https://bugs.kde.org/show_bug.cgi?id=471829
--- Comment #10 from Ninpo <ni...@qap.la> --- (In reply to Albert Astals Cid from comment #9) > > (In reply to Albert Astals Cid from comment #4) > > > Mirco can you please have a look to see if there's any value in these > > > claims? > > > > Might be a language barrier thing, but this comes across as rather > > condescending and hostile to people reporting a legitimate build failure and > > bug. Commenting as someone that's used KDE almost exclusively as a DE since > > 1999 I'm extremely disappointed. > > It is not a legitimate build failure, the reporter is adding -Werror on > their own and suffering because of it, noone forced them to have this > problem but themselves. Ummm...what? -Wall enables -Wstrict-aliasing=3 by default. From the gcc man page: > Level 3 (default for -Wstrict-aliasing): Should have very few false positives > and few false negatives. Slightly slower than levels 1 or 2 when > optimization is enabled. Takes care of the common pun+dereference pattern in > the front end: "*(int*)&some_float". If optimization is enabled, it also > runs in the back end, where it deals with multiple statement cases using > flow-sensitive points-to information. Only warns when the converted pointer > is dereferenced. Does not warn about incomplete types. These are typically things you want to catch and are legitimate bugs worth reporting. -Werror=strict-aliasing makes it easier to spot and report these rather than waiting for a random run time error where the user then has to get backtraces to report the same bug. Do you think turning the -Werror=strict-aliasing off makes the bug go away? I assure you it does not, here's my build log for the same package: > [with T = unsigned int; T min = 0; T max = 1; uchar = unsigned char; qint32 = > int]’: > ../kimageformats-5.107.0/src/imageformats/psd.cpp:1143:50: required from > here > ../kimageformats-5.107.0/src/imageformats/psd.cpp:828:23: warning: > dereferencing type-punned pointer will break strict-aliasing rules > [-Wstrict-aliasing] > 828 | auto ftmp = (*reinterpret_cast<float*>(&tmp) - double(min)) / > (double(max) - double(min)); It's a legitimate build failure (last I checked there's no DO NOT USE in the gcc man page for the flag) based off a legitimate strict-aliasing bug in the code. So it's been reported to upstream, as is the way with FOSS. You haven't answered why it's OK to violate strict aliasing in this code (making this a not legitimate bug). It's only an illegitimate build failure if the user turned -Werror=strict-aliasing on by accident which they clearly did not. -- You are receiving this mail because: You are watching all bug changes.
