January 2006 - Darwin-dev - lists.apple.com

Re: xcodebuild question
by j o a r 07 Jan '06

07 Jan '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com On 7 jan 2006, at 01.21, jmzorko(a)mac.com wrote: j o a r _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… This email sent to site_archiver(a)lists.apple.com ... now, i'm doing all of this on OSX Panther (10.3.9) so the messages about "temporarily downgrading it to version 39" intrigue me as something that may be a part of my problem (as well as the expected-a-string-but-found-an-array message). Do I need to specify something different on the "xcodebuild" command line to build using XCode 1.5 instead of the newer version? Or is there something in the vlc.pbproj file that is 10.4 specific perhaps? Or something else i'm not aware of? They probably upgraded it to Xcode 2.2, which only runs on Mac OS X 10.4. The release notes for Xcode 2.2 states that: "Once upgraded for Xcode 2.1 or later, a project file cannot be opened or built by Xcode 2.0 or earlier." smime.p7s

1 0

xcodebuild question
by jmzorko＠mac.com 07 Jan '06

07 Jan '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Hello, all ... Regards, John Falling You - exploring the beauty of voice and sound http://www.fallingyou.com _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… Normally i'm a *nix guy and use Apple's versions of the GNU tools when building apps that run on OSX. However, i'm trying to build the latest VLC (http://www.videolan.org) and it's failing toward the very end, when it tries to run the xcodebuild step. I tried running this step manually to narrow the problem down, and this is what I see: PugsleyButt:~/multimedia/vlc-trunk/tmp/extras/MacOSX jmzorko$ xcodebuild -target vlc 2006-01-06 16:15:53.063 xcodebuild[20453] ASSERTION FAILURE in /SourceCache/DevToolsBase/DevToolsBase-387/pbxcore/Target.subproj/ PBXBuildSettingsDictionary.m:106 (function=PBXBuildSettingsDictionary_getValueForKeyPath): value should be a string, but it is <NSCFArray:0x003713c0> 2006-01-06 16:15:53.077 xcodebuild[20453] Exception raised during posting of notification. Ignored. exception: ASSERTION FAILURE in /SourceCache/DevToolsBase/DevToolsBase-387/pbxcore/Target.subproj/ PBXBuildSettingsDictionary.m:106 (function=PBXBuildSettingsDictionary_getValueForKeyPath): value should be a string, but it is <NSCFArray:0x003713c0> (NOTE: project vlc was written by a newer version of Xcode (42) -- temporarily downgrading it to version 39 (without modifying project file)) 2006-01-06 16:15:53.165 xcodebuild[20453] ASSERTION FAILURE in /SourceCache/DevToolsBase/DevToolsBase-387/pbxcore/Target.subproj/ PBXBuildSettingsDictionary.m:106 (function=PBXBuildSettingsDictionary_getValueForKeyPath): value should be a string, but it is <NSCFArray:0x003713c0> ** INTERNAL ERROR: Uncaught Exception ** Exception: ASSERTION FAILURE in /SourceCache/DevToolsBase/DevToolsBase-387/pbxcore/Target.subproj/ PBXBuildSettingsDictionary.m:106 (function=PBXBuildSettingsDictionary_getValueForKeyPath): value should be a string, but it is <NSCFArray:0x003713c0> Stack: 0 0x94eb7cd0 -[TSException raise] 1 0x94eb0ca4 __TSAssertionFailureHandler 2 0x963041a0 _PBXBuildSettingsDictionary_getValueForKeyPath 3 0x96309cfc -[PBXJamString writeVariablesForSettings:usingPrefix:options:] 4 0x963813c8 -[PBXJamfileWriter _writeBuildSettingsForTarget:] 5 0x96331c20 -[PBXJamfileWriter jamfileContentsForTarget:buildOperation:] 6 0x963315c4 +[PBXJamfileWriter jamfileContentsForTarget:buildOperation: withAuxilliaryFilesDirectoryPath:] 7 0x9636c0e8 -[PBXTarget buildToolStdinDataForTargetBuildWithBuildOperation:] 8 0x96336670 -[PBXTarget createAndConfigureTaskForBuildingWithBuildOperation:synchronously:] 9 0x9632fe58 -[PBXTargetBuildOperation nextConfiguredTaskForRunningSynchronously:] 10 0x9632edd0 -[PBXBackgroundTaskOperation _launchNextBackgroundTaskOrThreadedOperation] 11 0x96402954 -[PBXBackgroundTaskOperation runSynchronously] 12 0x00007ed0 13 0x0000991c 14 0x00006850 15 0x8fe1a274 Abort trap PugsleyButt:~/multimedia/vlc-trunk/tmp/extras/MacOSX jmzorko$ ... now, i'm doing all of this on OSX Panther (10.3.9) so the messages about "temporarily downgrading it to version 39" intrigue me as something that may be a part of my problem (as well as the expected-a-string-but-found-an-array message). Do I need to specify something different on the "xcodebuild" command line to build using XCode 1.5 instead of the newer version? Or is there something in the vlc.pbproj file that is 10.4 specific perhaps? Or something else i'm not aware of? This email sent to site_archiver(a)lists.apple.com

