- Added a damage cap to CR_ACIDDEMONSTRATION of INT_MAX/2 to prevent overflows. It is not capped to INT_MAX because damage increasing skills would cause it to still over flow afterwards.

git-svn-id: https://rathena.svn.sourceforge.net/svnroot/rathena/trunk@8107 54d463be-8e91-2dee-dedb-b68131a5f0ec

2 files changed, 7 insertions, 0 deletions

diff --git a/Changelog-Trunk.txt b/Changelog-Trunk.txt
index 945d2ad21..06835a1c8 100644
--- a/Changelog-Trunk.txt
+++ b/Changelog-Trunk.txt
@@ -3,6 +3,9 @@ Date	Added
 AS OF SVN REV. 5091, WE ARE NOW USING TRUNK.  ALL UNTESTED BUGFIXES/FEATURES GO INTO TRUNK.

 IF YOU HAVE A WORKING AND TESTED BUGFIX PUT IT INTO STABLE AS WELL AS TRUNK.

 2006/08/03

+	* Added a damage cap to CR_ACIDDEMONSTRATION of INT_MAX/2 to prevent

+	  overflows. It is not capped to INT_MAX because damage increasing skills

+	  would cause it to still overflow afterwards. [Skotlex]

 	* Fixed NPC_CRITICALSLASH not being considered as a Critical hit by the

 	  code. [Skotlex]

 	* Added RUDE_ATTACKED_COUNT so you can easily adjust how many rude-attacks

diff --git a/src/map/battle.c b/src/map/battle.c
index 0e53bb9e8..7a7a1d58b 100644
--- a/src/map/battle.c
+++ b/src/map/battle.c
@@ -2620,6 +2620,10 @@ struct Damage  battle_calc_misc_attack(
 	case CR_ACIDDEMONSTRATION: // updated the formula based on a Japanese formula found to be exact [Reddozen]
 		md.damage = 7*tstatus->vit*sstatus->int_*sstatus->int_ / (10*(tstatus->vit+sstatus->int_));
 		if (tsd) md.damage>>=1;
+		if (md.damage < 0 || md.damage > INT_MAX>>1)
+	  	//Overflow prevention, will anyone whine if I cap it to a few billion?
+		//Not capped to INT_MAX to give some room for further damage increase.
+			md.damage = INT_MAX>>1;
 		break;
 	case NJ_ZENYNAGE:
 		md.damage = skill_get_zeny(skill_num ,skill_lv);