Eddie Hung [Mon, 15 Jul 2019 19:03:51 +0000 (12:03 -0700)]
$__ICE40_CARRY_LUT4 -> $__ICE40_FULL_ADDER as per @whitequark
Eddie Hung [Sat, 13 Jul 2019 08:11:00 +0000 (01:11 -0700)]
ice40_opt to $__ICE40_CARRY_LUT4 into $lut not SB_LUT
Eddie Hung [Fri, 12 Jul 2019 15:22:26 +0000 (08:22 -0700)]
Do not double count cells in abc
Eddie Hung [Fri, 12 Jul 2019 08:32:10 +0000 (01:32 -0700)]
Use Const::from_string() not its constructor...
Eddie Hung [Fri, 12 Jul 2019 08:17:53 +0000 (01:17 -0700)]
Off by one
Eddie Hung [Fri, 12 Jul 2019 08:15:22 +0000 (01:15 -0700)]
Fix spacing
Eddie Hung [Fri, 12 Jul 2019 08:08:48 +0000 (01:08 -0700)]
Remove double push
Eddie Hung [Fri, 12 Jul 2019 07:53:01 +0000 (00:53 -0700)]
Map to and from this box if -abc9
Eddie Hung [Fri, 12 Jul 2019 07:52:31 +0000 (00:52 -0700)]
ice40_opt to handle this box and opt back to SB_LUT4
Eddie Hung [Fri, 12 Jul 2019 07:52:19 +0000 (00:52 -0700)]
Add new box to cells_sim.v
Eddie Hung [Fri, 12 Jul 2019 07:51:37 +0000 (00:51 -0700)]
_ABC macro will map and unmap to this new box
Eddie Hung [Fri, 12 Jul 2019 07:50:42 +0000 (00:50 -0700)]
Combine SB_CARRY+SB_LUT into one $__ICE40_CARRY_LUT4 box
Eddie Hung [Thu, 11 Jul 2019 19:55:35 +0000 (12:55 -0700)]
Merge pull request #1182 from koriakin/xc6s-bram
synth_xilinx: Initial Spartan 6 block RAM inference support.
Eddie Hung [Thu, 11 Jul 2019 19:55:14 +0000 (12:55 -0700)]
Merge pull request #1185 from koriakin/xc-ff-init-vals
xilinx: Fix the default values for FDPE/FDSE INIT attributes to match ISE/Vivado.
Marcin Kościelnicki [Thu, 11 Jul 2019 19:13:12 +0000 (21:13 +0200)]
xilinx: Fix the default values for FDPE/FDSE INIT attributes to match ISE/Vivado.
Eddie Hung [Thu, 11 Jul 2019 15:49:06 +0000 (08:49 -0700)]
Enable &mfs for abc9, even if it only currently works for ice40
Marcin Kościelnicki [Tue, 2 Jul 2019 12:28:35 +0000 (14:28 +0200)]
synth_xilinx: Initial Spartan 6 block RAM inference support.
Clifford Wolf [Thu, 11 Jul 2019 05:25:52 +0000 (07:25 +0200)]
Merge pull request #1172 from whitequark/write_verilog-Sa-as-qmark
write_verilog: write RTLIL::Sa aka - as Verilog ?
Clifford Wolf [Thu, 11 Jul 2019 05:23:28 +0000 (07:23 +0200)]
Merge pull request #1179 from whitequark/attrmap-proc
attrmap: also consider process, switch and case attributes
Eddie Hung [Wed, 10 Jul 2019 21:38:13 +0000 (14:38 -0700)]
Merge pull request #1180 from YosysHQ/eddie/no_abc9_retime
Error out if -abc9 and -retime specified
Eddie Hung [Wed, 10 Jul 2019 21:38:00 +0000 (14:38 -0700)]
Merge pull request #1148 from YosysHQ/xc7mux
synth_xilinx to infer wide multiplexers using new '-widemux <min>' option
Eddie Hung [Wed, 10 Jul 2019 19:47:48 +0000 (12:47 -0700)]
Error out if -abc9 and -retime specified
Eddie Hung [Wed, 10 Jul 2019 19:32:33 +0000 (12:32 -0700)]
Add some spacing
Eddie Hung [Wed, 10 Jul 2019 19:20:04 +0000 (12:20 -0700)]
Add some ASCII art explaining mux decomposition
whitequark [Wed, 10 Jul 2019 12:28:32 +0000 (12:28 +0000)]
attrmap: also consider process, switch and case attributes.
