Full VGA/EGA support

🎯 Major Features & Improvements

VGA/EGA Graphics System Overhaul

• Complete VGA 256KB Support: Full implementation of VGA graphics modes with 256KB video memory using planar memory layout with 0xP3P2P1P0 byte packing
• EGA Graphics Modes: Added support for EGA 320x200x16 (0x0D), 640x200x16 (0x0E), and 640x350x16 (0x10) modes with proper plane expansion
• VGA Mode Implementation: Complete VGA 640x480x16 (0x12) and enhanced VGA 320x200x256 (0x13) rendering
• Chain4 Mode Support: Proper handling of VGA chain4 mode for 256-color graphics with chain4 flag in vga_cache_t
• Pixel Packing Optimization: Replaced spread4_u32 with ega_pack8_from_planes for improved clarity and code reuse across platforms
• Enhanced Memory Access: Added vga_mem_read16 implementation and optimized memory read/write operations with 32-bit latch system
• Plane Expansion LUT: Added 16-entry plane expansion lookup table for efficient bit-plane to pixel conversion

Cross-Platform Graphics Performance

• Unified Rendering Logic: Standardized graphics rendering across Windows, Linux, and Pico platforms
• Loop Unrolling: Optimized EGA/VGA rendering loops for improved performance
• Bitwise Operations: Replaced division with bitwise operations for better performance
• Memory Alignment: 4-byte alignment for PSRAM variables and video memory buffers

PSRAM & Memory Management

• Butter-PSRAM Support: Auto-detection and configuration for Butter-PSRAM hardware with dynamic GPIO pin detection (GPIO 8 for MURM20, GPIO 47 for PIMO, GPIO 19 default)
• Dynamic Memory Allocation: Runtime PSRAM size detection (16MB, 8MB, 4MB, 1MB) with validation through test patterns
• Memory Layout Optimization: 4-byte alignment for all PSRAM variables (RAM, UMB, HMA, EMS) with aligned(4) attribute
• Virtual Memory Enhancement: Enhanced XMS/EMS memory management with shared UMB support and proper alignment
• PSRAM Frequency Control: Configurable PSRAM frequency (166MHz default) with timing adjustments
• Memory Mapping: PSRAM mapped at 0x11000000 with optimized access patterns

Platform-Specific Improvements

Raspberry Pi Pico (RP2040/RP2350)

• CPU Frequency Handling: Dynamic frequency configuration - 504MHz for Pico2 with HDMI, 378MHz for other HDMI configurations, 500MHz default
• PSRAM GPIO Auto-Detect: Automatic detection of onboard PSRAM GPIO configurations with hardware-specific pinouts
• Flash Frequency Control: Configurable flash frequency (100MHz default) with build name integration (F100-P166-500MHz format)
• Build Configuration: Streamlined CMake configuration with platform-specific optimizations and conditional compilation
• Performance Optimization: Strategic register keyword usage and memory access optimizations with #pragma GCC optimize("Ofast")

Windows/Linux Host Builds

• MiniFB Integration: Improved MiniFB graphics interface with cleaner code
• All Video Modes: Complete support for PX3 video modes across all platforms
• Build Fixes: Resolved compilation issues across different host platforms

---

🐛 Bug Fixes & Stabilization

Video System Fixes

• 40x25 Text Modes: Fixed rendering issues in 40-column text modes
• CGA Rendering: Corrected pointer types and rendering logic
• Palette Handling: Fixed incorrect palette usage in 640x350 EGA mode
• Boundary Checks: Removed redundant checks while maintaining safety

Memory Management Fixes

• XMS Alignment: Fixed 4-byte alignment for UMB memory blocks
• EMS Integration: Improved Expanded Memory Specification handling
• Memory Leaks: Resolved memory allocation and deallocation issues
• Swap System: Enhanced virtual memory swapping mechanisms

Audio & I/O Fixes

• Speaker Sample Integration: Fixed sound mixing logic for PC speaker
• Sound Blaster Emulation: Resolved swap-related sound card issues
• Port Handling: Improved I/O port access and timing

---

⚡ Performance Optimizations

Rendering Performance

• Bit Spreading Optimization: Fast bit spreading for EGA/VGA pixel processing
• Loop Unrolling: Reduced instruction count in critical rendering paths
• Memory Access Patterns: Optimized video memory access for better cache utilization
• Reduced Redundancy: Eliminated unnecessary buffer writes and calculations

Memory Performance

• Aligned Memory Access: 4-byte alignment for improved ARM architecture performance
• PSRAM Optimization: Optimized external PSRAM access patterns
• Virtual Memory: Improved page file access and caching

CPU Optimization

• Register Usage: Strategic use of register keyword for frequently accessed variables
• Bitwise Operations: Replaced expensive operations with faster alternatives
• Function Inlining: Critical path optimizations for better performance

---

🔧 Technical Improvements

Code Quality & Architecture

• Const Correctness: Added const qualifiers throughout codebase for better type safety
• Type Safety: Improved type handling with proper pointer arithmetic and casting
• Code Cleanup: Removed unused variables (VIDEORAM, port3DA from graphics.h) and dead code paths
• Header Organization: Better organization of include files and dependencies
• Variable Naming: Consistent naming conventions (e.g., ega_planes instead of generic variables)

Memory Management Architecture

• Planar Memory Layout: VGA memory organized as 0xP3P2P1P0 (plane0 in lowest byte) for optimal 32-bit access
• 32-bit Latch System: Implemented VGA latch system for planar memory operations
• Memory Access Patterns: Optimized VIDEORAM indexing using bitwise AND instead of modulo operations
• Buffer Management: Unified buffer handling across graphics modes with proper alignment

Build System & Configuration

• CMake Streamlining: Removed complex conditional logic in favor of cleaner platform detection
• Build Naming: Enhanced build names with frequency information (F100-P166-500MHz format)
• Configuration Files: Added support for config.286 runtime configuration
• Compilation Flags: Platform-specific optimization flags and conditional compilation

Full Changelog: v1.1.0...v2.0.0