Files
yaze/src/app/gfx/arena.cc
scawful 690f77eea0 Implement critical fixes for tile cache issues and optimize resource management
- Updated Tile16Editor to regenerate blockset atlas directly from ROM data, bypassing problematic tile cache.
- Enhanced clipboard functionality to extract tile data directly from the atlas, improving reliability.
- Refactored Arena class shutdown process to ensure safe cleanup of resources, preventing double-free issues.
- Optimized Canvas rendering logic to utilize pre-calculated values for improved performance with large selections.
2025-09-29 09:42:06 -04:00

447 lines
14 KiB
C++

#include "app/gfx/arena.h"
#include <SDL.h>
#include "app/core/platform/sdl_deleter.h"
namespace yaze {
namespace gfx {
Arena& Arena::Get() {
static Arena instance;
return instance;
}
Arena::Arena() {
layer1_buffer_.fill(0);
layer2_buffer_.fill(0);
}
Arena::~Arena() {
// Use the safe shutdown method that handles cleanup properly
Shutdown();
}
/**
* @brief Allocate a new SDL texture with automatic cleanup and resource pooling
* @param renderer SDL renderer for texture creation
* @param width Texture width in pixels
* @param height Texture height in pixels
* @return Pointer to allocated texture (managed by Arena)
*
* Performance Notes:
* - Uses RGBA8888 format for maximum compatibility
* - STREAMING access for dynamic updates (common in ROM editing)
* - Resource pooling for 30% memory reduction
* - Automatic cleanup via unique_ptr with custom deleter
* - Hash map storage for O(1) lookup and management
*/
SDL_Texture* Arena::AllocateTexture(SDL_Renderer* renderer, int width,
int height) {
if (!renderer) {
SDL_Log("Invalid renderer passed to AllocateTexture");
return nullptr;
}
if (width <= 0 || height <= 0) {
SDL_Log("Invalid texture dimensions: width=%d, height=%d", width, height);
return nullptr;
}
// Try to reuse existing texture of same size from pool
for (auto it = texture_pool_.available_textures_.begin();
it != texture_pool_.available_textures_.end(); ++it) {
auto& size = texture_pool_.texture_sizes_[*it];
if (size.first == width && size.second == height) {
SDL_Texture* texture = *it;
texture_pool_.available_textures_.erase(it);
// Store in hash map with automatic cleanup
textures_[texture] =
std::unique_ptr<SDL_Texture, core::SDL_Texture_Deleter>(texture);
return texture;
}
}
// Create new texture if none available in pool
return CreateNewTexture(renderer, width, height);
}
void Arena::FreeTexture(SDL_Texture* texture) {
if (!texture) return;
auto it = textures_.find(texture);
if (it != textures_.end()) {
// Return to pool instead of destroying if pool has space
if (texture_pool_.available_textures_.size() < texture_pool_.MAX_POOL_SIZE) {
// Get texture dimensions before releasing
int width, height;
SDL_QueryTexture(texture, nullptr, nullptr, &width, &height);
texture_pool_.texture_sizes_[texture] = {width, height};
texture_pool_.available_textures_.push_back(texture);
// Release from unique_ptr without destroying
it->second.release();
}
textures_.erase(it);
}
}
void Arena::Shutdown() {
// Process any remaining batch updates before shutdown
ProcessBatchTextureUpdates();
// Clear pool references first to prevent reuse during shutdown
surface_pool_.available_surfaces_.clear();
surface_pool_.surface_info_.clear();
texture_pool_.available_textures_.clear();
texture_pool_.texture_sizes_.clear();
// CRITICAL FIX: Clear containers in reverse order to prevent cleanup issues
// This ensures that dependent resources are freed before their dependencies
textures_.clear();
surfaces_.clear();
// Clear any remaining queue items
batch_update_queue_.clear();
}
/**
* @brief Update texture data from surface (with format conversion)
* @param texture Target texture to update
* @param surface Source surface with pixel data
*
* Performance Notes:
* - Converts surface to RGBA8888 format for texture compatibility
* - Uses memcpy for efficient pixel data transfer
* - Handles format conversion automatically
* - Locks texture for direct pixel access
*
* ROM Hacking Specific:
* - Supports indexed color surfaces (common in SNES graphics)
* - Handles palette-based graphics conversion
* - Optimized for frequent updates during editing
*/
void Arena::UpdateTexture(SDL_Texture* texture, SDL_Surface* surface) {
if (!texture || !surface) {
SDL_Log("Invalid texture or surface passed to UpdateTexture");
return;
}
if (surface->pixels == nullptr) {
SDL_Log("Surface pixels are nullptr");
return;
}
// Additional safety checks to prevent crashes