1 0

Re: CrashReporter, signals and exception ports
by Steve Checkoway 06 Jan '06

06 Jan '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com This causes no end of frustration, doesn't it? *sigh* We do this just with a signal handler. We catch: static int signals[] = { SIGALRM, SIGBUS, SIGFPE, SIGHUP, SIGILL, SIGINT, SIGABRT, SIGQUIT, SIGSEGV, SIGTRAP, SIGTERM, SIGVTALRM, SIGXCPU, SIGXFSZ, #if defined(HAVE_DECL_SIGPWR) && HAVE_DECL_SIGPWR SIGPWR, #endif -1 }; And then we set up the signal handlers using: struct sigaction sa; sa.sa_flags = 0; if( p != NULL ) sa.sa_flags |= SA_ONSTACK; sa.sa_flags |= SA_NODEFER; sa.sa_flags |= SA_SIGINFO; sigemptyset(&sa.sa_mask); /* Set up our signal handlers. */ sa.sa_sigaction = SigHandler; for( int i = 0; signals[i] != -1; ++i ) sigaction( signals[i], &sa, NULL ); We didn't have to do anything with exception ports to get this to work. - Steve _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… This email sent to site_archiver(a)lists.apple.com On Jan 6, 2006, at 1:15 PM, Jean-Francois Im wrote: This all started when I noticed that an application's windows are not closed until the CrashReporter's close button is clicked. It makes sense for most applications to show a dialog instead of just magically vanishing and leaving a dialog box behind, however, when the application is a fullscreen window, the CrashReporter actually appears "behind" the application, with the side effect that the close button cannot be clicked. Simple enough, I thought, it's just a matter of adding a signal handler, catch the signal for the exceptions, close the window, then resend the signal to ourself, thus causing the CrashReporter to pop up. Unfortunately, CrashReporter sets up an exception port, thus intervening before the signal handler has a chance to do something(because the exception is not translated into a signal). where p is the pointer to our signal stack set up with sigaltstack. Doing this we can shut down the full screen graphics context and perform any other clean up. We actually give a stack trace (among other things) ourself and setting the handlers seems to prevent the Crash Reporter from running. Thus, I've searched for information on how to set up an exception port and have it rethrow the exception after closing the window, but documentation is quite scarce on the topic of exception ports. I have found a couple of pointers which sent me on the right track, but now I appear to have hit a wall. smime.p7s

1 0

CrashReporter, signals and exception ports
by Jean-Francois Im 06 Jan '06

06 Jan '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com User-agent: Mozilla Thunderbird 1.0.7 (Macintosh/20050923) Hi, (If desired, you can stop reading here.) Thanks in advance for any help on the topic. Jean-Francois Im PS: For posterity, I'll include the documentation I found useful : http://www.wodeveloper.com/omniLists/macosx-dev/2001/March/msg00849.html http://web.mit.edu/darwin/src/modules/xnu/osfmk/man/ http://www.opensource.apple.com/darwinsource/10.4.3/xnu-792.6.22/bsd/uxkern… _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… I am trying to execute some code when exceptions occur in my code(execution of trap, segment violations, etc.) while still keeping the normal flow of exceptions(ie. CrashReporter pops up). I have figured out after a while that I need to write an exception port, I am unable to figure out exactly how to do so. Currently, I have an exception port which traps the exception but does not rethrow it(I assume that it is possible). I'd paste the code here, but it's somewhat longish and it's not all that different from the ux_handler() function present in XNU. http://www.opensource.apple.com/darwinsource/10.4.3/xnu-792.6.22/bsd/uxkern… Ideally, I'd like to rethrow the exception while knowing which thread caused it, so that I can use the backtrace from the CrashReporter to help in debugging crashes. This all started when I noticed that an application's windows are not closed until the CrashReporter's close button is clicked. It makes sense for most applications to show a dialog instead of just magically vanishing and leaving a dialog box behind, however, when the application is a fullscreen window, the CrashReporter actually appears "behind" the application, with the side effect that the close button cannot be clicked. Simple enough, I thought, it's just a matter of adding a signal handler, catch the signal for the exceptions, close the window, then resend the signal to ourself, thus causing the CrashReporter to pop up. Unfortunately, CrashReporter sets up an exception port, thus intervening before the signal handler has a chance to do something(because the exception is not translated into a signal). Thus, I've searched for information on how to set up an exception port and have it rethrow the exception after closing the window, but documentation is quite scarce on the topic of exception ports. I have found a couple of pointers which sent me on the right track, but now I appear to have hit a wall. This email sent to site_archiver(a)lists.apple.com

