can you confirm you have no shorts left from triton route?

I don't see any errors reported by routing. Anything I should look for specifically?

check the summary.csv and make sure you have no shorts reported

In this case, no shorts reported (and thanks for pointer to the CSV!)

where's your config.tcl?

Here's what I'm currently working with:

# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#      <http://www.apache.org/licenses/LICENSE-2.0>
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# SPDX-License-Identifier: Apache-2.0

set script_dir [file dirname [file normalize [info script]]]
source $::env(SYNTH_DIR)/../../../caravel/openlane/user_project_wrapper_empty/fixed_wrapper_cfgs.tcl

set ::env(DESIGN_NAME) user_project_wrapper

set ::env(FP_PDN_CHECK_NODES) 0

set ::env(CLOCK_PORT) "wb_clk_i"

set ::env(CLOCK_PERIOD) "10"

set ::env(PL_OPENPHYSYN_OPTIMIZATIONS) 0
set ::env(DIODE_INSERTION_STRATEGY) 0

set ::env(MAGIC_WRITE_FULL_LEF) 1

# save some time
set ::env(RUN_KLAYOUT_XOR) 0
set ::env(RUN_KLAYOUT_DRC) 0

# magic drc checking on the sram block shows millions of false errors
set ::env(MAGIC_DRC_USE_GDS) 0

set ::env(PL_RANDOM_GLB_PLACEMENT) 1
set ::env(SYNTH_TOP_LEVEL) 0
set ::env(MACRO_PLACEMENT_CFG) $::env(SYNTH_DIR)/macro.cfg

set ::env(GLB_RT_OBS) "li1 0 0 2920 3520"
set ::env(GLB_RT_ALLOW_CONGESTION) 1

# Playing with different extraction settings...
set ::env(YOSYS_REWRITE_VERILOG) 0

set ::env(GLB_RT_ADJUSTMENT) 0.7
set ::env(GLB_RT_L2_ADJUSTMENT) 0.9
set ::env(GLB_RT_L3_ADJUSTMENT) 0.7

set ::env(GLB_RT_MAXLAYER) 5
set ::env(ROUTING_OPT_ITERS) 80

set ::env(CLOCK_TREE_SYNTH) 0
set ::env(ROUTING_CORES) 10
set ::env(PL_TARGET_DENSITY)     0.09


set ::env(FILL_INSERTION) 0
set ::env(TAP_DECAP_INSERTION) 0

can you upload the lvs?

Of course. Here's the .v, .spice, and .lvs.log

it looks to me like like the sram isn't getting connected

@Matt Venn, when you say the nodes look disconnected, which file are you looking at? Looking in the .v, I see .dout1 on the SRAMs is disconnected (has NC* nets), but the others appear to be connected.

so the .v is the post synthesis netlist

Oh... I saw that @Tim Edwards updated the pin sizes on the SRAMS in the main repo, and I hadn't yet updated.

I know that Matt G has updated them recently

Also good know that one error can cause a cascade.

absolutelyt

Yeah, I don't recall running into any LVS issues last time around. Not the case this spin...

@Matthew Ballance: The "disconnected node" messages probably aren't meaningful. They are just reflecting the fact that the subcell is an abstract view and the pins aren't connected to anything (I recently tried to suppress that message in the case of black-box views but there were issues that make it hard to correct, so it is still doing it, for now).

Thanks, @Tim Edwards. I notice a lot of power-pin mismatches at the beginning of the log, and believe I saw another similar report. Any clues there?

I pulled your log file you posted above, and I don't see anything in particular until the top level. There is definitely some issue with power connections, as the top level has separate nets for well and substrate than for power and ground. That suggests that you don't have tap cells in the layout.

Hmm... Looking up through the log (output from running the flow), I do see a warning "Skipping Tap/Decap Insertion" right after macro placement. Where would I look to understand why tap cells aren't in the layout?

That's a question to ask in #openlane . I don't have that detailed knowledge about the internal actions of OpenLane.

Will do, @Tim Edwards. Appreciate your help!

@Matthew Ballance maybe you could try a more simple openlane config? I noticed you have made quite a few changes to it that could be affecting things like this. Or are all the changes you've made to already solve other issues?

@Matt Venn, unfortunately there are lots and lots of knobs to turn... I started off with the config in caravel_user_project/openlane/user_project_wrapper. That one had Tap/Decap insertion turned off, which clearly is a problem in my situation. Good news is I'm learning new things all throughout the process 😉