Jump to content
The Dark Mod Forums

Beta testing 2.10


stgatilov

Recommended Posts

@Amadeus

The crash happens when the engine attempts to resize the vertex memory.

Maybe the fix for X-Ray views increased the memory usage?

Setting a higher default value alleviates the crash:

seta r_frameIndexMemory "8192"
seta r_frameVertexMemory "8192"

@cabalistic @stgatilov @duzenko

  • Like 1

Please visit TDM's IndieDB site and help promote the mod:

 

http://www.indiedb.com/mods/the-dark-mod

 

(Yeah, shameless promotion... but traffic is traffic folks...)

Link to comment
Share on other sites

Ok, changing shadow mode to "Maps" worked fine and the mission no longer crashes for me! I did try changing some video settings before posting here, but I didn't think about shadows.

As a side note, I did try executing TDM in GDB as suggested, and when I encounter the crash, the TDM window freezes but doesn't disappear, and it seems that a big part of my entire desktop freezes. I still can change GNOME 3's virtual desktops with Ctrl + Alt + Uparrow or Downarrow, but other things seems to be frozen, including the mouse, forcing me to reboot the PC with its On/Off button. I tried two times and this happened two times. Very strange. So I couldn't copy the contents of the GDB console, because I had to reboot. Anyway, this is not needed anymore, as the problem is resolved.

As another side note, even the first problem (the one about frozen aim during the charge of a shot) seems to be gone! I tried with different missions, even with my original settings and I couldn't reproduce it anymore. Computers will never stop surprising me :)

So, another time, really huge thank you. I hope my posts have been (or will be) useful in some ways for this game.

  • Like 3
Link to comment
Share on other sites

Oh this definitely helps!

I feel like this might be a 2.10 release blocker since it is possible that any large mission could cause vertex buffer resizing so we should handle that gracefully rather than crashing. That said, it's rare and we have 2 workarounds now so who knows.

  • Like 1

Please visit TDM's IndieDB site and help promote the mod:

 

http://www.indiedb.com/mods/the-dark-mod

 

(Yeah, shameless promotion... but traffic is traffic folks...)

Link to comment
Share on other sites

14 minutes ago, nbohr1more said:

Setting a higher default value alleviates the crash:

 

Just tried it, that seems to have fixed it. If xray views is the culprit then other FMs with xrays should probably be tested to using Stencil Shadows as well. 

@Dave the TafferSorry you had to deal with the crash! Hopefully you will enjoy the mission now though :)

Edited by Amadeus
  • Like 1
Link to comment
Share on other sites

7 hours ago, nbohr1more said:

I feel like this might be a 2.10 release blocker since it is possible that any large mission could cause vertex buffer resizing so we should handle that gracefully rather than crashing. That said, it's rare and we have 2 workarounds now so who knows.

The engine already can resize vertex buffer.
The problem is most likely elsewhere.

UPDATE: For me, the crash happens immediately on startup.
It does not happen with "r_useMultiDrawIndirect 1".

Link to comment
Share on other sites

Awwwww, on a Nth retry I was gifted with some debug context wisdom

Capture.PNG

This happens when index buffer resizes and some frame-temporary surfaces don't have valid VBO handles.

So what I think the difference between the new and old backends is - the old sends the client-side index data to the draw routine while the new ignores the error and still sends nullptr (with index buffer unbound) to driver.

I'm going to let @cabalistic and @stgatilov decide on it

Capture.thumb.PNG.d79d64abced00970019c2168fd9eb0b1.PNG

  • Like 3
Link to comment
Share on other sites

3 hours ago, duzenko said:

This happens when index buffer resizes and some frame-temporary surfaces don't have valid VBO handles.

So what I think the difference between the new and old backends is - the old sends the client-side index data to the draw routine while the new ignores the error and still sends nullptr (with index buffer unbound) to driver.

I'm going to let @cabalistic and @stgatilov decide on it

I see the error on the same call stack.

I don't understand why multi-draw-indirect code path works, and this one crashes. There is no check for invalid buffers in either of them.

Perhaps we can simply check for validity of vertex&index cache here:

	vertexCache.VertexPosition( attribBind == ATTRIB_REGULAR ? drawSurfs[0]->ambientCache : drawSurfs[0]->shadowCache, attribBind );
	vertexCache.IndexPosition( drawSurfs[0]->indexCache );

And skip drawing if any of the caches is invalid?

