PDA

View Full Version : New bug?



Catt
04-18-2003, 05:21 PM
I'm getting an error, right after zoning I get:

Zone: EntryCode: Server, Zone: poknowledge
Lost sync, relog or zone to reset
uncompress failed on 0x21f6: buffer error
Your player's id is 1452

I don't see that anyone else has this trouble, anyone got any ideas?

Catt
04-18-2003, 06:01 PM
A little more info on this issue.

It seems to be just 1 character, I've loaded others on the same account into the same zone, no troubles.

Most obvious difference aside from levels is the one with the trouble has his epic ??

mvern
04-18-2003, 08:44 PM
I think I have an idea of whats causing it, will probably require a new libEQ to fix it, still looking into it tho.

Amadeus
04-19-2003, 01:01 AM
I cannot tell you how cool it is to have someone like Mvern there, responding, and working deligently when things go wrong. And, yes, if I could think of any more ways of kissing his ass, I'd do so ..hehe :D

Catt
04-19-2003, 05:06 AM
Thanks for the replies and hope. :)

I've done a little more testing, had a friend with a very similar character (with epic, same class, couple levels higher, etc) sign on and he had no troubles.

Anything I might try to help figure it out, or even a hint what it's related to?

Stoneguy
04-19-2003, 05:49 AM
Greetings,

I'm having the same issue. I really appreciate your dedication to this project, and without you we would be in trouble. My character has his epic as well, if that makes any difference, although it is banked.

Thank you for your time.

Catt
04-19-2003, 06:38 AM
I found a work around for this, it seems to be tied to how much stuff a character has. I moved 3 bags (10 slot) out of bank to a mule, and now have no problem.

Soneguy, you a packrat too?

I tried to find the eror text in the code, it doesn't seem to be there, it coded into LibEQ?

Spaz
04-19-2003, 09:05 AM
I've moved things between characters rather extensively and some still crap out, others still do not.

Stripped one down quite a bit.

Some more info: It happens with accounts that are LOY enabled and accounts that are not, level doesn't seem to be an issue, I have a mid 30's, a mid 50's and a 65 that all do it. Also have a number that do not do it.

Stoneguy
04-19-2003, 12:16 PM
An update,

I only seem to not get the error when I zone naked, ie after I die and when I get rezzes.

That's all, thanks.

Stoneguy
04-20-2003, 02:30 AM
I think I've tracked the problem down to the number of total occupied slots in your character's inventory. At the moment I'm hovering around the limit, and if I don't keep stuff out of my inventory the bug re-manifests itself.

I hope this helps.

LordCrush
04-20-2003, 05:57 AM
Originally posted by Catt
I'm getting an error, right after zoning I get:

Zone: EntryCode: Server, Zone: poknowledge
Lost sync, relog or zone to reset
uncompress failed on 0x21f6: buffer error
Your player's id is 1452

I don't see that anyone else has this trouble, anyone got any ideas?

Same Error here and unknowns too - have LoY and my bank is nearly full ... i ll try to do some investigation after Easter ... RL issues ;)

/wave

-- LC

Spaz
04-20-2003, 07:40 AM
Stoneguy - what does the limit seem to be?

And do you mean char inventory top level slots, or bank slots, or a combination of both?

LordCrush
04-20-2003, 10:38 AM
I have tried a char lvl 9 with nothing in bank and 6 items in char inventory

I have nearly no unknown spawns

After 3 hrs

Thom_Merrilin
04-20-2003, 01:56 PM
With me, I'm getting some slightly different results...

I seem to be suffering from "warm up time". When SEQ starts, it gives me all those unknowns, and refuses to show ground spawn(dropped items and world containers). However, after a while of SEQ running, it somehow corrects itself, and works flawlessly from then on.

There, of course, might be other issues at hand, but this is what I seem to have noticed.

Mic
04-20-2003, 03:57 PM
So, I'm not sure if this error is same error, but rather then start a new thread. So, here goes.
seq loads up fine, but it doesen't decode and crashes everytime I zone.

Error: PlayerItemCode (6441) (dateLen:10 != sizeof(playerItemStruct):468)!
Segmentation fault

I'm currently useing qt 3.0.6, the new libEQ.a, and all the new seq files. Any clue on what could be the problem?

Catt
04-21-2003, 04:41 AM
Originally posted by Mic
So, I'm not sure if this error is same error, but rather then start a new thread. So, here goes.
seq loads up fine, but it doesen't decode and crashes everytime I zone.

