From: chaoskagami <chaos.kagami@gmail.com>
Date: Fri, 20 May 2016 20:07:13 +0000 (-0400)
Subject: Okay, I understand how to fix the sysmodule size to grow now. Ugh.
X-Git-Tag: stable-1~60
X-Git-Url: https://chaos.moe/g/?a=commitdiff_plain;h=9dd86897766e474e764a3c60554f5a239de290e9;p=corbenik%2Fcorbenik.git

Okay, I understand how to fix the sysmodule size to grow now. Ugh.
---

diff --git a/source/patch/module.c b/source/patch/module.c
index ccc6cd9..b3c0cc8 100644
--- a/source/patch/module.c
+++ b/source/patch/module.c
@@ -43,16 +43,16 @@ int patch_modules() {
 				uint32_t need_units = (module->contentSize - sysmodule->contentSize);
 				fprintf(stderr, "Module: Would grow %d units but NYI\n", need_units);
 				continue;
-				// FIXME - Adjust sysmodule section and FIRM NCCH. This is not correct.
-/*				fprintf(stderr, "Module: Grow %d units\n", module->contentSize - sysmodule->contentSize);
-				// Location to shuffle to.
-				uint8_t* move_to = ((uint8_t*)sysmodule + (module->contentSize + need_units) * 0x200);
-				uint8_t* move_from = ((uint8_t*)sysmodule + module->contentSize * 0x200);
 
-				uint32_t copy_size = 0x10000; // FIXME - Add a safe way to properly calculate the half of NCCH we need to copy. This is okay for now.
+				// TODO - so in a nutshell, the reason Luma works is because it relocates
+				// the sysmodule section to its final location while fixing the size.
+				// Once stuff is at the right place, it no longer cares about its' size.
 
-				memmove(move_to, move_from, copy_size); */
-				// TODO - This is hackish and possibly incorrect. It needs testing.
+				// Cakes performs an in-place resize and leaves the copy for later; essentially,
+				// this means I need to scale EVERYTHING after the increase, and adjust NCCH sections accordingly.
+
+				// It would be hella easier to just patch at the final location in memory;
+				// and this was actually mentioned on Luma's issues.
 			}
 
 			// Move the remaining modules closer