Re: next nightly build problem

[Bruno H. <br...@cl...>]

Hi Don,

> I think I can guess which ones work from this:
> -rw-rw-r--. 1 don don    5372 Mar  3 20:41 build-porting64-gcc-generational_gc-multithread_gc/cbcstep3.log
> -rw-rw-r--. 1 don don    5116 Mar  3 20:40 build-porting64-gcc-generational_gc-old_gc/cbcstep3.log
> -rw-rw-r--. 1 don don 3042294 Mar  3 20:22 build-porting64-gcc-generic64_heapcodes/cbcstep3.log
> -rw-rw-r--. 1 don don 3037594 Mar  3 20:17 build-porting64-gcc-heapcodes/cbcstep3.log
> -rw-rw-r--. 1 don don 3046443 Mar  3 20:32 build-porting64-gcc-heapcodes-spvw_mixed_blocks/cbcstep3.log
> -rw-rw-r--. 1 don don 3043261 Mar  3 20:26 build-porting64-gcc-heapcodes-spvw_pages/cbcstep3.log
> -rw-rw-r--. 1 don don    4860 Mar  3 20:39 build-porting64-gcc-multithread_gc/cbcstep3.log
> -rw-rw-r--. 1 don don    4604 Mar  3 20:38 build-porting64-gcc-old_gc/cbcstep3.log
> -rw-rw-r--. 1 don don    4636 Mar  3 20:42 build-porting64-gcc-safety0/cbcstep3.log
> -rw-rw-r--. 1 don don    4983 Mar  3 20:43 build-porting64-gcc-safety0-optimized/cbcstep3.log
> -rw-rw-r--. 1 don don    4636 Mar  3 20:14 build-porting64-gcc-safety3/cbcstep3.log
> -rw-rw-r--. 1 don don 3040627 Mar  3 20:37 build-porting64-gcc-spvw_pure_blocks/cbcstep3.log
> -rw-rw-r--. 1 don don 3041797 Mar  3 20:20 build-porting64-gcc-standard_heapcodes/cbcstep3.log
> -rw-rw-r--. 1 don don 3046009 Mar  3 20:34 build-porting64-gcc-typecodes-spvw_mixed_blocks/cbcstep3.log
> -rw-rw-r--. 1 don don 3042770 Mar  3 20:29 build-porting64-gcc-typecodes-spvw_pages/cbcstep3.log

Yup. Same thing here; the exact file sizes may vary a bit.

> Next steps?

Depends on what you want to do.

- The set of builds that failed gives you an understanding which features
  caused regressions. In this case it's evident: forcing HEAPCODES worked,
  forcing TYPECODES worked, but the default choice did not work.

- If you are interested in continuous integration, you may want to set up
  build scripts that use 'multibuild-porting64-gcc' and signal regressions
  in any of the build configurations. This would be useful for Sam and me.

- If you are interested in a build that works, choose one of those with a
  large cbcstep3.log.

- If you want to know about the performance: Since "make check" runs the
  benchmarks, in particular, the log file contains the benchmark results.
  You can compare them:
$ grep '^total .* scaled$' */cbcstep3.log | tr : '\t' | expand -t 65
build-porting32-gcc-heapcodes/cbcstep3.log                       total              9.93600 sec    12.48954 scaled
build-porting64-gcc-heapcodes/cbcstep3.log                       total             10.46000 sec    13.25959 scaled
build-porting64-gcc-heapcodes-spvw_mixed_blocks/cbcstep3.log     total             10.34000 sec    13.07827 scaled
build-porting64-gcc-heapcodes-spvw_pages/cbcstep3.log            total             12.67200 sec    16.27370 scaled
build-porting64-gcc-standard_heapcodes/cbcstep3.log              total             10.12000 sec    12.93405 scaled
build-porting64-gcc-typecodes-spvw_mixed_blocks/cbcstep3.log     total             10.59200 sec    13.69968 scaled
build-porting64-gcc-typecodes-spvw_pages/cbcstep3.log            total             13.44000 sec    17.30458 scaled

  What this tells me, is that
    - A 32-bit build is faster than a 64-bit build. (Not surprising:
      Larger pointers means more use of memory, and a factor S more
      memory means a factor S² more runtime - usual rule of thumb.)
    - SPVW_PAGES is significantly slower than SPVW_BLOCKS.
    - TYPECODES is a bit slower than HEAPCODES. (I find this surprising.)

> I guess you'd like them all to work, right?

Yes, this is the goal.

Bruno