mirror of
https://github.com/YosysHQ/yosys
synced 2025-04-13 20:38:44 +00:00
449 lines
19 KiB
ReStructuredText
449 lines
19 KiB
ReStructuredText
Minimizing failing (or bugged) designs
|
|
======================================
|
|
|
|
- how to guide
|
|
- assumes knowledge and familiarity with Yosys
|
|
- something is wrong with your design OR something is wrong with Yosys
|
|
|
|
+ how to work out which
|
|
|
|
- *read* the error message
|
|
- is it a Yosys error? (starts with ERROR:)
|
|
|
|
+ does it give you a line number from your design
|
|
|
|
- is it a runtime error, e.g. SEGFAULT
|
|
- are you using the latest release of Yosys
|
|
|
|
+ has your problem already been fixed
|
|
|
|
- is your input design valid?
|
|
|
|
+ if you're using Verilog, try load it with `iverilog`_ or `verilator`_
|
|
|
|
.. _iverilog: https://steveicarus.github.io/iverilog/
|
|
.. _verilator: https://www.veripool.org/verilator/
|
|
|
|
- make sure to back up your code (design source and yosys script(s)) before
|
|
making any modifications
|
|
|
|
+ even if the code itself isn't important, this can help avoid "losing" the
|
|
error while trying to debug it
|
|
|
|
|
|
.. _minimize your RTLIL:
|
|
|
|
Minimizing RTLIL designs with bugpoint
|
|
--------------------------------------
|
|
|
|
Yosys provides the `bugpoint` command for reducing a failing design to the
|
|
smallest portion of that design which still results in failure. While initially
|
|
developed for Yosys crashes, `bugpoint` can also be used for designs that lead
|
|
to non-fatal errors, or even failures in other tools that use the output of a
|
|
Yosys script.
|
|
|
|
Can I use bugpoint?
|
|
~~~~~~~~~~~~~~~~~~~
|
|
|
|
The first thing to be aware of is that `bugpoint` is not available in every
|
|
build of Yosys. Because the command works by invoking external processes, it
|
|
requires that Yosys can spawn executables. Notably this means `bugpoint` is not
|
|
able to be used in WebAssembly builds such as that available via YoWASP. The
|
|
easiest way to check your build of Yosys is by running ``yosys -qq -p '!echo
|
|
test'``. If this echoes ``test`` in the console, then `bugpoint` will work as
|
|
expected. If instead if it displays the text ``ERROR: Shell is not available.``
|
|
then `bugpoint` will not work either.
|
|
|
|
.. note::
|
|
|
|
The console command ``yosys -qq -p '!echo test'`` uses the ``-qq`` flag to
|
|
prevent Yosys from outputting non-error messages to the console. The ``-p``
|
|
option executes ``!echo test`` as a Yosys command, attempting to pass ``echo
|
|
test`` to the shell for execution. For more about the ``-p`` option and
|
|
common pitfalls, check out :ref:`getting_started/scripting_intro:script
|
|
parsing` in the :doc:`/getting_started/index` section.
|
|
|
|
.. TODO:: Add ``YOSYS_DISABLE_SPAWN`` check in ``bugpoint.cc``.
|
|
|
|
At least in the help text, so that ``yosys -h bugpoint`` will correctly
|
|
indicate if the command will work instead of this roundabout method.
|
|
|
|
Next you need to separate loading the design from the failure point; you should
|
|
be aiming to reproduce the failure by running ``yosys -s <load.ys> -s
|
|
<failure.ys>``. If the failure occurs while loading the design, such as during
|
|
`read_verilog` you will instead have to minimize the input design yourself.
|
|
Check out the instructions for :ref:`using_yosys/bugpoint:minimizing verilog
|
|
designs` below.
|
|
|
|
The commands in ``<load.ys>`` only need to be run once, while those in
|
|
``<failure.ys>`` will be run on each iteration of `bugpoint`. If you haven't
|
|
already, following the instructions for :ref:`using_yosys/bugpoint:minimizing
|
|
scripts` will also help with identifying exactly which commands are needed to
|
|
produce the failure and which can be safely moved to the loading script.
