[Issue 20143] ICE in optimizer on real 0/0 returned as double
https://issues.dlang.org/show_bug.cgi?id=20143 --- Comment #6 from Dlang Bot --- dlang/dmd pull request #14280 "merge stable" was merged into master: - 6d590f4c88d7ada0743e5bdb4223a7d044924322 by mhh: Fix Issue 20143 - Backend constant folding did not support division by real. This involved simply adding the correct case to the grand olde switch statement that does the evalu8tion https://github.com/dlang/dmd/pull/14280 --
[Issue 20143] ICE in optimizer on real 0/0 returned as double
https://issues.dlang.org/show_bug.cgi?id=20143 Dlang Bot changed: What|Removed |Added Status|NEW |RESOLVED Resolution|--- |FIXED --- Comment #5 from Dlang Bot --- dlang/dmd pull request #14192 "Fix Issue 20143 - Backend constant folding did not support division b…" was merged into stable: - 7e2ca1ba2f849ad4a1670ff351702a2806dfd89a by mhh: Fix Issue 20143 - Backend constant folding did not support division by real. This involved simply adding the correct case to the grand olde switch statement that does the evalu8tion https://github.com/dlang/dmd/pull/14192 --
[Issue 20143] ICE in optimizer on real 0/0 returned as double
https://issues.dlang.org/show_bug.cgi?id=20143 Dlang Bot changed: What|Removed |Added Keywords||pull --- Comment #4 from Dlang Bot --- @maxhaton updated dlang/dmd pull request #14192 "Fix Issue 20143 - Backend constant folding did not support division b…" fixing this issue: - Fix Issue 20143 - Backend constant folding did not support division by real. This involved simply adding the correct case to the grand olde switch statement that does the evalu8tion https://github.com/dlang/dmd/pull/14192 --
[Issue 20143] ICE in optimizer on real 0/0 returned as double
https://issues.dlang.org/show_bug.cgi?id=20143 mhh changed: What|Removed |Added CC||maxha...@gmail.com Severity|enhancement |normal --
[Issue 20143] ICE in optimizer on real 0/0 returned as double
https://issues.dlang.org/show_bug.cgi?id=20143 Iain Buclaw changed: What|Removed |Added Keywords||backend CC||ibuc...@gdcproject.org --
[Issue 20143] ICE in optimizer on real 0/0 returned as double
https://issues.dlang.org/show_bug.cgi?id=20143 FeepingCreature changed: What|Removed |Added CC||default_357-l...@yahoo.de --- Comment #3 from FeepingCreature --- This is one for Walter. The proximate cause is that in elem.d, the "0" on the left side of 0 / fun(0) is converted from a long double constant to a double constant sight unseen. (elem.d: shrinkLongDoubleConstantIfPossible) This leaves the operation post inlining standing as "double / long double", which is not a valid operation. --
[Issue 20143] ICE in optimizer on real 0/0 returned as double
https://issues.dlang.org/show_bug.cgi?id=20143 bachm...@yahoo.com changed: What|Removed |Added CC||bachm...@yahoo.com --- Comment #2 from bachm...@yahoo.com --- (In reply to Daniel Kozak from comment #1) > needs to be compiled with enabled inlineing (-inline) and it is not > important that return type of `bug` function is double it could be int, but > if we change fun tu return double it works as expected It actually has to be compiled with the combination -O -inline. Works fine if compiling with either -O or -inline but not both. --
[Issue 20143] ICE in optimizer on real 0/0 returned as double
https://issues.dlang.org/show_bug.cgi?id=20143 Daniel Kozak changed: What|Removed |Added CC||kozz...@gmail.com --- Comment #1 from Daniel Kozak --- needs to be compiled with enabled inlineing (-inline) and it is not important that return type of `bug` function is double it could be int, but if we change fun tu return double it works as expected --