![[logo]](/img/veripool_small.png)
Activity
From 09/13/2019 to 10/12/2019
10/12/2019
- 10:57 PM Issue #1548 (Closed): Fuzzer: Segfault when table has error
- Using version:
Verilator 4.019 devel rev v4.018-39-g239ef1ae
And command line:
verilator_bin --lint-only 1.s...
10/11/2019
- 08:24 PM Issue #1547 (Closed): Refactor Verilator building in Travis
- Currently, for verilator CI we are caching object files when building Verilator but for verilator_ext_tests we are do...
- 01:01 AM Issue #1363: CMake support
- Maarten is right, without setting the C++ version threading fails on the Travis compiler. I've gone back to if(THREAD...
10/10/2019
- 11:53 PM Issue #1363: CMake support
- Great. Thanks for separating, I pushed the test suite change part. Looking at rest...
- 11:35 PM Issue #1363: CMake support
- Ok, I've made a few more tweaks. Most of the tests work now with (despite) this patch:...
- 04:50 PM Issue #1363: CMake support
- Ok, cmake support will require the user to set the correct C++ standard and pass the correct SYSTEMC_CXX_FLAGS. It wi...
- 04:23 PM Issue #1363: CMake support
- I think we should assume C++11, so not muck with flags. At present Verilator complains on any non-C11 system and say...
- 03:55 PM Issue #1363: CMake support
- I was thinking we'd change the condition around setting C++11 to if(THREADING && !SYSTEMC) rather than if(THREADING)....
- 03:52 PM Issue #1363: CMake support
- I suggest we should not set any language flags, I think any other policy will just lead to trouble with libraries. I...
- 03:41 PM Issue #1363: CMake support
- Passing SYSTEMC_CXX_FLAGS will still fail the vltmt test because the C++11 flag of multithreaded will overrule the st...
- 02:59 PM Issue #1363: CMake support
- >Principally it looks like the tests where it compares the STDOUT to a golden file fail. I propose to just add a cmak...
- 02:51 PM Issue #1363: CMake support
- Patrick Stewart wrote:
> Principally it looks like the tests where it compares the STDOUT to a golden file fail. I p... - 02:07 PM Issue #1363: CMake support
- Principally it looks like the tests where it compares the STDOUT to a golden file fail. I propose to just add a cmake...
- 01:40 PM Issue #1363: CMake support
- Wilson Snyder wrote:
> 1. Perhaps we can just use the existing SYSTEMC_CXX_FLAGS? I was going to compare the gmake ... - 12:48 PM Issue #1363: CMake support
- 1. Perhaps we can just use the existing SYSTEMC_CXX_FLAGS? I was going to compare the gmake vs cmake options passed ...
- 12:29 PM Issue #1363: CMake support
- I should admit I did not test systemc with multithreaded verilator.
Hardcoding the c++ standard to 14 is imho a ba... - 02:13 AM Issue #1363: CMake support
- Wilson Snyder wrote:
> Very close.
>
> 1. Please use --generate-key mentioned previously.
>
> 2. Please remove...
10/09/2019
- 11:35 PM Issue #1363: CMake support
- Very close.
1. Please use --generate-key mentioned previously.
2. Please remove the unlink "$self->{obj_dir}/CM... - 10:54 PM Issue #1363: CMake support
- Got thinking about --protect-key in that test; I committed a fix to add a new option --generate-key as that seems use...
- 08:07 PM Issue #1363: CMake support
- I've added an additional change to use --protect-key in the cmake_protect_lib example, otherwise you can get problems...
- 04:57 PM Issue #1363: CMake support
- > I cleaned these and other minor stuff up, see attached patch. Please check this & commit.
Thanks, added
>
>... - 10:54 AM Issue #1490 (Resolved): Add an option to create a DPI protected library
- Squashed and pushed. Agreed that we can fix up --protect-lib under CMake when it's all in the trunk.
- 01:28 AM Issue #593 (WillNotFix): Support running make and creating top C file
- Verilating and making can be done with cmake which is soon to be completed.
Making a main.cpp needs timescales, wh... - 01:03 AM Issue #1185 (WillNotFix): Support for interfaces in top level ports
- This is unlikely to be supported in the medium term due to the complexities of mapping into C++. A wrapper is sugges...
- 12:41 AM Issue #1472 (NotEnoughInfo): Memory not updating in for loop
- Thanks for filing this, feel free to reopen if a complete test case can be provided.
10/08/2019
- 11:46 PM Issue #1363: CMake support
- Sorry I'm pestering you with these small batches; I'm learning as I use it.
Got some warnings:
test_regress... - 06:08 PM Issue #1363: CMake support
- Ok, I think it's correct now. I've also changed Maarten's email address as he requested.
