WorksButNotTested

> I will try something more later. I am also busy with doing some other work and I am currently in Singapore so we have quite a time difference. >...

> The remaining problem you can see here: > > > > It always detects only a map_size of 1 in my little test binary. My little test binary has...

> The map_size != 64kb comes from afl-compiler-rt.o.c > > ``` > #if MAP_SIZE #define MAP_INITIAL_SIZE 2097152 > #else > #define MAP_INITIAL_SIZE MAP_SIZE > #endif > ``` > > I...

> I'm not sure I follow. Can you elaborate please?

I'd suggest [here](https://github.com/AFLplusplus/AFLplusplus/blob/1f4b45b5bf006bfa1b3957c99bad3f0ec1c355dd/frida_mode/src/instrument/instrument.c#L291) is a good place to add... ```c if (__afl_map_size == MAP_SIZE) __afl_map_size = 0x10000; ``` This should hopefully correct the size of coverage map used when `afl-frida-trace.so`...

The code for collecting the range is [here](https://github.com/AFLplusplus/AFLplusplus/blob/40947508037b874020c8dd1251359fecaab04b9d/frida_mode/src/ranges.c#L594). If `AFL_FRIDA_VERBOSE` and `AFL_DEBUG_CHILD` is set, then you should see the ranges printed at each step to confirm the final calculated set...

I've not tried recently on ARM64, but if you build using... `make FRIDA_SOURCE=1 -C frida_mode/` then it should download and build the FRIDA gumjs devkit (on which `afl-frida-trace.so` depends) from...

> I digged a bit deeper into this. I copied all the code for `gum_process_enumerate_ranges` into an own dynamic library so that I can print out all the info by...

> I have reported the problem to frida-gum. > > What should I do about the other problem that I initially mentioned: > > ``` > instrument_previous_pc_addr = gum_memory_allocate_near( >...

Ole has just merge [this](https://github.com/frida/frida-gum/commit/42b95bb2383378e859267055e9757e1526c322c9) and released version 15.2.1. I have bumped the version in AFL++ FRIDA mode in [this PR](https://github.com/AFLplusplus/AFLplusplus/pull/1479) and also fixed the page alignment issue with `gum_memory_allocate_near`...