There are 7 messages in this thread.
You are currently looking at messages 0 to 7.
I have been using Active-HDL (AHDL) for over a year now and I am pretty happy with it. For one thing, it doesn't have the apparently unfix-able, tool crashing, memory leak that every version of Modelsim I have ever used seems to have. I find that as I go through the edit-compile-simulate cycle I use the toolbar buttons to 1) Restart the simulation, 2) recompile all modules 3) use a right click menu to clear the waveform display. But this mostly does not actually recompile the code or does not recompile it fully or does not use the recompiled code. After a source change I have to run around this cycle twice to get the new code to actually simulate. What am I missing? I find that if I *don't* recompile and try to run again, I get a dialog telling me the source is out of date and I can recompile if I want. If I say yes, it recompiles and *always* seems to do it correctly. I would say that was a clue, but I am not being Hercule Poirot today it seems. Rick
On Jul 2, 3:17=A0pm, rickman <gnu...@gmail.com> wrote: > I have been using Active-HDL (AHDL) for over a year now and I am > pretty happy with it. =A0For one thing, it doesn't have the apparently > unfix-able, tool crashing, memory leak that every version of Modelsim > I have ever used seems to have. > > I find that as I go through the edit-compile-simulate cycle I use the > toolbar buttons to 1) Restart the simulation, 2) recompile all modules > 3) use a right click menu to clear the waveform display. =A0But this > mostly does not actually recompile the code or does not recompile it > fully or does not use the recompiled code. =A0After a source change I > have to run around this cycle twice to get the new code to actually > simulate. > > What am I missing? =A0I find that if I *don't* recompile and try to run > again, I get a dialog telling me the source is out of date and I can > recompile if I want. =A0If I say yes, it recompiles and *always* seems > to do it correctly. > > I would say that was a clue, but I am not being Hercule Poirot today > it seems. > > Rick Don't know about AHDL, but in ModelSim you definitely need to do steps 1 and 2 in the reverse error to use the newly recompiled code. The last time I touched Aldec stuff was when Xilinx Foundation (4.1 and prior) had it bundled together. Regards, Gabor______________________________
On Jul 2, 3:49=A0pm, gabor <ga...@alacron.com> wrote: > On Jul 2, 3:17=A0pm, rickman <gnu...@gmail.com> wrote: > > > > > I have been using Active-HDL (AHDL) for over a year now and I am > > pretty happy with it. =A0For one thing, it doesn't have the apparently > > unfix-able, tool crashing, memory leak that every version of Modelsim > > I have ever used seems to have. > > > I find that as I go through the edit-compile-simulate cycle I use the > > toolbar buttons to 1) Restart the simulation, 2) recompile all modules > > 3) use a right click menu to clear the waveform display. =A0But this > > mostly does not actually recompile the code or does not recompile it > > fully or does not use the recompiled code. =A0After a source change I > > have to run around this cycle twice to get the new code to actually > > simulate. > > > What am I missing? =A0I find that if I *don't* recompile and try to run > > again, I get a dialog telling me the source is out of date and I can > > recompile if I want. =A0If I say yes, it recompiles and *always* seems > > to do it correctly. > > > I would say that was a clue, but I am not being Hercule Poirot today > > it seems. > > > Rick > > Don't know about AHDL, but in ModelSim you definitely need to > do steps 1 and 2 in the reverse error to use the newly > recompiled code. =A0The last time I touched Aldec stuff was > when Xilinx Foundation (4.1 and prior) had it bundled together. > > Regards, > Gabor Duh! Now that you say that, I see how obvious that is. I guess pushing buttons makes it so easy to forget what is going on. I was thinking that restart the simulation was ending it and the compile would start the new one, but obviously "restart" means restart! Would have thought that would happen??? That is what happens when you don't use a tool for a few months. I do still remember which end of the hammer to hold... Thanks, Rick
Rick, > 2) recompile all modules Sometimes "recompile all modules" may not work first time if the order of compilation has not been set properly. /Mikhail
MM wrote: > Sometimes "recompile all modules" may not work first time if the order of > compilation has not been set properly. With vhdl-mode, right-click, Speedbar, Generate Makefile right-click, Speedbar, Make does the trick. -- Mike Treseler
On Fri, 03 Jul 2009 09:28:49 -0700, Mike Treseler <m...@gmail.com> wrote: >MM wrote: > >> Sometimes "recompile all modules" may not work first time if the order of >> compilation has not been set properly. > >With vhdl-mode, > >right-click, Speedbar, Generate Makefile >right-click, Speedbar, Make > >does the trick. because VHDL sources contain all the information necessary to determine the (or rather, a ) correct order of compilation, and vhdl-mode simply translates it. There is really no excuse for other tools being unable to do the same. (Mixed language projects may be another matter) - Brian______________________________
On Jul 4, 7:19=A0am, Brian Drummond <brian_drumm...@btconnect.com> wrote: > On Fri, 03 Jul 2009 09:28:49 -0700, Mike Treseler <mtrese...@gmail.com> w= rote: > >MM wrote: > > >> Sometimes "recompile all modules" may not work first time if the order= of > >> compilation has not been set properly. > > >With vhdl-mode, > > >right-click, Speedbar, Generate Makefile > >right-click, Speedbar, Make > > >does the trick. > > because VHDL sources contain all the information necessary to determine t= he (or > rather, a ) correct order of compilation, and vhdl-mode simply translates= it. > > There is really no excuse for other tools being unable to do the same. > > (Mixed language projects may be another matter) > > - Brian The Lattice software has a mode of compiling that evaluates the order to compile the modules and remembers it for subsequent builds. That was not an issue with my build. I was just restarting my simulation before I recompiled. Rick______________________________