|
|
|
|
.. note::
|
|
|
|
You should also be able to run the two scripts separately, calling first
|
|
``yosys -s <load.ys> -p 'write_rtlil design.il'`` and then ``yosys -s
|
|
<failure.ys> design.il``. If this doesn't work then it may mean that the
|
|
failure isn't reproducible from RTLIL and `bugpoint` won't work either.
|
|
|
|
When we talk about failure points here, it doesn't just mean crashes or errors
|
|
in Yosys. The ``<failure.ys>`` script can also be a user-defined failure such
|
|
as the `select` command with one of the ``-assert-*`` options; an example where
|
|
this might be useful is when a pass is supposed to remove a certain kind of
|
|
cell, but there is some edge case where the cell is not removed. Another
|
|
use-case would be minimizing a design which fails with the `equiv_opt` command,
|
|
suggesting that the optimization in question alters the circuit in some way.
|
|
|
|
It is even possible to use `bugpoint` with failures *external* to Yosys, by
|
|
making use of the `exec` command in ``<failure.ys>``. This is especially useful
|
|
when Yosys is outputting an invalid design, or when some other tool is
|
|
incompatible with the design. Be sure to use the ``exec -expect-*`` options so
|
|
that the pass/fail can be detected correctly. Multiple calls to `exec` can be
|
|
made, or even entire shell scripts (e.g. ``exec -expect-return 1 -- bash
|
|
<script.sh>``).
|
|
|
|
Our final failure we can use with `bugpoint` is one returned by a wrapper
|
|
process, such as ``valgrind`` or ``timeout``. In this case you will be calling
|
|
something like ``<wrapper> yosys -s <failure.ys> design.il``. Here, Yosys is
|
|
run under a wrapper process which checks for some failure state, like a memory
|
|
leak or excessive runtime. Note however that unlike the `exec` command, there
|
|
is currently no way to check the return status or messages from the wrapper
|
|
process; only a binary pass/fail.
|
|
|
|
|
|
How do I use bugpoint?
|
|
~~~~~~~~~~~~~~~~~~~~~~
|
|
|
|
At this point you should have:
|
|
|
|
1. either an RTLIL file containing the design to minimize (referred to here as
|
|
``design.il``), or a Yosys script, ``<load.ys>``, which loads it; and
|
|
2. a Yosys script, ``<failure.ys>``, which produces the failure and returns a
|
|
non-zero return status.
|
|
|
|
Now call ``yosys -qq -s <failure.ys> design.il`` and take note of the error(s)
|
|
that get printed. A template script, ``<bugpoint.ys>``, is provided here which
|
|
you can use. Make sure to configure it with the correct filenames and use only
|
|
one of the methods to load the design. Fill in the ``-grep`` option with the
|
|
error message printed just before. If you are using a wrapper process for your
|
|
failure state, add the ``-runner "<wrapper>"`` option to the `bugpoint` call.
|
|
For more about the options available, check ``help bugpoint`` or
|
|
:doc:`/cmd/bugpoint`.
|
|
|
|
.. code-block:: yoscrypt
|
|
:caption: ``<bugpoint.ys>`` template script
|
|
|
|
# Load design
|
|
read_rtlil design.il
|
|
## OR
|
|
script <load.ys>
|
|
|
|
# Call bugpoint with failure
|
|
bugpoint -script <failure.ys> -grep "<string>"
|
|
|
|
# Save minimized design
|
|
write_rtlil min.il
|
|
|
|
.. note::
|
|
|
|
Using ``-grep "<string>"`` with `bugpoint` is optional, but helps to ensure
|
|
that the minimized design is reproducing the right error, especially when
|
|
``<failure.ys>`` contains more than one command. Unfortunately this does not
|
|
work with runtime errors such as a ``SEGFAULT`` as it is only able to match
|
|
strings from the log file.
|
|
|
|
.. TODO:: Consider checking ``run_command`` return value for runtime errors.
|
|
|
|
Currently ``BugpointPass::run_yosys`` returns ``run_command(yosys_cmdline) ==
|
|
0``, so it shouldn't be too hard to add an option for it. Could also be
|
|
used with the ``-runner`` option, which might give it a bit more flexibility.