1 0

Re: Win32 cross-compiler for OSX?
by Mitch Murphy 06 Jan '06

06 Jan '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com cheers, Mitch --------- #! /bin/bash #----------------------------------------------------- # # BEGIN USER SETTINGS # # You need to review and adjust the macros that follow # #----------------------------------------------------- # What are your make flags? (ie. -j 8 for a parallel build) MAKEFLAGS='-j 1' # What flavor of GCC cross-compiler are we building? TARGET=mingw32 # What directory will the cross-compiler be built in? # This is the directory into which source archives will # be downloaded, expanded, compiled, etc. You need to # have write-access to this directory. If you leave it # blank, it defaults to the current directory. BUILDDIR= # Where does the cross-compiler go? # This should be the directory into which your cross-compiler # will be installed. Remember that if you set this to a directory # that only root has write access to, you will need to run this # script as root. PREFIX=/Developer/SDKs/gcc-4.0.2-mingw.sdk # Purge anything and everything already in the $PREFIX #(also known as the destination or installation) directory? # Set to "Y" to purge, any other value omits the purge step. PURGE_DIR="Y" # Set the following to the files from the current MinGW release # (or whichever MinGW release you wish to build and install) # You need to set both the URL they will be downloaded from # and the exact name of the individual component files. MINGW_URL="http://heanet.dl.sourceforge.net/sourceforge/mingw" GCC_URL="http://mirrors.rcn.net/pub/sourceware/gcc/releases/gcc-4.0.2" # GCC_CORE is required; the other components are optional. # Set any you don't want to "". You need binutils, # mingw runtime and w32api; do not ever set those to "". GCC_CORE_ARCHIVE="gcc-core-4.0.2.tar.gz" # GCC_GPP_ARCHIVE="gcc-g++-4.0.2.tar.gz" GCC_OBJC_ARCHIVE="gcc-objc-4.0.2.tar.gz" # GCC_G77_ARCHIVE="gcc-g77-3.4.2-20040916-1-src.tar.gz" # GCC_OBJC_ARCHIVE="gcc-objc-3.4.2-20040916-1-src.tar.gz" # GCC_JAVA_ARCHIVE="gcc-java-3.4.2-20040916-1-src.tar.gz" GCC_ADA_ARCHIVE="" GCC_PATCH="" # "The specified version of binutils did not build for me (on neither # Debian amd64 nor i386), whereas binutils-2.16.91-20050827-1-src.tar.gz # built fine on both. " # # - JoelDice # # BINUTILS_ARCHIVE="binutils-2.15.91-20040904-1-src.tar.gz" BINUTILS_ARCHIVE="binutils-2.16.91-20050827-1-src.tar.gz" MINGW_ARCHIVE="mingw-runtime-3.9.tar.gz" W32API_ARCHIVE="w32api-3.5.tar.gz" #----------------------------------------------------- # # END USER SETTINGS # # The remainder of the script should not neet any edits # #----------------------------------------------------- # Make sure these are initialized as we want them GCC_CORE="" BINUTILS="" GCC_LANGS="c" # Set our build directory and where our sources will go if [ "x$BUILDDIR" = "x" ]; then # Default to the current directory BUILDDIR=$(pwd) fi SRCDIR="$BUILDDIR/source" # Need install directory first on the path so gcc can find binutils PATH="$PREFIX/bin:$PATH" #----------------------------------------------------- # # Functions that do most of the work # #----------------------------------------------------- function download_files { # Download a file from a given url, only if it is not present mkdir -p "$SRCDIR" # Make sure wget is installed if test "x`which wget`" = "x" ; then echo "You need to install wget." exit 1 fi download_file "$GCC_CORE_ARCHIVE" "$GCC_URL" if [ "x$GCC_GPP_ARCHIVE" != "x" ]; then download_file "$GCC_GPP_ARCHIVE" "$GCC_URL" fi if [ "x$GCC_G77_ARCHIVE" != "x" ]; then download_file "$GCC_G77_ARCHIVE" "$GCC_URL" fi