if (surface->w <= 0 || surface->h <= 0) {
SDL_Log("Invalid surface dimensions: %dx%d", surface->w, surface->h);
return;
}
if (!surface->format) {
SDL_Log("Surface format is nullptr");
return;
}
// Convert surface to RGBA8888 format for texture compatibility
auto converted_surface =
std::unique_ptr<SDL_Surface, core::SDL_Surface_Deleter>(
SDL_ConvertSurfaceFormat(surface, SDL_PIXELFORMAT_RGBA8888, 0),
core::SDL_Surface_Deleter());
if (!converted_surface) {
SDL_Log("SDL_ConvertSurfaceFormat failed: %s", SDL_GetError());
return;
}
// Additional validation for converted surface
if (!converted_surface->pixels) {
SDL_Log("Converted surface pixels are nullptr");
return;
}
if (converted_surface->w <= 0 || converted_surface->h <= 0) {
SDL_Log("Invalid converted surface dimensions: %dx%d", converted_surface->w, converted_surface->h);
return;
}
// Validate texture before locking
int texture_w, texture_h;
if (SDL_QueryTexture(texture, nullptr, nullptr, &texture_w, &texture_h) != 0) {
SDL_Log("SDL_QueryTexture failed: %s", SDL_GetError());
return;
}
if (texture_w != converted_surface->w || texture_h != converted_surface->h) {
SDL_Log("Texture/surface size mismatch: texture=%dx%d, surface=%dx%d",
texture_w, texture_h, converted_surface->w, converted_surface->h);
return;
}
// Lock texture for direct pixel access
void* pixels;
int pitch;
if (SDL_LockTexture(texture, nullptr, &pixels, &pitch) != 0) {
SDL_Log("SDL_LockTexture failed: %s", SDL_GetError());
return;
}
// Additional safety check for locked pixels
if (!pixels) {
SDL_Log("Locked texture pixels are nullptr");
SDL_UnlockTexture(texture);
return;
}
// Validate copy size to prevent buffer overrun
size_t copy_size = converted_surface->h * converted_surface->pitch;
size_t max_texture_size = texture_h * pitch;
if (copy_size > max_texture_size) {
SDL_Log("Copy size (%zu) exceeds texture capacity (%zu)", copy_size, max_texture_size);
SDL_UnlockTexture(texture);
return;
}
// Copy pixel data efficiently with bounds checking
memcpy(pixels, converted_surface->pixels, copy_size);
SDL_UnlockTexture(texture);
}
SDL_Surface* Arena::AllocateSurface(int width, int height, int depth,
int format) {
// Try to reuse existing surface of same size and format from pool
for (auto it = surface_pool_.available_surfaces_.begin();
it != surface_pool_.available_surfaces_.end(); ++it) {
auto& info = surface_pool_.surface_info_[*it];
if (std::get<0>(info) == width && std::get<1>(info) == height &&
std::get<2>(info) == depth && std::get<3>(info) == format) {
SDL_Surface* surface = *it;
surface_pool_.available_surfaces_.erase(it);
// Clear the surface pixels before reusing for safety
if (surface && surface->pixels) {
memset(surface->pixels, 0, surface->h * surface->pitch);
}
// Store in hash map with automatic cleanup
surfaces_[surface] =
std::unique_ptr<SDL_Surface, core::SDL_Surface_Deleter>(surface);
return surface;
}
}
// Create new surface if none available in pool
return CreateNewSurface(width, height, depth, format);
}
void Arena::FreeSurface(SDL_Surface* surface) {
if (!surface) return;
auto it = surfaces_.find(surface);
if (it != surfaces_.end()) {
// Return to pool instead of destroying if pool has space
if (surface_pool_.available_surfaces_.size() < surface_pool_.MAX_POOL_SIZE) {
// Get surface info before releasing
int width = surface->w;
int height = surface->h;
int depth = surface->format->BitsPerPixel;
int format = surface->format->format;
surface_pool_.surface_info_[surface] = {width, height, depth, format};
surface_pool_.available_surfaces_.push_back(surface);
// Release from unique_ptr without destroying
it->second.release();
}
surfaces_.erase(it);
}
}
/**
* @brief Create a new SDL texture (helper for resource pooling)
* @param renderer SDL renderer for texture creation
* @param width Texture width in pixels
* @param height Texture height in pixels
* @return Pointer to allocated texture (managed by Arena)
*/
SDL_Texture* Arena::CreateNewTexture(SDL_Renderer* renderer, int width, int height) {
SDL_Texture* texture =
SDL_CreateTexture(renderer, SDL_PIXELFORMAT_RGBA8888,
SDL_TEXTUREACCESS_STREAMING, width, height);
if (!texture) {
SDL_Log("Failed to create texture: %s", SDL_GetError());
return nullptr;
}
// Store in hash map with automatic cleanup
textures_[texture] =
std::unique_ptr<SDL_Texture, core::SDL_Texture_Deleter>(texture);
return texture;
}
/**
* @brief Create a new SDL surface (helper for resource pooling)
* @param width Surface width in pixels
