context->MakeCurrent();
renderer->Context().MakeCurrent();
}
void GPUSynch::ReleaseContext() {
context->DoneCurrent();
renderer->Context().DoneCurrent();
}
void GPUSynch::PushGPUEntries(Tegra::CommandList&& entries) {

View File

@@ -19,8 +19,7 @@ namespace VideoCommon {
/// Implementation of GPU interface that runs the GPU synchronously
class GPUSynch final : public Tegra::GPU {
public:
explicit GPUSynch(Core::System& system, std::unique_ptr<VideoCore::RendererBase>&& renderer,
std::unique_ptr<Core::Frontend::GraphicsContext>&& context);
explicit GPUSynch(Core::System& system);
~GPUSynch() override;
void Start() override;
@@ -36,9 +35,6 @@ public:
protected:
void TriggerCpuInterrupt([[maybe_unused]] u32 syncpoint_id,
[[maybe_unused]] u32 value) const override {}
private:
std::unique_ptr<Core::Frontend::GraphicsContext> context;
};
} // namespace VideoCommon

View File

@@ -24,7 +24,7 @@ void HLE_771BB18C62444DA0(Engines::Maxwell3D& maxwell3d, const std::vector<u32>&
maxwell3d.regs.index_array.first = parameters[4];
if (maxwell3d.ShouldExecute()) {
maxwell3d.GetRasterizer().Draw(true, true);
maxwell3d.Rasterizer().Draw(true, true);
}
maxwell3d.regs.index_array.count = 0;
maxwell3d.mme_draw.instance_count = 0;
@@ -42,7 +42,7 @@ void HLE_0D61FC9FAAC9FCAD(Engines::Maxwell3D& maxwell3d, const std::vector<u32>&
maxwell3d.mme_draw.instance_count = count;
if (maxwell3d.ShouldExecute()) {
maxwell3d.GetRasterizer().Draw(false, true);
maxwell3d.Rasterizer().Draw(false, true);
}
maxwell3d.regs.vertex_buffer.count = 0;
maxwell3d.mme_draw.instance_count = 0;
@@ -65,7 +65,7 @@ void HLE_0217920100488FF7(Engines::Maxwell3D& maxwell3d, const std::vector<u32>&
maxwell3d.regs.draw.topology.Assign(
static_cast<Tegra::Engines::Maxwell3D::Regs::PrimitiveTopology>(parameters[0]));
if (maxwell3d.ShouldExecute()) {
maxwell3d.GetRasterizer().Draw(true, true);
maxwell3d.Rasterizer().Draw(true, true);
}
maxwell3d.regs.reg_array[0x446] = 0x0; // vertex id base?
maxwell3d.regs.index_array.count = 0;

View File

@@ -34,7 +34,6 @@ void MacroInterpreterImpl::Execute(const std::vector<u32>& parameters, u32 metho
this->parameters = std::make_unique<u32[]>(num_parameters);
}
std::memcpy(this->parameters.get(), parameters.data(), num_parameters * sizeof(u32));
this->num_parameters = num_parameters;
// Execute the code until we hit an exit condition.
bool keep_executing = true;