Error: PlayerItemCode (6441) (dateLen:10 != sizeof(playerItemStruct):468)!
Segmentation fault

I'm currently useing qt 3.0.6, the new libEQ.a, and all the new seq files. Any clue on what could be the problem?

That sounds like a bad libEQ.a get the current one get rid of all others on the whole PC, do a clean make.

snoopman41
04-22-2003, 05:55 AM
same error


Lost sync, relog or zone to reset
uncompress failed on 0x21f6: buffer error


- with a char i have always this error and nothing is decoded

- with a another one i do not get this error, but nothing is decoded except static ground spawn (like yew leaves), other players in the zone are well decoded, but own char position is not showed on the map.

LordCrush
05-23-2003, 06:11 AM
I still get this error


decoding 0x7039 with 0x91852b6b
clean uncompress on 0x3039:
decoding 0x7154 with 0x6efadcd2
clean uncompress on 0x3154:
decoding 0x7154 with 0xee4a74b0
clean uncompress on 0x3154:
SpellItem 'Tangling Weeds' finished.
clean uncompress on 0x101f:
clean uncompress on 0x101f:
clean uncompress on 0x101f:
decoding 0x7154 with 0x0d7a9621
Lost sync, relog or zone to reset
uncompress failed on 0x3154: data error
clean uncompress on 0x1287:
SPELL: You begin casting Tangling Weeds. Current Target is spawn(30277631)
selfStartSpellCast - id=2591
SpellItem 'Tangling Weeds' finished.
clean uncompress on 0x1287:
clean uncompress on 0x101f:


after that all new spawns show up as unknown...

From the code the Error-Message comes from libEQ.cpp, and i remeber that mvern posted he perhaps had an idea what it would come from ...

Has somebody an idea ?

Thanx in advance =)

-- LC

codepig666
05-23-2003, 10:52 PM
When compress fails with a buffer error, it means the buffer was too small. I fixed this in last weeks patch.

The error you are getting now is different LordCrush =*(

The "data error" is more sinister. basically, if a decompression error occurs. SEQ is set to give up on trying to decode until you zone again.

The heart of the problem is the very first decompression failure. I have not yet experienced this problem firsthand, so I am hoping it is isolated.


Anyone else with the same error as in Lord Crush's post just prior to this one please post it .

Thanks
piggy

LordCrush
05-26-2003, 02:55 AM
piggy thank you for response :)

if i can help debugging this - i am not so experienced in coding, but it will be good enough to get out some more information ... Do you have a hint where to start ?

Thnx

-- LC

Bubba69
05-26-2003, 05:15 PM
If this has anything to bank slots, could the 2 new (account shared) bank slot that they are adding affect it? The slots are already on test atm. Just a shot in the dark

TrainSpotter
05-26-2003, 06:05 PM
I dont think it has to do with the new bank slots. They would cause everyone to get that problem, not just a few.

Then again, i could be wrong.

codepig666
05-27-2003, 06:45 PM
Lord Crush:

Are you still getting the data error on decompress?

If so, can you do a complete purge of your seq and pull it all down from cvs again. Be sure to delete any and all libEQ.a files you have too.

If you do a completely clean install after purging everything, you will see a billion alerts about how it is set to not decrypt, but it will still decrypt just fine.

If you do all that and still get the error, post again please.

--piggy

TrainSpotter
05-27-2003, 07:36 PM
Codepigg,

I am getting that same error. I will try to purge and re-install/cvs the latest versions and see where that will get me. As of now, nothing is uncompressing for me. Map, character, nothing...

will let ya know in like 20 min if this works.

Zoolander
05-27-2003, 10:40 PM
I'm not seeing any data errors decompressing, but I do have a repeatable crash with only one of my characters. The information printed by the Player::backfill() function is wrong. The character name (which I modified) has the first 2 characters cut off, the character has no last name, the level should be 25, the player money and bank money are both incorrect, and the expAA value seems wrong since the character is level 25. I have other characters with the same race, class, level, and number of letters in their name that don't cause the crash although the information printed by Player::backfill() is also incorrect for them in a similar way. The struct length warning is also duplicated on other characters that work correctly.

Here's a log and backtrace. This is from a checkout into a brand new directory today. I deleted the /usr/local/share/showeq directory before I did 'make install' and there is no libEQ.a on my machine.

