[Bug]: SM crash PVKII version 1.0 (32bit Windows)

[ambaca]

Prerequisites

• [x] I have checked that my issue doesn't exist yet in the issue tracker

Operating System and Version

Windows 10 Home

Game / AppID and Version

pvkii(17570) version 1.0.0.1

SourceMod Version

1.12.0.7210

Metamod:Source Version

2.0.0-dev+1363

Version Verification

• [x] I have updated SourceMod to the latest version and the issue persists
• [x] I have updated SourceMod to the latest snapshot and the issue persists
• [x] I have updated Metamod:Source to the latest snapshot and the issue persists

Updated SourceMod Version

1.13.0.7246

Updated Metamod:Source Version

2.0.0-dev+1363

Description

PVKII 32bit SRCDS will run with MM:S.

• But SRCDS hangs and crash on startup after few seconds, when loading SM.

Can't get any common logs.

Steps to Reproduce

No response

Relevant Log Output

**DebugView++**

Using Breakpad minidump system. Version: 0 AppID: 17570
Loaded 1638 VPK file hashes from G:\server\pirates, vikings and knights ii\pvkii\pvkii.vpk for pure server operation.
Loaded 1638 VPK file hashes from G:\server\pirates, vikings and knights ii\pvkii\pvkii.vpk for pure server operation.
Loaded 1222 VPK file hashes from G:\server\pirates, vikings and knights ii\hl2\hl2_textures.vpk for pure server operation.
Loaded 383 VPK file hashes from G:\server\pirates, vikings and knights ii\hl2\hl2_sound_misc.vpk for pure server operation.
Loaded 447 VPK file hashes from G:\server\pirates, vikings and knights ii\hl2\hl2_misc.vpk for pure server operation.
Loaded 5 VPK file hashes from G:\server\pirates, vikings and knights ii\platform\platform_misc.vpk for pure server operation.
Convar r_rimlight has conflicting FCVAR_CHEAT flags (child: FCVAR_CHEAT, parent: no FCVAR_CHEAT, parent wins)
Convar r_rimlight has conflicting FCVAR_CHEAT flags (child: FCVAR_CHEAT, parent: no FCVAR_CHEAT, parent wins)
Convar r_rimlight has conflicting FCVAR_CHEAT flags (child: FCVAR_CHEAT, parent: no FCVAR_CHEAT, parent wins)
Convar r_rimlight has conflicting FCVAR_CHEAT flags (child: FCVAR_CHEAT, parent: no FCVAR_CHEAT, parent wins)
server.dll loaded for "PVKII"
<process started at 17:48:39.472 has terminated with 0xc0000005 (EXCEPTION_ACCESS_VIOLATION)>

[Spirrwell]

I'm working on putting together updates to make SourceMod/MetaMod functional again with PVKII. The HL2SDK branch for pvkii is going to need to be updated.

We recently ported the game to 64-bit, and it would probably be easiest to update if we copied the more recent tf2 branch and updated the interfaces accordingly.

Is there a preference in the way that would be done in a PR? Would one big commit replacing the content of the pvkii branch with the content of the tf2 branch and subsequent smaller commits with changes be acceptable?

[psychonic]

I'm working on putting together updates to make SourceMod/MetaMod functional again with PVKII. The HL2SDK branch for pvkii is going to need to be updated.

We recently ported the game to 64-bit, and it would probably be easiest to update if we copied the more recent tf2 branch and updated the interfaces accordingly.

Is there a preference in the way that would be done in a PR? Would one big commit replacing the content of the pvkii branch with the content of the tf2 branch and subsequent smaller commits with changes be acceptable?

Yes, that would be fine. Thanks!

[MihanDeNoob]

I'm working on putting together updates to make SourceMod/MetaMod functional again with PVKII. The HL2SDK branch for pvkii is going to need to be updated.

We recently ported the game to 64-bit, and it would probably be easiest to update if we copied the more recent tf2 branch and updated the interfaces accordingly.

Is there a preference in the way that would be done in a PR? Would one big commit replacing the content of the pvkii branch with the content of the tf2 branch and subsequent smaller commits with changes be acceptable?

Thank you for bricking sourcemod for pvk with the release of 64bit version that didnt improve anything, and even made things worse at certain points.

[Spirrwell]

I will say that if you're desperate for SourceMod and have the know-how and understanding to build it yourself, a lot of the work is done here:

https://github.com/Octoshark/sourcemod-pvkii https://github.com/Octoshark/metamod-source-pvkii https://github.com/Octoshark/hl2sdk-pvkii/tree/pvkii-64 https://github.com/Octoshark/hl2sdk-manifests

But, I still needed to check the 32-bit build and update the signatures/offsets there. There has also been at least one update to the game since I started this work, so I am unsure if that may have broken anything. I also haven't done a containerized build of the tier1/mathlib for the hl2sdk yet.

Something I noticed is that we still used unsigned long for entity handles, and I'm not sure if that will cause problems. But, I do plan to update PVKII to use a uint32 for those.

If you enjoy saysounds on your server, this soundlib fork uses the ambuild system and will require minimal tweaking to get working when the above work gets merged in at some point: https://github.com/Adrianilloo/soundlib

I work full time, and I simply cannot keep up with the pace of the game's development. So, sorry.

[MihanDeNoob]

@Spirrwell the real problem was releasing an untested versions of the game and breaking stuff along the way. 64bit update does absolutely nothing besides braking stuff, for example:

Worse client performance with 64bit (at least from my tests, -15fps avarge) DXVK on 64bit doesnt work (the most important performance feature for source games) General worse performance after update, even on 32bit version (fps drops down to 38)

Mind you, same was not happening with my cpu in tf2, even before its 64bit update that actually improved things.

My criticism is simple - dont release something if its not tested.