- 04:42 PM Issue #1363: CMake support
- Apologies I mucked up keeping the changes separate after changing the indentation. I'll check in new version soon, I ...
- 04:29 PM Issue #1363: CMake support
- Tried your branch quickly.
0. Got
CMake Error at verilator/verilator-config.cmake:311:
Parse error. Exp... - 02:24 PM Issue #1363: CMake support
> Suggest we standardize on two spaces in cmake files per indent level as that seems what others recommend and is w...- 02:20 AM Issue #1363: CMake support
- Few more things:
Suggest we standardize on two spaces in cmake files per indent level as that seems what others re... - 10:15 PM Issue #1490: Add an option to create a DPI protected library
- In this
+++ b/bin/verilator
+encrypted RTL (i.e. IEEE P1735). See examples/protect_lib in the
+dist... - 02:15 PM Issue #1490: Add an option to create a DPI protected library
- I believe this covers all the feedback:
https://github.com/toddstrader/verilator-dev/tree/protect-lib
Also, per s... - 01:52 AM Issue #1489: Python support for Verilated designs
- > In `src/V3EmitMk.cpp`, I change the default target to PREFIX__PYTHON when python support is enabled.
> Has this br...
10/07/2019
- 11:06 PM Issue #1534: Fuzzer script
- I'd suggest putting the files under the "nodist" directory, this is for things that are developer only and don't need...
- 05:31 PM Issue #1534: Fuzzer script
- The contributor agreement is fine.
At this point, I don't have a real bird's eye view of the how the project wor... - 11:03 PM Issue #1489: Python support for Verilated designs
- Patrick Stewart wrote:
> Maarten: The python gmake support was missing a rule to actually build the module.
> I've ... - 05:41 PM Issue #1489: Python support for Verilated designs
- I've just pushed a new version to github that should address the above issues. I've stripped out the unrelated format...
- 05:46 PM Issue #1363: CMake support
- I've renamed the examples in the PR.
> Finally, I'm not that up on CMake. I'm hoping as cmake-related issues arise... - 12:04 PM Issue #1545: Warning-CASEOVERLAP is not triggering for signals wider than 12
- Thanks for confirming and providing the workaround.
I was suspecting this kind of lookup table to be the limiting fa... - 11:38 AM Issue #1545 (Confirmed): Warning-CASEOVERLAP is not triggering for signals wider than 12
- Currently the algorithm is very stupid, namely build a lookup table and check for duplicates.
I changed this defin... - 09:33 AM Issue #1545 (Confirmed): Warning-CASEOVERLAP is not triggering for signals wider than 12
- Hi,
I noticed that I'm not getting the CASEOVERLAP warnings if the value is wider than 12 bits.
In the example be...
10/06/2019
- 05:25 PM Issue #1521 (Resolved): Add --protect-ids to enhance --protect-lib obfuscation
- Pushed to git towards 4.022.
Todd, you'll want to presumably make this new option also turn on when --protect-lib ... - 01:30 PM Issue #1521: Add --protect-ids to enhance --protect-lib obfuscation
- Calling it --protect-ids; nearly there.
- 02:56 PM Issue #1535 (Assigned): Double quotes in -f option file
- Great, awaiting next patch.
- 02:36 PM Issue #1544 (Closed): Improve readme
- The existing readme document is really an install document.
Readme should be more modern, github-like. This probab... - 02:33 PM Issue #1489: Python support for Verilated designs
- Note examples have been renamed, see last msg in bug1363.
- 02:32 PM Issue #1363 (Assigned): CMake support
- I renamed the old examples to have "make" in the name so users can tell make vs cmake. Can you please rename yours t...
- 02:08 PM Issue #1469 (Closed): VPI module
- In 4.020. Thanks for reporting this; if there are additional related problems, please open a new issue.
- 02:08 PM Issue #1536 (Closed): Misoptimization of case with string compares
- In 4.020. Thanks for reporting this; if there are additional related problems, please open a new issue.
- 02:08 PM Issue #1528 (Closed): VPI: Let port vars return variable type, not vpiPort
- In 4.020. Thanks for reporting this; if there are additional related problems, please open a new issue.
- 02:07 PM Issue #1525 (Closed): Improve error report when erroneously using interface in expression
- In 4.020. Thanks for reporting this; if there are additional related problems, please open a new issue.
- 02:07 PM Issue #1512 (Closed): Make callCbs public (again) and return if callbacks were called
- In 4.020. Thanks for reporting this; if there are additional related problems, please open a new issue.
- 02:07 PM Issue #1511 (Closed): Add --public-flat-rw flag
- In 4.020. Thanks for reporting this; if there are additional related problems, please open a new issue.
