fusee_cpp: tweaks, now completes SecureInitialize on hardware

fusee_cpp/loader_stub/source/fusee_loader_main.cpp

@@ -26,16 +26,11 @@ namespace ams::nxboot::loader {
         constexpr size_t ProgramImageSizeMax = ProgramImageEnd - ProgramImageBase;
 
         void CopyBackwards(void *dst, const void *src, size_t size) {
-            /* We want to copy 32-bits at a time from destination to source. */
+                  u8 *dst_8 = static_cast<u8 *>(dst) + size;
-            const size_t words = util::DivideUp(size, sizeof(u32));
+            const u8 *src_8 = static_cast<const u8 *>(src) + size;
 
-            /* Convert to 32-bit pointers. */
+            for (size_t i = 0; i < size; ++i) {
-                  u32 *dst_32 = static_cast<u32 *>(dst) + words;
+                *(--dst_8) = *(--src_8);
-            const u32 *src_32 = static_cast<const u32 *>(src) + words;
-
-            /* Copy data. */
-            for (size_t i = 0; i < words; ++i) {
-                *(--dst_32) = *(--src_32);
             }
         }
 
@@ -43,7 +38,7 @@ namespace ams::nxboot::loader {
 
     NORETURN void UncompressAndExecute(const void *program, size_t program_size) {
         /* Relocate the compressed binary to a place where we can safely decompress it. */
-        void *relocated_program = reinterpret_cast<void *>(util::AlignDown(ProgramImageEnd - program_size, sizeof(u32)));
+        void *relocated_program = reinterpret_cast<void *>(ProgramImageEnd - program_size);
         if (relocated_program != program) {
             CopyBackwards(relocated_program, program, program_size);
         }

fusee_cpp/loader_stub/source/fusee_loader_uncompress.cpp

@@ -87,7 +87,9 @@ namespace ams::nxboot::loader {
                 }
 
                 void Copy(size_t size) {
-                    __builtin_memcpy(this->dst + this->dst_offset, this->src + this->src_offset, size);
+                    for (size_t i = 0; i < size; ++i) {
+                        this->dst[this->dst_offset + i] = this->src[this->src_offset + i];
+                    }
                     this->dst_offset += size;
                     this->src_offset += size;
                 }

fusee_cpp/program/program.ld

@@ -11,8 +11,10 @@ SECTIONS
 
 	.crt0 :
 	{
+        FILL(0x00000000)
 		KEEP (*(.crt0 .crt0.*))
-		. = ALIGN(8);
+        . = ORIGIN(main) + 0xC0 - 1;
+        BYTE(00);
 	} >main AT>glob
 
 	.text :
@@ -22,37 +24,31 @@ SECTIONS
 		*(.text.startup .text.startup.*)
 		*(.text.hot .text.hot.*)
 		*(.text .stub .text.* .gnu.linkonce.t.*)
-		. = ALIGN(8);
 	} >main AT>glob
 
 	.init :
 	{
 		KEEP( *(.init) )
-		. = ALIGN(8);
 	} >main AT>glob
 
 	.plt :
 	{
 		*(.plt)
 		*(.iplt)
-		. = ALIGN(8);
 	} >main AT>glob
 
 	.fini :
 	{
 		KEEP( *(.fini) )
-		. = ALIGN(8);
 	} >main AT>glob
 
 
 	/* =========== RODATA section =========== */
-	. = ALIGN(8);
 	__rodata_start = . ;
 
 	.rodata :
 	{
 		*(.rodata .rodata.* .gnu.linkonce.r.*)
-		. = ALIGN(8);
 	} >main AT>glob
 
 	.eh_frame_hdr     : { __eh_frame_hdr_start = .; *(.eh_frame_hdr) *(.eh_frame_entry .eh_frame_entry.*) __eh_frame_hdr_end = .; } >main AT>glob
@@ -63,7 +59,6 @@ SECTIONS
 	.hash              : { *(.hash) } >main AT>glob
 
 	/* =========== DATA section =========== */
-	. = ALIGN(8);
 	__data_start = . ;
 
 	.eh_frame : ONLY_IF_RW { KEEP (*(.eh_frame)) *(.eh_frame.*) } >main AT>glob
@@ -71,14 +66,14 @@ SECTIONS
 	.gnu_extab : ONLY_IF_RW { *(.gnu_extab*) } >main AT>glob
 	.exception_ranges : ONLY_IF_RW { *(.exception_ranges .exception_ranges*) } >main AT>glob
 