UPDATE: Ok, I think multi-draw-indirect goes away because: "Note that indices stored in client memory are not supported. If no buffer is bound to the GL_ELEMENT_ARRAY_BUFFER binding, an error will be generated." Ordinary draw calls support client memory, so the driver treats zero offset as a pointer, leading to crash.

  • Like 1
Link to comment
Share on other sites

1 hour ago, stgatilov said:

I see the error on the same call stack.

I don't understand why multi-draw-indirect code path works, and this one crashes. There is no check for invalid buffers in either of them.

Perhaps we can simply check for validity of vertex&index cache here:

	vertexCache.VertexPosition( attribBind == ATTRIB_REGULAR ? drawSurfs[0]->ambientCache : drawSurfs[0]->shadowCache, attribBind );
	vertexCache.IndexPosition( drawSurfs[0]->indexCache );

And skip drawing if any of the caches is invalid?

UPDATE: Ok, I think multi-draw-indirect goes away because: "Note that indices stored in client memory are not supported. If no buffer is bound to the GL_ELEMENT_ARRAY_BUFFER binding, an error will be generated." Ordinary draw calls support client memory, so the driver treats zero offset as a pointer, leading to crash.

I'm not sure why indirect path does not crash - maybe its group's first handle is valid, and the invalid handles don't unbind the index VBO

Link to comment
Share on other sites

1 hour ago, stgatilov said:

beta210-08 is available.

Please everyone who had crashes with previous version (e.g. on Written In Stone), try again.

Hi everyone,

I installed 2.10 beta 8 updating over beta 7, and now even with shadow mode set to "Stencil" Written in Stone no longer crashes near the start on my system.

I add that yesterday evening I played and almost finished Written in Stone with beta 7 and shadow mode set to "Maps" without any problem.

  • Like 1
Link to comment
Share on other sites

22 minutes ago, AluminumHaste said:

Which version were you updating from? 2.10b7? Or another version?

b6, hadn't got around to looking at b7.

Will try a two step upgrade (b6->b7->b8) and see if that works.

Edit: Yes, that seems to have done the trick.

Hmm.  Feels like a bug, or at least a misfeature, in the installer?

Edited by Araneidae
Link to comment
Share on other sites

1 hour ago, Araneidae said:

Alas:

fail.png

Most likely you tried to update while new version was being synchronized to mirrors.
Just try again: it will more likely work.

To be honest, I thought mirrors will temporarily disable HTTP serving while rsync is in process...

  • Like 2
Link to comment
Share on other sites

Hello Folks 😀

This is my first Post in the forum (but I´m your Fan for much longer )

and i just Registered Cause there is some ennoying stuff with the Engine or your Host-Servers !

To clean my Computer and to prepare a Virgin TDM Database for my Brother ,

Yesterday i made a totally new Tdm Installation with 2.09b and a Second ! with beta210-07

than i started to download the Missions from the ingame-online-Archive 

This worked realy Horrible :

To clear a Mission as Downloaded (and not anymore as  available , so you know you have the newest stuff )

one has to Download it mostly 3 sometimes 4/5 times !!!!!!

1.:  Download "Mission-name"

2.: Download "Mission-name *"

3. Download "Mission.name #"

sometimes these where enough , sometimes I had to do it again ,

Sometimes it was cleared fromm the available (But reappeared by the next start of the Game)

Sometimes instad of cleared it appeared 2or3 times in the List .

This Happend with both Installation 2.09b / beta210-07

and it seemes also to be with my fomer 2.05 updated to 2.08 Install

For a Totally NEW TDM-Install ONE Mission-Download of the newest Mission-Version should be good !!!

Maybe You all have the Missions updated time by time at the Moment a new version had been out ,

so nobody realised this  or your Download-Server gives them one by one : but if so :

a pure Mission Update shouldend be as big  as the Original Mission .

Hopfully you can look whats the reason of this behaveiour and solve it for other/new Users .

     Happy Greetings  To all of you 

You all earned at least a GOLD-Medal  

 

P.S. :  I just made a Test with my old Tdm 2.05 install 

in The Ingame Downloader appear all of the new Missions 

so I loaded some of the Newer :  Deadeye / The Factory Heist 

afterwords loading does , install goes -- but starting remains Black !

