Repository navigation
Commit 203ec08
committed
Update the remaining rules for the max fanout backstop
Two more designs fail their metadata checks for the same reason as the four
already updated; they were missed on the first pass. Their build 8 numbers
match build 7 to the decimal, so this is the backstop's steady cost rather
than run to run noise.
Regenerated with genRuleFile.py --failing from the metadata.json of
PR-4462-head build 8, so no passing rule is tightened. aes-block's
canonical netlist hash rule is left alone: the backstop runs in rsz, well
after synthesis, so a netlist change there is not this branch's to rebaseline.
| Design | Metric | Old | New |
| ------ | ------ | --- | --- |
| asap7/aes-block | cts__timing__setup__tns | -1890.0 | -3080.0 |
| | globalroute__timing__setup__tns | -206.0 | -331.0 |
| | finish__timing__setup__tns | -418.0 | -694.0 |
| nangate45/aes | globalroute__timing__setup__tns | -0.785 | -0.963 |
Signed-off-by: minjukim55 <mkim@precisioninno.com>1 parent be4e71e commit 203ec08
2 files changed
Lines changed: 4 additions & 4 deletions
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
42 | 42 | | |
43 | 43 | | |
44 | 44 | | |
45 | | - | |
| 45 | + | |
46 | 46 | | |
47 | 47 | | |
48 | 48 | | |
| |||
62 | 62 | | |
63 | 63 | | |
64 | 64 | | |
65 | | - | |
| 65 | + | |
66 | 66 | | |
67 | 67 | | |
68 | 68 | | |
| |||
94 | 94 | | |
95 | 95 | | |
96 | 96 | | |
97 | | - | |
| 97 | + | |
98 | 98 | | |
99 | 99 | | |
100 | 100 | | |
| |||
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
62 | 62 | | |
63 | 63 | | |
64 | 64 | | |
65 | | - | |
| 65 | + | |
66 | 66 | | |
67 | 67 | | |
68 | 68 | | |
| |||
0 commit comments