Icenowy Zheng
7854d5ba21
anlogic: fix dbits of Anlogic Eagle DRAM16X4
...
The dbits of DRAM16X4 is wrong set to 2, which leads to waste of DRAM
bits.
Fix the dbits number in the RAM configuration.
Signed-off-by: Icenowy Zheng <icenowy@aosc.io>
2018-12-18 14:38:44 +08:00
Icenowy Zheng
d53a2bd1d3
anlogic: add support for Eagle Distributed RAM
...
The MSLICEs on the Eagle series of FPGA can be configured as Distributed
RAM.
Enable to synthesis to DRAM.
As the Anlogic software suite doesn't support any 'bx to exist in the
initializtion data of DRAM, do not enable the initialization support of
the inferred DRAM.
Signed-off-by: Icenowy Zheng <icenowy@aosc.io>
2018-12-17 23:20:40 +08:00
Icenowy Zheng
634d7d1c14
Revert "Leave only real black box cells"
...
This reverts commit 43030db5ff
.
For a synthesis tool, generating EG_LOGIC cells are a good choice, as
they can be furtherly optimized when PnR, although sometimes EG_LOGIC is
not as blackbox as EG_PHY cells (because the latter is more close to the
hardware implementation).
Signed-off-by: Icenowy Zheng <icenowy@aosc.io>
2018-12-17 23:20:40 +08:00
Clifford Wolf
5fa5dbbdda
Rename "fine:" label to "map:" in "synth_ice40"
...
Signed-off-by: Clifford Wolf <clifford@clifford.at>
2018-12-16 16:36:19 +01:00
Clifford Wolf
2a681909df
Merge pull request #724 from whitequark/equiv_opt
...
equiv_opt: new command, for verifying optimization passes
2018-12-16 15:54:26 +01:00
Clifford Wolf
ceffa66dbd
Merge pull request #730 from smunaut/ffssr_dont_touch
...
ice40: Honor the "dont_touch" attribute in FFSSR pass
2018-12-16 15:50:42 +01:00
Clifford Wolf
0c69f1d777
Merge pull request #725 from olofk/ram4k-init
...
Only use non-blocking assignments of SB_RAM40_4K for yosys
2018-12-16 15:42:04 +01:00
Sylvain Munaut
add6ab9b2a
ice40: Honor the "dont_touch" attribute in FFSSR pass
...
This is useful if you want to place FF manually ... can't merge SR in those
because it might make the manual placement invalid
Signed-off-by: Sylvain Munaut <tnt@246tNt.com>
2018-12-08 22:46:28 +01:00
whitequark
7ff5a9db2d
equiv_opt: pass -D EQUIV when techmapping.
...
This allows avoiding techmap crashes e.g. because of large memories
in white-box cell models.
2018-12-07 17:20:34 +00:00
Olof Kindgren
889297c62a
Only use non-blocking assignments of SB_RAM40_4K for yosys
...
In an initial statement, blocking assignments are normally used
and e.g. verilator throws a warning if non-blocking ones are used.
Yosys cannot however properly resolve the interdependencies if
blocking assignments are used in the initialization of SB_RAM_40_4K
and thus this has been used.
This patch will change to use non-blocking assignments only for yosys
2018-12-06 21:45:59 +01:00
whitequark
6e559ee3c7
synth_ice40: split `map_gates` off `fine`.
2018-12-06 12:04:39 +00:00
whitequark
d9fa4387c9
synth_ice40: add -noabc option, to use built-in LUT techmapping.
...
This should be combined with -relut to get sensible results.
2018-12-05 17:13:46 +00:00
whitequark
9ef078848a
gate2lut: new techlib, for converting Yosys gates to FPGA LUTs.
2018-12-05 17:13:27 +00:00
whitequark
12596b5003
Fix typo.
2018-12-05 17:13:27 +00:00
Clifford Wolf
e115303129
Merge pull request #713 from Diego-HR/master
...
Changes in GoWin synth commands and ALU primitive support
2018-12-05 09:08:30 -08:00
Clifford Wolf
1a260ce89b
Merge pull request #712 from mmicko/anlogic-support
...
Initial support for Anlogic FPGA
2018-12-05 09:08:04 -08:00
whitequark
45cb6200af
opt_lut: add -dlogic, to avoid disturbing logic such as carry chains.
2018-12-05 16:30:37 +00:00
whitequark
ea4870b126
synth_ice40: add -relut option, to run ice40_unlut and opt_lut.
2018-12-05 16:30:37 +00:00
whitequark
1719aa88ac
Extract ice40_unlut pass from ice40_opt.
...
Currently, `ice40_opt -unlut` would map SB_LUT4 to $lut and convert
them back to logic immediately. This is not desirable if the goal
is to operate on $lut cells. If this is desirable, the same result
as `ice40_opt -unlut` can be achieved by running simplemap and opt
after ice40_unlut.
2018-12-05 16:30:24 +00:00
Diego H
819ca73096
Changes in GoWin synth commands and ALU primitive support
2018-12-03 20:08:35 -06:00
Miodrag Milanovic
43030db5ff
Leave only real black box cells
2018-12-02 11:57:50 +01:00
Miodrag Milanovic
83bce9f59c
Initial support for Anlogic FPGA
2018-12-01 18:28:54 +01:00
Sylvain Munaut
3e5ab50a73
ice40: Add option to only use CE if it'd be use by more than X FFs
...
