Ivan Maidanski

> Is there possibly some way to intercept the call that allocates on the fake stack? Of course, it would be better to have a way to get the address...

> Is there possibly some way to intercept the call that allocates on the fake stack? @kmatheussen, could you please check if there's a way to intercept or figure out...

> (void*)__asan_addr_is_in_fake_stack((void*)__asan_get_current_fake_stack(), &allocated[0], &allocated[1], &allocated[2]) > And now the start and end addresses have been put into allocated[1] and allocated[2] start/end of fake frame or fake stack? The doc says...

Looks like fake stack grows up, right? Or every frame is allocated in the heap separately?

> Looks like fake stack grows up, right? Or every frame is allocated in the heap separately? According to the ASan source code, there's a thread-local pool and each frame...

Looking into main branch of llvm: internal lsan_common.h declares `GetThreadExtraStackRangesLocked(thread_id)`.

Related failure. Source: master (63c90f040c) Build: https://github.com/ivmai/bdwgc-fork/actions/runs/20021730022/job/57409870381 Host: Ubuntu 24.04/x86_64 Config: `ubuntu-latest configure --enable-cplusplus --enable-gc-assertions --enable-static clang -fno-common -fno-omit-frame-pointer -fsanitize=address` Output: ``` GC_USE_GETWRITEWATCH=0 build/gctest shell: /usr/bin/bash -e {0} Running on...

Source: master (e92dc51a7) Build: https://github.com/bdwgc/bdwgc/actions/runs/17787824831/job/50558558655

How about renaming the repository to e.g. `bdwgc-rust` after transfer? (we already have `bdwgc-d` repository for D language bindings, so it seem "bdwgc-\" is a good naming convention).

Note: after transfer please do not reuse "bdwgc-alloc" repository in your private space - this would allow Github to automatically redirect https://github.com/raviqqe/bdwgc-alloc to new project URL (e.g. https://github.com/bdwgc/bdwgc-rust). The private...