Opcode Logging Mask: 0 0 0
Client Detected: 192.168.1.100
GuildMgr: new guildsfile written
GuildMgr: guildsfile loaded
Zone: EntryCode: Client
WARNING: CharProfileCode (006b) (dataLen:3926 != sizeof(charProfileStruct):14302)!
Player::backfill():
Player: Name: 'cdefg' Last: ''
Player: Level: 493
Player: PlayerMoney: P=0 G=0 S=0 C=0
Player: BankMoney: P=0 G=0 S=0 C=0
Player: Exp =18,415,616
Player: ExpAA =115

Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 16384 (LWP 31519)]
0x402e85da in QTextString::toString(QMemArray<QTextStringChar> const&, bool) () from /usr/qt/3/lib/libqt-mt.so.3
(gdb) bt
#0 0x402e85da in QTextString::toString(QMemArray<QTextStringChar> const&, bool) () from /usr/qt/3/lib/libqt-mt.so.3
#1 0x402e8694 in QTextString::toString(bool) const () from /usr/qt/3/lib/libqt-mt.so.3
#2 0x4031b4ca in QLineEdit::text() const () from /usr/qt/3/lib/libqt-mt.so.3
#3 0x4031b104 in QLineEdit::setText(QString const&) () from /usr/qt/3/lib/libqt-mt.so.3
#4 0x4036dee6 in QSpinBox::updateDisplay() () from /usr/qt/3/lib/libqt-mt.so.3
#5 0x4036d85e in QSpinBox::setValue(int) () from /usr/qt/3/lib/libqt-mt.so.3
#6 0x40563f17 in QSpinBox::qt_invoke(int, QUObject*) () from /usr/qt/3/lib/libqt-mt.so.3
#7 0x4026b419 in QObject::activate_signal(QConnectionList*, QUObject*) () from /usr/qt/3/lib/libqt-mt.so.3
#8 0x4026b4f5 in QObject::activate_signal(int, int) () from /usr/qt/3/lib/libqt-mt.so.3
#9 0x080f73a7 in Map::panXChanged(int) (this=0xaaaaaaab, t0=-1431655765) at qmetaobject.h:261
#10 0x080ef2a0 in Map::mapUnloaded() (this=0xbfffe288) at mapcore.h:101
#11 0x080f74f4 in Map::qt_invoke(int, QUObject*) (this=0x82f2e88, _id=1080471908, _o=0xbfffe288) at m_map.cpp:1146
#12 0x4026b419 in QObject::activate_signal(QConnectionList*, QUObject*) () from /usr/qt/3/lib/libqt-mt.so.3
#13 0x4026b2bd in QObject::activate_signal(int) () from /usr/qt/3/lib/libqt-mt.so.3
#14 0x080f664e in MapMgr::mapUnloaded() (this=0xaaaaaaab) at qmetaobject.h:261
#15 0x080d4115 in MapMgr::zoneBegin(QString const&) (this=0x4066b164, shortZoneName=@0xaaaaaaab) at map.cpp:299
#16 0x080f66d3 in MapMgr::qt_invoke(int, QUObject*) (this=0x83535f8, _id=-1073749100, _o=0x4066b164) at qucom_p.h:449
#17 0x4026b390 in QObject::activate_signal(QConnectionList*, QUObject*) () from /usr/qt/3/lib/libqt-mt.so.3
#18 0x4026b6e5 in QObject::activate_signal(int, QString) () from /usr/qt/3/lib/libqt-mt.so.3
#19 0x08149f17 in ZoneMgr::zoneBegin(QString const&) (this=0xaaaaaaab, t0=@0xaaaaaaab) at qmetaobject.h:261
#20 0x08149a80 in ZoneMgr::zoneEntry(ServerZoneEntryStruct const*, unsigned, unsigned char) (this=0xbfffe4b8,
zsentry=0x832f880, len=2863311531, dir=2 '\002') at zonemgr.cpp:153
#21 0x0814a506 in ZoneMgr::qt_invoke(int, QUObject*) (this=0x832f880, _id=5, _o=0xbfffe600) at qucom_p.h:312
#22 0x4026b390 in QObject::activate_signal(QConnectionList*, QUObject*) () from /usr/qt/3/lib/libqt-mt.so.3
#23 0x08092cfc in EQPacket::zoneEntry(ServerZoneEntryStruct const*, unsigned, unsigned char) (this=0x828e1f0,
t0=0xaaaaaaab, t1=398, t2=2 '\002') at m_packet.cpp:1786
#24 0x0808adb2 in EQPacket::dispatchZoneData(unsigned, unsigned char*, unsigned char) (this=0x828e1f0, len=398,
data=0x81da8a0 "T\002x\223Ä", dir=2 '\002') at packet.cpp:2453
#25 0x0808863c in EQPacket::dispatchPacket(int, unsigned char*) (this=0x828e1f0, size=181,
buffer=0xaaaaaaab <Address 0xaaaaaaab out of bounds>) at packet.h:429
#26 0x08087c9e in EQPacket::processPackets() (this=0x68a) at packet.cpp:787
#27 0x08094a4c in EQPacket::qt_invoke(int, QUObject*) (this=0x828e1f0, _id=2, _o=0xbfffeee0) at m_packet.cpp:2177
#28 0x4026b419 in QObject::activate_signal(QConnectionList*, QUObject*) () from /usr/qt/3/lib/libqt-mt.so.3
#29 0x4026b2bd in QObject::activate_signal(int) () from /usr/qt/3/lib/libqt-mt.so.3
#30 0x4054dd0b in QTimer::timeout() () from /usr/qt/3/lib/libqt-mt.so.3
#31 0x4028c4e2 in QTimer::event(QEvent*) () from /usr/qt/3/lib/libqt-mt.so.3
#32 0x4020d544 in QApplication::internalNotify(QObject*, QEvent*) () from /usr/qt/3/lib/libqt-mt.so.3
#33 0x4020cabb in QApplication::notify(QObject*, QEvent*) () from /usr/qt/3/lib/libqt-mt.so.3
#34 0x401e80d5 in QEventLoop::activateTimers() () from /usr/qt/3/lib/libqt-mt.so.3
#35 0x401c58b8 in QEventLoop::processEvents(unsigned) () from /usr/qt/3/lib/libqt-mt.so.3
#36 0x40221386 in QEventLoop::enterLoop() () from /usr/qt/3/lib/libqt-mt.so.3
#37 0x40221228 in QEventLoop::exec() () from /usr/qt/3/lib/libqt-mt.so.3
#38 0x4020d771 in QApplication::exec() () from /usr/qt/3/lib/libqt-mt.so.3
---Type <return> to continue, or q <return> to quit---
#39 0x080654ca in main (argc=1, argv=0xbffff764) at main.cpp:930
#40 0x408137a7 in __libc_start_main () from /lib/libc.so.6