|
|
|
|
By default, `bugpoint` is able to remove any part of the design. In order to
|
|
keep certain parts, for instance because you already know they are related to
|
|
the failure, you can use the ``bugpoint_keep`` attribute. This can be done with
|
|
``(* bugpoint_keep *)`` in Verilog, ``attribute \bugpoint_keep 1`` in RTLIL, or
|
|
``setattr -set bugpoint_keep 1 [selection]`` from a Yosys script. It is also
|
|
possible to limit `bugpoint` to only removing certain *kinds* of objects, such
|
|
as only removing entire modules or cells (instances of modules). For more about
|
|
the options available, check ``help bugpoint`` or :doc:`/cmd/bugpoint`.
|
|
|
|
In some situations, it may also be helpful to use `setenv` before `bugpoint` to
|
|
set environment variables for the spawned processes. An example of this is
|
|
``setenv UBSAN_OPTIONS halt_on_error=1`` for where you are trying to raise an
|
|
error on undefined behaviour but only want the child process to halt on error.
|
|
|
|
.. note::
|
|
|
|
Using `setenv` in this way may or may not affect the current process. For
|
|
instance the ``UBSAN_OPTIONS halt_on_error`` here only affects child
|
|
processes, as does the :doc:`Yosys environment variable</appendix/env_vars>`
|
|
``ABC`` because they are only read on start-up. While others, such as
|
|
``YOSYS_NOVERIFIC`` and ``HOME``, are evaluated each time they are used.
|
|
|
|
Once you have finished configuration, you can now run ``yosys <bugpoint.ys>``.
|
|
The first thing `bugpoint` will do is test the input design fails. If it
|
|
doesn't, make sure you are using the right ``yosys`` executable; unless the
|
|
``-yosys`` option is provided, it will use whatever the shell defaults to. If
|
|
you are using the ``-runner`` option, try replacing the `bugpoint` command with
|
|
``write_rtlil test.il`` and then on a new line, ``!<wrapper> yosys -s
|
|
<failure.ys> test.il`` to check it works as expected and returns a non-zero
|
|
status.
|
|
|
|
Depending on the size of your design, and the length of your ``<failure.ys>``,
|
|
`bugpoint` may take some time; remember, it will run ``yosys -s <failure.ys>``
|
|
on each iteration of the design. The bigger the design, the more iterations.
|
|
The longer the ``<failure.ys>``, the longer each iteration will take. As the
|
|
design shrinks and `bugpoint` converges, each iteration should take less and
|
|
less time. Once all simplifications are exhausted and there are no more objects
|
|
that can be removed, the script will continue and the minimized design can be
|
|
saved.
|
|
|
|
|
|
What do I do with the minimized design?
|
|
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
|
|
|
|
- check minimized design still fails, especially if not using `write_rtlil`
|
|
- e.g. if you ran :ref:`bugpoint_script` below, then calling ``yosys -s
|
|
<failure.ys> min.v`` should still fail in the same way
|
|
|
|
.. code-block:: yoscrypt
|
|
:caption: example `bugpoint` minimizer
|
|
:name: bugpoint_script
|
|
|
|
read_verilog design.v
|
|
bugpoint -script <failure.ys>
|
|
write_verilog min.v
|
|
|
|
- `write_rtlil` is more reliable since `bugpoint` will have run that exact
|
|
code through the failing script; other ``write_*`` commands convert from the
|
|
RTLIL and then back again during the ``read_*`` which can result in
|
|
differences which mean the design no longer fails
|
|
- check out :ref:`using_yosys/bugpoint:identifying issues` for more on what to
|
|
do next
|
|
|
|
|
|
.. _minimize your script:
|
|
|
|
Minimizing scripts
|
|
------------------
|
|
|
|
- reminder to back up original code before modifying it
|
|
- if you're using command line, convert it to a script
|
|
- if you're using one of the :doc:`/using_yosys/synthesis/synth`, replace it
|
|
with its contents
|
|
|
|
+ can also do this piece-wise with the ``-run`` option
|
|
+ e.g. replacing ``synth -top <top> -lut`` with :ref:`replace_synth`
|
|
+ the options ``-top <top> -lut`` can be provided to each `synth` step, or
|
|
to just the step(s) where it is relevant, as done here
|
|
|
|
.. code-block:: yoscrypt
|
|
:caption: example replacement script for `synth` command
|
|
:name: replace_synth
|
|
|
|
synth -top <top> -run :coarse
|
|
synth -lut -run coarse:fine
|
|
synth -lut -run fine:check