- 02:07 PM Issue #1510 (Closed): VPI: Get vpiType string
- In 4.020. Thanks for reporting this; if there are additional related problems, please open a new issue.
- 02:07 PM Issue #1509 (Closed): Fix variable cast in VPI
- In 4.020. Thanks for reporting this; if there are additional related problems, please open a new issue.
- 02:07 PM Issue #1507 (Closed): Linting error: backward indexing
- In 4.020. Thanks for reporting this; if there are additional related problems, please open a new issue.
- 02:07 PM Issue #1499 (Closed): Have V3EmitC::emitcSyms honor --output-split-cfuncs
- In 4.020. Thanks for reporting this; if there are additional related problems, please open a new issue.
- 02:07 PM Issue #1496 (Closed): Support additional file I/O functions $fseek, $ftell, $frewind
- In 4.020. Thanks for reporting this; if there are additional related problems, please open a new issue.
- 02:06 PM Issue #1494 (Closed): verilator 4.018 build fails at 'make test' in examples/hello_world_c
- In 4.020. Thanks for reporting this; if there are additional related problems, please open a new issue.
- 02:05 PM Verilator 4.020 Released
- Verilator 4.020 2019-10-06
*** Support $fseek, $ftell, $frewind, bug1496. [Howard Su]
*** Add --public-flat... - 01:23 PM Issue #1543 (Feature): Improve performance of icache
- 01:14 PM Issue #1543 (Feature): Improve performance of icache
- Feature tracking bug.
The biggest performance cost presently is icache pressure. We should figure out techniques ... - 01:05 PM Issue #1542 (Feature): Improve performance of bit operations by vectorizing
- Feature tracking bug.
Verilator at present only vectorizes some very simple assignments of the form
assign a... - 01:01 PM Issue #1541 (Feature): Support unpacked structs
- Feature tracking bug.
As a step towards class support (bug377), support unpacked structures.
- 01:00 PM Issue #1540 (Feature): Improve internal elaboration algorithm
- Feature tracking bug.
Verilator implements elaboration using an algorithm that goes back to its birth and predates... - 12:54 PM Issue #1539 (Feature): Improve internals of inliner
- Feature tracking bug.
The current module inliner code works by operating on modules before the complete scoped hie... - 12:34 PM Issue #1538 (Feature): Support full UVM parsing and XML dump
- Feature tracking bug.
Currently Verilator parses a subset of SystemVerilog and reports many unsupported errors at ... - 12:31 PM Issue #1537 (Feature): Support FST structure tracing w/GTKwave
- Feature tracking bug.
Currently Verilator dumps structs as packed values in FST. Ideally it would allow the user ...
10/05/2019
- 05:40 PM Issue #1536: Misoptimization of case with string compares
- Confirmed fixed. Thanks Wilson for the quick fix!
- 01:13 AM Issue #1489: Python support for Verilated designs
- I'm getting confused by all the layers. Do you propose
1. cmake goes to master, 2. everything in pydpi?
Or 1. ...
10/04/2019
- 11:42 PM Issue #1363: CMake support
- Ok, thanks for rebasing. Looked through this again and a few things:
examples/cmake -> rename to examples/c... - 11:20 PM Issue #1363: CMake support
- Note to self - this is: https://github.com/verilator/verilator/pull/2
- 07:40 PM Issue #1363: CMake support
- I've added a rebased version as a pull request on your new(?) github repo. I'm hoping we can get the CMake in at leas...
- 11:20 PM Issue #1489: Python support for Verilated designs
- Note to self this is also https://github.com/verilator/verilator/pull/3
- 11:03 PM Issue #1489: Python support for Verilated designs
- I have added DPI support for Python in a pydpi branch.
These commits should go on top of the python branch.
(which ... - 07:47 PM Issue #1489: Python support for Verilated designs
- I've rebased this work onto master, added a couple of minor fixes and added coverage support here: https://github.com...
- 11:14 PM Issue #1536 (Resolved): Misoptimization of case with string compares
- Thanks for the good test case. What's going on is when the case statement is getting optimized it treated the differ...
- 09:32 AM Issue #1536 (Closed): Misoptimization of case with string compares
- I have (simplified) code like this:...
10/03/2019
- 09:38 AM Issue #1521: Add --protect-ids to enhance --protect-lib obfuscation
- #4 makes sense to me. Naively, it seems that s/name()/nameHashed()/ in V3EmitC would be pretty close to what we need...
10/02/2019
- 10:48 PM Issue #1469: VPI module
- bug1469-1 pushed.
