warning: dereferencing pointer does break strict-aliasing rules
in [C++]
|
From: tf on 2 Apr 2010 10:25 Seungbeom Kim <musiphil(a)bawi.org> wrote: > On 2010-04-01 14:56, Paul Bibbings wrote: >> Seungbeom Kim<musiphil(a)bawi.org> writes: >> >>> I get this warning while compiling the program below with g++-4.4.3: >>> >>> $ g++ -O2 -Wall test.cc >>> test.cc: In function ???int main()???: >>> test:8: warning: dereferencing pointer ???<anonymous>??? does break >>> strict-aliasing rules >>> /usr/include/c++/4.4/bits/stl_list.h:214: note: initialized from here > > Turning off that optimization could be an answer, but I???d like to > know whether the warning is just a false positive or it really means > I risk some undefined behaviour and I should do something (turn the > optimization off or change the code somehow). "does" means you *are* violating the rules. g++ says "may" (IIRC, maybe "can") when it can't figure out for sure. I think there are different levels of strict aliasing checking. See the manpage. FWIW, I just disable strict aliasing on all C++ code I write. AFAICT, strict aliasing was for C, g++ is just (incorrectly?) applying it to C++. Further, these aliasing concerns are exactly why we have the valarray header, and so it seems dumb in C++: if it was critical for performance that my data didn't alias, I would use valarray. If I'm not, then why would my compiler assume my variables didn't alias? I've never brought this up on gcc-help; it's possible I'm missing something. -tom -- [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ]
From: Mathias Gaunard on 3 Apr 2010 05:34 On 1 avr, 15:08, Seungbeom Kim <musip...(a)bawi.org> wrote: > I get this warning while compiling the program below with g++-4.4.3: > > $ g++ -O2 -Wall test.cc > test.cc: In function �int main()�: > test:8: warning: dereferencing pointer �<anonymous>� does break strict-aliasing > rules > /usr/include/c++/4.4/bits/stl_list.h:214: note: initialized from here > > I don�t see anything suspicious in the following program, but there > should be a reason for the warning. Can anyone explain this? There is no cast, therefore there should be no problem. Report this issue to the GCC guys. -- [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ]
From: Seungbeom Kim on 3 Apr 2010 11:15 On 2010-04-03 13:34, Mathias Gaunard wrote: > On 1 avr, 15:08, Seungbeom Kim<musip...(a)bawi.org> wrote: >> I get this warning while compiling the program below with g++-4.4.3: >> >> $ g++ -O2 -Wall test.cc >> test.cc: In function �int main()�: >> test:8: warning: dereferencing pointer �<anonymous>� does break strict-aliasing >> rules >> /usr/include/c++/4.4/bits/stl_list.h:214: note: initialized from here >> >> I don�t see anything suspicious in the following program, but there >> should be a reason for the warning. Can anyone explain this? > > There is no cast, therefore there should be no problem. Actually there is one at stl_list.h:214, though that is just a downcast and I doubt it should actually be a problem. -- Seungbeom Kim [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ]
From: Joshua Maurice on 5 Apr 2010 09:17
On Apr 2, 6:25 pm, t...(a)shigeru.sci.utah.edu wrote: > Seungbeom Kim <musip...(a)bawi.org> wrote: > > On 2010-04-01 14:56, Paul Bibbings wrote: > >> Seungbeom Kim<musip...(a)bawi.org> writes: > > >>> I get this warning while compiling the program below with g++-4.4.3: > > >>> $ g++ -O2 -Wall test.cc > >>> test.cc: In function ???int main()???: > >>> test:8: warning: dereferencing pointer ???<anonymous>??? does break > >>> strict-aliasing rules > >>> /usr/include/c++/4.4/bits/stl_list.h:214: note: initialized from here > > > Turning off that optimization could be an answer, but I???d like to > > know whether the warning is just a false positive or it really means > > I risk some undefined behaviour and I should do something (turn the > > optimization off or change the code somehow). > > "does" means you *are* violating the rules. g++ says "may" (IIRC, maybe > "can") when it can't figure out for sure. > > I think there are different levels of strict aliasing checking. See > the manpage. > > FWIW, I just disable strict aliasing on all C++ code I write. AFAICT, > strict aliasing was for C, g++ is just (incorrectly?) applying it to > C++. INCITS+ISO+IEC+14882-2003 C++03 standard. 3.10 Lvalues and rvalues / 15 If a program attempts to access the stored value of an object through an lvalue of other than one of the following types the behavior is undefined48): � the dynamic type of the object, � a cv-qualified version of the dynamic type of the object, � a type that is the signed or unsigned type corresponding to the dynamic type of the object, � a type that is the signed or unsigned type corresponding to a cv- qualified version of the dynamic type of the object, � an aggregate or union type that includes one of the aforementioned types among its members (including, recursively, a member of a subaggregate or contained union), � a type that is a (possibly cv-qualified) base class type of the dynamic type of the object, � a char or unsigned char type. __________________ 48) The intent of this list is to specify those circumstances in which an object may or may not be aliased. -- [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] |