riscv
/
yosys
mirror of https://github.com/YosysHQ/yosys.git
1
0
Fork 0
Code Issues Projects Releases Wiki Activity
yosys/tests/arch/intel_alm/adffs.ys

49 lines
2.0 KiB
Plaintext
Raw Permalink Normal View History

synth_intel_alm: alternative synthesis for Intel FPGAs By operating at a layer of abstraction over the rather clumsy Intel primitives, we can avoid special hacks like `dffinit -highlow` in favour of simple techmapping. This also makes the primitives much easier to manipulate, and more descriptive (no more cyclonev_lcell_comb to mean anything from a LUT2 to a LUT6).
2019-11-19 04:19:00 -06:00
read_verilog ../common/adffs.v
design -save read
hierarchy -top adff
proc
intel_alm: Add global buffer insertion Signed-off-by: gatecat <gatecat@ds0.me>
2021-05-15 08:34:48 -05:00
equiv_opt -async2sync -assert -map +/intel_alm/common/alm_sim.v -map +/intel_alm/common/dff_sim.v synth_intel_alm -family cyclonev -noiopad -noclkbuf # equivalency check
synth_intel_alm: alternative synthesis for Intel FPGAs By operating at a layer of abstraction over the rather clumsy Intel primitives, we can avoid special hacks like `dffinit -highlow` in favour of simple techmapping. This also makes the primitives much easier to manipulate, and more descriptive (no more cyclonev_lcell_comb to mean anything from a LUT2 to a LUT6).
2019-11-19 04:19:00 -06:00
design -load postopt # load the post-opt design (otherwise equiv_opt loads the pre-opt design)
cd adff # Constrain all select calls below inside the top module
select -assert-count 1 t:MISTRAL_FF
select -assert-count 1 t:MISTRAL_NOT
select -assert-none t:MISTRAL_FF t:MISTRAL_NOT %% t:* %D
design -load read
hierarchy -top adffn
proc
intel_alm: Add global buffer insertion Signed-off-by: gatecat <gatecat@ds0.me>
2021-05-15 08:34:48 -05:00
equiv_opt -async2sync -assert -map +/intel_alm/common/alm_sim.v -map +/intel_alm/common/dff_sim.v synth_intel_alm -family cyclonev -noiopad -noclkbuf # equivalency check
synth_intel_alm: alternative synthesis for Intel FPGAs By operating at a layer of abstraction over the rather clumsy Intel primitives, we can avoid special hacks like `dffinit -highlow` in favour of simple techmapping. This also makes the primitives much easier to manipulate, and more descriptive (no more cyclonev_lcell_comb to mean anything from a LUT2 to a LUT6).
2019-11-19 04:19:00 -06:00
design -load postopt # load the post-opt design (otherwise equiv_opt loads the pre-opt design)
cd adffn # Constrain all select calls below inside the top module
select -assert-count 1 t:MISTRAL_FF
select -assert-none t:MISTRAL_FF %% t:* %D
design -load read
hierarchy -top dffs
proc
intel_alm: Add global buffer insertion Signed-off-by: gatecat <gatecat@ds0.me>
2021-05-15 08:34:48 -05:00
equiv_opt -async2sync -assert -map +/intel_alm/common/alm_sim.v -map +/intel_alm/common/dff_sim.v synth_intel_alm -family cyclonev -noiopad -noclkbuf # equivalency check
synth_intel_alm: alternative synthesis for Intel FPGAs By operating at a layer of abstraction over the rather clumsy Intel primitives, we can avoid special hacks like `dffinit -highlow` in favour of simple techmapping. This also makes the primitives much easier to manipulate, and more descriptive (no more cyclonev_lcell_comb to mean anything from a LUT2 to a LUT6).
2019-11-19 04:19:00 -06:00
design -load postopt # load the post-opt design (otherwise equiv_opt loads the pre-opt design)
cd dffs # Constrain all select calls below inside the top module
select -assert-count 1 t:MISTRAL_FF
select -assert-count 1 t:MISTRAL_ALUT2
select -assert-none t:MISTRAL_FF t:MISTRAL_ALUT2 %% t:* %D
design -load read
hierarchy -top ndffnr
proc
intel_alm: Add global buffer insertion Signed-off-by: gatecat <gatecat@ds0.me>
2021-05-15 08:34:48 -05:00
equiv_opt -async2sync -assert -map +/intel_alm/common/alm_sim.v -map +/intel_alm/common/dff_sim.v synth_intel_alm -family cyclonev -noiopad -noclkbuf # equivalency check
synth_intel_alm: alternative synthesis for Intel FPGAs By operating at a layer of abstraction over the rather clumsy Intel primitives, we can avoid special hacks like `dffinit -highlow` in favour of simple techmapping. This also makes the primitives much easier to manipulate, and more descriptive (no more cyclonev_lcell_comb to mean anything from a LUT2 to a LUT6).
2019-11-19 04:19:00 -06:00
design -load postopt # load the post-opt design (otherwise equiv_opt loads the pre-opt design)
cd ndffnr # Constrain all select calls below inside the top module
select -assert-count 1 t:MISTRAL_FF
Replace opt_rmdff with opt_dff.
2020-07-20 16:19:51 -05:00
select -assert-count 2 t:MISTRAL_NOT
synth_intel_alm: alternative synthesis for Intel FPGAs By operating at a layer of abstraction over the rather clumsy Intel primitives, we can avoid special hacks like `dffinit -highlow` in favour of simple techmapping. This also makes the primitives much easier to manipulate, and more descriptive (no more cyclonev_lcell_comb to mean anything from a LUT2 to a LUT6).
2019-11-19 04:19:00 -06:00
Replace opt_rmdff with opt_dff.
2020-07-20 16:19:51 -05:00
select -assert-none t:MISTRAL_FF t:MISTRAL_NOT %% t:* %D
intel_alm: add Cyclone 10 GX tests
2020-07-05 12:53:14 -05:00