Loading ...
Sorry, an error occurred while loading the content.
 

Re: [jallist] Re: latest jalv2.4q

Expand Messages
  • Bert van Dam
    Hi Rob, Kyle, ... I agree. As far as printed books are concerned this construct is never used, explained or promoted so removing it doesn t hurt backwards
    Message 1 of 33 , Jun 1, 2013
      Hi Rob, Kyle,

      >> Finally, you can tell the compiler per-variable:
      >>
      >> var byte foo shared at 0x88
      >>
      >> regardless of any `pragma data' or `pragma shared' lines, the compiler will assume `foo' is shared.
      >>
      >
      > This possibility of 'shared' is a useless, even undesirable feature! The
      > compiler won't care to set the banking bits, thus when address 0x88 is
      > not in a shared memory region at one time it may access the right byte,
      > another time it will access the byte at address 0x08 or 0x108 or 0x188.
      > Very difficult to debug! So as far as I'm concerned the 'shared'
      > option with 'var' should be dropped.

      I agree. As far as printed books are concerned this construct is never
      used, explained or promoted so removing it doesn't hurt backwards
      compatibility.

      > Furthermore I have pre-allocated some variables (in particular bytes for
      > port shadowing) at locations in a shared memory region (as far as
      > available). This requires some memory management which is otherwise done
      > by the compiler. I would call this 'system programming' in contrast to
      > 'application programming'.

      I will not be following that for the Starterpack, I will leave that up
      to the compiler.

      Have fun,
      Bert


      www.vandam.tk
    • Matthew Schinkel
      I have no idea what the problem was, and I had no way of troubleshooting more then that since I don t know ASM. Maybe that s a question for Kyle although it
      Message 33 of 33 , Jun 10, 2013
        I have no idea what the problem was, and I had no way of troubleshooting more then that since I don't know ASM. Maybe that's a question for Kyle although it doesn't matter much now that Kyle fixed it.

        Matt.

        --- In jallist@yahoogroups.com, "Gordon Williams" <gwilliams@...> wrote:
        >
        > Hi Matt,
        >
        > What would be the difference between the pseudo-variable in the stopwatch
        > library and the function below. I'm wondering what is the actual problem.
        > Could you provide a better description?
        >
        > thanks,
        >
        > Gordon Williams
        >
        >
        >
        >
        >
        > ----- Original Message -----
        > From: "Matthew Schinkel" <mattschinkel@...>
        > To: <jallist@yahoogroups.com>
        > Sent: Friday, June 07, 2013 8:30 PM
        > Subject: [jallist] Re: latest jalv2.4q
        >
        >
        > > Seems to be just the stopwatch lib/sample.
        > >
        > > Matt.
        > >
        > > --- In jallist@yahoogroups.com, "Gordon Williams" <gwilliams@> wrote:
        > > >
        > > > Hi Matthew,
        > > >
        > > > Does this apply to all 'get functions that no return value is provided
        > if
        > > > used pragma inline? Is this a stopwatch problem or a more general
        > problem?
        > > >
        > > > Would (from the device include file)
        > > >
        > > > function port_A_low'get return byte is
        > > > pragma inline
        > > > return _PORTA & 0x0F
        > > > end function
        > > >
        > > > be affected in the same way?
        > > >
        > > > thanks,
        > > >
        > > > Gordon Williams
        > > >
        > > >
        > > >
        > > > ----- Original Message -----
        > > > From: "Matthew Schinkel" <mattschinkel@>
        > > > To: <jallist@yahoogroups.com>
        > > > Sent: Thursday, June 06, 2013 8:13 PM
        > > > Subject: [jallist] Re: latest jalv2.4q
        > > >
        > > >
        > > > > Hi Kyle, I submitted a bugreport for stopwatch.
        > > > >
        > > > >
        > > >
        > http://jallib.googlecode.com/svn/trunk/casualties/bugreport_20130606/18f67j50_stopwatch.zip
        > > > >
        > > > > no return value for stopwatch'get() & stopwatch_cycles'get() in
        > > > stopwatch.jal if pragma inline is used.
        > > > >
        > > > > Thanks,
        > > > >
        > > > > --- In jallist@yahoogroups.com, "Matthew Schinkel" <mattschinkel@>
        > > > wrote:
        > > > > >
        > > > > > Kyle,
        > > > > >
        > > > > > I found a bug relating to pragma inline and pseudo variables. I'll
        > make
        > > > a package and put it up on SVN as soon as I get a chance. More info
        > here:
        > > > > >
        > > > > >
        > https://groups.google.com/forum/?fromgroups=#!topic/jallib/NaOJ1CDYDNE
        > > > > >
        > > > > > With pragma inline, I am only getting a return value of 0.
        > > > > >
        > > > > > Thanks,
        > > > > > Matt.
        > > > > >
        > > > > > --- In jallist@yahoogroups.com, "zz" <kyle@> wrote:
        > > > > > >
        > > > > > > Greets,
        > > > > > >
        > > > > > > I just put up the latest jalv2.4q-beta. This should be very close
        > to
        > > > release so please try it out & let me know.
        > > > > > >
        > > > > > > http://www.casadeyork.com/jalv2/archive/20130522-jalv24q-beta.zip
        > > > > > >
        > http://www.casadeyork.com/jalv2/archive/20130522-jalv24q-betasrc.zip
        > > > > > >
        > > > > > > I also removed links to jalv2.4p from the webpage as that version
        > > > really shouldn't be used.
        > > > > > >
        > > > > > > If you find issues, let me know, otherwise I'll make this release
        > > > official in the next week or so.
        > > > > > >
        > > > > > > --kyle
        > > > > > >
        > > > > >
        > > > >
        > > > >
        > > > >
        > > > >
        > > > > ------------------------------------
        > > > >
        > > > > Yahoo! Groups Links
        > > > >
        > > > >
        > > > >
        > > >
        > >
        > >
        > >
        > >
        > > ------------------------------------
        > >
        > > Yahoo! Groups Links
        > >
        > >
        > >
        >
      Your message has been successfully submitted and would be delivered to recipients shortly.