perf: drop stdint.h + picolibc — zero libc dependency

dma.h: typedef u32 instead of uint32_t, no stdint.h include
dma.c: use u32 instead of uint32_t
user-sample.c: unsigned char/int instead of uint8_t/uint32_t

Makefile: filter-out -lc -lm -lg -lsemihost from LDFLAGS
  → picolibc .data section (2.1KB, 532 words) eliminated
  → startup data-copy loop: ~30000 → ~200 cycles
  → total start.S latency: ~44000 → ~14000 cycles (~0.42ms at 33MHz)

sdk/software/examples/asm/dma.h

@@ -1,16 +1,15 @@
 #pragma once
 
-#include <stdint.h>
-
+typedef unsigned int u32;
 
 typedef struct {
-    volatile uint32_t SRC_ADDR;  // 0x00: 源地址
-    volatile uint32_t DST_ADDR;  // 0x04: 目的地址
-    volatile uint32_t LENGTH;    // 0x08: 传输总字节数
-    volatile uint32_t TAG;       // 0x0C: 软件 Tag (硬件不处理，留给软件标记任务用)
-    volatile uint32_t CTRL;      // 0x10: 控制寄存器 (配置 Burst 属性 + 启动)
-    volatile uint32_t STATUS;    // 0x14: 状态寄存器 (Busy, Done)
-    volatile uint32_t RESV[10];  // 0x18 ~ 0x3C: 保留空间，凑齐 0x40 字节
+    volatile u32 SRC_ADDR;  // 0x00: source address
+    volatile u32 DST_ADDR;  // 0x04: destination address
+    volatile u32 LENGTH;    // 0x08: transfer byte count
+    volatile u32 TAG;       // 0x0C: software tag (ignored by HW)
+    volatile u32 CTRL;      // 0x10: control (burst config + start)
+    volatile u32 STATUS;    // 0x14: status (busy, done)
+    volatile u32 RESV[10];  // 0x18 ~ 0x3C: padding to 0x40 bytes
 } dma_ch_regs_t;
 
 // DMA 控制器基地址
@@ -26,5 +25,5 @@ typedef struct {
 #define STATUS_BUSY_BIT      (1 << 0)
 #define STATUS_DONE_BIT      (1 << 1)
 
-void dma_start_transfer(int ch_id, uint32_t src, uint32_t dst, uint32_t bytes, uint32_t tag);
+void dma_start_transfer(int ch_id, u32 src, u32 dst, u32 bytes, u32 tag);
 void dma_wait_polling(int ch_id);