Not a big Deal -- these are made for newer Versions of TDM 

But not everybody updates to the Newest Version

So it would be fine to make it known (visible before Download)

in the Game Discription underneath : Which TDM-Version is needed to play

Edited by Bergante
  • Like 2
Link to comment
Share on other sites

A few players have reported guards being weirdly alert to every little noise in my map Newford Road.

I have done some experimenting and I can confirm that it is an issue with the beta, and it seems to be OS dependent.

On a hunch I dug out an old dual boot machine, with one partition running Mint 20.3 and the other running Windows 10. I installed the TDM2.10b8 on both and played the mission. On Linux the map behaved as I expected. On Windows, dropping the glass sheet on the clock tower made every guard in the map go berserk.

I then downgraded to TDM2.09b on the Windwows partition and tried again; although the map uses some 2.10 assets and features it will run OK for a while and I could confirm that the guards do NOT freak out at every tiny sound.

So whatever the issue is, it was introduced at some point between 2.09b and 2.10b5, since during beta testing one of the testers reported the problem as early as the fifth beta. I just could never replicate it until now.

edit: 2.10 beta 1 is is affected too.

Here are the specs for the potato I was running these tests on, in case there is a clue in there:

Spoiler
System:    Kernel: 5.4.0-100-generic x86_64 bits: 64 compiler: gcc v: 9.3.0 Desktop: MATE 1.26.0 
           wm: marco dm: LightDM Distro: Linux Mint 20.3 Una base: Ubuntu 20.04 focal 
Machine:   Type: Desktop Mobo: Acer model: Aspire XC-730 serial: <filter> 
           UEFI: American Megatrends v: R01-A3 date: 07/11/2017 
CPU:       Topology: Quad Core model: Intel Pentium J4205 bits: 64 type: MCP arch: Goldmont rev: 9 
           L2 cache: 1024 KiB 
           flags: lm nx pae sse sse2 sse3 sse4_1 sse4_2 ssse3 vmx bogomips: 11980 
           Speed: 2396 MHz min/max: 800/2600 MHz Core speeds (MHz): 1: 2396 2: 2432 3: 2581 
           4: 2471 
Graphics:  Device-1: Intel Celeron N3350/Pentium N4200/Atom E3900 Series Integrated Graphics 
           vendor: Acer Incorporated ALI driver: i915 v: kernel bus ID: 00:02.0 chip ID: 8086:5a84 
           Display: x11 server: X.Org 1.20.13 driver: modesetting unloaded: fbdev,vesa 
           compositor: marco resolution: 1920x1080~60Hz 
           OpenGL: renderer: Mesa Intel HD Graphics 505 (APL 3) v: 4.6 Mesa 21.2.6 
           direct render: Yes 
Audio:     Device-1: Intel Celeron N3350/Pentium N4200/Atom E3900 Series Audio Cluster 
           vendor: Acer Incorporated ALI driver: snd_hda_intel v: kernel bus ID: 00:0e.0 
           chip ID: 8086:5a98 
           Sound Server: ALSA v: k5.4.0-100-generic 
Network:   Device-1: Realtek RTL8111/8168/8411 PCI Express Gigabit Ethernet 
           vendor: Acer Incorporated ALI driver: r8169 v: kernel port: e000 bus ID: 01:00.0 
           chip ID: 10ec:8168 
           IF: enp1s0 state: down mac: <filter> 
           IF-ID-1: usb0 state: unknown speed: N/A duplex: N/A mac: <filter> 
Drives:    Local Storage: total: 931.51 GiB used: 31.05 GiB (3.3%) 
           ID-1: /dev/sda vendor: Toshiba model: DT01ACA100 size: 931.51 GiB speed: 6.0 Gb/s 
           serial: <filter> 
Partition: ID-1: / size: 357.10 GiB used: 31.00 GiB (8.7%) fs: ext4 dev: /dev/sda6 
USB:       Hub: 1-0:1 info: Full speed (or root) Hub ports: 8 rev: 2.0 chip ID: 1d6b:0002 
           Device-1: 1-1:2 info: SHARKOON 2.4G Mouse type: Mouse driver: hid-generic,usbhid 
           rev: 1.1 chip ID: 1ea7:0064 
           Device-2: 1-2:3 info: Samsung Galaxy series misc. (tethering mode) 
           type: Bluetooth,CDC-Data driver: rndis_host rev: 2.0 chip ID: 04e8:6863 
           Device-3: 1-3:4 info: SiGma Micro Keyboard TRACER Gamma Ivory type: Keyboard,HID 
           driver: hid-generic,usbhid rev: 1.1 chip ID: 1c4f:0002 
           Device-4: 1-8:5 info: Realtek 3-in-1 (SD/SDHC/SDXC) Card Reader type: Mass Storage 
           driver: ums-realtek rev: 2.0 chip ID: 0bda:0153 
           Hub: 2-0:1 info: Full speed (or root) Hub ports: 7 rev: 3.0 chip ID: 1d6b:0003 
Sensors:   System Temperatures: cpu: 49.0 C mobo: N/A 
           Fan Speeds (RPM): N/A 
Repos:     No active apt repos in: /etc/apt/sources.list 
           Active apt repos in: /etc/apt/sources.list.d/official-package-repositories.list 
           1: deb http: //packages.linuxmint.com una main upstream import backport #id:linuxmint_main
           2: deb http: //archive.ubuntu.com/ubuntu focal main restricted universe multiverse
           3: deb http: //archive.ubuntu.com/ubuntu focal-updates main restricted universe multiverse
           4: deb http: //archive.ubuntu.com/ubuntu focal-backports main restricted universe multiverse
           5: deb http: //security.ubuntu.com/ubuntu/ focal-security main restricted universe multiverse
           6: deb http: //archive.canonical.com/ubuntu/ focal partner
Info:      Processes: 175 Uptime: 7m Memory: 7.62 GiB used: 648.3 MiB (8.3%) Init: systemd v: 245 
           runlevel: 5 Compilers: gcc: 9.3.0 alt: 9 Client: Unknown python3.8 client inxi: 3.0.38 

