Do not modify original schedule in qcompile
- Quantify version: develop
Disclaimer: I am mocking up some things, apologies for not providing executable code.
@AdriaanRol @adamorangeqs @jvoven @kel85uk please mind this bug!!!
Description
Run this kind of cell twice (or more in the same python kernel)
_, progs = qcompile(
schedule=sched_abs,
device_cfg=ss.SPIN_DEVICE_CFG,
hardware_mapping=ss.SPIN_HARDWARE_MAPPING
)
for seq_idx in range(2):
fn = progs["qcm0"][f"seq{seq_idx}"]["seq_fn"]
print(fn, ":\n")
with open(fn, "r") as file:
result = json.load(file)
print(result["program"])this will end up producing something like this the first time:
/Users/Victor/quantify-data/schedules/20210414-131320-554-b4e8bc/RP_baseband_sequencer_cfg.json :
move 1,R0
wait_sync 4 #sync
start: set_mrk 1
set_awg_gain -13107,-13107 #Set gain
play 0,1,20
set_awg_gain 0,0 #Set gain
play 0,1,20
set_awg_gain 13107,13107 #Set gain
play 0,1,20
set_awg_gain -13107,-13107 #Set gain
play 0,1,20
set_awg_gain 0,0 #Set gain
play 0,1,20
set_awg_gain 13107,13107 #Set gain
play 0,1,20
set_awg_gain -13107,-13107 #Set gain
play 0,1,20
set_awg_gain 0,0 #Set gain
play 0,1,20
set_awg_gain 13107,13107 #Set gain
play 0,1,20
wait 8 #Sync with other sequencers
loop R0,@start
stop BUT the next time(s) you tun the same cel it will produce unexpected instructions:
/Users/Victor/quantify-data/schedules/20210414-130759-630-9dd3aa/RP_baseband_sequencer_cfg.json :
move 1,R0
wait_sync 4 #sync
start: set_mrk 1
set_awg_gain -13107,-13107 #Set gain
play 0,1,0
set_awg_gain -13107,-13107 #Set gain
play 0,1,0
set_awg_gain -13107,-13107 #Set gain
play 0,1,0
set_awg_gain -13107,-13107 #Set gain
play 0,1,20
set_awg_gain 0,0 #Set gain
play 0,1,0
set_awg_gain 0,0 #Set gain
play 0,1,0
set_awg_gain 0,0 #Set gain
play 0,1,0
set_awg_gain 0,0 #Set gain
play 0,1,20
set_awg_gain 13107,13107 #Set gain
play 0,1,0
set_awg_gain 13107,13107 #Set gain
play 0,1,0
set_awg_gain 13107,13107 #Set gain
play 0,1,0
set_awg_gain 13107,13107 #Set gain
play 0,1,20
set_awg_gain -13107,-13107 #Set gain
play 0,1,0
set_awg_gain -13107,-13107 #Set gain
play 0,1,0
set_awg_gain -13107,-13107 #Set gain
play 0,1,0
set_awg_gain -13107,-13107 #Set gain
play 0,1,20
set_awg_gain 0,0 #Set gain
play 0,1,0
set_awg_gain 0,0 #Set gain
play 0,1,0
set_awg_gain 0,0 #Set gain
play 0,1,0
set_awg_gain 0,0 #Set gain
play 0,1,20
set_awg_gain 13107,13107 #Set gain
play 0,1,0
set_awg_gain 13107,13107 #Set gain
play 0,1,0
set_awg_gain 13107,13107 #Set gain
play 0,1,0
set_awg_gain 13107,13107 #Set gain
play 0,1,20
set_awg_gain -13107,-13107 #Set gain
play 0,1,0
set_awg_gain -13107,-13107 #Set gain
play 0,1,0
set_awg_gain -13107,-13107 #Set gain
play 0,1,0
set_awg_gain -13107,-13107 #Set gain
play 0,1,20
set_awg_gain 0,0 #Set gain
play 0,1,0
set_awg_gain 0,0 #Set gain
play 0,1,0
set_awg_gain 0,0 #Set gain
play 0,1,0
set_awg_gain 0,0 #Set gain
play 0,1,20
set_awg_gain 13107,13107 #Set gain
play 0,1,0
set_awg_gain 13107,13107 #Set gain
play 0,1,0
set_awg_gain 13107,13107 #Set gain
play 0,1,0
set_awg_gain 13107,13107 #Set gain
play 0,1,20
wait 8 #Sync with other sequencers
loop R0,@start
stop What Was Expected
Running the qcompile repeated times would not modify the original schedule. Or would raise some kind of exception in the worse case (probably not desired for UX reasons)
What Actually Happened
It seems that the original dict holding the schedule information gets modified and this leads to very unexpected program output