KAPATEJIb

Bug confirmed with rev 9867

turn off grid unloading

please use code=diff tag next time diff --git a/src/game/Group.cpp b/src/game/Group.cpp index 3af15a6..004d999 100644 --- a/src/game/Group.cpp +++ b/src/game/Group.cpp @@ -980,9 +980,9 @@ void Group::SendUpdate() // guess size WorldPacket data(SMSG_GROUP_LIST, (1+1+1+1+8+4+GetMembersCount()*20)); data << uint8(m_groupType); // group type (flags in 3.3) - data << uint8(isBGGroup() ? 1 : 0); // 2.0.x, isBattleGroundGroup? data << uint8(citr->group); // groupid data << uint8(citr->assistant ? 0x01 : 0x00); // 0x2 main assist, 0x4 main tank + data << uint8(isBGGroup() ? 1 : 0); // 2.0.x, isBattleGroundGroup? if(m_groupType & GROUPTYPE_LFD) { data << uint8(0); you also should post it to Under Review section, or I can do it for you...

check after 9857

Bug confirmed. patch from ru-mangos.ru by laise diff --git a/src/game/Unit.cpp b/src/game/Unit.cpp index 801c51d..02ddcaf 100644 --- a/src/game/Unit.cpp +++ b/src/game/Unit.cpp @@ -7080,8 +7080,8 @@ bool Unit::HandleDummyAuraProc(Unit *pVictim, uint32 damage, Aura* triggeredByAu if (dummySpell->Id == 49005) { // TODO: need more info (cooldowns/PPM) - triggered_spell_id = 61607; - break; + target->CastSpell(target, 61607, true, NULL, triggeredByAura); + return true; } // Vendetta if (dummySpell->SpellFamilyFlags & UI64LIT(0x0000000000010000))

i was reverted this in my core, but the bug is appears again i think it's not fixed For example one of DK from my server was killed paladin with http://www.wowhead.com/spell=45477 for 2870781277 damage

hack-fix, but it should help with problem diff --git a/src/game/Map.cpp b/src/game/Map.cpp index acb66ec..7f7f9cc 100644 --- a/src/game/Map.cpp +++ b/src/game/Map.cpp @@ -2410,7 +2410,7 @@ bool InstanceMap::Add(Player *player) sLog.outError("GroupBind save players: %d, group count: %d", groupBind->save->GetPlayerCount(), groupBind->save->GetGroupCount()); else sLog.outError("GroupBind save NULL"); - ASSERT(false); + player->RepopAtGraveyard(); } // if the group/leader is permanently bound to the instance // players also become permanently bound when they enter And another diff --git a/src/game/Map.cpp b/src/game/Map.cpp index fc7dab7..1766310 100644 --- a/src/game/Map.cpp +++ b/src/game/Map.cpp @@ -2214,7 +2214,7 @@ bool InstanceMap::Add(Player *player) if(playerBind->save != mapSave) { sLog.outError("InstanceMap::Add: player %s(%d) is permanently bound to instance %d,%d,%d,%d,%d,%d but he is being put in instance %d,%d,%d,%d,%d,%d", player->GetName(), player->GetGUIDLow(), playerBind->save->GetMapId(), playerBind->save->GetInstanceId(), playerBind->save->GetDifficulty(), playerBind->save->GetPlayerCount(), playerBind->save->GetGroupCount(), playerBind->save->CanReset(), mapSave->GetMapId(), mapSave->GetInstanceId(), mapSave->GetDifficulty(), mapSave->GetPlayerCount(), mapSave->GetGroupCount(), mapSave->CanReset()); - ASSERT(false); + player->RepopAtGraveyard(); } } else

it's because of ASSERT(spawns.IsSpawnedObject<T>(obj->guid)); it's return "false", it's mean Object doesn't spawned. I think need to respawn it if object isn't spawned, that what i do in my patch. ofcourse this is just in theory, i didn't test this yet, because i don't know how to reproduce this crash.

