> | For reasons that are not clear, the new expander triggered
> | the problem through an existing test.

I think the problem was made evident by the previous commit "fix
optimizer's single-use tracking, especially on inlining" (
https://github.com/plt/racket/commit/bfc2b27d653de50ce0e111962f59e1a3be521956
), not by the new expander.

The travis tests are failing since that commit (
https://travis-ci.org/plt/racket/builds/71256073 ,
https://travis-ci.org/gus-massa/racket/builds/71319704 ). Specially
with the clang compiler (???). And in each run it creates a bogus
error in a different unrelated test. (I retried a few times.) This
looks like a typical bad coordinate shift for me. (I sadly have a lot
of experience with that.)

Gustavo





On Thu, Jul 16, 2015 at 9:19 PM,  <[email protected]> wrote:
> mflatt has updated `master' from b72dceb865 to aaee824f68.
>   http://git.racket-lang.org/plt/b72dceb865..aaee824f68
>
> =====[ 2 Commits ]======================================================
> Directory summary:
>   61.0% pkgs/racket-test-core/tests/racket/
>   38.9% racket/src/racket/src/
>
> ~~~~~~~~~~
>
> e65beb2 Matthew Flatt <[email protected]> 2015-07-16 17:06
> :
> | adjust a core test to avoid `mscheme`
> :
>   M pkgs/racket-test-core/tests/racket/stx.rktl | 4 ++--
>
> ~~~~~~~~~~
>
> aaee824 Matthew Flatt <[email protected]> 2015-07-16 18:12
> :
> | optimizer: fix coordinate shift when a `letrec` is split
> |
> | The combination of splitting a `letrec` and optimizing
> | the resulting `(let ([x <proc>]) x)` to just `<proc>`
> | used a bad coordinate shift, which made property testing
> | incorrect, etc.
> |
> | For reasons that are not clear, the new expander triggered
> | the problem through an existing test.
> :
>   M racket/src/racket/src/optimize.c | 6 +++---
>
> =====[ Overall Diff ]===================================================
> [...]

-- 
You received this message because you are subscribed to the Google Groups 
"Racket Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To post to this group, send email to [email protected].
To view this discussion on the web visit 
https://groups.google.com/d/msgid/racket-dev/CAPaha9PHmOgs3g0Oxb2U6ZPM5NtD9d0jwzQrh2aq46-daQSEug%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.

Reply via email to