Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Tweak M7 uArch model: MAC issue slot constraint + MAC forwarding path #138

Draft
wants to merge 3 commits into
base: main
Choose a base branch
from
Draft

Tweak M7 uArch model: MAC issue slot constraint + MAC forwarding path #138

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
e9733d4
define missing SlothySelfCheckException
mkannwischer Dec 15, 2024
eaa2e75
Cortex-M7: Fix MAC instructions to one issue slot
dop-amin Nov 29, 2024
1795e7d
M7: Adjust MAC forwarding paths
mkannwischer Dec 20, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
3 changes: 3 additions & 0 deletions slothy/core/core.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -55,6 +55,9 @@
class SlothyException(Exception):
"""Generic exception thrown by SLOTHY"""

class SlothySelfCheckException(SlothyException):
"""Exception thrown by SLOTHY during tht selfcheck """

class Result(LockAttributes):
"""The results of a one-shot SLOTHY optimization run"""

Expand Down
18 changes: 14 additions & 4 deletions slothy/targets/arm_v7m/cortex_m7.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -56,11 +56,17 @@ def add_further_constraints(slothy):
add_st_hazard(slothy)

add_dsp_slot_constraint(slothy)
add_mac_slot_constraint(slothy)

def add_dsp_slot_constraint(slothy):
slothy.restrict_slots_for_instructions_by_class(
[pkhbt, pkhtb, pkhbt_shifted, ubfx_imm, uadd16, usub16, sadd16, ssub16], [0])

def add_mac_slot_constraint(slothy):
slothy.restrict_slots_for_instructions_by_class(
[mul, mul_short, smull, smlal, mla, mls, smulwb, smulwt, smultb, smultt,
smulbb, smlabt, smlabb, smlatt, smlad, smladx, smuad, smuadx, smmulr], [1])

# TODO: this seems incorrect
def add_slot_constraints(slothy):
slothy.restrict_slots_for_instructions_by_class(
Expand Down Expand Up @@ -285,12 +291,16 @@ def get_latency(src, out_idx, dst):
latency = lookup_multidict(default_latencies, src)

# Forwarding path to MAC instructions
if instclass_dst in [mla, mls, smlabb, smlabt, smlatt, smlatb] and src.args_out[0] == dst.args_in[2]:
if instclass_dst in [mla, mls, smlabb, smlabt, smlatt, smlatb] and dst.args_in[2] in (src.args_out + src.args_in_out):
latency = latency - 1

if instclass_dst in [smlal] and \
(src.args_out[0] == dst.args_in_out[0] or src.args_out[0] == dst.args_in_out[1]):
latency = latency - 1
if instclass_dst in [smlal]:
if len(src.args_out) > 1:
if (src.args_out[0] == dst.args_in_out[0] or src.args_out[0] == dst.args_in_out[1]):
latency = latency - 1
elif len(src.args_in_out) > 1:
if (src.args_in_out[0] == dst.args_in_out[0] or src.args_in_out[0] == dst.args_in_out[1]):
latency = latency - 1

# Multiply accumulate chain latency is 1
if instclass_src in [smlal] and instclass_dst in [smlal] and \
Expand Down
Loading