Error when sideloading apps (EXC_BAD_ACCESS) about altstore

[eni9889]

This is due to unsigned framework binaries and there is a pending PR to fix it: rileytestut/AltSign#1

[Smethan]

I have tried pulling in your PR and building, now the freshly compiled AltStore just crashes on my phone, says something about a CoreData error

[eni9889]

Yea same happens to me, don’t know why the PR would cause that will investigate. It’s only an issue on iOS 13 no problems on iOS 12

[Smethan]

Seems like its not an issue with your PR, I built it with a clean pull of the repo and I still got the crash, so it must be something else, and it does seem to be isolated to iOS 13, my iOS 12 device worked fine with it...

[eni9889]

@Smethan it turns out you need to use Xcode 10.3 to build the project. Xcode 11 has issues.

[Smethan]

Good to know. I’ll build there I suppose

[Smethan]

Alrighty, I've compiled it successfully on Xcode 10.3 and ran it on my phone, it works mostly, but I still get the same EXC_BAD_ACCESS error when trying to sign the same apps

[Smethan]

I've looked into it and apparently its an issue with dereferencing a null pointer at line 79 in ldid.cpp
uint32_t data = mach_header.Swap(signature->dataoff);
The Xcode static analyzer gave the error "Access to field 'dataoff' results in a dereference of a null pointer (loaded from variable 'signature')"

I don't quite know how I would fix this, as I'm not exactly a C++ developer, but maybe it could help someone else out.

[Smethan]

Update: I bypassed that error by always making signature non-null, but then I got another similar error, so I don't quite know what the real issue here is. Its looking ldid might just have some issues on iOS 13, or just with A12 devices

[eni9889]

Did you use my branch for both altsign and ldid?

_{Sent with GitHawk}

[Smethan]

Yes. I pulled the regular AltStore repo, deleted the altsign sub module, and then checked yours out along with ldid

[Smethan]

So, it has been a few days, and it seems to me that the Swap() command may be the culprit, as the previous error happened there, and it happens again on the swap command if I bypass that one. I know for a fact that it's a problem within the Entitlements section of the non-dependency ldid, might not be getting the entitlements correctly?