I'm trying to find a way how to fix these annoying crashes... #0 0x0061f410 in __kernel_vsyscall () No symbol table info available. #1 0x004d8df0 in raise () from /lib/libc.so.6 No symbol table info available. #2 0x004da701 in abort () from /lib/libc.so.6 No symbol table info available. #3 0x004d226b in __assert_fail () from /lib/libc.so.6 No symbol table info available. #4 0x08390723 in PoolGroup<Creature>::SpawnObject (this=0x988e0e8, spawns=@0xd3715a4, limit=1, triggerFrom=119807, instantly=false) at ../../../src/game/PoolManager.cpp:279 obj = (PoolObject *) 0xa50c0b0 i = 0 lastDespawned = 0 count = 1 __PRETTY_FUNCTION__ = "void PoolGroup<T>::SpawnObject(SpawnedPoolData&, uint32, uint32, bool) [with T = Creature]" __FUNCTION__ = "SpawnObject" #5 0x0838d201 in PoolManager::SpawnPoolGroup<Creature> (this=0xd371528, pool_id=5676, db_guid=119807, instantly=false) at ../../../src/game/PoolManager.cpp:726 No locals. #6 0x08390b3c in PoolManager::UpdatePool<Creature> (this=0xd371528, pool_id=5676, db_guid_or_pool_id=119807) at ../../../src/game/PoolManager.cpp:787 motherpoolid = <value optimized out> #7 0x081aff15 in Creature::Update (this=0x273f9bd8, diff=129) at ../../../src/game/Creature.cpp:407 poolid = 5676 #8 0x0829b07a in VisitorHelper<MaNGOS::ObjectUpdater, Creature> ( v=@0xaa6ee280, c=@0x251c8384) at ../../../src/game/GridNotifiersImpl.h:48 No locals. #9 0x0829cbd1 in VisitorHelper<MaNGOS::ObjectUpdater, Creature, TypeList<DynamicObject, TypeList<Corpse, TypeNull> > > (v=@0xaa6ee280, c=@0x251c8384) at ../../../src/game/../framework/GameSystem/TypeContainerVisitor.h:57 No locals. #10 0x08296c3d in Map::Update (this=0x14cbb8b8, t_diff=@0xaa6ee2b8) at ../../../src/game/../framework/GameSystem/TypeContainerVisitor.h:80 plr = <value optimized out> area = {right_offset = 1, left_offset = 2, upper_offset = 1, lower_offset = 2} updater = {i_timeDiff = 129} __FUNCTION__ = "Update" __PRETTY_FUNCTION__ = "virtual void Map::Update(const uint32&)" #11 0x082a8a35 in MapManager::Update (this=0x130b8140, diff=32) at ../../../src/game/MapManager.cpp:262 I doesn't know if it works, but seems it's a typo in Poolmanager.cpp need some tests. diff --git a/src/game/PoolManager.cpp b/src/game/PoolManager.cpp index d94381d..cef9ab9 100644 --- a/src/game/PoolManager.cpp +++ b/src/game/PoolManager.cpp @@ -276,9 +276,9 @@ void PoolGroup<T>::SpawnObject(SpawnedPoolData& spawns, uint32 limit, uint32 tri if (obj->guid == triggerFrom) { + ReSpawn1Object(obj); ASSERT(spawns.IsSpawnedObject<T>(obj->guid)); ASSERT(spawns.GetSpawnedObjects(poolId) > 0); - ReSpawn1Object(obj); triggerFrom = 0; continue; } If you have ideas how to fix it with more generic way - post it.

Fixed crash with PerformIndoorCheck(). Thanks to przemratajczak diff --git a/src/game/Player.cpp b/src/game/Player.cpp index 988d19c..a1537e2 100644 --- a/src/game/Player.cpp +++ b/src/game/Player.cpp @@ -1384,7 +1384,7 @@ void Player::Update( uint32 p_time ) UpdateEnchantTime(p_time); UpdateHomebindTime(p_time); - if(sWorld.getConfig(CONFIG_UINT32_VMAP_INDOOR_INTERVAL) && + if(IsInWorld() && sWorld.getConfig(CONFIG_UINT32_VMAP_INDOOR_INTERVAL) && (m_IndoorCheckTimer+=p_time) >= sWorld.getConfig(CONFIG_UINT32_VMAP_INDOOR_INTERVAL)) { PerformIndoorCheck();

this is wrong see DIRECT_DAMAGE = 0, // used for normal weapon damage (not for class abilities or spells)

What bug does the patch fix? What features does the patch add? Impemented effect of spell 66118 and difficulty variations. This ability used by Anub'arak in TOC instance For which repository revision was the patch created? 9797 Is there a thread in the bug report section or at lighthouse? If yes, please add a link to the thread. can't find any Who has been writing this patch? Please include either forum user names or email addresses. original author is MaS0n, rewritten by me diff --git a/src/game/SpellAuras.cpp b/src/game/SpellAuras.cpp index d31b27f..f6d48fd 100644 --- a/src/game/SpellAuras.cpp +++ b/src/game/SpellAuras.cpp @@ -8131,6 +8131,30 @@ void Aura::PeriodicDummyTick() case 2: m_target->CastSpell(m_target, 55739, true); break; } return; + case 66118: // Leeching Swarm 10 man + case 68646: + { + int32 damage = (m_modifier.m_amount * m_target->GetHealth()) / 100; + if (damage < 250) + damage = 250; + int32 heal = damage * 68 / 100; + m_target->CastCustomSpell(m_target, 66240, &damage, NULL, NULL, true, NULL, this); + if (Unit* caster = GetCaster()) + m_target->CastCustomSpell(caster, 66125, &heal, NULL, NULL, true, NULL, this); + return; + } + case 67630: // Leeching Swarm 25 man + case 68647: + { + int32 damage = (m_modifier.m_amount * m_target->GetHealth()) / 100; + if (damage < 250) + damage = 250; + int32 heal = damage * 155 / 100; + m_target->CastCustomSpell(m_target, 66240, &damage, NULL, NULL, true, NULL, this); + if (Unit* caster = GetCaster()) + m_target->CastCustomSpell(caster, 66125, &heal, NULL, NULL, true, NULL, this); + return; + } // Exist more after, need add later default: break;

