Verilator

07/23/2008

12:52 PM Verilator 3.670 Released

Verilator 3.670 2008/07/23 ** Add --x-assign=fast option, and make it the default. This chooses performance over reset debugging. See the manual. ** Add --autoflush, for flushing streams after $display. [Steve Tong] ** Add C... Wilson Snyder

11:54 AM Issue #14 (Closed): Verilator Doesn't catch duplicate declaration of signal

As you noted, this is ugly, but very hard to fix at this point as everything is tracked by signal, not bit. Someday. Wilson Snyder

07:22 AM Issue #26: wrong circular logic detect

The circular logic detector needs to consider index of wire. The attached file is a simple example to demo it. Although the design is not circular, verilator detect it as circular design. Thank you. cy wang

07/22/2008

06:29 PM Issue #23 (Closed): prefix problem

Thanks for following up. Wilson Snyder

03:17 PM Issue #25 (Closed): Public function output compile error "non-lvalue statement"

Thanks for the bug, I've just fixed this in the public git respository, and it will be in the next release. Wilson Snyder

09:09 AM Issue #25: Public function output compile error "non-lvalue statement"

<pre> The attached file is the original verilog code. verilator --cc my_tt.v cd obj_dir make -f Vmy_tt.mk the error from gcc non-lvalue statement. I trace the generated code. I guess the left express and right express should be exchange... cy wang

12:57 AM Issue #24 (Rejected): synopsys translate_off and synopsys translate_on

Absolutely not for religious reasons. translate_offs are horrors and I'm not going to encourage them; you should be using ifdefs instead. See for example my paper http://www.veripool.org/papers/TenIPEdits_SNUGBos07_paper.pdf Wilson Snyder

12:49 AM Issue #24: synopsys translate_off and synopsys translate_on

<pre> //synopsys translate_off non-synthesizable code //synopsys translate on I comment this region by hand. If verilator can automatically do it, It will reduce manual effort. </pre> cy wang

07/21/2008

03:12 PM Issue #23 (AskedReporter): prefix problem

What sort of problems? What's the error message? Do you see the files in the installed locations? Also try it without setting VERILATOR_ROOT, that shouldn't be needed if installing. Wilson Snyder

07/20/2008

11:43 AM Issue #23: prefix problem

./configure --prefix=/XX/local;make;make install the following files are installed. --> /XX/local/bin/verilator /XX/local/bin/verilator_bin /XX/local/share/verilator/include /XX/local/share/verilator/src setenv VERILATOR_ROOT /XX/local/sh... cy wang

07/14/2008

09:16 PM Issue #21 (Closed): $fopen() & friends don't work as expected...

There is still a fix here... I updated HEAD to warn if the $fopen etc descriptor isn't 64 bits so these errors can be caught. For MCD support please file another bug, thanks. Wilson Snyder

09:14 PM Issue #22 (Rejected): no STMTDLY warning for always #1 -- blocks

#1 isn't one simulation time step, it's one time unit, they're different things. You're making an assumption about the upper level c file that you shouldn't, and can't be known to Verilator. Besides, people that are silly enough to put #1 in a... Wilson Snyder

08:44 PM Issue #22: no STMTDLY warning for always #1 -- blocks

The STMTDLY warning is not necessairy for 'always # 1'-blocks: <pre> /* verilator lint_off STMTDLY */ always # 1 begin clk = ~clk; end </pre> in both cases the inner block should get evaluated every simulation time step, Verilat... Holger Wächtler

06:03 PM Issue #21 (AskedReporter): $fopen() & friends don't work as expected...

Works for me. Perhaps try `verilator_file_descriptor infile, outfile; instead of a integer. Wilson Snyder

05:41 PM Issue #20 (Closed): install man page in $prefix fails

Fixed in HEAD. Wilson Snyder

05:18 PM Issue #19 (Closed): $stime()

I'm generally of the force-people-to-update-code mindset, but given it's just a few lines. Fixed in the HEAD. Wilson Snyder

04:13 PM Issue #21: $fopen() & friends don't work as expected...

file test.vt: <pre> module testbench; integer infile, outfile, count, a; initial begin infile = $fopen("test.in", "r"); outfile = $fopen("test.out", "w"); $display("count == ... Holger Wächtler

03:54 PM Issue #20: install man page in $prefix fails

<pre> /bin/sh ./mkinstalldirs /Users/holger/verilog-stuff/3rdparty/share/man/man1 for p in verilator.1 ; do \ /usr/bin/install -c $p /Users/holger/verilog-stuff/3rdparty/share/man/man1/$p; \ done install: verilator.1: No such file or direct... Holger Wächtler

01:45 PM Issue #15 (Closed): $fscanf() and $sscanf()

Wilson Snyder

01:45 PM Issue #18 (Rejected): make $dumpfile(), $dumpvars() no-op when --trace is disabled

Er, $dumpfile etc isn't ever supported and I think making them error only when tracing is on will just confuse people; enabling waves to debug something shouldn't generate new errors. Unfortunately, using $dumpfile etc to control tracing won't ... Wilson Snyder

01:41 PM Issue #19 (AskedReporter): $stime()

I had assumed $stime is really a backward compatibility thing that code shouldn't be using any more. Can you give me an example where $stime is really needed and $time isn't more proper? Wilson Snyder

01:36 PM Issue #17 (Closed): git build fails due to missing config_rev.h

This was ok in the HEAD, but as you also noted in a separate mail, the git HEAD wasn't properly visible outside. I fixed the permissions on the repository, and a "git fetch" will now properly get this change along with the others. Wilson Snyder

07/05/2008

05:03 PM Issue #19: $stime()

report simulation time as 32bit integer Holger Wächtler

02:25 PM Issue #18: make $dumpfile(), $dumpvars() no-op when --trace is disabled

or, alternatively, use occurences of $dumpfile(), $dumpvars() to let verilator decide whether to generate the trace code or not? Right now the user has to touch two switches (the command line and the source) to enable tracing. Holger Wächtler

07/04/2008

01:12 PM Issue #17: git build fails due to missing config_rev.h

Hi, since config_rev.h is generated from the SVN versioning number by config_rev.pl, this approach fails when using the git tree: the .svn directories are missing. I'm not sure about the correct fix, should a special git tag or number be used? b... Holger Wächtler

07/01/2008

06:24 PM Issue #16 (Closed): list of file-i/o system tasks, now only partially implemented

$fscanf,$sscanf,$fgetc,$fgets,$fflush now work in the GIT repository. I've yet to see any code that cared about $ferror, $fread, $fseek, $ftell, $rewind, $ungetc so will wait until code really needs them. $fopenr,$fopenw,$fopena,$fputc are n... Wilson Snyder

06:21 PM Issue #15 (Resolved): $fscanf() and $sscanf()

$fscanf, $fgets and $sscanf are in the git repository. Please give these a try, and close the bug if it works ok for you. Wilson Snyder

06/26/2008

05:17 PM Issue #16: list of file-i/o system tasks, now only partially implemented

Hi, here the list of file-i/o system tasks, some of them are missing to fulfil the 2001 spec. Most of them map directly to libc function calls. $fopenr $fopenw $fopena $fflush $fputc $fgetc $fgets $ungetc $ftell $fseek $rewind Holger Wächtler

05:07 PM Issue #15: $fscanf() and $sscanf()

In order to implement testcases, it would be nice to have $fscanf(), $fgets() and $sscanf(). Icarus Verilog implements them in vpi/sys_scanf.c and vpi/sys_fileio.c, maybe some code can get shared if the licenses are compatible. best Regards, Holger Wächtler