- 01:59 PM Issue #1469: VPI module
- Hi,
sorry, even worse, I deleted the branch and the comment because I realized it could be done more elegantly: I ... - 11:16 AM Issue #1469: VPI module
- I suspect you forgot to push bug1469-1.
- 01:58 AM Issue #1469 (Resolved): VPI module
- Great, looks good now thanks for taking all the feedback.
Pushed to git towards eventual 4.020 release.
- 04:15 PM Issue #1535: Double quotes in -f option file
- Hi Wilson,
As for big-3 simulators, i have tested the following cases:... - 04:01 PM Issue #1535: Double quotes in -f option file
- I acknowledge this and future contributions are made under the Developer Certificate of Origin (https://developercert...
- 03:07 PM Issue #1535 (AskedReporter): Double quotes in -f option file
- Thanks for your issue & looking at how to fix this.
As the old comment noted, most other simulators don't support ... - 02:50 PM Issue #1535 (Closed): Double quotes in -f option file
- SystemVerilog allows string parameters.
Using -f option file should support quoted strings.
Please find, attached... - 02:04 AM Issue #1501: Support bind statements within generate blocks
- Thanks!! It's not a show-stopper for us. Our workaround for anyone watching is to clock-gate based on the bind param...
- 01:59 AM Issue #1501: Support bind statements within generate blocks
- Haven't forgotten about this. Got the exact case provided working, but based on where in the tree the bind is there's...
10/01/2019
- 10:51 AM Issue #1521: Add --protect-ids to enhance --protect-lib obfuscation
- Got an experiment in place and realized changing v3Name to make this work
doesn't really work, as it currently adds ... - 10:45 AM Issue #1513 (Closed): CONTRIBUTING file, internals, and FAQ cleanup
- Thanks all.
- 10:06 AM Issue #1513: CONTRIBUTING file, internals, and FAQ cleanup
- I've added a blurb in internals.adoc about the CI setup.
- 05:16 AM Issue #1469: VPI module
- Hi again,
sorry, I shouldn't develop, test and prepare patches while running ORConf, branches got mixed up. Everyt... - 01:53 AM Issue #1534 (Feature): Fuzzer script
- This is much appreciated.
1. First legal stuff, do you agree what you attach and future contributions are open sou... - 12:50 AM Issue #1534: Fuzzer script
- As far as I know there isn't an option to keep the inputs ASCII.
There is an option to use a collection of inter... - 12:09 AM Issue #1534: Fuzzer script
- Sweet. Always wanted to try AFL but never got to it.
Ran it for 2 minutes (which know is nothing the the grand sch... - 12:09 AM Issue #1533 (Closed): Fuzzer: buffer overflow detected
- Fixed in git. As this is unlikely to be hit by users, closing immediately rather than waiting for a release.
09/30/2019
- 11:48 PM Issue #1530 (Closed): Fuzzer: terminate called after throwing an instance of 'std::length_error'
- Fixed in git. As this is unlikely to be hit by users, closing immediately rather than waiting for a release.
- 08:06 PM Issue #1530 (Closed): Fuzzer: terminate called after throwing an instance of 'std::length_error'
- Running the attached test case with "verilator_bin --cc 2.v" produces:...
- 11:36 PM Issue #1531 (Duplicate): Fuzzer: Hang on short testcase
- Same problem/solution as bug1530.
- 08:16 PM Issue #1531 (Duplicate): Fuzzer: Hang on short testcase
- Running "verilator_bin --cc 3.v" with the attaced testcase does not terminate in a reasonable amount of time. It doe...
- 11:22 PM Issue #1529 (Closed): Fuzzer: Sefault on pattern testcase
- Fixed in git. As this is unlikely to be hit by users, closing immediately rather than waiting for a release.
- 10:47 PM Issue #1529: Fuzzer: Sefault on pattern testcase
- I have filed a bug with instructions on how to run the fuzzer here:
https://www.veripool.org/issues/1534-Verilator... - 09:21 PM Issue #1529: Fuzzer: Sefault on pattern testcase
- I take it this and the other issues are from a fuzzer? Which one? Would you be willing to contribute the configurat...
- 07:59 PM Issue #1529 (Closed): Fuzzer: Sefault on pattern testcase
- Running "verilator_bin --cc 1.v" with the attached testcase produces:...
- 10:50 PM Issue #1532 (Closed): Fuzzer: fatal flex scanner internal error--end of buffer missed
- Thanks for the testcase. Flex EOF detection wasn't properly handled.
Fixed in git. As this is unlikely to be hit b... - 08:32 PM Issue #1532 (Closed): Fuzzer: fatal flex scanner internal error--end of buffer missed
- Running "verilator_bin --cc 5.v" produces:...
