Memory limit is too restrictive - every possible binaries showing same info.. looks like there is some other issue. about e9afl HOT 4 CLOSED

It is probably just the memory limit issue. E9AFL is based on the E9Patch static binary rewriting system, and E9Patch emits binaries that consume large amounts of virtual address space (although actual physical memory usage is proportional to the binary file size). This seems to mess with AFL's virtual address space limits, which are imposed in order to detect oversized allocation bugs. A similar problem exists when combining AFL wth ASAN:

afl-fuzz tries to protect you from processes that go off-rails and start consuming all the available memory in a vain attempt to parse a malformed input file.

Rather than using -m none, there should be some finite limit (larger than 850) that will also work.

from e9afl.

Do the instrumented binaries run normally without afl-fuzz?

from e9afl.

Yes, it works fine without afl-fuzz.

from e9afl.

afl-fuzz worked with -m none option.
afl-fuzz -i input/ -o output/ -m none -- ./esif_ufd.afl -a @@

Thanks.

from e9afl.