Signed-off-by: Sylvain Munaut <tnt@246tNt.com>
2018-11-27 21:50:42 +01:00
Clifford Wolf
dbc4cb8f4a
Merge pull request #697 from eddiehung/xilinx_ps7
...
Add support for PS7 block for Xilinx
2018-11-12 09:09:22 +01:00
Clifford Wolf
317cc9c2b7
Merge pull request #695 from daveshah1/ecp5_bb
...
ecp5: Adding some blackbox cells
2018-11-12 09:08:49 +01:00
Eddie Hung
99a14b0e37
Add support for Xilinx PS7 block
2018-11-10 12:45:07 -08:00
David Shah
fae3e645a2
ecp5: Add 'fake' DCU parameters
...
Signed-off-by: David Shah <dave@ds0.me>
2018-11-09 18:25:42 +00:00
David Shah
960c8794fa
ecp5: Add blackboxes for ancillary DCU cells
...
Signed-off-by: David Shah <dave@ds0.me>
2018-11-09 15:18:30 +00:00
David Shah
1f51332808
ecp5: Adding some blackbox cells
...
Signed-off-by: David Shah <dave@ds0.me>
2018-11-07 14:56:38 +00:00
Clifford Wolf
d084fb4c3f
Fix sf2 LUT interface
...
Signed-off-by: Clifford Wolf <clifford@clifford.at>
2018-10-31 15:36:53 +01:00
Clifford Wolf
cf79fd4376
Basic SmartFusion2 and IGLOO2 synthesis support
...
Signed-off-by: Clifford Wolf <clifford@clifford.at>
2018-10-31 15:28:57 +01:00
David Shah
b65932edc4
ecp5: Remove DSP parameters that don't work
...
Signed-off-by: David Shah <davey1576@gmail.com>
2018-10-22 16:20:38 +01:00
David Shah
101f5234ff
ecp5: Add DSP blackboxes
...
Signed-off-by: David Shah <davey1576@gmail.com>
2018-10-21 19:27:02 +01:00
David Shah
d29b517fef
ecp5: Sim model fixes
...
Signed-off-by: David Shah <dave@ds0.me>
2018-10-19 15:16:40 +01:00
David Shah
677b8ed3ca
ecp5: Add latch inference
...
Signed-off-by: David Shah <dave@ds0.me>
2018-10-19 15:16:40 +01:00
Clifford Wolf
24a5c65856
Merge pull request #657 from mithro/xilinx-vpr
...
xilinx: Still map LUT7/LUT8 to Xilinx specific primitives when using `-vpr`
2018-10-18 10:54:03 +02:00
David Shah
df4bfa0ad6
ecp5: Disable LSR inversion
...
Signed-off-by: David Shah <dave@ds0.me>
2018-10-16 12:48:39 +01:00
David Shah
812538a036
BRAM improvements
...
Signed-off-by: David Shah <dave@ds0.me>
2018-10-12 14:22:21 +01:00
David Shah
bdfead8c64
ecp5: Adding BRAM maps for all size options
...
Signed-off-by: David Shah <dave@ds0.me>
2018-10-10 17:18:17 +01:00
David Shah
983fb7ff88
ecp5: First BRAM type maps successfully
...
Signed-off-by: David Shah <dave@ds0.me>
2018-10-10 16:35:19 +01:00
David Shah
2ef1af8b58
ecp5: Script for BRAM IO connections
...
Signed-off-by: David Shah <dave@ds0.me>
2018-10-10 16:11:00 +01:00
David Shah
346cbbdbdc
ecp5: Adding BRAM initialisation and config
...
Signed-off-by: David Shah <dave@ds0.me>
2018-10-09 14:19:04 +01:00
Tim 'mithro' Ansell
b111ea1228
xilinx: Still map LUT7/LUT8 to Xilinx specific primitives.
...
Then if targeting vpr map all the Xilinx specific LUTs back into generic
Yosys LUTs.
2018-10-08 16:52:12 -07:00
David Shah
31e22c8b96
ecp5: Add blackbox for DP16KD
...
Signed-off-by: David Shah <dave@ds0.me>
2018-10-05 11:35:59 +01:00
Clifford Wolf
5f1fea08d5
Add inout ports to cells_xtra.v
...
Signed-off-by: Clifford Wolf <clifford@clifford.at>
2018-10-04 11:30:55 +02:00
Tim Ansell
ad975fb694
xilinx: Adding missing inout IO port to IOBUF
2018-10-03 16:38:32 -07:00
Clifford Wolf
76baae4b94
Merge pull request #645 from daveshah1/ecp5_dram_fix
...
ecp5: Don't map ROMs to DRAM
2018-10-02 10:00:10 +02:00
David Shah
fcd39e1398
ecp5: Don't map ROMs to DRAM
...
Signed-off-by: David Shah <davey1576@gmail.com>
2018-10-01 18:34:41 +01:00
Clifford Wolf
51f1bbeeb0
Add iCE40 SB_SPRAM256KA simulation model
...
Signed-off-by: Clifford Wolf <clifford@clifford.at>
2018-09-10 11:57:24 +02:00
Henner Zeller
3aa4484a3c
Consistent use of 'override' for virtual methods in derived classes.
...
o Not all derived methods were marked 'override', but it is a great
feature of C++11 that we should make use of.
o While at it: touched header files got a -*- c++ -*- for emacs to
provide support for that language.
o use YS_OVERRIDE for all override keywords (though we should probably
use the plain keyword going forward now that C++11 is established)
2018-07-20 23:51:06 -07:00