- 10:47 PM Issue #1534 (Closed): Fuzzer script
- In this bug: https://www.veripool.org/issues/1529-Verilator-Sefault-on-short-testcase
There was a request to file ... - 08:43 PM Issue #1533 (Closed): Fuzzer: buffer overflow detected
- Running the attached testcase with "verilator_bin --cc 6.v" produces:...
09/29/2019
- 09:14 PM Issue #1513 (Assigned): CONTRIBUTING file, internals, and FAQ cleanup
- Sigh. Keeping this open as a reminder.
- 09:14 PM Issue #1513 (Closed): CONTRIBUTING file, internals, and FAQ cleanup
- Committed CONTRIBUTING.adoc and conversion of internals.pod to internals.adoc.
Switched to asciidoc over markdown ... - 06:49 PM Issue #1469: VPI module
- Wilson Snyder wrote:
> Before, and worse after. I thought my small changes might have done it, but seemed to get th... - 06:44 PM Issue #1469: VPI module
- Before, and worse after. I thought my small changes might have done it, but seemed to get the same with the patch yo...
- 06:25 PM Issue #1469: VPI module
- Thanks, Wilson. Just to confirm: The tests are broken with the dangling pointer check in and not before, right?
- 06:01 PM Issue #1469: VPI module
- If you want to post github links the easiest thing to do is make a personal branch (e.g. called bug1469) and point to...
- 03:11 PM Issue #1469: VPI module
- Hi,
thanks, I was thinking the same, but not sure about the impact. As people probably rely on name() returning th... - 03:25 PM Issue #1528: VPI: Let port vars return variable type, not vpiPort
- Just for completeness. When it return vpiPort, the proper sequence would be something like:...
09/28/2019
- 10:58 PM Issue #1513: CONTRIBUTING file, internals, and FAQ cleanup
- Todd, Good point, I agree we should mention Travis, can you please suggest what we should add? I suggest at bullet p...
- 10:34 PM Issue #1513: CONTRIBUTING file, internals, and FAQ cleanup
- This all looks good to me. I run perldoc on the .pod files, but agree that it is probably best to move them to .md t...
- 03:42 PM Issue #1513: CONTRIBUTING file, internals, and FAQ cleanup
- FAQ cleanup seemed obvious improvement, wiki now points at most recent release's bin/verilator FAQ.
- 03:08 PM Issue #1513 (Assigned): CONTRIBUTING file, internals, and FAQ cleanup
- 1. Currently the internals.pod file is used to make HTML/PDF/etc files. I doubt these are useful, is anyone using th...
- 04:11 PM Issue #1469: VPI module
- Ok, only remaining problem I think is the new decodeName function by doing a strdup will leak more on each call. One...
- 03:18 PM Issue #1469: VPI module
- Hi,
thanks a lot for pointing it out. I discovered that icarus (and other simulators I assume) actually return the... - 01:11 AM Issue #1469: VPI module
- Thanks, very close.
+++ b/test_regress/t/t_vpi_module.cpp
+ CHECK_RESULT_CSTR(name, "t.mod_a.mod_c__02E"...
09/27/2019
- 04:45 PM Issue #1469: VPI module
- Hi,
thanks for your input. I have updated the patch with your input and also test the escaped identifier.
Best,... - 12:36 PM Issue #1490: Add an option to create a DPI protected library
- Since you're neutral on lib vs library, if we use --protect-lib I believe your paper/this feature will be top google ...
- 08:37 AM Issue #1490: Add an option to create a DPI protected library
- As discussed, I sliced of the XSim support and new AST nodes as two separate commits and pushed them.
> Multitop s... - 03:36 AM Issue #1490: Add an option to create a DPI protected library
- Multitop should make just one top V/Mk file, I still think this will work fine.
- 04:02 AM Issue #1528 (Resolved): VPI: Let port vars return variable type, not vpiPort
- Not sure I follow what this helps, but don't see what it breaks either and tests don't cover it, so will assume it he...
09/26/2019
- 08:18 PM Issue #1490: Add an option to create a DPI protected library
- I'm ambivalent towards the choice between --dpi-lib and --dpi-library but I do think one of those would be better tha...
- 12:12 PM Issue #1490: Add an option to create a DPI protected library
- >I'm going to suggest --protect-lib.
Like that, let's sleep on it.
Is protect-library better or worse? I'm alwa... - 11:17 AM Issue #1490: Add an option to create a DPI protected library
- Also, re: testing this specific error:...
- 07:34 AM Issue #1490: Add an option to create a DPI protected library
- > perhaps the makefile by default should make both versions?
Yeah, the only reason I have this as two separate Ver... - 04:51 AM Issue #1490: Add an option to create a DPI protected library
- Also . . .
