Re: else ladders practice

On 26/11/2024 12:29, Tim Rentsch wrote:

Bart <bc@freeuk.com> writes:
 

On 25/11/2024 18:49, Tim Rentsch wrote:
>

Bart <bc@freeuk.com> writes:
>

It's funny how nobody seems to care about the speed of compilers
(which can vary by 100:1), but for the generated programs, the 2:1
speedup you might get by optimising it is vital!

>
I think most people would rather take this path (these times
are actual measured times of a recently written program):
>
      compile time:       1 second
      program run time:   ~7 hours
>
than this path (extrapolated using the ratios mentioned above):
>
      compile time:       0.01 second
      program run time:   ~14 hours

>
I'm trying to think of some computationally intensive app that would
run non-stop for several hours without interaction.

 The conclusion is the same whether the program run time
is 7 hours, 7 minutes, or 7 seconds.

Funny you should mention 7 seconds. If I'm working on single source file called sql.c for example, that's how long it takes for gcc to create an unoptimised executable:
   c:\cx>tm gcc sql.c            #250Kloc file
   TM: 7.38
Testing it might only take a second:
   c:\cx>type input
   select 2+2;
   c:\cx>sql <input
   4
With a different compiler then the edit-run cycle can be a lot nippier:
   c:\cx>tm cc sql
   Compiling sql.c to sql.exe
   TM: 0.27
If that is still onerous, I can try interpreting:
   c:\cx>tm cc -i sql <input
   Compiling sql.c to sql.(int)
   4
   TM: 0.19
So compiling to IL, then interpreting that IL (which is 40 times slower than native code), /and/ running my test, takes 1/5th of a second in all.
That's 40 times faster than the equivalent with gcc-O0 (despite the interpreted part being 40 times slower!):
   c:\cx\tm test.bat
   c:\cx>gcc sql.c -osql.exe   && sql  0<input
   4
   TM: 7.74
And 200 times faster than gcc-O2 which everyone here seems to be recommending:
   c:\cx>tm test.bat
   c:\cx>gcc sql.c -O2 -osql.exe   && sql  0<input
   4
   TM: 38.60
Some might advise not working with such a large single source module at all, but that is the task here. If trying to investigate why my cc product is failing, I might put in tracing statements into the source, and compile and run with both compilers to compare the outputs. For such a purpose, -O2 or -O3 is utterly pointless.