Performance Factors in C++
in [C++]
|
From: Mathias Gaunard on 31 Aug 2009 11:10 On 31 ao�t, 02:55, Thomas Maeder <mae...(a)glue.ch> wrote: > Provided the iterators are pointers (or close relatives to pointers), > which is the prerequisite for considiering the usage of memmove() or > memcpy(), std::copy can certainly determine whether the areas overlap. No, it can't. The compiler could do it, by doing alias analysis, which is a hard (NP- hard even) problem -- which is why you can't rely on the compiler doing it -- but a function certainly cannot. -- [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ]
From: cpp4ever on 31 Aug 2009 21:10 Sada wrote: > hi how can I improve the performance of a C++ program? > > In my program I am using strcmp() strcpy() functions heavily , are > they costly? if so how can i avoid them?? > Is memcpy() is better than strcpy()? > > Please suggest me the factors that affect the performance of a C++ > program. > Avoid allocating and deallocating memory, or passing large objects by value. Both of these will impair performance. JB -- [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ]
From: James Kuyper on 1 Sep 2009 22:02 Mathias Gaunard wrote: > On 31 août, 02:55, Thomas Maeder <mae...(a)glue.ch> wrote: > >> Provided the iterators are pointers (or close relatives to pointers), >> which is the prerequisite for considiering the usage of memmove() or >> memcpy(), std::copy can certainly determine whether the areas overlap. > > No, it can't. > The compiler could do it, by doing alias analysis, which is a hard (NP- > hard even) problem -- which is why you can't rely on the compiler > doing it -- but a function certainly cannot. Could you explain that in more detail? It seems to me that std::copy<T*>(p,q,n) could certainly make use of std::less<T*> to check whether [p,p+n) overlaps [q,q+n). What's the barrier to performing such a comparison? -- [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ]
From: slobasso on 2 Sep 2009 23:26 On Sep 2, 6:02 am, James Kuyper <jameskuy...(a)verizon.net> wrote: > Mathias Gaunard wrote: > > On 31 ao�t, 02:55, Thomas Maeder <mae...(a)glue.ch> wrote: > > >> Provided the iterators are pointers (or close relatives to pointers), > >> which is the prerequisite for considiering the usage of memmove() or > >> memcpy(), std::copy can certainly determine whether the areas overlap. > > > No, it can't. > > The compiler could do it, by doing alias analysis, which is a hard (NP- > > hard even) problem -- which is why you can't rely on the compiler > > doing it -- but a function certainly cannot. > > Could you explain that in more detail? It seems to me that > std::copy<T*>(p,q,n) could certainly make use of std::less<T*> to check > whether [p,p+n) overlaps [q,q+n). What's the barrier to performing such > a comparison? There are multiple problems here: 1) std::copy() is an algorithm that uses iterators. Iterators are not raw pointers, but raw pointers can be iterators. Some types of iterators cannot be compared for 'less than' or 'greater than' (non random access iterators). It would be possible to write std::copy to handle this situation, but only for random access iterators, but this would be in appropriate due to the second problem (see below). 2) Most copies are non-overlapping. If I add code inside std::copy to make it handle overlapping copies, it would slow down all the non- overlapping copies. If you look at the standard implementation where std::copy is used in various implementations, the fact that the range is overlapping or not becomes obvious. And if it is overlapping, the copy direction is also obvious without explicitly testing the range in many cases. All this improves performance for everyone rather than making all code suffer from worst case scenarios. None of this stops you from implementing an overlapping copy in terms of std::copy or std::copy_backward that only accepts random access iterators. But I would suggest you look at your code and you may find that information implicit. -- [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ]
From: David Abrahams on 3 Sep 2009 01:27
on Tue Sep 01 2009, cpp4ever <n2xssvv.g02gfr12930-AT-ntlworld.com> wrote: > Sada wrote: > >> hi how can I improve the performance of a C++ program? >> >> In my program I am using strcmp() strcpy() functions heavily , are >> they costly? if so how can i avoid them?? >> Is memcpy() is better than strcpy()? >> >> Please suggest me the factors that affect the performance of a C++ >> program. >> > Avoid allocating and deallocating memory, or passing large objects by > value. Both of these will impair performance. Not necessarily. See http://cpp-next.com/archive/2009/08/want-speed-pass-by-value/ -- Dave Abrahams BoostPro Computing http://www.boostpro.com [ See http://www.gotw.ca/resources/clcm.htm for info about ] [ comp.lang.c++.moderated. First time posters: Do this! ] |