First crash for two days of test and it's related to creature pool spawn system, old crash from clean mangos... average online 150-240. CentOS 5.3 x86. 9797 rev with mods.

seems this patch is outdated after 9798. i was tried to fix it by self, but it makes spell more buggy (does ubnormal amounts of damage) diff --git a/src/game/SpellEffects.cpp b/src/game/SpellEffects.cpp index d629562..ac1725d 100644 --- a/src/game/SpellEffects.cpp +++ b/src/game/SpellEffects.cpp @@ -777,7 +777,10 @@ void Spell::EffectSchoolDMG(SpellEffectIndex effect_idx) else if (m_spellInfo->Id == 54158) { // [1 + 0.25 * SPH + 0.16 * AP] - damage += int32(m_caster->GetTotalAttackPowerValue(BASE_ATTACK) * 0.16f); + float ap = m_caster->GetTotalAttackPowerValue(BASE_ATTACK); + int32 holy = m_caster->SpellBaseDamageBonusDone(GetSpellSchoolMask(m_spellInfo)) + + unitTarget->SpellBaseDamageBonusTaken(GetSpellSchoolMask(m_spellInfo)); + damage += int32(ap * 0.16f) + int32(holy * 25 / 100); } break; } I still can't understand how to correctly update some patches after this commit Any updates?

There is no crashes, lags or freezes for 20 hours with online 100-250. Seems it's a record for my server after 3.3.3 switch it still have problems with compilation at some platforms as you can see... maybe i will make a copy-branch of 333 with this patch. it's easy to merge, use part from Theirs at Map.cpp, use "Mine before theirs" at World.h . It's simple if you using TortoiseGIT tool.

I has 6 hours server uptime now (after restart) with ~180 ppl online. My players live at BG's (because of new random BG system). Can't find any bug related to new Vmaps. I think this can't be called *BETA* version, it works like *release* CentOS 5.3, 9797 rev (own branch with ~130 custom patches), YTDB 543

yes it's fixed, there is no freezes at using combination that i posted before

now when i know how to reproduce i'm going to test it on clean+your patch to be 100% sure. I will post result... EDIT: YES! It freezes! now i'm start thinking it's related to CLEAN core without your patch, but i can't check it because already deleted original Vmaps :cool:

i was pushed LATEST repo as i already said, recompiled many times and has same result. And i can't start your extractor even if i installed MS visual C++ 2010 Redist package from link that posted TOM_RUS, extractor show me same erros - missing files MSVCP100.dll and some others. And I am already tried to find this DLL on some DLL sharing sites, but they won't work. about freezes: i was disabled some mob spells with knockback effect (33822 and 55715, both uses spell effect 98) freezes is gone away... as i already said i can't be sure if this freezes related to your patch. But i recommend you check them too. How to easy reproduce my freeze: .npc add 19394 select him .cast 33822

I'm testing faramir's patch at live Linux-based server, what i can say... i didn't see any bugs related to new vmaps. All works ok, only experienced some server freezes related to SPELL_EFFECT_98, but i can't say if they related to your patch, because i'm installed this http://ru-mangos.ru/showthread.php?p=5222#post5222 too and updated from 9796 to 9799. My server uses CentOS 5.3 with ~150 online.

this means i should have less amount of files in vmap folder? Pushed latest changes from your repo, recompiled, deleted old vmap and buildings folder, re-extracted with vmapExtractor3 then vmap_assembler but has SAME amount of files! 8885 files in vmap folder. Still has many small files (<1 kbyte) anyway i will try to use it on my Linux-based live server, will see what can happen :eek:

i don't know... it's difficult to find anything through this FLOOD! Confirmed, only first time when i started server. Then i used .modify scale 0.5 on my character and enter to this room ^^. Later when i was restarted server i was going through doorways with normal size without any problem And i think this would be better to use vmap.indoorCheckInterval = 1000 as default value to not confuse noobs like me I will test your latest update today or tomorrow...

ok, unmounting now works forgot about _default_ value (it's disabled)... but still have flood in server console, it's difficult to find any useful information about events or commands used in console with this source of flood here + printf("Got AreaInfo: flag %u, adt %i, root %i, group %i\\n", mogpFlags, adtId, rootId, groupId); + printf("Got WMOAreaTableEntry! flag %u, areaid %u\\n", wmoEntry->Flags, wmoEntry->areaId); + printf("Got AreaTableEntry\\n"); i think "pringf" shouldn't be used here, maybe sLog.outDebug?

I'm still getting same flood in console with faramir's branch without any other mods. Vmaps are extracted and replaced ofcourse. Unmounting still not work how can this happen? Maybe i need to rename ruRU folder to enGB or something like this? Or it just a problem with extractor that compiled in VS 2008 and only VS2010 users get _normal_ Vmap data? king48488, can you check it with extractor or vmaps that i was upload to filekeeper and posted on prev page? Or just upload your Vmap folder to something like filekeeper?

yes, i am using patch with my repo. I will try clean now, but i think my core mods NOT related to that problem. This can happen because of Ru-ru locale?