 

 

Edited by thebigh

My missions:           Stand-alone                                                      Duncan Lynch series                              

                                      Down and Out on Newford Road              the Factory Heist

                                                                                                  A House Call

                                                                                                  The House of deLisle                                                                                                  

                              

Link to comment
Share on other sites

5 hours ago, Bergante said:

To clear a Mission as Downloaded (and not anymore as  available , so you know you have the newest stuff ) one has to Download it mostly 3 sometimes 4/5 times !!!!!!

Yes, indeed it sucks.

The second download (*) is update, although I don't understand why it needs update immediately download, and the third one is translations, although I don't know why aren't they downloaded immediately (in fact, I don't know well how they are supposed to work).

Quote

 

P.S. :  I just made a Test with my old Tdm 2.05 install
in The Ingame Downloader appear all of the new Missions so I loaded some of the Newer :  Deadeye / The Factory Heist afterwords loading does , install goes -- but starting remains Black !

 

Mapper can set "minimum required version" on his mission.
I checked Deadeye: minimum version is not set there.

Note that we don't support old versions of TDM in general, and it is understandable that mappers don't want to spend time for going back 5 versions to see which exact version is the first that does not work.

4 hours ago, thebigh said:

A few players have reported guards being weirdly alert to every little noise in my map Newford Road.
I have done some experimenting and I can confirm that it is an issue with the beta, and it seems to be OS dependent.

And... how to reproduce it?
Does it happen on other missions?

Link to comment
Share on other sites

Link to comment
Share on other sites

Guest
This topic is now closed to further replies.

  • Recent Status Updates

    • Petike the Taffer

      I've finally managed to log in to The Dark Mod Wiki. I'm back in the saddle and before the holidays start in full, I'll be adding a few new FM articles and doing other updates. Written in Stone is already done.
      · 4 replies
    • nbohr1more

      TDM 15th Anniversary Contest is now active! Please declare your participation: https://forums.thedarkmod.com/index.php?/topic/22413-the-dark-mod-15th-anniversary-contest-entry-thread/
       
      · 0 replies
    • JackFarmer

      @TheUnbeholden
      You cannot receive PMs. Could you please be so kind and check your mailbox if it is full (or maybe you switched off the function)?
      · 1 reply
    • OrbWeaver

      I like the new frob highlight but it would nice if it was less "flickery" while moving over objects (especially barred metal doors).
      · 4 replies
    • nbohr1more

      Please vote in the 15th Anniversary Contest Theme Poll
       
      · 0 replies
×
×
  • Create New...