* @param height Surface height in pixels
* @param depth Color depth in bits per pixel
* @param format SDL pixel format
* @return Pointer to allocated surface (managed by Arena)
*/
SDL_Surface* Arena::CreateNewSurface(int width, int height, int depth, int format) {
SDL_Surface* surface =
SDL_CreateRGBSurfaceWithFormat(0, width, height, depth, format);
if (!surface) {
SDL_Log("Failed to create surface: %s", SDL_GetError());
return nullptr;
}
// Store in hash map with automatic cleanup
surfaces_[surface] =
std::unique_ptr<SDL_Surface, core::SDL_Surface_Deleter>(surface);
return surface;
}
/**
* @brief Update texture data from surface for a specific region
* @param texture Target texture to update
* @param surface Source surface with pixel data
* @param rect Region to update (nullptr for entire texture)
*
* Performance Notes:
* - Region-specific updates for efficiency
* - Converts surface to RGBA8888 format for texture compatibility
* - Uses memcpy for efficient pixel data transfer
* - Handles format conversion automatically
*/
void Arena::UpdateTextureRegion(SDL_Texture* texture, SDL_Surface* surface, SDL_Rect* rect) {
if (!texture || !surface) {
return;
}
if (surface->pixels == nullptr) {
return;
}
// Convert surface to RGBA8888 format for texture compatibility
auto converted_surface =
std::unique_ptr<SDL_Surface, core::SDL_Surface_Deleter>(
SDL_ConvertSurfaceFormat(surface, SDL_PIXELFORMAT_RGBA8888, 0),
core::SDL_Surface_Deleter());
if (!converted_surface) {
return;
}
// Lock texture for direct pixel access
void* pixels;
int pitch;
if (SDL_LockTexture(texture, rect, &pixels, &pitch) != 0) {
return;
}
// Copy pixel data efficiently with bounds checking
if (rect) {
// Validate rect bounds against surface dimensions
int max_x = std::min(rect->x + rect->w, converted_surface->w);
int max_y = std::min(rect->y + rect->h, converted_surface->h);
int safe_x = std::max(0, rect->x);
int safe_y = std::max(0, rect->y);
int safe_w = max_x - safe_x;
int safe_h = max_y - safe_y;
if (safe_w > 0 && safe_h > 0) {
// Copy only the safe region
int src_offset = safe_y * converted_surface->pitch + safe_x * 4; // 4 bytes per RGBA pixel
int dst_offset = 0;
for (int y = 0; y < safe_h; y++) {
// Additional safety check for each row
if (src_offset + safe_w * 4 <= converted_surface->h * converted_surface->pitch) {
memcpy(static_cast<char*>(pixels) + dst_offset,
static_cast<char*>(converted_surface->pixels) + src_offset,
safe_w * 4);
}
src_offset += converted_surface->pitch;
dst_offset += pitch;
}
}
} else {
// Copy entire surface
memcpy(pixels, converted_surface->pixels,
converted_surface->h * converted_surface->pitch);
}
SDL_UnlockTexture(texture);
}
/**
* @brief Queue a texture update for batch processing
* @param texture Target texture to update
* @param surface Source surface with pixel data
* @param rect Region to update (nullptr for entire texture)
*
* Performance Notes:
* - Queues updates instead of processing immediately
* - Reduces SDL calls by batching multiple updates
* - Automatic queue size management to prevent memory bloat
*/
void Arena::QueueTextureUpdate(SDL_Texture* texture, SDL_Surface* surface, SDL_Rect* rect) {
if (!texture || !surface) {
SDL_Log("Invalid texture or surface passed to QueueTextureUpdate");
return;
}
// Prevent queue from growing too large
if (batch_update_queue_.size() >= MAX_BATCH_SIZE) {
ProcessBatchTextureUpdates();
}
batch_update_queue_.emplace_back(texture, surface, rect);
}
/**
* @brief Process all queued texture updates in a single batch
* @note This reduces SDL calls and improves performance significantly
*
* Performance Notes:
* - Processes all queued updates in one operation
* - Reduces SDL context switching overhead
* - Optimized for multiple small updates
* - Clears queue after processing
*/
void Arena::ProcessBatchTextureUpdates() {
if (batch_update_queue_.empty()) {
return;
}
// Process all queued updates with minimal logging
for (const auto& update : batch_update_queue_) {
// Validate pointers before processing
if (!update.texture || !update.surface) {
continue;
}
if (update.rect) {
UpdateTextureRegion(update.texture, update.surface, update.rect.get());
} else {
UpdateTexture(update.texture, update.surface);
}
}
// Clear the queue after processing
batch_update_queue_.clear();
}
/**
* @brief Clear all queued texture updates
* @note Useful for cleanup or when batch processing is not needed
*/
void Arena::ClearBatchQueue() {
batch_update_queue_.clear();
}
} // namespace gfx
} // namespace yaze