View File

@@ -14,11 +14,15 @@
namespace Tegra {
MemoryManager::MemoryManager(Core::System& system, VideoCore::RasterizerInterface& rasterizer)
: system{system}, rasterizer{rasterizer}, page_table(page_table_size) {}
MemoryManager::MemoryManager(Core::System& system_)
: system{system_}, page_table(page_table_size) {}
MemoryManager::~MemoryManager() = default;
void MemoryManager::BindRasterizer(VideoCore::RasterizerInterface& rasterizer_) {
rasterizer = &rasterizer_;
}
GPUVAddr MemoryManager::UpdateRange(GPUVAddr gpu_addr, PageEntry page_entry, std::size_t size) {
u64 remaining_size{size};
for (u64 offset{}; offset < size; offset += page_size) {
@@ -217,7 +221,7 @@ void MemoryManager::ReadBlock(GPUVAddr gpu_src_addr, void* dest_buffer, std::siz
// Flush must happen on the rasterizer interface, such that memory is always synchronous
// when it is read (even when in asynchronous GPU mode). Fixes Dead Cells title menu.
rasterizer.FlushRegion(src_addr, copy_amount);
rasterizer->FlushRegion(src_addr, copy_amount);
system.Memory().ReadBlockUnsafe(src_addr, dest_buffer, copy_amount);
}
@@ -266,7 +270,7 @@ void MemoryManager::WriteBlock(GPUVAddr gpu_dest_addr, const void* src_buffer, s
// Invalidate must happen on the rasterizer interface, such that memory is always
// synchronous when it is written (even when in asynchronous GPU mode).
rasterizer.InvalidateRegion(dest_addr, copy_amount);
rasterizer->InvalidateRegion(dest_addr, copy_amount);
system.Memory().WriteBlockUnsafe(dest_addr, src_buffer, copy_amount);
}
@@ -312,10 +316,10 @@ void MemoryManager::CopyBlockUnsafe(GPUVAddr gpu_dest_addr, GPUVAddr gpu_src_add
WriteBlockUnsafe(gpu_dest_addr, tmp_buffer.data(), size);
}
bool MemoryManager::IsGranularRange(GPUVAddr gpu_addr, std::size_t size) {
bool MemoryManager::IsGranularRange(GPUVAddr gpu_addr, std::size_t size) const {
const auto cpu_addr{GpuToCpuAddress(gpu_addr)};
if (!cpu_addr) {
return {};
return false;
}
const std::size_t page{(*cpu_addr & Core::Memory::PAGE_MASK) + size};
return page <= Core::Memory::PAGE_SIZE;

View File

@@ -68,9 +68,12 @@ static_assert(sizeof(PageEntry) == 4, "PageEntry is too large");
class MemoryManager final {
public:
explicit MemoryManager(Core::System& system, VideoCore::RasterizerInterface& rasterizer);
explicit MemoryManager(Core::System& system);
~MemoryManager();
/// Binds a renderer to the memory manager.
void BindRasterizer(VideoCore::RasterizerInterface& rasterizer);
std::optional<VAddr> GpuToCpuAddress(GPUVAddr addr) const;
template <typename T>
@@ -109,7 +112,7 @@ public:
/**
* IsGranularRange checks if a gpu region can be simply read with a pointer.
*/
bool IsGranularRange(GPUVAddr gpu_addr, std::size_t size);
bool IsGranularRange(GPUVAddr gpu_addr, std::size_t size) const;
GPUVAddr Map(VAddr cpu_addr, GPUVAddr gpu_addr, std::size_t size);
GPUVAddr MapAllocate(VAddr cpu_addr, std::size_t size, std::size_t align);
@@ -141,7 +144,7 @@ private:
Core::System& system;
VideoCore::RasterizerInterface& rasterizer;
VideoCore::RasterizerInterface* rasterizer = nullptr;
std::vector<PageEntry> page_table;
};

View File

@@ -9,7 +9,9 @@
namespace VideoCore {
RendererBase::RendererBase(Core::Frontend::EmuWindow& window) : render_window{window} {
RendererBase::RendererBase(Core::Frontend::EmuWindow& window_,
std::unique_ptr<Core::Frontend::GraphicsContext> context_)
: render_window{window_}, context{std::move(context_)} {
RefreshBaseSettings();
}

View File

@@ -15,7 +15,8 @@
namespace Core::Frontend {
class EmuWindow;
}
class GraphicsContext;
} // namespace Core::Frontend
namespace VideoCore {
@@ -25,14 +26,15 @@ struct RendererSettings {
// Screenshot
std::atomic<bool> screenshot_requested{false};
void* screenshot_bits;
void* screenshot_bits{};
std::function<void()> screenshot_complete_callback;
Layout::FramebufferLayout screenshot_framebuffer_layout;
};
class RendererBase : NonCopyable {
public:
explicit RendererBase(Core::Frontend::EmuWindow& window);
explicit RendererBase(Core::Frontend::EmuWindow& window,
std::unique_ptr<Core::Frontend::GraphicsContext> context);
virtual ~RendererBase();
/// Initialize the renderer
@@ -68,6 +70,14 @@ public:
return *rasterizer;
}
Core::Frontend::GraphicsContext& Context() {
return *context;
}
const Core::Frontend::GraphicsContext& Context() const {
return *context;
}
Core::Frontend::EmuWindow& GetRenderWindow() {
return render_window;
}
@@ -94,6 +104,7 @@ public:
protected:
Core::Frontend::EmuWindow& render_window; ///< Reference to the render window handle.
std::unique_ptr<RasterizerInterface> rasterizer;
std::unique_ptr<Core::Frontend::GraphicsContext> context;
f32 m_current_fps = 0.0f; ///< Current framerate, should be set by the renderer
int m_current_frame = 0; ///< Current frame, should be set by the renderer

