x86: Rework Global Descriptor Table loading

The inline assembler is ugly and uses hard coded magic numbers. Make it more
elegant to allow cleaner implementation of future GDT related patches. The
compiler seems smart enough to generate the same code anyway

--
Changes for v2:
 - Rebased against revised patch #3
 - Use GDT size define instead of magic number
 - Added commit message
diff --git a/arch/x86/cpu/cpu.c b/arch/x86/cpu/cpu.c
index 209ff29..8102fb9 100644
--- a/arch/x86/cpu/cpu.c
+++ b/arch/x86/cpu/cpu.c
@@ -55,35 +55,39 @@
 	u32 ptr;
 } __packed;
 
-static void reload_gdt(void)
+static void load_ds(u32 segment)
 {
-	/*
-	 * There are machines which are known to not boot with the GDT
-	 * being 8-byte unaligned.  Intel recommends 16 byte alignment
-	 */
-	static const u64 boot_gdt[] __attribute__((aligned(16))) = {
-		/* CS: code, read/execute, 4 GB, base 0 */
-		[X86_GDT_ENTRY_32BIT_CS] = GDT_ENTRY(0xc09b, 0, 0xfffff),
-		/* DS: data, read/write, 4 GB, base 0 */
-		[X86_GDT_ENTRY_32BIT_DS] = GDT_ENTRY(0xc093, 0, 0xfffff),
-		/* 16-bit CS: code, read/execute, 64 kB, base 0 */
-		[X86_GDT_ENTRY_16BIT_CS] = GDT_ENTRY(0x109b, 0, 0x0ffff),
-		/* 16-bit DS: data, read/write, 64 kB, base 0 */
-		[X86_GDT_ENTRY_16BIT_DS] = GDT_ENTRY(0x1093, 0, 0x0ffff),
-	};
-	static struct gdt_ptr gdt;
+	asm volatile("movl %0, %%ds" : : "r" (segment * X86_GDT_ENTRY_SIZE));
+}
+
+static void load_es(u32 segment)
+{
+	asm volatile("movl %0, %%es" : : "r" (segment * X86_GDT_ENTRY_SIZE));
+}
+
+static void load_fs(u32 segment)
+{
+	asm volatile("movl %0, %%fs" : : "r" (segment * X86_GDT_ENTRY_SIZE));
+}
+
+static void load_gs(u32 segment)
+{
+	asm volatile("movl %0, %%gs" : : "r" (segment * X86_GDT_ENTRY_SIZE));
+}
+
+static void load_ss(u32 segment)
+{
+	asm volatile("movl %0, %%ss" : : "r" (segment * X86_GDT_ENTRY_SIZE));
+}
 
-	gdt.len = sizeof(boot_gdt)-1;
-	gdt.ptr = (u32)&boot_gdt;
+static void load_gdt(const u64 *boot_gdt, u16 num_entries)
+{
+	struct gdt_ptr gdt;
+
+	gdt.len = (num_entries * 8) - 1;
+	gdt.ptr = (u32)boot_gdt;
 
-	asm volatile("lgdtl %0\n" \
-		     "movl $((2+1)*8), %%ecx\n" \
-		     "movl %%ecx, %%ds\n" \
-		     "movl %%ecx, %%es\n" \
-		     "movl %%ecx, %%fs\n" \
-		     "movl %%ecx, %%gs\n" \
-		     "movl %%ecx, %%ss" \
-		     : : "m" (gdt) : "ecx");
+	asm volatile("lgdtl %0\n" : : "m" (gdt));
 }
 
 int x86_cpu_init_f(void)
@@ -113,7 +117,27 @@
 	    "movl	%%eax, %%cr0\n"
 	    "wbinvd\n" : : "i" (nw_cd_rst) : "eax");
 
+	/*
+	 * There are machines which are known to not boot with the GDT
+	 * being 8-byte unaligned. Intel recommends 16 byte alignment
+	 */
+	static const u64 boot_gdt[] __aligned(16) = {
+		/* CS: code, read/execute, 4 GB, base 0 */
+		[X86_GDT_ENTRY_32BIT_CS] = GDT_ENTRY(0xc09b, 0, 0xfffff),
+		/* DS: data, read/write, 4 GB, base 0 */
+		[X86_GDT_ENTRY_32BIT_DS] = GDT_ENTRY(0xc093, 0, 0xfffff),
+		/* 16-bit CS: code, read/execute, 64 kB, base 0 */
+		[X86_GDT_ENTRY_16BIT_CS] = GDT_ENTRY(0x109b, 0, 0x0ffff),
+		/* 16-bit DS: data, read/write, 64 kB, base 0 */
+		[X86_GDT_ENTRY_16BIT_DS] = GDT_ENTRY(0x1093, 0, 0x0ffff),
+	};
+
-	reload_gdt();
+	load_gdt(boot_gdt, X86_GDT_NUM_ENTRIES);
+	load_ds(X86_GDT_ENTRY_32BIT_DS);
+	load_es(X86_GDT_ENTRY_32BIT_DS);
+	load_fs(X86_GDT_ENTRY_32BIT_DS);
+	load_gs(X86_GDT_ENTRY_32BIT_DS);
+	load_ss(X86_GDT_ENTRY_32BIT_DS);
 
 	/* Initialize core interrupt and exception functionality of CPU */
 	cpu_init_interrupts();