Clifford Wolf [Wed, 10 Jul 2019 06:48:20 +0000 (08:48 +0200)]
Merge pull request #1177 from YosysHQ/clifford/async
Fix clk2fflogic adff reset semantic to negative hold time on reset
Eddie Hung [Wed, 10 Jul 2019 04:26:38 +0000 (21:26 -0700)]
Call muxpack and pmux2shiftx before cmp2lut
Eddie Hung [Tue, 9 Jul 2019 21:29:58 +0000 (14:29 -0700)]
Restore opt_clean back to original place
Eddie Hung [Tue, 9 Jul 2019 21:28:54 +0000 (14:28 -0700)]
Restore missing techmap -map +/cmp2lut.v with LUT_WIDTH=6
David Shah [Tue, 9 Jul 2019 21:26:10 +0000 (22:26 +0100)]
synth_ecp5: Fix typo in copyright header
Signed-off-by: David Shah <dave@ds0.me>
Clifford Wolf [Tue, 9 Jul 2019 20:59:51 +0000 (22:59 +0200)]
Merge pull request #1174 from YosysHQ/eddie/fix1173
Increment _TECHMAP_BITS_CONNMAP_ by one since counting from zero
Clifford Wolf [Tue, 9 Jul 2019 20:51:25 +0000 (22:51 +0200)]
Merge pull request #1175 from whitequark/write_verilog-fix-case-attr-position
write_verilog: fix placement of case attributes
Clifford Wolf [Tue, 9 Jul 2019 20:44:39 +0000 (22:44 +0200)]
Fix tests/various/async FFL test
Signed-off-by: Clifford Wolf <clifford@clifford.at>
Clifford Wolf [Tue, 9 Jul 2019 20:21:25 +0000 (22:21 +0200)]
Improve tests/various/async, disable failing ffl test
Signed-off-by: Clifford Wolf <clifford@clifford.at>
Eddie Hung [Tue, 9 Jul 2019 19:35:41 +0000 (12:35 -0700)]
Extend using A[1] to preserve don't care
Eddie Hung [Tue, 9 Jul 2019 19:19:40 +0000 (12:19 -0700)]
Merge pull request #1171 from YosysHQ/revert-1166-eddie/synth_keepdc
Revert "Add "synth -keepdc" option"
Eddie Hung [Tue, 9 Jul 2019 19:16:33 +0000 (12:16 -0700)]
Merge remote-tracking branch 'origin/eddie/fix1173' into xc7mux
whitequark [Tue, 9 Jul 2019 19:14:03 +0000 (19:14 +0000)]
write_verilog: fix placement of case attributes. NFC.
Eddie Hung [Tue, 9 Jul 2019 19:14:00 +0000 (12:14 -0700)]
Increment _TECHMAP_BITS_CONNMAP_ by one since counting from zero
Clifford Wolf [Tue, 9 Jul 2019 18:58:59 +0000 (20:58 +0200)]
Add tests/various/async.{sh,v}
Signed-off-by: Clifford Wolf <clifford@clifford.at>
Clifford Wolf [Tue, 9 Jul 2019 18:58:28 +0000 (20:58 +0200)]
Improve tests/various/run-test.sh
Signed-off-by: Clifford Wolf <clifford@clifford.at>
Clifford Wolf [Tue, 9 Jul 2019 18:58:01 +0000 (20:58 +0200)]
Add tests/simple_abc9/.gitignore
Signed-off-by: Clifford Wolf <clifford@clifford.at>
whitequark [Tue, 9 Jul 2019 18:30:24 +0000 (18:30 +0000)]
write_verilog: write RTLIL::Sa aka - as Verilog ?.
