| |
| |
| |
|
Page: 1 2
Comments:
<0> hello
<0> if i have a statement like this a+b;
<0> now is it specified whether a is first evaluated or b
<0> a and b are expressions
<1> I set a breakpoint on a function and can 'cont' all day long in GDB with it with no crashes. But if I do a bt on this breakpoint I can only see a few frames back until I hit a ??. Why can't I see further? Am I not compiling something right?... example: http://s162582763.onlinehome.us/bt.txt
<2> is there anyway to replace struct keyword with some kind of macro, such that i could have struct and safe_struct for example, to allow using GCC to add some fields (a signature) to every struct in the code, except specific ones
<3> howdy folks
<3> i'm trying to compile some c functions that i want to call from some Ada
<3> how do i get them to compile together?
<3> i have gps installed on my mac and i thought the IDE would be smart enough to do this for me
<3> i don't even know where to start looking in the gcc docs for help
<3> if someone could just point me in the proper direction i would greatly appreciate it
<4> hello, why the gcc take the initial stack pointer of main() to the end address of sram?
<5> is there a way to specify to unroll a certain for loop?
<6> morning
<4> all slept?
<4> anyone ?
<5> hi
<4> hi
<4> u don't sleep? geck?
<4> what r u doing?
<5> hi
<5> working on something
<4> me too
<4> school project?
<5> nope
<4> hobby project :)
<5> yeah
<5> http://people.freedesktop.org/~geckosenator/t.png
<5> that's a screenshot
<4> what's that?
<4> earth?
<7> when I have a variable declaration right after a switch, why does it give a syntax error? the error goes away if I put a ; before the declaration
<5> yeah
<5> hagabaka: example?
<7> someone pointed me to this, which explains it. example's there too http://www.thescripts.com/forum/thread214730.html
<5> hmm, it said gcc 3.2.2
<7> but it's based on C99 syntax
<7> the same happens with gcc 3.4.5 for me
<5> hmm
<5> dude
<5> that's perfectly normal
<8> i'm trying to build a dynamic library that should statically link a lib i made and dynamically link some system libs... i'm trying this on darwin right now, but i also have to compile it on linux, couldn't figure it out there either. i have the flags -dynamiclib (or -shared on linux) to create the dynamic lib then -static -lmystaticlib -dynamic/-shared -lsystemlibs. i get "conflicting code...
<8> ...gen style switches are used".
<5> you can never declare variables in a switch...
<7> you can declare variables in switch, by C99
<7> the problem is you can't declare right after any label
<5> really?
<5> then that is a bug
<7> that's what the thread says, and i tested it too
<5> I thought you still couldn't put them there even with c99
<7> it's dictated by the grammar
<7> it's "the standard", so you probably can't call it a bug
<5> er what?
<5> the standard says you can't declare a variable right after a case?
<7> the grammar published by the standard doesn't allow it
<7> read the thread
<4> hello, what 's actually happenned with "inline" function?
<5> hmm that is interesting
<4> what gcc does with inline function??
<4> does inline function has prologue/epilogue?
<4> ??
<4> hello?
<5> hi
<5> do you know what it does?
<9> Hello! Is there any documentation describing the optimization of different gcc versions for amd64? More specifically - are there significant advances post 3.4.5?
<9> http://www.devx.com/amd/Article/32207 claims "significant improvements", but it's hard for me to tell whether they are likely to affect me.
<10> hello there
<10> is a difference between cc and gcc please ?
<6> Bert-, try $ file `which cc`
<11> hello main () { int i = 0 ; i = i++ ; printf ("%d", i);} gives 1 as output .. for gcc but similar code in java gives 0 ... any ideas ?
<12> i = i++ is not sensible
<11> noshadow, it is when you are trying to understand various compiler's behaviours ;)
<11> what happens when you use i++ as a rvalue ?
<12> try reading 6.5.16 of the C standard
<11> okay
<6> _chax, take a look at -Wsequence-point
<11> sm0ketst, will do
<11> noshadow, sm0ketst thx
<13> hi all,
<13> any official channel for gdb
<12> mamohidi: try ##workingset
<14> hello, howd you fix line: int fpsize; fpbuff = malloc(sizeof(fpsize)); which dies on error: invalid conversion from void* to unsigned char* ?
<12> xrelay: compile with a C compiler instead
<14> gcc ?
<14> tnx
<12> are you sure you are using gcc and not g++?
<14> g++ :)
<12> in C++ you must cast, or use new or whatever it is called instead of malloc.