View File

@@ -4,16 +4,17 @@
#include "common/assert.h"
#include <glad/glad.h>
#include "video_core/renderer_opengl/gl_buffer_cache.h"
#include "video_core/renderer_opengl/gl_fence_manager.h"
namespace OpenGL {
GLInnerFence::GLInnerFence(u32 payload, bool is_stubbed)
: VideoCommon::FenceBase(payload, is_stubbed), sync_object{} {}
GLInnerFence::GLInnerFence(u32 payload, bool is_stubbed) : FenceBase(payload, is_stubbed) {}
GLInnerFence::GLInnerFence(GPUVAddr address, u32 payload, bool is_stubbed)
: VideoCommon::FenceBase(address, payload, is_stubbed), sync_object{} {}
: FenceBase(address, payload, is_stubbed) {}
GLInnerFence::~GLInnerFence() = default;

View File

@@ -5,7 +5,6 @@
#pragma once
#include <memory>
#include <glad/glad.h>
#include "common/common_types.h"
#include "video_core/fence_manager.h"

View File

@@ -214,20 +214,20 @@ std::optional<std::vector<ShaderDiskCacheEntry>> ShaderDiskCacheOpenGL::LoadTran
// Skip games without title id
const bool has_title_id = system.CurrentProcess()->GetTitleID() != 0;
if (!Settings::values.use_disk_shader_cache.GetValue() || !has_title_id) {
return {};
return std::nullopt;
}
Common::FS::IOFile file(GetTransferablePath(), "rb");
if (!file.IsOpen()) {
LOG_INFO(Render_OpenGL, "No transferable shader cache found");
is_usable = true;
return {};
return std::nullopt;
}
u32 version{};
if (file.ReadBytes(&version, sizeof(version)) != sizeof(version)) {
LOG_ERROR(Render_OpenGL, "Failed to get transferable cache version, skipping it");
return {};
return std::nullopt;
}
if (version < NativeVersion) {
@@ -235,12 +235,12 @@ std::optional<std::vector<ShaderDiskCacheEntry>> ShaderDiskCacheOpenGL::LoadTran
file.Close();
InvalidateTransferable();
is_usable = true;
return {};
return std::nullopt;
}
if (version > NativeVersion) {
LOG_WARNING(Render_OpenGL, "Transferable shader cache was generated with a newer version "
"of the emulator, skipping");
return {};
return std::nullopt;
}
// Version is valid, load the shaders
@@ -249,7 +249,7 @@ std::optional<std::vector<ShaderDiskCacheEntry>> ShaderDiskCacheOpenGL::LoadTran
ShaderDiskCacheEntry& entry = entries.emplace_back();
if (!entry.Load(file)) {
LOG_ERROR(Render_OpenGL, "Failed to load transferable raw entry, skipping");
return {};
return std::nullopt;
}
}
@@ -290,12 +290,12 @@ std::optional<std::vector<ShaderDiskCachePrecompiled>> ShaderDiskCacheOpenGL::Lo
ShaderCacheVersionHash file_hash{};
if (!LoadArrayFromPrecompiled(file_hash.data(), file_hash.size())) {
precompiled_cache_virtual_file_offset = 0;
return {};
return std::nullopt;
}
if (GetShaderCacheVersionHash() != file_hash) {
LOG_INFO(Render_OpenGL, "Precompiled cache is from another version of the emulator");
precompiled_cache_virtual_file_offset = 0;
return {};
return std::nullopt;
}
std::vector<ShaderDiskCachePrecompiled> entries;
@@ -305,15 +305,16 @@ std::optional<std::vector<ShaderDiskCachePrecompiled>> ShaderDiskCacheOpenGL::Lo
if (!LoadObjectFromPrecompiled(entry.unique_identifier) ||
!LoadObjectFromPrecompiled(entry.binary_format) ||
!LoadObjectFromPrecompiled(binary_size)) {
return {};
return std::nullopt;
}
entry.binary.resize(binary_size);
if (!LoadArrayFromPrecompiled(entry.binary.data(), entry.binary.size())) {
return {};
return std::nullopt;
}
}
return entries;
return std::move(entries);
}
void ShaderDiskCacheOpenGL::InvalidateTransferable() {