-	.preinit_array ALIGN(8) :
+	.preinit_array :
 	{
 		PROVIDE (__preinit_array_start = .);
 		KEEP (*(.preinit_array))
 		PROVIDE (__preinit_array_end = .);
 	} >main AT>glob
 
-	.init_array ALIGN(8) :
+	.init_array :
 	{
 		PROVIDE (__init_array_start = .);
 		KEEP (*(SORT(.init_array.*)))
@@ -86,7 +81,7 @@ SECTIONS
 		PROVIDE (__init_array_end = .);
 	} >main AT>glob
 
-	.fini_array ALIGN(8) :
+	.fini_array :
 	{
 		PROVIDE (__fini_array_start = .);
 		KEEP (*(.fini_array))
@@ -94,7 +89,7 @@ SECTIONS
 		PROVIDE (__fini_array_end = .);
 	} >main AT>glob
 
-	.ctors ALIGN(8) :
+	.ctors :
 	{
 		KEEP (*crtbegin.o(.ctors)) /* MUST be first -- GCC requires it */
 		KEEP (*(EXCLUDE_FILE (*crtend.o) .ctors))
@@ -102,7 +97,7 @@ SECTIONS
 		KEEP (*(.ctors))
 	} >main AT>glob
 
-	.dtors ALIGN(8) :
+	.dtors :
 	{
 		KEEP (*crtbegin.o(.dtors))
 		KEEP (*(EXCLUDE_FILE (*crtend.o) .dtors))
@@ -117,30 +112,26 @@ SECTIONS
 
 	__got_end__ = .;
 
-	.data ALIGN(8) :
+	.data :
 	{
 		*(.data .data.* .gnu.linkonce.d.*)
 		SORT(CONSTRUCTORS)
 	} >main AT>glob
 
 	__bss_start__ = .;
-	.bss ALIGN(8) :
+	.main.fill :
 	{
+        FILL(0x00000000)
 		*(.dynbss)
 		*(.bss .bss.* .gnu.linkonce.b.*)
 		*(COMMON)
-		. = ALIGN(16);
+        . = ORIGIN(main) + LENGTH(main) - 1;
+        BYTE(0x00);
 	} >main AT>glob
 
 	__bss_end__ = .;
 	__main_end__ = ABSOLUTE(.) ;
 
-    .main.fill : {
-        FILL(0x00000000)
-        . = ORIGIN(main) + LENGTH(main) - 1;
-        BYTE(0x00);
-    } >main AT>glob
-
 	/* ==================
 	   ==== Metadata ====
 	   ================== */

fusee_cpp/program/source/fusee_crt0.cpp

@@ -28,7 +28,7 @@ namespace ams::nxboot::crt0 {
 
     }
 
-    void Initialize(uintptr_t bss_start, uintptr_t bss_end) {
+    void Initialize() {
         /* TODO: Collect timing information? */
 
         /* Setup exception vectors. */
@@ -43,9 +43,6 @@ namespace ams::nxboot::crt0 {
             SetExceptionVector(7, reinterpret_cast<uintptr_t>(::ams::nxboot::ExceptionHandler7));
         }
 
-        /* Clear bss. */
-        std::memset(reinterpret_cast<void *>(bss_start),      0, bss_end - bss_start);
-
         /* Call init array. */
         __libc_init_array();
     }

fusee_cpp/program/source/fusee_start.s

@@ -62,9 +62,7 @@ _ZN3ams6nxboot5StartEv:
     ldr lr, =_ZN3ams6nxboot16ExceptionHandlerEv
 
     /* Perform runtime initialization. */
-    ldr r0, =__bss_start__
+    bl _ZN3ams6nxboot4crt010InitializeEv
-    ldr r1, =__bss_end__
-    bl _ZN3ams6nxboot4crt010InitializeEjj
 
     /* Perform nx boot procedure. */
     bl _ZN3ams6nxboot4MainEv

libraries/config/arch/arm/cpu/arm7tdmi/cpu.mk

@@ -1,5 +1,5 @@
 export ATMOSPHERE_DEFINES  += -DATMOSPHERE_CPU_ARM7TDMI
-export ATMOSPHERE_SETTINGS += -march=armv4t -mtune=arm7tdmi -mthumb -mthumb-interwork
+export ATMOSPHERE_SETTINGS += -march=armv4t -mtune=arm7tdmi -mthumb -mthumb-interwork -fstrict-volatile-bitfields
 export ATMOSPHERE_CFLAGS   +=
 export ATMOSPHERE_CXXFLAGS +=
 export ATMOSPHERE_ASFLAGS  +=