@thephd Isn’t that the domain of specialised tools like Intel Inspector?

@IT_Fettchen kann ich in der entsprechenden Handreichung nicht finden: https://www.protokoll-inland.de/SharedDocs/downloads/Webs/PI/DE/Allgemeines/Anschriften.pdf?__blob=publicationFile&v=1

@dendrobatus_azureus Also ‘make check’ before ‘make install’ might tell you about how wise installing actually is.

@zwol @tedmielczarek @b0rk @leftpaddotpy Variables that were only (temporarily) set in the environment on the other hand are not easily available when attempting to reconstruct what happened in a previous build.

@zwol @tedmielczarek @b0rk @leftpaddotpy arguments get passed again during recheck (`./config.status --recheck`). That's not particularly relevant when just building an autoconf-using package, but quite valuable to persist build settings while developing the software in question.

@resuna Well, it probably is a legit proof of artificial+natural stupidity.

@levitte @bagder Well, at least C effectively mandates long long and therefore 64bit integers in practice since C99. So someone is trying to build on a platform not updated since the nineties.

@profoundlynerdy Perhaps start here: https://begriffs.com/posts/2018-11-15-c-portability.html

@mjd I always found mmv to be useful and widely available for this task.

@maub But why do we spend so much time in phase 3?

@jrose @jann @thephd I Wonder if there’s some lesson in #Perl and #fortran90 solving this issue decades ago, but both are languages everyone seems to rather make fun of than learn from.

@klyx @leyrer Wusste gar nicht, dass da auch Nekromanten eine Rolle spielen.

@thephd Shoud perhaps note that POSIX only demands this since 2018, previously only systems with the XSI option made that guarantee.

@vulpine Which continent?

@nonproductive @Nonilex Seems like confusing a necessary prerequisite and the thing itself.

@rmartinnielsen @timbray @b0rk @Xavier Having . in PATH is a rooky mistake many have come to regret on multi-user machines.

@bishop It's almost as if private replies to usenet posts used smtp instead of nntp for a valid reason.

@freiefunken @bureaux Tatsächlich ist die aktuelle Situation nicht ganz so rosig: https://doku.tid.dfn.de/de:dfnpki:tcsfaq:aktuellesituation und die Nachfolgelösung mit HARICA ist noch nicht ganz "fertig".

@truls46 @hhyan If you think sequential, unit-stride memory access does not matter for performance, you've been asleep since hardware prefetching and cache were introduced in the mid-nineties.

@vitaut Isn't nvc the nvhpc C compiler? And nvcc compiles this weird incomplete C++ dialect that's CUDA? Did you perhaps intend to invoke nvc++ (or whatever it's called on Windows)?