View File

@@ -403,7 +403,7 @@ void CachedSurface::DecorateSurfaceName() {
LabelGLObject(GL_TEXTURE, texture.handle, GetGpuAddr(), params.TargetName());
}
void CachedSurfaceView::DecorateViewName(GPUVAddr gpu_addr, std::string prefix) {
void CachedSurfaceView::DecorateViewName(GPUVAddr gpu_addr, const std::string& prefix) {
LabelGLObject(GL_TEXTURE, main_view.handle, gpu_addr, prefix);
}

View File

@@ -90,7 +90,7 @@ public:
Tegra::Texture::SwizzleSource z_source,
Tegra::Texture::SwizzleSource w_source);
void DecorateViewName(GPUVAddr gpu_addr, std::string prefix);
void DecorateViewName(GPUVAddr gpu_addr, const std::string& prefix);
void MarkAsModified(u64 tick) {
surface.MarkAsModified(true, tick);

View File

@@ -313,10 +313,11 @@ public:
}
};
RendererOpenGL::RendererOpenGL(Core::Frontend::EmuWindow& emu_window, Core::System& system,
Core::Frontend::GraphicsContext& context)
: RendererBase{emu_window}, emu_window{emu_window}, system{system}, context{context},
program_manager{device}, has_debug_tool{HasDebugTool()} {}
RendererOpenGL::RendererOpenGL(Core::System& system_, Core::Frontend::EmuWindow& emu_window_,
Tegra::GPU& gpu_,
std::unique_ptr<Core::Frontend::GraphicsContext> context_)
: RendererBase{emu_window_, std::move(context_)}, system{system_},
emu_window{emu_window_}, gpu{gpu_}, program_manager{device}, has_debug_tool{HasDebugTool()} {}
RendererOpenGL::~RendererOpenGL() = default;
@@ -384,7 +385,7 @@ void RendererOpenGL::SwapBuffers(const Tegra::FramebufferConfig* framebuffer) {
if (has_debug_tool) {
glBindFramebuffer(GL_DRAW_FRAMEBUFFER, 0);
Present(0);
context.SwapBuffers();
context->SwapBuffers();
}
}

View File

@@ -56,8 +56,9 @@ class FrameMailbox;
class RendererOpenGL final : public VideoCore::RendererBase {
public:
explicit RendererOpenGL(Core::Frontend::EmuWindow& emu_window, Core::System& system,
Core::Frontend::GraphicsContext& context);
explicit RendererOpenGL(Core::System& system, Core::Frontend::EmuWindow& emu_window,
Tegra::GPU& gpu,
std::unique_ptr<Core::Frontend::GraphicsContext> context);
~RendererOpenGL() override;
bool Init() override;
@@ -93,9 +94,9 @@ private:
bool Present(int timeout_ms);
Core::Frontend::EmuWindow& emu_window;
Core::System& system;
Core::Frontend::GraphicsContext& context;
Core::Frontend::EmuWindow& emu_window;
Tegra::GPU& gpu;
const Device device;
StateTracker state_tracker{system};
@@ -120,7 +121,7 @@ private:
std::vector<u8> gl_framebuffer_data;
/// Used for transforming the framebuffer orientation
Tegra::FramebufferConfig::TransformFlags framebuffer_transform_flags;
Tegra::FramebufferConfig::TransformFlags framebuffer_transform_flags{};
Common::Rectangle<int> framebuffer_crop_rect;
/// Frame presentation mailbox

View File

@@ -237,8 +237,10 @@ std::string BuildCommaSeparatedExtensions(std::vector<std::string> available_ext
} // Anonymous namespace
RendererVulkan::RendererVulkan(Core::Frontend::EmuWindow& window, Core::System& system)
: RendererBase(window), system{system} {}
RendererVulkan::RendererVulkan(Core::System& system_, Core::Frontend::EmuWindow& emu_window,
Tegra::GPU& gpu_,
std::unique_ptr<Core::Frontend::GraphicsContext> context)
: RendererBase{emu_window, std::move(context)}, system{system_}, gpu{gpu_} {}
RendererVulkan::~RendererVulkan() {
ShutDown();