Currently, the only ways (determined by grepping for regex \bSa\b) to
end up with RTLIL::Sa in a netlist is by reading a Verilog constant
with ? in it as a part of case, or by running certain FSM passes.
Both of these cases should be round-tripped back to ? in Verilog.
Eddie Hung [Tue, 9 Jul 2019 17:59:37 +0000 (10:59 -0700)]
Extend during mux decomposition with 1'bx
Eddie Hung [Tue, 9 Jul 2019 17:38:07 +0000 (10:38 -0700)]
Fix typo and comments
Eddie Hung [Tue, 9 Jul 2019 17:22:57 +0000 (10:22 -0700)]
Merge pull request #1170 from YosysHQ/eddie/fix_double_underscore
Rename __builtin_bswap32 -> bswap32
Eddie Hung [Tue, 9 Jul 2019 17:22:49 +0000 (10:22 -0700)]
Merge remote-tracking branch 'origin/master' into xc7mux
Eddie Hung [Tue, 9 Jul 2019 17:21:54 +0000 (10:21 -0700)]
synth_xilinx to call commands of synth -coarse directly
Eddie Hung [Tue, 9 Jul 2019 17:15:02 +0000 (10:15 -0700)]
Revert "synth_xilinx to call "synth -run coarse" with "-keepdc""
This reverts commit
7f964859ec99500e471853f5914b6e5b7c35a031.
Eddie Hung [Tue, 9 Jul 2019 17:14:23 +0000 (10:14 -0700)]
Revert "Add "synth -keepdc" option"
Eddie Hung [Tue, 9 Jul 2019 16:35:09 +0000 (09:35 -0700)]
Rename __builtin_bswap32 -> bswap32
Eddie Hung [Tue, 9 Jul 2019 16:22:12 +0000 (09:22 -0700)]
Fix spacing
Eddie Hung [Tue, 9 Jul 2019 16:16:00 +0000 (09:16 -0700)]
Fix spacing
Clifford Wolf [Tue, 9 Jul 2019 14:59:43 +0000 (16:59 +0200)]
Merge pull request #1168 from whitequark/bugpoint-processes
Add support for processes in bugpoint
Clifford Wolf [Tue, 9 Jul 2019 14:59:18 +0000 (16:59 +0200)]
Merge pull request #1169 from whitequark/more-proc-cleanups
A new proc_prune pass
Clifford Wolf [Tue, 9 Jul 2019 14:57:16 +0000 (16:57 +0200)]
Merge pull request #1163 from whitequark/more-case-attrs
More support for case rule attributes
Clifford Wolf [Tue, 9 Jul 2019 14:56:29 +0000 (16:56 +0200)]
Merge pull request #1162 from whitequark/rtlil-case-attrs
Allow attributes on individual switch cases in RTLIL
Clifford Wolf [Tue, 9 Jul 2019 14:49:08 +0000 (16:49 +0200)]
Merge pull request #1167 from YosysHQ/eddie/xc7srl_cleanup
Cleanup synth_xilinx SRL inference, make more consistent
whitequark [Tue, 9 Jul 2019 08:14:52 +0000 (08:14 +0000)]
proc_prune: promote assigns to module connections when legal.
This can pave the way for further transformations by exposing
identities that were previously hidden in a process to any pass that
uses SigMap. Indeed, this commit removes some ad-hoc logic from
proc_init that appears to have been tailored to the output of
genrtlil in favor of using `SigMap.apply()`. (This removal is not
optional, as the ad-hoc logic cannot cope with the result of running
proc_prune; a similar issue was fixed in proc_arst.)
whitequark [Mon, 8 Jul 2019 15:19:01 +0000 (15:19 +0000)]
proc_prune: new pass.
The proc_prune pass is similar in nature to proc_rmdead pass: while
proc_rmdead removes branches that never become active because another
branch preempts it, proc_prune removes assignments that never become
active because another assignment preempts them.
Genrtlil contains logic similar to the proc_prune pass, but their
purpose is different: genrtlil has to prune assignments to adapt
the semantics of blocking assignments in HDLs (latest assignment
wins) to semantics of assignments in RTLIL processes (assignment in
the most specific case wins). On the other hand proc_prune is
a general purpose RTLIL simplification that benefits all frontends,
even those not using the Yosys AST library.