<14> i'll keep that in mind, tnx noshadow :)
<15> How can I add an include (or additional libraries) before the default includes/libraries to ./configure ? (so it will find correct version of some dependencies)
<15> Sorry for this easy (?) question, but I'm not used to compile from source and I google does not provide me the answer (or I'm asking the wrong questions)
<15> I have set the CPPFLAGS and LDFLAGS
<16> ouch, gcc 2.95.3 does not compile anymore ;S i wonder why... probably coz of glibc
<17> Not GCC related, but would someone know if there is -m32 equivalent in ICC? To generate x86 code on x86-64. (manual doesn't mention anything, release notes do mention support for both i32 and i32em on i32em).
<16> ouch #2 , it cannot link statically against glibc....
<16> undefined ref to _Unwind_Resume
<16> aww, and autoconf b0rks up with '/work/gcc-2.95.3/texinfo/configure : line 798 `echo $ac_n" (cached) $ac_c" 1>&6
<18> hey
<18> i got a standard gcc (4.1.1) install and i got some problems when i wanted to compile g-wrap which want to use libffi
<18> as i know libffi is part of gcc at todays
<18> but. my gcc install is a default install no tricks or patches or anything else.
<18> g-wrap finds ffi.h in /usr/include but can not use that because ffi.h wants to include ffitarget.h
<18> this ffitarget.h is at /usr/lib/gcc/x86_64-unknown-linux-gnu/4.1.1/include/libffi/ffitarget.h
<18> so my question is why gcc default install dont ship this ffitarget.h to /usr/include instead of gcc/CARCH-*/.. dir ?
<18> is there some special thing for ship to that dir and not to /usr/include ?
<19> hello, I am trying to compil a linux host based cross-compiler for solaris9
<19> has anyone experienced this?
<19> gcc keeps trying to find includes in /usr/include instead of `pwd`/include as suggested in the configure of binutils --with-headers=`pwd`/include
<20> why do you need such an option when buikding bintutils ?
<20> besides `pwd`/include does not look good, it should be spmetjhing lioke --with-headers=$prefix/sparc-solaris/include
<20> when building gcc
<19> velco: I have the same problem with your method
<19> velco: trying to compile gcc4
<19> my mistake, binutils
<19> when I try to compile gcc 2.95 I have another problem which is reorg.c:384: error: invalid lvalue in increment
<19> velco: for binutils, I need that option to get a working binutil for solaris9 (do I :) ?)
<20> I don't think you need it for binutils
<20> for binutils configure --prefix=XXX --target=sparc-solaris should be ok
<20> or i686-solaris or whatever
<19> well same error
<19> strerror.c:461: error: conflicting types for 'sys_errlist'
<19> /usr/include/bits/sys_errlist.h:28: error: previous declaration of 'sys_errlist' was here
<20> building what ?
<19> binutils
<19> binutils-2.9.1
<20> hehe
<20> what is that antique ?
<19> hehe, I used the version I've found here: http://is.rice.edu/~welsh/cross-compiler/#HeaderFiles
<20> I've just built binutils 2.16.1 with no problems
<19> ok trying with your version
<20> and with my command line :)
<19> it's compiling atm
<19> and it gets further than my old buddy :)
<21> hi
<21> is there a sane way to tell gcc's inline asm handling to ***ign me a scratch register?
<21> also, does gcc implicitly ***ume that inputs are clobbered?
<21> because I have a tight loop that I'd like gcc to generate in the hope of avoiding having to ***ign a label name in inline asm, however I'd prefer for it not to reload the input operands
<20> GyrosGeier: yes, no.
<21> not good
<20> is the inline asm snippet big ?
<21> because I'm running out of registers if gcc thinks it needs to take copies of everything
<21> velco, an unrolled loop of "shift one to the right setting carry, set GPIO to one if carry set, clear GPIO if carry clear, set other GPIO, repeat"
<20> paste it somewhere
<21> sec
<21> hmm
<21> I've tried three pastebins now, they all take an awfully long time
<21> http://pastebin.ca/168083
<20> "mov %0, %0, RRX \n"
<20> what is this doing >?
<21> that is the rotate right with carry
Return to
#gcc
or
Go to some related
logs:
#perl
setting the clock fluxbox
$_sessions don't work
installing RARubuntu
perlcoc
#php
gtkpod segfault fc5
raid 5/10/50 defined
sourceomatic for ubuntu
why does ps2 mouse lag on gentoo
|
|
