But linked lists are worse on performance than realloc. One way to use realloc better is to have a current length and max size and also rellocate more memory than need at the needed time. This cause better locality than a linked list. And you can always use realloc to make the list to the current length. Linked lists cause many problems when it comes to prefetching and fast memory access so they are slow when you are accessing things down a list. You can do a tree using one list also, by having an inverse tree. Graphs can uses multiple lists instead of linked ones.
Also linked lists are very prone to leaking and very hard to figure out which one to free first.
Note defining who frees what and when is a different issue than what memory manager you are you using. Defining an API is good but unless you define when things can be released/retained you are f**d so the who issue really is more a social one.
Actually all Macs are defined that way, the hardware clock is defined that way. Little know fact (or widely known) almost all Macs will reset to January 1, 1969 if the batter is removed.
Sun should have invited us GCC developers also to help out with this because most of us want a way to do Inter modular optimizations but we have the FSF looking over our shoulder on how we implement it, right now (the mainline) you have to compile all the source files at the same time to get IMA to work correctly and you have to say to produce an.o file first.
He missed the best: GNUStep. GNUStep uses Objective-C and is a clone of the OpenStep API's and is pretty stable. To write a simple Application you do not have to write that much code any more.
You should look at this bug which was filed, PR11842. People calm down, this is not really big news as FSF has done this before with Apple and other people so this should have not come as a big surprise.
Actually 3.3.2 will be faster than 3.3.1 if someone accepts my patch. The person who ran these, had reran these after my patch for 3.4 went in, so in fact 3.4 is faster than what it was.
I do not work on binutils at all and I am a Mac OS X user, not really a big GNU/Linux user so my performance is many to speed up Powerpc code and not x86, if my changes speed up x86 that was just an extra benefit.
Re:Java dead code removal?
on
GCC 3.3.1 Released
·
· Score: 5, Informative
The dead code removal has to be done in the linker, not in GCC. Note the linker is part of the binutils project, not GCC, complain to them, not us, GCC.
Also static linking anything will grow your executable more than needed.
Dead code removal in your code, not libraries is done by GCC already and is being improved still.
Re:I don't get it? Updates to an obseleted arch?
on
GCC 3.3.1 Released
·
· Score: 1
Also m68k-elf was just fixed on the mainline, it was having trouble building because improvements with respect to adding builtin defines.
Re:Oops, and there goes varargs.h...
on
GCC 3.3.1 Released
·
· Score: 3, Informative
But most of the time to get these programs to run correctly on any GCC you would have to update it any way because aliasing rules and other reasons. Remember most of the time when you use -traditional you really wanted the traditional preprocessor rather than the traditional c compiler so you can still use -traditional-cpp.
The reason why GCC removed -traditional because ISO C89 is already 13 years old and it was getting to hard to maintain those features.
Update the code to ISO C90 is not that hard any way because most of the time for varargs it just a replace with stdards and such.
The Constitution (and the Amendments) is above any Federal Law, in fact any Federal Law that violates the Constitution can be stuck down (we all should know that, DMCA and such go against the Constitution). The Constitution gives powers to the Federal Government, in fact right now the Federal Government is over stepping it bounds in terms of powers with the DMCA and the PATRIOT Act.
PPC did not exist until Apple, IBM, and MOT got together to design it. On the other hand Power (or RS6000) did exist and is what PPC is based on which was running AIX. Linux on PPC did not exist until Apple added support for it that is what I was trying to say; IBM was not the one who first added support for it, Apple was and that was back in Summer of 1998 see www.mklinux.org.
Slashdot Mirror
But linked lists are worse on performance than realloc.
One way to use realloc better is to have a current length and max size and also rellocate more memory than need at the needed time. This cause better locality than a linked list. And you can always use realloc to make the list to the current length. Linked lists cause many problems when it comes to prefetching and fast memory access so they are slow when you are accessing things down a list. You can do a tree using one list also, by having an inverse tree. Graphs can uses multiple lists instead of linked ones.
Also linked lists are very prone to leaking and very hard to figure out which one to free first.
Note defining who frees what and when is a different issue than what memory manager you are you using. Defining an API is good but unless you define when things can be released/retained you are f**d so the who issue really is more a social one.
It is called realloc, that is the real way that people should use to self-manage memory allocation, and something that detects leaks is also needed.
Actually all Macs are defined that way, the hardware clock is defined that way.
Little know fact (or widely known) almost all Macs will reset to January 1, 1969 if the batter is removed.
Sun should have invited us GCC developers also to help out with this because most of us want a way to do Inter modular optimizations but we have the FSF looking over our shoulder on how we implement it, right now (the mainline) you have to compile all the source files at the same time to get IMA to work correctly and you have to say to produce an .o file first.
You know a GNUStep OS (not just another distro) would be a really good thing, maybe use HURD as the kernel.
Easierly, s390 is 31 bits (yes 31) and GCC supports that.
s390x is 64bit though.
This is the same fuss M$ had when AOL locked other people out of their network so M$ is being hypocritical here.
He missed the best: GNUStep.
GNUStep uses Objective-C and is a clone of the OpenStep API's and is pretty stable.
To write a simple Application you do not have to write that much code any more.
You should look at this bug which was filed, PR11842.
People calm down, this is not really big news as FSF has done this before with Apple and other people so this should have not come as a big surprise.
Actually 3.3.2 will be faster than 3.3.1 if someone accepts my patch. The person who ran these, had reran these after my patch for 3.4 went in, so in fact 3.4 is faster than what it was.
I do not work on binutils at all and I am a Mac OS X user, not really a big GNU/Linux user so my performance is many to speed up Powerpc code and not x86, if my changes speed up x86 that was just an extra benefit.
The dead code removal has to be done in the linker, not in GCC. Note the linker is part of the binutils project, not GCC, complain to them, not us, GCC.
Also static linking anything will grow your executable more than needed.
Dead code removal in your code, not libraries is done by GCC already and is being improved still.
Also m68k-elf was just fixed on the mainline, it was having trouble building because improvements with respect to adding builtin defines.
But most of the time to get these programs to run correctly on any GCC you would have to update it any way because aliasing rules and other reasons.
Remember most of the time when you use -traditional you really wanted the traditional preprocessor rather than the traditional c compiler so you can still use -traditional-cpp.
The reason why GCC removed -traditional because ISO C89 is already 13 years old and it was getting to hard to maintain those features.
Update the code to ISO C90 is not that hard any way because most of the time for varargs it just a replace with stdards and such.
The Constitution (and the Amendments) is above any Federal Law, in fact any Federal Law that violates the Constitution can be stuck down (we all should know that, DMCA and such go against the Constitution). The Constitution gives powers to the Federal Government, in fact right now the Federal Government is over stepping it bounds in terms of powers with the DMCA and the PATRIOT Act.
Even more blast from the past the co-operation between Sun and NeXT for OpenSTEP and such.
Any newer ID game will work.
Any blizzard game will work also.
Isn't it "Dude where is my car?"
Solaris did, that is how.
I had meant early 1996, not summer of 1998.
PPC did not exist until Apple, IBM, and MOT got together to design it. On the other hand Power (or RS6000) did exist and is what PPC is based on which was running AIX. Linux on PPC did not exist until Apple added support for it that is what I was trying to say; IBM was not the one who first added support for it, Apple was and that was back in Summer of 1998 see www.mklinux.org.
Most of the original 32bit PPC support was done by Apple, not IBM (for MKLinux). So how can SCO claim that one.
It was not listed on the web site so I thought it was not out yet sorry for that.
Not out yet :(.
I missed one, hppa GNU/Linux.