The proc_prune pass is added to the proc script after proc_rmdead,
since it gives better results with fewer branches.
whitequark [Tue, 9 Jul 2019 09:08:38 +0000 (09:08 +0000)]
bugpoint: add -assigns and -updates options.
whitequark [Tue, 9 Jul 2019 08:57:57 +0000 (08:57 +0000)]
proc_clean: add -quiet option.
This is useful for other passes that call it often, like bugpoint.
Eddie Hung [Tue, 9 Jul 2019 06:51:13 +0000 (23:51 -0700)]
Decompose mux inputs in delay-orientated (rather than area) fashion
Eddie Hung [Tue, 9 Jul 2019 06:49:16 +0000 (23:49 -0700)]
Do not call opt -mux_undef (part of -full) before muxcover
Eddie Hung [Tue, 9 Jul 2019 06:05:48 +0000 (23:05 -0700)]
Add one more comment
Eddie Hung [Tue, 9 Jul 2019 06:02:57 +0000 (23:02 -0700)]
Less thinking
Eddie Hung [Tue, 9 Jul 2019 05:56:19 +0000 (22:56 -0700)]
Reword
Eddie Hung [Tue, 9 Jul 2019 04:43:16 +0000 (21:43 -0700)]
Merge pull request #1166 from YosysHQ/eddie/synth_keepdc
Add "synth -keepdc" option
Eddie Hung [Tue, 9 Jul 2019 02:26:43 +0000 (19:26 -0700)]
Merge branch 'master' of github.com:YosysHQ/yosys
Eddie Hung [Tue, 9 Jul 2019 02:23:24 +0000 (19:23 -0700)]
synth_xilinx to call "synth -run coarse" with "-keepdc"
Eddie Hung [Tue, 9 Jul 2019 02:21:53 +0000 (19:21 -0700)]
Merge remote-tracking branch 'origin/eddie/synth_keepdc' into xc7mux
Eddie Hung [Tue, 9 Jul 2019 02:21:21 +0000 (19:21 -0700)]
Clarify script -scriptwire doc
Eddie Hung [Tue, 9 Jul 2019 02:15:37 +0000 (19:15 -0700)]
Add synth -keepdc to CHANGELOG
Eddie Hung [Tue, 9 Jul 2019 02:15:07 +0000 (19:15 -0700)]
Clarify 'wreduce -keepdc' doc
Eddie Hung [Tue, 9 Jul 2019 02:14:54 +0000 (19:14 -0700)]
Add synth -keepdc option
Eddie Hung [Tue, 9 Jul 2019 00:06:35 +0000 (17:06 -0700)]
Map $__XILINX_SHIFTX in a more balanced manner
Eddie Hung [Tue, 9 Jul 2019 00:06:22 +0000 (17:06 -0700)]
Capitalisation
Eddie Hung [Tue, 9 Jul 2019 00:04:39 +0000 (17:04 -0700)]
Add synth_xilinx -widemux recommended value
Eddie Hung [Mon, 8 Jul 2019 21:34:37 +0000 (14:34 -0700)]
Merge pull request #1164 from YosysHQ/eddie/muxcover_mux2
Add muxcover -mux2=cost option
David Shah [Mon, 8 Jul 2019 20:04:33 +0000 (21:04 +0100)]
Merge pull request #1160 from ZirconiumX/cyclone_v
synth_intel: Warn about untested Quartus backend
Eddie Hung [Mon, 8 Jul 2019 19:50:59 +0000 (12:50 -0700)]
Update muxcover doc as per @ZirconiumX
Eddie Hung [Mon, 8 Jul 2019 19:03:38 +0000 (12:03 -0700)]
Fixes for 2:1 muxes
Eddie Hung [Mon, 8 Jul 2019 18:29:21 +0000 (11:29 -0700)]
synth_xilinx -widemux=2 is minimum now
Eddie Hung [Mon, 8 Jul 2019 18:08:20 +0000 (11:08 -0700)]
Parametric muxcover costs as per @daveshah1
Eddie Hung [Mon, 8 Jul 2019 18:00:31 +0000 (11:00 -0700)]
Merge remote-tracking branch 'origin/eddie/muxcover_mux2' into xc7mux
Eddie Hung [Mon, 8 Jul 2019 18:00:06 +0000 (11:00 -0700)]
atoi -> stoi
Eddie Hung [Mon, 8 Jul 2019 17:59:12 +0000 (10:59 -0700)]
Add muxcover -mux2=cost option
Eddie Hung [Mon, 8 Jul 2019 17:48:10 +0000 (10:48 -0700)]
atoi -> stoi as per @daveshah1
Eddie Hung [Mon, 8 Jul 2019 17:46:08 +0000 (10:46 -0700)]
Merge remote-tracking branch 'origin/master' into xc7mux
whitequark [Mon, 8 Jul 2019 15:11:29 +0000 (15:11 +0000)]
verilog_backend: dump attributes on SwitchRule.
