Add ecall-based syscall entry point

Architecture-specific implementations require direct linkage to
override weak symbols. Archives extract objects only when symbols
are unresolved, skipping strong overrides when weak symbols satisfy
references.

Introduce trap-based syscall entry using ecall instruction and
modify build system to link entry point before archive, ensuring
architecture override takes precedence at link time.

Author: HeatCrab

Makefile

@@ -29,7 +29,7 @@ deps += $(LIB_OBJS:%.o=%.o.d)
 APPS := coop echo hello mqueues semaphore mutex cond \
         pipes pipes_small pipes_struct prodcons progress \
         rtsched suspend test64 timer timer_kill \
-        cpubench test_libc
+        cpubench test_libc umode
 
 # Output files for __link target
 IMAGE_BASE := $(BUILD_DIR)/image
@@ -66,9 +66,9 @@ $(APPS): %: rebuild $(BUILD_APP_DIR)/%.o linmo
 # Link target - creates all output files
 __link: $(IMAGE_FILES)
 
-$(IMAGE_BASE).elf: $(BUILD_APP_DIR)/*.o $(BUILD_DIR)/liblinmo.a
+$(IMAGE_BASE).elf: $(BUILD_APP_DIR)/*.o $(BUILD_DIR)/liblinmo.a $(ENTRY_OBJ)
 	$(VECHO) "  LD\t$@\n"
-	$(Q)$(LD) $(LDFLAGS) -T$(LDSCRIPT) -Map $(IMAGE_BASE).map -o $@ $(BUILD_APP_DIR)/*.o -L$(BUILD_DIR) -llinmo
+	$(Q)$(LD) $(LDFLAGS) -T$(LDSCRIPT) -Map $(IMAGE_BASE).map -o $@ $(BUILD_APP_DIR)/*.o $(ENTRY_OBJ) -L$(BUILD_DIR) -llinmo
 
 $(IMAGE_BASE).lst: $(IMAGE_BASE).elf
 	$(VECHO) "  DUMP\t$@\n"

arch/riscv/build.mk

@@ -74,6 +74,14 @@ HAL_OBJS := boot.o hal.o muldiv.o
 HAL_OBJS := $(addprefix $(BUILD_KERNEL_DIR)/,$(HAL_OBJS))
 deps += $(HAL_OBJS:%.o=%.o.d)
 
+# Architecture-specific syscall entry point requiring direct linkage.
+# Archives only extract objects when symbols are unresolved. Since the generic
+# syscall dispatcher provides a weak symbol, the archive mechanism would skip
+# the strong override. Direct linking ensures the architecture-specific
+# implementation takes precedence at link time.
+ENTRY_OBJ := $(BUILD_KERNEL_DIR)/entry.o
+deps += $(ENTRY_OBJ).d
+
 $(BUILD_KERNEL_DIR)/%.o: $(ARCH_DIR)/%.c | $(BUILD_DIR)
 	$(VECHO) "  CC\t$@\n"
 	$(Q)$(CC) $(CFLAGS) -o $@ -c -MMD -MF $@.d $<

arch/riscv/entry.c

@@ -0,0 +1,42 @@
+/* RISC-V Kernel Entry Points
+ *
+ * This file implements architecture-specific entry mechanisms into the kernel,
+ * primarily the system call trap interface using the RISC-V ecall instruction.
+ *
+ * System Call Calling Convention (RISC-V ABI):
+ * - a7 (x17): System call number
+ * - a0 (x10): Argument 1 / Return value
+ * - a1 (x11): Argument 2
+ * - a2 (x12): Argument 3
+ *
+ * The ecall instruction triggers an environment call exception that transfers
+ * control to the M-mode exception handler (hal.c), which then dispatches to
+ * the appropriate system call implementation via the syscall table.
+ */
+
+#include <sys/syscall.h>
+
+/* Architecture-specific syscall implementation using ecall trap.
+ * This overrides the weak symbol defined in kernel/syscall.c.
+ */
+int syscall(int num, void *arg1, void *arg2, void *arg3)
+{
+    register int a0 asm("a0") = (int) arg1;
+    register int a1 asm("a1") = (int) arg2;
+    register int a2 asm("a2") = (int) arg3;
+    register int a7 asm("a7") = num;
+
+    /* Execute ecall instruction to trap into M-mode.
+     * The M-mode exception handler will:
+     * 1. Save the current task context
+     * 2. Dispatch to the syscall handler based on a7
+     * 3. Place the return value in a0
+     * 4. Restore context and return to user mode via mret
+     */
+    asm volatile("ecall"
+                 : "+r"(a0) /* a0 is both input (arg1) and output (retval) */
+                 : "r"(a1), "r"(a2), "r"(a7)
+                 : "memory", "cc");
+
+    return a0;
+}