|
|
synth -run check:
|
|
|
|
- remove everything *after* the error occurs
|
|
- can use `log` command to print messages to help locate the failure point
|
|
- `echo` can also help (``echo on``)
|
|
|
|
+ if you used a ``-run`` option like in :ref:`replace_synth` above, you can
|
|
now replace the failing step with its contents and repeat the above if
|
|
needed
|
|
+ checking the log should tell you the last command that ran which can make
|
|
this easier
|
|
+ say we ran :ref:`replace_synth` and were able to remove the ``synth -run
|
|
check:`` and still got our error, then we check the log and we see the last
|
|
thing before the error was ``7.2. Executing MEMORY_MAP pass (converting
|
|
memories to logic and flip-flops).``
|
|
+ we can then update our script to the following:
|
|
|
|
.. code-block:: yoscrypt
|
|
:caption: example replacement script for `synth` when `memory_map` is failing
|
|
|
|
synth -top <top> -run :fine
|
|
opt -fast -full
|
|
memory_map
|
|
|
|
- try ``write_rtlil <design.il>; design -reset; read_rtlil <design.il>;`` before
|
|
the failure point
|
|
|
|
+ ideally you now have a single command that is producing an error and can
|
|
`minimize your RTLIL`_ with the ``<design.il>`` output
|
|
+ if not, try to move the write/reset/read earlier in the script until you can
|
|
reproduce the error
|
|
+ if you have no idea where exactly you should put the reset, the best way is
|
|
to use a "binary search" type approach, reducing the possible options by
|
|
half after each attempt
|
|
|
|
* for example, your script has 16 commands in it before failing on the 17th
|
|
* if resetting immediately before the 17th doesn't reproduce the error, try
|
|
between the 8th and 9th (8 is half of the total 16)
|
|
* if that produces the error then you can remove everything before the
|
|
`read_rtlil` and try reset again in the middle of what's left, making sure
|
|
to use a different name for the output file so that you don't overwrite
|
|
what you've already got
|
|
* if the error isn't produced then you need to go earlier still, so in this
|
|
case you would do between the 4th and 5th (4 is half of the previous 8)
|
|
* repeat this until you can't reduce the remaining commands any further
|
|
|
|
.. TODO:: is it possible to dump scratchpad?
|
|
|
|
is there anything else in the yosys/design state that doesn't get included in
|
|
`write_rtlil`?
|
|
|
|
- you can also try to remove or comment out commands prior to the failing
|
|
command; just because the first and last commands are needed doesn't mean that
|
|
every command between them is
|
|
|
|
|
|
Minimizing Verilog designs
|
|
--------------------------
|
|
|
|
- manual process
|
|
- made easier if the error message is able to identify the source line or name
|
|
of the object
|
|
- reminder to back up original code before modifying it
|
|
- if a specific module is causing the problem, try to set that as the top
|
|
module, you can then remove
|
|
|
|
+ if the problem is parameter specific you may be able to change the default
|
|
parameters so that they match the problematic configuration
|
|
|
|
- as with `minimize your script`_, if you have no idea what is or is not
|
|
relevant, try to follow a "binary search" type approach where you remove (or
|
|
comment out) roughly half of what's left at a time
|
|
- focusing on one type of object at a time simplifies the process, removing as
|
|
many as you can until the error disappears if any of the remaining objects are
|
|
removed
|
|
- periodically check if anything is totally disconnected (ports, wires, etc), if
|
|
it is then it can be removed too
|
|
- start by removing cells (instances of modules)
|
|
|
|
+ if a module has no more instances, remove it entirely
|
|
|
|
- then processes
|
|
- try to remove or reduce assignments and operations
|
|
|
|
+ are there any wires/registers which get read but never written?
|
|
|
|
* try removing the signal declaration and replacing references to it with
|
|
``'0`` or ``'x``
|
|
* try this with constants too
|
|
|
|
+ can you replace strings with numeric values?
|
|
+ are you able to simplify any operations? like replacing ``a & '0`` with
|
|
``'0``
|
|
+ if you have enable or reset logic, does the error still happen without that?