This appears to be an omission.
whitequark [Mon, 8 Jul 2019 13:18:18 +0000 (13:18 +0000)]
proc_mux: consider \src attribute on CaseRule.
whitequark [Mon, 8 Jul 2019 12:48:50 +0000 (12:48 +0000)]
verilog_backend: dump attributes on CaseRule, as comments.
Attributes are not permitted in that position by Verilog grammar.
whitequark [Mon, 8 Jul 2019 12:29:08 +0000 (12:29 +0000)]
genrtlil: emit \src attribute on CaseRule.
whitequark [Mon, 8 Jul 2019 11:34:58 +0000 (11:34 +0000)]
Allow attributes on individual switch cases in RTLIL.
The parser changes are slightly awkward. Consider the following IL:
process $0
<point 1>
switch \foo
<point 2>
case 1'1
assign \bar \baz
<point 3>
...
case
end
end
Before this commit, attributes are valid in <point 1>, and <point 3>
iff it is immediately followed by a `switch`. (They are essentially
attached to the switch.) But, after this commit, and because switch
cases do not have an ending delimiter, <point 3> becomes ambiguous:
the attribute could attach to either the following `case`, or to
the following `switch`. This isn't expressible in LALR(1) and results
in a reduce/reduce conflict.
To address this, attributes inside processes are now valid anywhere
inside the process: in <point 1> and <point 3> a part of case body,
and in <point 2> as a separate rule. As a consequence, attributes
can now precede `assign`s, which is made illegal in the same way it
is illegal to attach attributes to `connect`.
Attributes are tracked separately from the parser state, so this
does not affect collection of attributes at all, other than allowing
them on `case`s. The grammar change serves purely to allow attributes
in more syntactic places.
Dan Ravensloft [Sun, 7 Jul 2019 15:00:38 +0000 (16:00 +0100)]
synth_intel: Warn about untested Quartus backend
Clifford Wolf [Fri, 5 Jul 2019 09:57:41 +0000 (11:57 +0200)]
Merge pull request #1159 from btut/fix/1090_segfault_cell_and_wire
Throw runtime exception when trying to convert inexistend C++ object to Python
Benedikt Tutzer [Thu, 4 Jul 2019 12:20:13 +0000 (14:20 +0200)]
Throw runtime exception when trying to convert a c++-pointer to a
python-object in case the pointer is a nullptr to avoid a segfault.
Fixes #1090
Eddie Hung [Wed, 3 Jul 2019 16:43:00 +0000 (09:43 -0700)]
Merge pull request #1156 from YosysHQ/eddie/fix_abc9_unknown_cell
write_xaiger to treat unknown cell connections as keep-s
Clifford Wolf [Wed, 3 Jul 2019 10:30:37 +0000 (12:30 +0200)]
Merge pull request #1147 from YosysHQ/clifford/fix1144
Improve specify dummy parser
Clifford Wolf [Wed, 3 Jul 2019 09:25:05 +0000 (11:25 +0200)]
Fix tests/various/specify.v
Signed-off-by: Clifford Wolf <clifford@clifford.at>