orfs: bump

[oharboe]

@maliberty @jeffng-or FYI

Running times are very well contained even when all repair is enabled, but WNS is 10% worse than it used to be, -3300 vs -2700.

Stage: cts

Variant	base	1	2
Description	Base settings, provides macro placement from hierarchical synthesis	Flattend, timing driven placement, CTS timing repair enabled	Flattend, timing driven placement, CTS timing repair enabled, last gasp timing repair
Buffer	263659	118586	118586
Clock buffer	23267	12762	12762
Clock inverter	7223	3654	3654
Inverter	143507	70385	70385
Macro	72	72	72
Multi-Input combinational cell	1359180	740180	740180
Sequential cell	239698	118443	118443
Tie cell	2575	60	60
Timing Repair Buffer	95644	46437	46437
Total	2134825	1110579	1110579
slack	-6015.186035	-3330.928955	-3330.928955
tns	-508683392.0	-93219328.0	-93219328.0
GLOBAL_ROUTE_ARGS	-congestion_iterations 30 -congestion_report_iter_step 5 -verbose
GPL_TIMING_DRIVEN		1	1
MACRO_PLACEMENT_TCL		$(location write_macro_placement)	$(location write_macro_placement)
SKIP_CTS_REPAIR_TIMING		0	0
SKIP_LAST_GASP			0
SYNTH_HIERARCHICAL		0	0
dissolve
previous_stage			floorplan: BoomTile_1_synth
2_1_floorplan.log	577	139	139
2_2_floorplan_io.log	38	20	20
2_3_floorplan_macro.log	1503	23	22
2_4_floorplan_tapcell.log	35	19	19
2_5_floorplan_pdn.log	1028	1021	1019
3_1_place_gp_skip_io.log	1636	1986	1988
3_2_place_iop.log	51	29	29
3_3_place_gp.log	6311	7519	7507
3_4_place_resized.log	613	330	329
3_5_place_dp.log	1581	858	855
4_1_cts.log	543	661	656

Base configuration variables

Variable	Value
CORE_AREA	2 2 1998 1998
DIE_AREA	0 0 2000 2000
FILL_CELLS
GPL_ROUTABILITY_DRIVEN	1
GPL_TIMING_DRIVEN	0
HOLD_SLACK_MARGIN	-200
IO_CONSTRAINTS	$(location :io-boomtile)
MACRO_PLACE_HALO	19 19
MAX_ROUTING_LAYER	M7
MIN_ROUTING_LAYER	M2
PDN_TCL	$(PLATFORM_DIR)/openRoad/pdn/BLOCKS_grid_strategy.tcl
PLACE_DENSITY	0.24
PLACE_PINS_ARGS	-annealing
ROUTING_LAYER_ADJUSTMENT	0.45
SDC_FILE	$(location :constraints-boomtile)
SETUP_SLACK_MARGIN	-1300
SKIP_CTS_REPAIR_TIMING	1
SKIP_INCREMENTAL_REPAIR	1
SKIP_LAST_GASP	1
SKIP_REPORT_METRICS	1
SYNTH_HIERARCHICAL	1
TAPCELL_TCL
TNS_END_PERCENT	0

[oharboe]

@jeffng-or FYI, same CI error as you see in #201

[oharboe]

@maliberty @jeffng-or Run times are now almost identical when GPL_TIMING_DRIVEN=1, SKIP_CTS_REPAIR_TIMING=0 and SKIP_LAST_GASP=0, so I think we can just make that the default now.

Was anything done on purpose to make it so, or is it just lucky initial conditions?

The only thing I can think of is The-OpenROAD-Project/OpenROAD-flow-scripts#2540

[oharboe]

@maliberty @jeffng-or Looking at grt for this PR.

Is there an easy way to see which side the pins are on for a macro?

Here I believe the pins are on the right sight, which is good:

Here I believe the pnis are at the bottom, which would be terrible:

[oharboe]

@maliberty I tried somewhat recently to use read_def on macros to see where pins are, but ran into bugs(filed with openroad).

[oharboe]

Ah, via "ITerms" I can highlight pins. They are at the top 😌

[oharboe]

MX means mirrored about the X axis, which means flipped up-side down and mirrored.

The macro has the pins at the bottom:

After MX, we expect to find the pins at the top:

Yet, from ITerms, it looks like the pins are at the top(which is probably good for this design):

[oharboe]

@AcKoucher FYI, macro placement looks good with megaboom on latest ORFS & OpenROAD.

[oharboe]

@maliberty @jeffng-or I think the global route DRC failures are better than they used to be after 5 iterations, all the errors are in the middle and not on some fringe macros(which could be indicative of some problems in macro placement).