Path: csiph.com!fu-berlin.de!uni-berlin.de!individual.net!not-for-mail From: Ian Collins Newsgroups: comp.lang.c++ Subject: Re: Examples of current platforms/architectures where sizeof(void*) > sizeof(int*) ? Date: Sun, 5 Sep 2021 11:24:06 +1200 Lines: 23 Message-ID: References: <87a6kuhchx.fsf@nosuchdomain.example.com> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8; format=flowed Content-Transfer-Encoding: 7bit X-Trace: individual.net 7NN8ycp62HhxkKfGQd0ciw5kvGcZeiZ/hZiUfEHjPALRBiIIUl Cancel-Lock: sha1:LX6aCXQpJn0uWXtSkZu8yRaMWoA= User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:78.0) Gecko/20100101 Thunderbird/78.11.0 In-Reply-To: Content-Language: en-US Xref: csiph.com comp.lang.c++:81049 On 05/09/2021 04:44, Bart wrote: > > tcc builds it 100 times faster than gcc -O2, and on a single core (of > the two on my AMD cpu). And about 30 times faster than even gcc -O0. > > However the gcc-code runs the interpreter twice as fast as tcc-code. So > the trade-off is clear: use tcc for most development for a rapid > edit-run cycle, and gcc for occasional better analysis, or for > production binaries. > > The latter can be done overnight so no one cares how long it takes; the > product is debugged and working at this point. This approach isn't uncommon, in our case we build with both clang++ and g++ with a comprehensive set of warnings on each build and a daily run with clang-tidy which takes up to ten times longer due to the build being confined to one machine. The problem with this, as is often the case with out of line builds, is getting people to check the warnings! -- Ian.