> The generated code seems wrongly indented (not indented) for some reason.
Isn't this because I'm u... - 04:25 AM Issue #1490: Add an option to create a DPI protected library
- Thanks again. Also, regarding names, I hate naming things. And yeah, I'm already pretty sure that --dpi-protect is ...
- 01:41 AM Issue #1490: Add an option to create a DPI protected library
- >Is dpiProtectLibName() the same thing as the new v3Global.opt function you're suggesting? Does >dpiProtectLibName() ...
- 01:00 PM Issue #1528 (Closed): VPI: Let port vars return variable type, not vpiPort
- VerilatedVpioVar does not provide any of the other properties of the vpiPort, but those of the referenced "vpiLowConn...
- 12:39 PM Issue #1470: VPI systemtf
- The first part is much easier than I thought, because @Verilated::fatalOnVpiError(false);@ does the job. Now I need t...
- 11:55 AM Issue #1521: Add --protect-ids to enhance --protect-lib obfuscation
- Ease of use is a good point so agree --dpi-protect (or whatever renamed) should set --protect-symbols by default; if ...
- 04:09 AM Issue #1521: Add --protect-ids to enhance --protect-lib obfuscation
- Do you see uses for this beyond --dpi-protect? I'd imagine, but of course can't say for sure, that IP vendors wouldn...
- 02:44 AM Issue #1521: Add --protect-ids to enhance --protect-lib obfuscation
- Proposed new flags to implement this feature for discussion:
--protect-symbols = When specified, Verilator wi... - 01:54 AM Issue #1525 (Resolved): Improve error report when erroneously using interface in expression
- Thanks for your issue. I was unable to get the exact error message you did, but got other ugly messages when misusin...
09/25/2019
- 08:21 PM Issue #1490: Add an option to create a DPI protected library
- Thanks for all the feedback. I've got a bunch of the items cleaned up and posted to my branch. But I've got a coupl...
- 12:15 AM Issue #1490: Add an option to create a DPI protected library
- ... think noe that you have AstText you can just hash those in the SV to fingerprint.
- 12:09 AM Issue #1490: Add an option to create a DPI protected library
- Forgot another. The generated V should pass a fingerprint of the info used to make the V to the DPI create for checki...
- 07:04 AM Issue #1525 (Closed): Improve error report when erroneously using interface in expression
- When an interface ("din" in this example) is erroneously used in an expression like this:...
09/24/2019
- 11:53 PM Issue #1490: Add an option to create a DPI protected library
- Forgot one of your questions, it's "DESCR" "IPTION" as I have personal scripts that check every source file has a des...
- 11:51 PM Issue #1490: Add an option to create a DPI protected library
- Good stuff. Review comments attached.
Main comment would be we should heave sensitivity lists based on what's need... - 04:05 PM Issue #1490: Add an option to create a DPI protected library
- I believe this is ready to go now:
https://github.com/toddstrader/verilator-dev/tree/dpi-compile-ast2
A couple no... - 02:44 PM Issue #1522 (Feature): Support mutable top-level parameters for --protect-lib
- Parameters that create data types or generate if/for subcells are unlikely to ever be "wire"able, but perhaps others ...
- 12:47 PM Issue #1522 (Feature): Support mutable top-level parameters for --protect-lib
- This may also have applications outside of --dpi-protect. Possible solutions:
* Automatically convert parameters to... - 02:42 PM Issue #1521 (Feature): Add --protect-ids to enhance --protect-lib obfuscation
- If anyone starts on this let me know, as I at some point might also take a look at it.
- 12:41 PM Issue #1521 (Closed): Add --protect-ids to enhance --protect-lib obfuscation
- Signal names and some other design details are visible via library symbols and possibly strings. We can have Verilat...
- 02:41 PM Issue #1520 (Feature): Improve --protect-lib performance
- 12:37 PM Issue #1520 (Feature): Improve --protect-lib performance
- Some possibilities:
* Detect cones of logic and only fetch the outputs that are affected by a given input
* Detect ... - 02:41 PM Issue #1519 (Feature): Benchmark --protect-lib runtime
- 12:34 PM Issue #1519 (Closed): Benchmark --protect-lib runtime
- There's obviously overhead in shuttling data back and forth with --dpi-protect. We should use some large-ish design ...
- 02:41 PM Issue #1518 (Feature): Protect against --protect-lib Verilator runtime incompatibility
- 12:16 PM Issue #1518: Protect against --protect-lib Verilator runtime incompatibility
- Some items such as the mutexes for multithreading, tracing, VPI name lookup, and coverage are assumed to be common ac...
