diff --git a/source/alignedseqmemcpy.h b/source/alignedseqmemcpy.h
new file mode 100644
index 0000000..dc915bf
--- /dev/null
+++ b/source/alignedseqmemcpy.h
@@ -0,0 +1,28 @@
+/*
+*   This file is part of Luma3DS
+*   Copyright (C) 2016-2019 Aurora Wright, TuxSH
+*
+*   This program is free software: you can redistribute it and/or modify
+*   it under the terms of the GNU General Public License as published by
+*   the Free Software Foundation, either version 3 of the License, or
+*   (at your option) any later version.
+*
+*   This program is distributed in the hope that it will be useful,
+*   but WITHOUT ANY WARRANTY; without even the implied warranty of
+*   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+*   GNU General Public License for more details.
+*
+*   You should have received a copy of the GNU General Public License
+*   along with this program.  If not, see <http://www.gnu.org/licenses/>.
+*
+*   Additional Terms 7.b and 7.c of GPLv3 apply to this file:
+*       * Requiring preservation of specified reasonable legal notices or
+*         author attributions in that material or in the Appropriate Legal
+*         Notices displayed by works containing it.
+*       * Prohibiting misrepresentation of the origin of that material,
+*         or requiring that modified versions of such material be marked in
+*         reasonable ways as different from the original version.
+*/
+#include "types.h"
+
+void *alignedseqmemcpy(void *dst, const void *src, u32 len);
diff --git a/source/alignedseqmemcpy.s b/source/alignedseqmemcpy.s
new file mode 100644
index 0000000..6784b69
--- /dev/null
+++ b/source/alignedseqmemcpy.s
@@ -0,0 +1,65 @@
+@   This file is part of Luma3DS
+@   Copyright (C) 2016-2019 Aurora Wright, TuxSH
+@
+@   This program is free software: you can redistribute it and/or modify
+@   it under the terms of the GNU General Public License as published by
+@   the Free Software Foundation, either version 3 of the License, or
+@   (at your option) any later version.
+@
+@   This program is distributed in the hope that it will be useful,
+@   but WITHOUT ANY WARRANTY; without even the implied warranty of
+@   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+@   GNU General Public License for more details.
+@
+@   You should have received a copy of the GNU General Public License
+@   along with this program.  If not, see <http://www.gnu.org/licenses/>.
+@
+@   Additional Terms 7.b and 7.c of GPLv3 apply to this file:
+@       * Requiring preservation of specified reasonable legal notices or
+@         author attributions in that material or in the Appropriate Legal
+@         Notices displayed by works containing it.
+@       * Prohibiting misrepresentation of the origin of that material,
+@         or requiring that modified versions of such material be marked in
+@         reasonable ways as different from the original version.
+
+.section    .text.alignedseqmemcpy, "ax", %progbits
+.arm
+@ Align on cache line boundaries & make sure the loops don't cross them.
+.align      5
+.global     alignedseqmemcpy
+.type       alignedseqmemcpy, %function
+alignedseqmemcpy:
+    @ src=r1 and dst=r0 are expected to be 4-byte-aligned
+    push    {r4-r10, lr}
+
+    lsrs    r12, r2, #5
+    sub     r2, r2, r12, lsl #5
+    beq     2f
+
+1:
+    ldmia   r1!, {r3-r10}
+    stmia   r0!, {r3-r10}
+    subs    r12, #1
+    bne     1b
+
+2:
+    lsrs    r12, r2, #2
+    sub     r2, r2, r12, lsl #2
+    beq     4f
+
+3:
+    ldr     r3, [r1], #4
+    str     r3, [r0], #4
+    subs    r12, #1
+    bne     3b
+
+4:
+    tst     r2, #2
+    ldrneh  r3, [r1], #2
+    strneh  r3, [r0], #2
+
+    tst     r2, #1
+    ldrneb  r3, [r1], #1
+    strneb  r3, [r0], #1
+
+    pop     {r4-r10, pc}
diff --git a/source/cache.s b/source/cache.s
index 996608e..0990022 100644
--- a/source/cache.s
+++ b/source/cache.s
@@ -22,7 +22,7 @@
 @         or requiring that modified versions of such material be marked in
 @         reasonable ways as different from the original version.
 
-.text
+.section .text.cache, "ax", %progbits
 .arm
 .align 4
 
diff --git a/source/crypto.c b/source/crypto.c
index 74e69a1..300df4b 100755
--- a/source/crypto.c
+++ b/source/crypto.c
@@ -36,6 +36,7 @@
 #include "emunand.h"
 #include "strings.h"
 #include "utils.h"
+#include "alignedseqmemcpy.h"
 #include "fatfs/sdmmc/sdmmc.h"
 
 /****************************************************************
@@ -288,24 +289,18 @@ void sha(void *res, const void *src, u32 size, u32 mode)
     sha_wait_idle();
     *REG_SHA_CNT = mode | SHA_CNT_OUTPUT_ENDIAN | SHA_NORMAL_ROUND;
 
-    const u32 *src32 = (const u32 *)src;
-    int i;
+    const u8 *src8 = (const u8 *)src;
     while(size >= 0x40)
     {
         sha_wait_idle();
-        for(i = 0; i < 4; ++i)
-        {
-            *REG_SHA_INFIFO = *src32++;
-            *REG_SHA_INFIFO = *src32++;
-            *REG_SHA_INFIFO = *src32++;
-            *REG_SHA_INFIFO = *src32++;
-        }
+        alignedseqmemcpy((void *)REG_SHA_INFIFO, src8, 0x40);
 
+        src8 += 0x40;
         size -= 0x40;
     }
 
     sha_wait_idle();
-    memcpy((void *)REG_SHA_INFIFO, src32, size);
+    alignedseqmemcpy((void *)REG_SHA_INFIFO, src8, size);
 
     *REG_SHA_CNT = (*REG_SHA_CNT & ~SHA_NORMAL_ROUND) | SHA_FINAL_ROUND;
 
diff --git a/source/start.s b/source/start.s
index 1af74dd..aa5e37f 100644
--- a/source/start.s
+++ b/source/start.s
@@ -22,7 +22,7 @@
 @         or requiring that modified versions of such material be marked in
 @         reasonable ways as different from the original version.
 
-.section .text.start
+.section .text.start, "ax", %progbits
 .align 4
 .global _start
 _start: