Re: Query related to escaped character in verilog

From: Clifford E. Cummings (cliffc@sunburst-design.com)
Date: Wed Feb 09 2005 - 18:20:51 PST

  • Next message: Steven Sharp: "Re: Query related to escaped character in verilog"

    Thanks to Steve and Neil for looking at this.

    Comments below.

    At 04:22 PM 2/9/2005, Steven Sharp wrote:

    > >Attached is a uuencoded tar-file with escape-test-code and the outputs from
    > >VCS 7.2 and ModelSim 6.0
    > >
    > >Hopefully the uuencoding will preserve all the interesting characters.
    >
    >My uudecode won't decode something containing a NUL character, so I
    >can't read your file.

    An if I embed the characters, they will probably disappear. Can you use
    winzip to read the files to a PC?

    > >The octal-13 character was also different (line feed or ^M), as were a few
    > >other of the obscure-variety characters.
    >
    >Were you on a PC? This might be a classic carriage-return/linefeed
    >versus linefeed issue.

    I was on a PC when I ran the example.

    > >I don't know if I agree that we should match Verilog-XL if we think it is
    > >doing something wrong. Seems like we cleaned up a couple of Verilog-XL-isms
    > >for IEEE Verilog-1995 and if we think printing a space when a
    > >null-character was requested is wrong, I am inclined to say so.
    >
    >And do you have any real argument that this is wrong?

    Not really. I just thought that printing a null character was the same as
    printing a non-existent character, but I don't do a lot of work with null
    characters so my opinion is probably not important :-)

    >Note that NUL characters tend to mess up any application written in C.
    >And if you tried to specify that NUL characters be printed out,
    >simulators with I/O libraries written in C might start truncating
    >outputs in a variety of unexpected ways. Have you ever tried printing
    >out NUL characters using %c in the middle of a complex format? NC can
    >do it, but I don't think most other simulators can.

    We must be some-type-of geeks to spend this much time running simulations
    and C-programs to enhance our knowledge of NUL characters. :-) :-)

    I think I won't share these experiments and results with my wife and
    daughters! :-)

    > > I have not
    > >tried this with C yet. Any takers?
    >
    >You can't print out an embedded NUL with %s in C, because a NUL will
    >terminate the string.
    >
    >
    > >I must admit that I would never base a Verilog-simulator-buying decision on
    > >this, but we should probably clean it up.
    >
    >The current specification in the LRM is lacking, and should be clarified.
    >This assumes that we can agree on what to specify. What XL does makes
    >reasonable sense, and is the de facto standard. I don't see a reason to
    >specify anything else.

    If everyone or near-everyone agrees, I see no reason to oppose.

    >Steven Sharp
    >sharp@cadence.com

    Regards - Cliff

    ----------------------------------------------------
    Cliff Cummings - Sunburst Design, Inc.
    14314 SW Allen Blvd., PMB 501, Beaverton, OR 97005
    Phone: 503-641-8446 / FAX: 503-641-8486
    cliffc@sunburst-design.com / www.sunburst-design.com
    Expert Verilog, SystemVerilog, Synthesis and Verification Training

    This archive was generated by hypermail 2.1.4 : Wed Feb 09 2005 - 18:05:59 PST and
    sponsored by Boyd Technology, Inc.