Fix naming for afl++

docs/Experiments.txt

@@ -40,7 +40,7 @@ you have exported CC=symcc, CXX=sym++ and SYMCC_NO_SYMBOLIC_INPUT=1, first
 download the code, then build it using its own build system, finally unset
 SYMCC_NO_SYMBOLIC_INPUT and analyze the program in concert with AFL (which
 requires building a second time for AFL, see docs/Fuzzing.txt). We used AFL
-2.56b and built the targets with AFL_USE_ASAN=1. Note that the fuzzing helper is
+2.56b and built the targets with AFL_USE_ASan=1. Note that the fuzzing helper is
 already installed in the Docker container.
 
 OpenJPEG [4]: we used revision 1f1e9682, built with CMake as described in the
@@ -61,7 +61,7 @@ tcpdump: we built both tcpdump [7] and libpcap [8]; in order to make the former
   and analyzed "tcpdump/tcpdump -e -r @@"; the corpus consisted of just a single
   dummy file containing the character "A".
 
-All experiments used one AFL master process, one secondary AFL process, and one
+All experiments used one AFL main process, one secondary AFL process, and one
 SymCC process. We let them run for 24 hours and repeated each of them 30 times
 to create the graphs in the paper; AFL map density was extracted from the
 secondary AFL process' "plot_data" file, column "map_size".

docs/Fuzzing.txt

@@ -50,7 +50,7 @@ $ tcpdump -e -r <pcap_file>
 Compile tcpdump and libpcap, the library it uses for pcap reading, once with
 SymCC and once with one of AFL's compiler wrappers (e.g., afl-clang). In order
 to detect memory corruptions, enable address sanitizer in the AFL-instrumented
-version by exporting AFL_USE_ASAN=1 before compiling:
+version by exporting AFL_USE_ASan=1 before compiling:
 
 $ git clone https://github.com/the-tcpdump-group/libpcap.git
 $ git clone https://github.com/the-tcpdump-group/tcpdump.git
@@ -66,7 +66,7 @@ $ make
 $ cd ..
 
 $ mkdir afl_build; cd afl_build
-$ export AFL_USE_ASAN=1
+$ export AFL_USE_ASan=1
 $ cp -r ../{libpcap,tcpdump} .
 $ cd libpcap
 $ CC=/path/to/afl-clang ./configure
@@ -88,10 +88,10 @@ AFL:
 $ mkdir corpus
 $ echo A > corpus/dummy
 
-Then launch one AFL master and one AFL secondary instance, both writing their
+Then launch one AFL main and one AFL secondary instance, both writing their
 outputs to the arbitrarily named directory "afl_out":
 
-$ afl-fuzz -M afl-master -i corpus -o afl_out -m none -- afl_build/tcpdump/tcpdump -e -r @@
+$ afl-fuzz -M afl-main -i corpus -o afl_out -m none -- afl_build/tcpdump/tcpdump -e -r @@
 $ afl-fuzz -S afl-secondary -i corpus -o afl_out -m none -- afl_build/tcpdump/tcpdump -e -r @@
 
 For simplicity, we disable memory limits (with "-m none"); be sure to read AFL's
@@ -109,7 +109,7 @@ you should see the counter "imported" in the "path geometry" section increase
 after a short time - this means that the fuzzer instances and SymCC are
 exchanging inputs. Crashes will be stored in afl_out/*/crashes as usual.
 
-It is possible to run SymCC with only an AFL master or only a secondary AFL
+It is possible to run SymCC with only an AFL main or only a secondary AFL
 instance; see the AFL docs for the implications. Moreover, the number of fuzzer
 and SymCC instances can be increased - just make sure that each has a unique
 name.