- 12:08 PM Issue #1518 (Feature): Protect against --protect-lib Verilator runtime incompatibility
- Currently all --dpi-protect'ed modules and the top-level design share the same Verilator runtime symbols. Some optio...
- 02:40 PM Issue #1517 (Feature): Add support for additional simulators in --protect-lib
- 12:02 PM Issue #1517 (Feature): Add support for additional simulators in --protect-lib
- So far XSim has been tested and works with a shared library. We need to determine what we need to be able to build l...
- 02:40 PM Issue #1523 (Feature): Add waveform replay tool
- Would ideally like to see something that works on an arbitrary design without requiring --dpi-protect, and ideally us...
- 12:51 PM Issue #1523 (Feature): Add waveform replay tool
- When simulating with a --dpi-protect'ed module, the end user cannot see the internals and therefore a waveform captur...
- 02:24 PM Issue #1524 (Assigned): Support sensitivity to DPI function outputs
- I need to look into this more, but I believe that Verilator does schedule processes based on the outputs of DPI funct...
09/23/2019
- 11:27 PM Issue #1469: VPI module
- Thanks for working on this. I think your general approach is fine.
I suggest you add a test where there's some bac... - 09:55 PM Issue #1469: VPI module
- This has been more complicated than I hoped. Attached I am sending you my proposal for review. It is probably not the...
- 10:44 PM Issue #1490: Add an option to create a DPI protected library
- Thanks, I think this looks much better & reasonable. I'd suggest using AstComment where you can instead of text with...
- 06:35 PM Issue #1490: Add an option to create a DPI protected library
- This is very much a WIP, but I want to make sure we're on the same page:
https://github.com/toddstrader/verilator-de... - 12:04 PM Issue #1511 (Resolved): Add --public-flat-rw flag
- Thanks all.
Pushed to git towards eventual 4.020 release.
09/21/2019
- 06:36 PM Issue #1514 (Feature): Switch for file to read public signals from
- Agreed, and suggest this should be handled by adding a config file parse rule to the bottom of verilog.y.
Ideally ... - 05:47 PM Issue #1514 (Feature): Switch for file to read public signals from
- We should have a switch that lets the user give a file with names of signals (including whitespaces) that are treated...
- 06:26 PM Issue #1515 (Feature): VPI: Log all variables that were accessed
- 05:49 PM Issue #1515 (Feature): VPI: Log all variables that were accessed
- The idea is to add the capability to the VPI runtime to log all variable accesses in a format compatible with #1514. ...
- 05:43 PM Issue #1513 (Closed): CONTRIBUTING file, internals, and FAQ cleanup
- We should add a CONTRIBUTING file on toplevel that describes the required technical and legal steps to submit patches...
- 05:41 PM Issue #1511: Add --public-flat-rw flag
- Wilson Snyder wrote:
> Ok, given the trade-offs seems reasonable.
>
> If this starts getting a lot of use we migh... - 12:51 PM Issue #1511: Add --public-flat-rw flag
- Ok, given the trade-offs seems reasonable.
If this starts getting a lot of use we might want an option to at runti... - 01:24 PM Issue #1490: Add an option to create a DPI protected library
- >It also slightly modifies the emitted Makefile, but I'm going to assume we're not going to represent the Makefile in...
- 11:45 AM Issue #1512 (Resolved): Make callCbs public (again) and return if callbacks were called
- Your description seems reasonable. One reason why I was asking was if this should be documented in the main docs, bu...
09/20/2019
- 03:40 PM Issue #1470: VPI systemtf
- Wilson Snyder wrote:
> Perhaps when suppressed this should be setting the error return code so vpi_chk_error returns... - 03:32 PM Issue #1470: VPI systemtf
- Perhaps when suppressed this should be setting the error return code so vpi_chk_error returns something sane e.g. lev...
- 02:30 PM Issue #1470: VPI systemtf
- Okay, here is one that is equally hacky and elegant on both sides. Instead of making the symbol weak, how about just ...
- 04:25 AM Issue #1470: VPI systemtf
- Thanks, Wilson, I was already thinking about the issue with other compilers but only came up with LLVM :) I think it ...
- 01:47 PM Issue #1512: Make callCbs public (again) and return if callbacks were called
- Hi,
conceptually the VPI library is generic in cocotb and it used some init hooks to install the callbacks. But Ve... - 01:30 PM Issue #1490: Add an option to create a DPI protected library
- More questions about the in-AST refactor:
Currently "--dpi-protect secret" creates foo.sv and foo.cpp. It also sl... - 05:01 AM Issue #1511: Add --public-flat-rw flag
- Thanks for your insights, I am a bit torn apart on this one. But I believe overall the usability wins in this case. T...
