On Thu, May 24, 2018 at 12:27 PM Steffen M��rcker <merkste@web.de> wrote:

Hi,

now I've observed the same issue. It might be related to context��
switching, since introducing a delay has a similar effect. Consider:

�� | FE_TONEAREST FE_DOWNWARD FE_UPWARD FE_TOWARDZERO |
�� FE_TONEAREST�� := 16r0000.
�� FE_DOWNWARD�� := 16r0400.
�� FE_UPWARD�� := 16r0800.
�� FE_TOWARDZERO := 16r0C00.
�� "For some reasons we have to call fegetround once."
�� "c := LibC new fegetround."
�� LibC new fesetround: FE_DOWNWARD.
�� (Delay forSeconds: 1) wait.
�� v1 := 1.0/10.0.
�� LibC new fesetround: FE_UPWARD.
�� v2 := 1.0/10.0.
�� LibC new fesetround: FE_TONEAREST.
�� v1 < v2.

If the delay is inserted, the script evaluates to false. Using the same��
LibC-instance or creating a new one does not seem to change anything here.��
Interestingly, a similar approach in VisualWorks does not show this issue��
yet.

���Ok, so maybe we need to use put evaluation in a block and use valueNoContextSwitch ?

���Maybe use an API like the one you propose before : 

Double roundToMinusInfWhile: [ ... ] 

��

Actually, I expect the FE_* macros to be platform/implementation��
dependent, as suggested here:
http://www.enseignement.polytechnique.fr/informatique/INF478/docs/Cpp/en/c/numeric/fenv/FE_round.html

���Ok.

Can we try to pack everything in a PR for Pharo 7.0 ?

���Thank you.

Serge Stinckwich
UMI UMMISCO 209 (SU/IRD/UY1)

"Programs must be written for people to read, and only incidentally for machines to execute."

http://www.doesnotunderstand.org/