Fix fortitude's "nemesis" streaming crash

[tyackman]

cout fortitude crash.log

[who-care123]

E8 ? ? ? ? 48 8B C8 48 8B F0 48 8B 10 FF 52 ? 48 8B CE add(1).rip()
if (auto ptr = *(uint64_t*)(a1 + 0xC10); !ptr) return 0;

[maybegreat48]

That should work, but I'd like to analyze their syncs and block it through the can apply data hook. I'm busy working on something else right now, but I'll get back to this project soon enough

[tyackman]

added the sync logs
cout fort.log

[who-care123]

That should work, but I'd like to analyze their syncs and block it through the can apply data hook. I'm busy working on something else right now, but I'll get back to this project soon enough

I just figured out what the crash is.
RDR has also the invalid trailer attachment crash(trailer attached to a target not a vehicle).
We can check it in CPhysicalAttachDataNode.
But as I said in Yimmenu GTA, I think we also need to check sync or creation(object->GetGameObject)
Btw, I think netObject + 0xB0 is GetGameObject

[who-care123]

That should work, but I'd like to analyze their syncs and block it through the can apply data hook. I'm busy working on something else right now, but I'll get back to this project soon enough

I just figured out what the crash is. RDR has also the invalid trailer attachment crash(trailer attached to a target not a vehicle). We can check it in CPhysicalAttachDataNode. But as I said in Yimmenu GTA, I think we also need to check sync or creation(object->GetGameObject) Btw, I think netObject + 0xB0 is GetGameObject

Or maybe we can check the attached object type

[elcapone25]

you right getGameObject is at 0xB0. 48 89 5C 24 ? 48 89 6C 24 ? 48 89 74 24 ? 57 41 54 41 55 41 56 41 57 48 83 EC 70 48 8B B9 ? ? ? ? here its the signature for the node. + the class class IDK_NODE{
public:
char pad_0[0x10];
uint32_t m_unk; //0x10
char pad_14[0xC];
uint32_t m_unk1; //0x20
uint32_t m_unk2; //0x24
uint32_t m_unk3; //0x28

}; if someone know the name would be good for block this crash

[SkiddyToast]

you right getGameObject is at 0xB0. 48 89 5C 24 ? 48 89 6C 24 ? 48 89 74 24 ? 57 41 54 41 55 41 56 41 57 48 83 EC 70 48 8B B9 ? ? ? ? here its the signature for the node. + the class class IDK_NODE{ public: char pad_0[0x10]; uint32_t m_unk; //0x10 char pad_14[0xC]; uint32_t m_unk1; //0x20 uint32_t m_unk2; //0x24 uint32_t m_unk3; //0x28

}; if someone know the name would be good for block this crash

The class is much larger than that. It's the vehicle gadget node btw.

[elcapone25]

you right getGameObject is at 0xB0. 48 89 5C 24 ? 48 89 6C 24 ? 48 89 74 24 ? 57 41 54 41 55 41 56 41 57 48 83 EC 70 48 8B B9 ? ? ? ? here its the signature for the node. + the class class IDK_NODE{ public: char pad_0[0x10]; uint32_t m_unk; //0x10 char pad_14[0xC]; uint32_t m_unk1; //0x20 uint32_t m_unk2; //0x24 uint32_t m_unk3; //0x28
}; if someone know the name would be good for block this crash

The class is much larger than that. It's the vehicle gadget node btw.

thanks for the information :)

[SkiddyToast]

This should be a sufficient class:

class CGadgetData
{
public:
  uint32_t m_type;
  uint8_t m_data[0xF3];
};
static_assert(sizeof(CGadgetData) == 0xF8);

class CVehicleGadgetNodeData
{
public:
  bool m_has_position;
  char pad_0001[15];
  float m_position[4];
  uint32_t m_num_gadgets;
  CGadgetData m_gadgets[2];
};
static_assert(sizeof(CVehicleGadgetNodeData) == 0x214); // Needs to be 0x220, prob from packing at 16.

[Rxann]

E8 ? ? ? ? 48 8B C8 48 8B F0 48 8B 10 FF 52 ? 48 8B CE add(1).rip()

if (auto ptr = *(uint64_t*)(a1 + 0xC10); !ptr) return 0;

This produces false positives + Game Crashes. I am not even sure this will block the crash.

[tyackman]

Fortnut crash may be fixed by #148. Needs further testing. EDIT: Game still eventually shits, though the trailer attachment protection seems to help some.
6-27-24 fort crash.log

[Rxann]

Nemesis crash seems to be ineffective after @Mr-X-GTA 's PR which allows the menu to not crash. I tested this and did not crash 2 times it was used on me. I will leave this issue open for a bit to see if anyone has any differing opinions on my conclusion.