09/19/2019
- 11:01 PM Issue #1512 (Assigned): Make callCbs public (again) and return if callbacks were called
- I'm ok applying this conceptually, but do you think this is the API you really want to use instead of abstracting thr...
- 08:17 PM Issue #1512 (Closed): Make callCbs public (again) and return if callbacks were called
- For the cocotb support we need VPI callbacks of all sorts. The access to callCbs() was public before but has not been...
- 10:52 PM Issue #1511 (Assigned): Add --public-flat-rw flag
- I really don't like expanding the userfullness use of public. There's two problems:
1. Users will probably set it... - 05:16 AM Issue #1511: Add --public-flat-rw flag
- This is a compromise between --public (deprecated) and requiring the designer to add the comments. By restricting it ...
- 05:04 AM Issue #1511 (Closed): Add --public-flat-rw flag
- Hi,
the purpose of these patches is to add a new switch --public-flat-rw that changes the behavior as if /*verilat... - 10:21 PM Issue #1470: VPI systemtf
- Weak seems somewhat hacky, but I can live with hacky. You do however need to do not break non-GCC compilers e.g. MSV...
- 09:00 PM Issue #1470: VPI systemtf
- Hi,
sorry, took much longer as I hoped to even look into it. Attached please find a first, rather trivial patch, t...
09/18/2019
- 04:24 PM Issue #1510 (Resolved): VPI: Get vpiType string
- Great. Pushed to git for eventual release 4.020.
- 11:45 AM Issue #1510: VPI: Get vpiType string
- Patch updated with test. Actually I found vpi_var the better test as it tests vpi_get_str too.
- 11:19 AM Issue #1510: VPI: Get vpiType string
- Yes, I am at it right now. Sorry, slipped me :)
- 11:18 AM Issue #1510 (Assigned): VPI: Get vpiType string
- Thanks for the patch, makes sense, could you also please add a test to t_vpi_get.cpp or one of the other tests? If n...
- 11:09 AM Issue #1510 (Closed): VPI: Get vpiType string
- vpi_get_str: Get vpiType string
- 11:25 AM Issue #1509 (Resolved): Fix variable cast in VPI
- Thanks for the good patch. Applied & pushed to git for eventual 4.016 release.
- 11:14 AM Issue #1509: Fix variable cast in VPI
- Updated patch for current HEAD
- 10:00 AM Issue #1509 (Closed): Fix variable cast in VPI
- Fix cast from Var to base type when getting handle in VPI
09/17/2019
- 08:09 PM Issue #1507: Linting error: backward indexing
- It worked fine. Thanks!
- 07:17 PM Issue #1507: Linting error: backward indexing
- Sorry, not having it in master was my fault, try again.
- 06:55 PM Issue #1507: Linting error: backward indexing
- Let me know if it is better to open a new issue, since it is unrelated topic.
I am on the master branch right now,... - 06:15 PM Issue #1507: Linting error: backward indexing
- It's not released (until next monthish), use master branch, that is don't use any tag.
- 06:02 PM Issue #1507: Linting error: backward indexing
- Wilson Snyder wrote:
> You can get it now from git, see the Install section. Typically releases (tarballs) are ever... - 05:25 PM Issue #1507: Linting error: backward indexing
- You can get it now from git, see the Install section. Typically releases (tarballs) are every month or so.
- 04:45 PM Issue #1507: Linting error: backward indexing
- Wilson Snyder wrote:
> Thanks for the report, simple enough to fix.
>
> Fixed in git towards 4.420.
Thanks for... - 01:18 AM Issue #1507 (Resolved): Linting error: backward indexing
- Thanks for the report, simple enough to fix.
Fixed in git towards 4.420.
- 12:58 AM Issue #1507 (Closed): Linting error: backward indexing
- I think I met a false error in linting:
Slice selection '[2:2]' has backward indexing versus data type's '[0:3]'
Sl... - 02:53 PM Issue #1508 (NoFixNeeded): CI is failing on CPAN installs
- OK, great. I didn't look closely enough at the log before, but now I see:...
- 01:28 PM Issue #1508: CI is failing on CPAN installs
- Note CPAN listed Parallel::Forker but it didn't seem to find it in Travis. This morning I tried pushing a new Parall...
- 01:13 PM Issue #1508 (NoFixNeeded): CI is failing on CPAN installs
- Travis jobs are failing while installing CPAN modules:...
09/15/2019
- 09:09 PM Issue #1506 (NoFixNeeded): Compile with uvm-systemc
- UVM-SystemC as I understand it is just another library which should not care about regular SystemC verilated modules....
- 07:15 PM Issue #1506 (NoFixNeeded): Compile with uvm-systemc
- Is there any way or example to compile the verilator with uvm-systemc?
Also available in: Atom