Jel321
05-28-2003, 06:50 AM
Thats pretty much the same crash Im seeing Zoo and the backtrace i see is also very close. Of course it happens mostly on my main character ;(

Sometimes its instaneous on zone, sometimes Ill get the map loaded and rarely ill get a full decode before it happens.

I thought it might be something QT related (base was Redhat 9 with QT311 from the OS install) so i went back and loaded up QT305, grabbed a fresh CVS copy and recompiled, same result.

Do you get the Backfill information by default or did you have to uncomment some stuff to see that?

J

Zoolander
05-28-2003, 08:56 PM
I didn't change anything to get the backfill info. It happens every time when this particular character zones and sometimes when other characters zone. Backfill sounds like something causes the packets to get queued for a bit until something happens and then they are parsed and applied.

codepig666
05-29-2003, 12:17 PM
The crash when processing the player packet is because I didn't clean that struct up yet. sorry.

It's just random luck as to (probably) what is in your inventory when you zone.

Instead of fixing the struct, I just padded it out to the right length.

Most of the data is correct (exp, name, spawnId,skills, etc).

The stuff that is wrong (INT,WIS,STR,etc) is just hitting an invalid value and blowing up.

I'll try to clean it up tonight. It's a big friggin packet, so no promises.

-piggy

Jel321
05-29-2003, 12:53 PM
Hey thanks a ton for letting us know that its a code issue not something we are doing bad heheh

No appologies needed bud :)

and yeah i looked at the struct and was like OMFG!

LordCrush
05-29-2003, 04:09 PM
Originally posted by codepig666
Lord Crush:

Are you still getting the data error on decompress?

If so, can you do a complete purge of your seq and pull it all down from cvs again. Be sure to delete any and all libEQ.a files you have too.

If you do a completely clean install after purging everything, you will see a billion alerts about how it is set to not decrypt, but it will still decrypt just fine.

If you do all that and still get the error, post again please.

--piggy

Hi Piggy,

i think i have fixed it now ... I have been playing the whole evening today without any unknown spawns :D :D

I replaced the hub where the SEQ-Box and the 2 EQ-PCs were connected to. ( The old hub did work fine for the last 1,5 years with SEQ and i have no clue what is wrong with it ... )

I am going on a business trip for the next 2 weeks and cannot investigate more but i will look deeper into this when i am back.

Thank you for offering help and for fixing SEQ - works great - WOOOOT

-- LC