[arch][arm64] save the boot arg registers (x0-x3) in a temporary spot in the boot path

Much of the start.S path avoids using these registers up until now to
avoid trashing any state, but its getting fairly difficult and error
prone to keep this up. Save the args as soon as its known that its the
boot cpu in a temporary place prior to calling lk_main. Wastes 32 bytes
of memory but should be more solid.

arch/arm64/start.S

@@ -32,6 +32,7 @@ attr                    .req x27
 FUNCTION(_start)
 .globl arm_reset
 arm_reset:
+    /* if we came in at higher than EL1, drop down to EL1 */
     bl      arm64_elX_to_el1
 
 #if WITH_KERNEL_VM
@@ -59,11 +60,21 @@ arm_reset:
     mrs     cpuid, mpidr_el1
     ubfx    cpuid, cpuid, #0, #SMP_CPU_ID_BITS
     cbnz    cpuid, .Lmmu_enable_secondary
-#endif
 
-    mov     tmp, #0
+    /* this path forward until .Lmmu_enable_secondary is the primary cpu only */
 
+#endif /* WITH_SMP */
+#endif /* WITH_KERNEL_VM */
+
+    /* save a copy of the boot args so x0-x3 are available for use */
+    adrp    tmp, arm64_boot_args
+    add     tmp, tmp, :lo12:arm64_boot_args
+    stp     x0, x1, [tmp], #16
+    stp     x2, x3, [tmp]
+
+#if WITH_KERNEL_VM
     /* walk through all the entries in the translation table, setting them up */
+    mov     tmp, #0
 .Lclear_top_page_table_loop:
     str     xzr, [page_table1, tmp, lsl #3]
     add     tmp, tmp, #1
@@ -81,7 +92,7 @@ arm_reset:
 
     tbzmask tmp, MMU_INITIAL_MAPPING_FLAG_DYNAMIC, .Lnot_dynamic
     adr     paddr, _start
-    mov     size, x0
+    mov     size, x0 /* use the arg passed through from platform_reset */
     str     paddr, [mmu_initial_mapping, #__MMU_INITIAL_MAPPING_PHYS_OFFSET]
     str     size, [mmu_initial_mapping, #__MMU_INITIAL_MAPPING_SIZE_OFFSET]
 
@@ -339,6 +350,7 @@ arm_reset:
 #endif
 #endif /* WITH_KERNEL_VM */
 
+    /* load the stack pointer */
     ldr tmp, =__stack_end
     mov sp, tmp
 
@@ -356,6 +368,12 @@ arm_reset:
     cbnz    tmp2, .L__bss_loop
 .L__bss_loop_done:
 
+    /* load the boot args we had saved previously */
+    adrp    tmp, arm64_boot_args
+    add     tmp, tmp, :lo12:arm64_boot_args
+    ldp     x0, x1, [tmp], #16
+    ldp     x2, x3, [tmp]
+
     bl  lk_main
     b   .
 
@@ -386,16 +404,24 @@ arm_reset:
 
 .ltorg
 
+.data
+    .balign 8
+LOCAL_DATA(arm64_boot_args)
+    .skip (4 * 8)
+END_DATA(arm64_boot_args)
+
 #if WITH_SMP
 .data
 DATA(page_tables_not_ready)
     .long       1
+END_DATA(page_tables_not_ready)
 #endif
 
 .section .bss.prebss.stack
     .align 4
 DATA(__stack)
     .skip ARCH_DEFAULT_STACK_SIZE * SMP_MAX_CPUS
+END_DATA(__stack)
 DATA(__stack_end)
 
 #if WITH_KERNEL_VM
@@ -403,4 +429,5 @@ DATA(__stack_end)
 .align 3 + MMU_PAGE_TABLE_ENTRIES_IDENT_SHIFT
 DATA(tt_trampoline)
     .skip 8 * MMU_PAGE_TABLE_ENTRIES_IDENT
+END_DATA(tt_trampoline)
 #endif