|
|
+ can you reduce an ``if .. else`` to a single case?
|
|
|
|
- if you're planning to share the minimized code:
|
|
|
|
+ make sure there is no sensitive or proprietary data in the design
|
|
+ instead of a long string of numbers and letters that had some meaning (or
|
|
were randomly or sequentially generated), can you give it a single character
|
|
name like ``a`` or ``x``
|
|
+ please try to keep things in English, using the letters a-z and numbers 0-9
|
|
(unless the error is arising because of the names used)
|
|
|
|
|
|
Identifying issues
|
|
------------------
|
|
|
|
- does the failing command indicate limited support, or does it mention some
|
|
other command that needs to be run first?
|
|
- if you're able to, try to match the minimized design back to its original
|
|
context
|
|
|
|
+ could you achieve the same thing a different way?
|
|
+ and if so, does this other method have the same issue?
|
|
|
|
- try to change the design in small ways and see what happens
|
|
|
|
+ `bugpoint` can reduce and simplify a design, but it doesn't *change* much
|
|
+ what happens if you change operators, for example a left shift (or `$shl`)
|
|
to a right shift (or `$shr`)?
|
|
+ is the issue tied to specific parameters, widths, or values?
|
|
|
|
- if you're familiar with C/C++ you might try to have a look at the source
|
|
code of the command that's failing
|
|
|
|
+ even if you can't fix the problem yourself, it can be very helpful for
|
|
anyone else investigating if you're able to identify where exactly the
|
|
issue is
|
|
+ if you're using a fuzzer to find issues in Yosys, you should be prepared to
|
|
do this step
|
|
|
|
.. warning::
|
|
|
|
In the event that you are unable to identify the root cause of a fuzzer
|
|
generated issue, **do not** open more than one issue at a time. You have no
|
|
way of being able to tell if multiple fuzzer generated issues are simply
|
|
different cases of the same problem, and opening multiple issues for the same
|
|
problem means more time is spent on triaging and diagnosing bug reports and
|
|
less on fixing the problem. If you are found to be doing this, your issues
|
|
may be closed without further investigation.
|
|
|
|
- search `the existing issues`_ and see if someone has already made a bug report
|
|
|
|
+ this is where changing the design and finding the limits of what causes the
|
|
failure really comes in handy
|
|
+ if you're more familiar with how the problem can arise, you may be able to
|
|
find a related issue more easily
|
|
+ if an issue already exists for one case of the problem but you've found
|
|
other cases, you can comment on the issue and help get it solved
|
|
|
|
.. _the existing issues: https://github.com/YosysHQ/yosys/issues
|
|
|
|
- if there are no existing or related issues already, then check out the steps
|
|
for :ref:`using_yosys/bugpoint:creating an issue on github`
|
|
|
|
|
|
Creating an issue on GitHub
|
|
---------------------------
|
|
|
|
- "Reproduction Steps" is ideally a code-block (starting and ending with triple
|
|
backquotes) containing the minimized design (Verilog or RTLIL), followed by a
|
|
code-block containing the minimized yosys script OR a command line call to
|
|
yosys with code-formatting (starting and ending with single backquotes)
|
|
|
|
.. code-block:: markdown
|
|
|
|
min.v
|
|
```verilog
|
|
// minimized Verilog design
|
|
```
|
|
|
|
min.ys
|
|
```
|
|
read_verilog min.v
|
|
# minimum sequence of commands to reproduce error
|
|
```
|
|
|
|
OR
|
|
|
|
`yosys -p ': minimum sequence of commands;' min.v`
|
|
|
|
- alternatively can provide a single code-block which includes the minimized
|
|
design as a "here document" followed by the sequence of commands which
|
|
reproduce the error
|
|
|
|
+ see :doc:`/using_yosys/more_scripting/load_design` for more on heredocs.
|
|
|
|
.. code-block:: markdown
|
|
|
|
```
|
|
read_rtlil <<EOF
|
|
# minimized RTLIL design
|
|
EOF
|
|
# minimum sequence of commands
|
|
```
|
|
|
|
- any environment variables or command line options should also be mentioned in
|
|
the "Reproduction Steps"
|