View File

@@ -38,7 +38,9 @@ struct VKScreenInfo {
class RendererVulkan final : public VideoCore::RendererBase {
public:
explicit RendererVulkan(Core::Frontend::EmuWindow& window, Core::System& system);
explicit RendererVulkan(Core::System& system, Core::Frontend::EmuWindow& emu_window,
Tegra::GPU& gpu,
std::unique_ptr<Core::Frontend::GraphicsContext> context);
~RendererVulkan() override;
bool Init() override;
@@ -58,6 +60,7 @@ private:
void Report() const;
Core::System& system;
Tegra::GPU& gpu;
Common::DynamicLibrary library;
vk::InstanceDispatch dld;

View File

@@ -3,6 +3,7 @@
// Refer to the license.txt file included.
#include <memory>
#include "common/logging/log.h"
#include "core/core.h"
#include "core/settings.h"
@@ -16,37 +17,46 @@
#include "video_core/video_core.h"
namespace {
std::unique_ptr<VideoCore::RendererBase> CreateRenderer(Core::Frontend::EmuWindow& emu_window,
Core::System& system,
Core::Frontend::GraphicsContext& context) {
std::unique_ptr<VideoCore::RendererBase> CreateRenderer(
Core::System& system, Core::Frontend::EmuWindow& emu_window, Tegra::GPU& gpu,
std::unique_ptr<Core::Frontend::GraphicsContext> context) {
switch (Settings::values.renderer_backend.GetValue()) {
case Settings::RendererBackend::OpenGL:
return std::make_unique<OpenGL::RendererOpenGL>(emu_window, system, context);
return std::make_unique<OpenGL::RendererOpenGL>(system, emu_window, gpu,
std::move(context));
#ifdef HAS_VULKAN
case Settings::RendererBackend::Vulkan:
return std::make_unique<Vulkan::RendererVulkan>(emu_window, system);
return std::make_unique<Vulkan::RendererVulkan>(system, emu_window, gpu,
std::move(context));
#endif
default:
return nullptr;
}
}
} // Anonymous namespace
namespace VideoCore {
std::unique_ptr<Tegra::GPU> CreateGPU(Core::Frontend::EmuWindow& emu_window, Core::System& system) {
std::unique_ptr<Tegra::GPU> gpu;
if (Settings::values.use_asynchronous_gpu_emulation.GetValue()) {
gpu = std::make_unique<VideoCommon::GPUAsynch>(system);
} else {
gpu = std::make_unique<VideoCommon::GPUSynch>(system);
}
auto context = emu_window.CreateSharedContext();
const auto scope = context->Acquire();
auto renderer = CreateRenderer(emu_window, system, *context);
auto renderer = CreateRenderer(system, emu_window, *gpu, std::move(context));
if (!renderer->Init()) {
return nullptr;
}
if (Settings::values.use_asynchronous_gpu_emulation.GetValue()) {
return std::make_unique<VideoCommon::GPUAsynch>(system, std::move(renderer),
std::move(context));
}
return std::make_unique<VideoCommon::GPUSynch>(system, std::move(renderer), std::move(context));
gpu->BindRenderer(std::move(renderer));
return gpu;
}
u16 GetResolutionScaleFactor(const RendererBase& renderer) {

View File

@@ -635,6 +635,11 @@ void Config::ReadCpuValues() {
ReadSetting(QStringLiteral("cpuopt_misc_ir"), true).toBool();
Settings::values.cpuopt_reduce_misalign_checks =
ReadSetting(QStringLiteral("cpuopt_reduce_misalign_checks"), true).toBool();
Settings::values.cpuopt_unsafe_unfuse_fma =
ReadSetting(QStringLiteral("cpuopt_unsafe_unfuse_fma"), true).toBool();
Settings::values.cpuopt_unsafe_reduce_fp_error =
ReadSetting(QStringLiteral("cpuopt_unsafe_reduce_fp_error"), true).toBool();
}
qt_config->endGroup();
@@ -1132,6 +1137,11 @@ void Config::SaveCpuValues() {
WriteSetting(QStringLiteral("cpuopt_misc_ir"), Settings::values.cpuopt_misc_ir, true);
WriteSetting(QStringLiteral("cpuopt_reduce_misalign_checks"),
Settings::values.cpuopt_reduce_misalign_checks, true);
WriteSetting(QStringLiteral("cpuopt_unsafe_unfuse_fma"),
Settings::values.cpuopt_unsafe_unfuse_fma, true);
WriteSetting(QStringLiteral("cpuopt_unsafe_reduce_fp_error"),
Settings::values.cpuopt_unsafe_reduce_fp_error, true);
}
qt_config->endGroup();

