Experiment: Simplify x25519 optimization script

paper/scripts/slothy_x25519.sh

@@ -115,29 +115,6 @@ i=2
 
 echo "*** Step 5"
 i=3
- ${SLOTHY_DIR}/slothy-cli Arm_AArch64 Arm_Cortex_A55 \
- ${OPT_DIR}/neon/X25519-AArch64-simple_unfold_process${i}.s \
- -o ${OPT_DIR}/neon/X25519-AArch64-simple_unfold_process$((${i}+1)).s \
- -r x25519_scalarmult_alt_unfold_process${i},x25519_scalarmult_alt_unfold_process$((${i}+1)) \
- -c inputs_are_outputs -c outputs="[x0]" \
- -s mainloop -e end_label \
- -c variable_size \
- -c max_solutions=512 \
- -c timeout=240 \
- -c constraints.stalls_first_attempt=32 \
- -c split_heuristic \
- -c split_heuristic_region="[0.3,1]" \
- -c objective_precision=0.1 \
- -c constraints.move_stalls_to_top \
- -c split_heuristic_stepsize=0.05 \
- -c split_heuristic_factor=5 \
- -c split_heuristic_repeat=3 \
- -c split_heuristic_abort_cycle_at=8 \
- -c constraints.model_latencies=False \
- $REDIRECT_OUTPUT
-
-echo "*** Step 6"
-i=4
  ${SLOTHY_DIR}/slothy-cli Arm_AArch64 Arm_Cortex_A55 \
  ${OPT_DIR}/neon/X25519-AArch64-simple_unfold_process${i}.s \
  -o ${OPT_DIR}/neon/X25519-AArch64-simple_unfold_process$((${i}+1)).s \
@@ -155,75 +132,31 @@ i=4
  -c split_heuristic_stepsize=0.05 \
  -c split_heuristic_factor=5 \
  -c split_heuristic_repeat=3 \
- -c split_heuristic_abort_cycle_at=5 \
+ -c split_heuristic_abort_cycle_at=8 \
  -c constraints.model_latencies=False \
  $REDIRECT_OUTPUT
 
 # Finally, also consider latencies
 
-echo "** Step 7-9: Consider latencies"
-echo "*** Step 7"
-i=5
- ${SLOTHY_DIR}/slothy-cli Arm_AArch64 Arm_Cortex_A55 \
- ${OPT_DIR}/neon/X25519-AArch64-simple_unfold_process${i}.s \
- -o ${OPT_DIR}/neon/X25519-AArch64-simple_unfold_process$((${i}+1)).s \
- -r x25519_scalarmult_alt_unfold_process${i},x25519_scalarmult_alt_unfold_process$((${i}+1)) \
- -c inputs_are_outputs -c outputs="[x0]" \
- -s mainloop -e end_label \
- -c variable_size \
- -c max_solutions=512 \
- -c timeout=300 \
- -c constraints.stalls_first_attempt=32 \
- -c split_heuristic \
- -c split_heuristic_region="[0,1]" \
- -c objective_precision=0.1 \
- -c split_heuristic_stepsize=0.05 \
- -c split_heuristic_optimize_seam=10 \
- -c split_heuristic_factor=8 \
- -c split_heuristic_repeat=10 \
- $REDIRECT_OUTPUT
-
-echo "*** Step 8"
-i=6
- ${SLOTHY_DIR}/slothy-cli Arm_AArch64 Arm_Cortex_A55 \
- ${OPT_DIR}/neon/X25519-AArch64-simple_unfold_process${i}.s \
- -o ${OPT_DIR}/neon/X25519-AArch64-simple_unfold_process$((${i}+1)).s \
- -r x25519_scalarmult_alt_unfold_process${i},x25519_scalarmult_alt_unfold_process$((${i}+1)) \
- -c inputs_are_outputs -c outputs="[x0]" \
- -s mainloop -e end_label \
- -c variable_size \
- -c max_solutions=512 \
- -c timeout=300 \
- -c constraints.stalls_first_attempt=32 \
- -c split_heuristic \
- -c split_heuristic_region="[0,1]" \
- -c objective_precision=0.1 \
- -c split_heuristic_stepsize=0.05 \
- -c split_heuristic_optimize_seam=10 \
- -c split_heuristic_factor=8 \
- -c split_heuristic_repeat=3 \
- $REDIRECT_OUTPUT
-
-echo "*** Step 9"
-i=7
+echo "** Step 6: Consider latencies"
+i=4
  ${SLOTHY_DIR}/slothy-cli Arm_AArch64 Arm_Cortex_A55 \
- ${OPT_DIR}/neon/X25519-AArch64-simple_unfold_process${i}.s \
+ ${OPT_DIR}/neon/X25519-AArch64-simple_unfold_process${i}.s  \
  -o ${OPT_DIR}/neon/X25519-AArch64-simple_opt.s \
- -r x25519_scalarmult_alt_unfold_process${i},x25519_scalarmult_opt \
+ -r x25519_scalarmult_alt_unfold_process${i},x25519_scalarmult_opt  \
  -c inputs_are_outputs -c outputs="[x0]" \
  -s mainloop -e end_label \
  -c variable_size \
  -c max_solutions=512 \
  -c timeout=300 \
  -c constraints.stalls_first_attempt=32 \
  -c split_heuristic \
- -c split_heuristic_region="[0,1]" \
  -c objective_precision=0.1 \
  -c split_heuristic_stepsize=0.05 \
  -c split_heuristic_optimize_seam=10 \
  -c split_heuristic_factor=8 \
- -c split_heuristic_repeat=3 \
+ -c split_heuristic_repeat=8 \
  -c constraints.move_stalls_to_top \
  $REDIRECT_OUTPUT
 
-cd "${0%/*}"
+ cd "${0%/*}"