View File

@@ -19,6 +19,8 @@ ConfigureCpu::ConfigureCpu(QWidget* parent) : QWidget(parent), ui(new Ui::Config
connect(ui->accuracy, qOverload<int>(&QComboBox::activated), this,
&ConfigureCpu::AccuracyUpdated);
connect(ui->accuracy, qOverload<int>(&QComboBox::currentIndexChanged), this,
&ConfigureCpu::UpdateGroup);
}
ConfigureCpu::~ConfigureCpu() = default;
@@ -28,6 +30,12 @@ void ConfigureCpu::SetConfiguration() {
ui->accuracy->setEnabled(runtime_lock);
ui->accuracy->setCurrentIndex(static_cast<int>(Settings::values.cpu_accuracy));
UpdateGroup(static_cast<int>(Settings::values.cpu_accuracy));
ui->cpuopt_unsafe_unfuse_fma->setEnabled(runtime_lock);
ui->cpuopt_unsafe_unfuse_fma->setChecked(Settings::values.cpuopt_unsafe_unfuse_fma);
ui->cpuopt_unsafe_reduce_fp_error->setEnabled(runtime_lock);
ui->cpuopt_unsafe_reduce_fp_error->setChecked(Settings::values.cpuopt_unsafe_reduce_fp_error);
}
void ConfigureCpu::AccuracyUpdated(int index) {
@@ -38,14 +46,21 @@ void ConfigureCpu::AccuracyUpdated(int index) {
QMessageBox::Yes | QMessageBox::No);
if (result == QMessageBox::No) {
ui->accuracy->setCurrentIndex(static_cast<int>(Settings::CPUAccuracy::Accurate));
return;
UpdateGroup(static_cast<int>(Settings::CPUAccuracy::Accurate));
}
}
}
void ConfigureCpu::UpdateGroup(int index) {
ui->unsafe_group->setVisible(static_cast<Settings::CPUAccuracy>(index) ==
Settings::CPUAccuracy::Unsafe);
}
void ConfigureCpu::ApplyConfiguration() {
Settings::values.cpu_accuracy =
static_cast<Settings::CPUAccuracy>(ui->accuracy->currentIndex());
Settings::values.cpuopt_unsafe_unfuse_fma = ui->cpuopt_unsafe_unfuse_fma->isChecked();
Settings::values.cpuopt_unsafe_reduce_fp_error = ui->cpuopt_unsafe_reduce_fp_error->isChecked();
}
void ConfigureCpu::changeEvent(QEvent* event) {

View File

@@ -26,6 +26,7 @@ private:
void RetranslateUI();
void AccuracyUpdated(int index);
void UpdateGroup(int index);
void SetConfiguration();

View File

@@ -38,6 +38,11 @@
<string>Accurate</string>
</property>
</item>
<item>
<property name="text">
<string>Unsafe</string>
</property>
</item>
<item>
<property name="text">
<string>Enable Debug Mode</string>
@@ -62,6 +67,53 @@
</item>
</layout>
</item>
<item>
<layout class="QVBoxLayout">
<item>
<widget class="QGroupBox" name="unsafe_group">
<property name="title">
<string>Unsafe CPU Optimization Settings</string>
</property>
<layout class="QVBoxLayout">
<item>
<widget class="QLabel">
<property name="wordWrap">
<bool>1</bool>
</property>
<property name="text">
<string>These settings reduce accuracy for speed.</string>
</property>
</widget>
</item>
<item>
<widget class="QCheckBox" name="cpuopt_unsafe_unfuse_fma">
<property name="text">
<string>Unfuse FMA (improve performance on CPUs without FMA)</string>
</property>
<property name="toolTip">
<string>
&lt;div&gt;This option improves speed by reducing accuracy of fused-multiply-add instructions on CPUs without native FMA support.&lt;/div&gt;
</string>
</property>
</widget>
</item>
<item>
<widget class="QCheckBox" name="cpuopt_unsafe_reduce_fp_error">
<property name="text">
<string>Faster FRSQRTE and FRECPE</string>
</property>
<property name="toolTip">
<string>
&lt;div&gt;This option improves the speed of some approximate floating-point functions by using less accurate native approximations.&lt;/div&gt;
</string>
</property>
</widget>
</item>
</layout>
</widget>
</item>
</layout>
</item>
<item>
<spacer name="verticalSpacer">
<property name="orientation">