if [ "x$GCC_OBJC_ARCHIVE" != "x" ]; then download_file "$GCC_OBJC_ARCHIVE" "$GCC_URL" fi if [ "x$GCC_JAVA_ARCHIVE" != "x" ]; then download_file "$GCC_JAVA_ARCHIVE" "$GCC_URL" fi if [ "x$GCC_ADA_ARCHIVE" != "x" ]; then download_file "$GCC_ADA_ARCHIVE" "$GCC_URL" fi download_file "$BINUTILS_ARCHIVE" "$MINGW_URL" download_file "$MINGW_ARCHIVE" "$MINGW_URL" download_file "$W32API_ARCHIVE" "$MINGW_URL" } function download_file { cd "$SRCDIR" if test ! -f $1 ; then echo "Downloading $1" wget "$2/$1" if test ! -f $1 ; then echo "Could not download $1" exit 1 fi else echo "Found $1 in the srcdir $SRCDIR" fi cd "$BUILDDIR" } function purge_existing_install { echo "Purging the existing files in $PREFIX" if cd "$PREFIX"; then rm -rf * fi cd "$BUILDDIR" } function install_libs { echo "Installing cross libs and includes" mkdir -p "$PREFIX/$TARGET" cd "$PREFIX" ln -s "$TARGET" usr cd "$PREFIX/$TARGET" tar -xzf "$SRCDIR/$MINGW_ARCHIVE" tar -xzf "$SRCDIR/$W32API_ARCHIVE" cd "$BUILDDIR" } function extract_binutils { cd "$SRCDIR" BINUTILS=`tar -tzf "$SRCDIR/$BINUTILS_ARCHIVE" | head -n 1` rm -rf "$BINUTILS" echo "Extracting binutils" tar -xzf "$SRCDIR/$BINUTILS_ARCHIVE" cd "$BUILDDIR" } cd "$BUILDDIR" } function extract_gcc { cd "$SRCDIR" GCC=`tar -tzf "$SRCDIR/$GCC_CORE_ARCHIVE" | head -n 1` rm -rf "$GCC" echo "Extracting gcc" tar -xzf "$SRCDIR/$GCC_CORE_ARCHIVE" if [ "x$GCC_GPP_ARCHIVE" != "x" ]; then GCC_LANGS=${GCC_LANGS}",c++" tar -xzf "$SRCDIR/$GCC_GPP_ARCHIVE" fi if [ "x$GCC_G77_ARCHIVE" != "x" ]; then GCC_LANGS=${GCC_LANGS}",f77" tar -xzf "$SRCDIR/$GCC_G77_ARCHIVE" fi if [ "x$GCC_OBJC_ARCHIVE" != "x" ]; then GCC_LANGS=${GCC_LANGS}",objc" tar -xzf "$SRCDIR/$GCC_OBJC_ARCHIVE" fi if [ "x$GCC_JAVA_ARCHIVE" != "x" ]; then GCC_LANGS=${GCC_LANGS}",java" tar -xzf "$SRCDIR/$GCC_JAVA_ARCHIVE" fi if [ "x$GCC_ADA_ARCHIVE" != "x" ]; then GCC_LANGS=${GCC_LANGS}",ada" tar -xzf "$SRCDIR/$GCC_ADA_ARCHIVE" fi cd "$BUILDDIR" } function patch_gcc { if [ "$GCC_PATCH" != "" ]; then echo "Patching gcc" cd "$SRCDIR/$GCC" patch -p1 < "$SRCDIR/$GCC_PATCH" cd "$BUILDDIR" fi } # --disable-shared \ cd "$BUILDDIR" } function final_tweaks { echo "Finalizing installation" # remove gcc build headers rm -rf "$PREFIX/$TARGET/sys-include" # make cc and c++ symlinks to gcc and g++ # if [ ! -f "$PREFIX/$TARGET/bin/g++" ]; then # ln "$PREFIX/bin/$TARGET-g++" "$PREFIX/$TARGET/bin/g++" # fi # if [ ! -f "$PREFIX/$TARGET/bin/cc" ]; then # ln -s "gcc" "$PREFIX/$TARGET/bin/cc" # fi # if [ ! -f "$PREFIX/$TARGET/bin/c++" ]; then # ln -s "g++" "$PREFIX/$TARGET/bin/c++" # fi echo "Installation complete!" } # # Main part of the script # download_files if [ "x$PURGE_DIR" = "xY" ]; then purge_existing_install fi install_libs extract_binutils configure_binutils build_binutils install_binutils extract_gcc patch_gcc configure_gcc build_gcc install_gcc final_tweaks # # End # _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… I would like to make Win32 executables from my PowerBook running Panther (Tiger eventually). I know I can do this from a Debian box using GNU gcc, but I was wondering if I can do it using Apple's gcc? here the script that worked for me, slightly modified from what i found at http://www.mingw.org/MinGWiki/index.php/BuildMingwCross. see http://sources.redhat.com/ml/crossgcc/2006-01/msg00005.html for how it did for darwin->linux cross-compiler using crosstool (http:// kegel.com/crosstool/crosstool-0.38/doc/crosstool-howto.html) # This is my script for building a complete MinGW cross-compiler toolchain # that runs under Linux to produce executables that run under Windows. It # probably works (or can easily be adapted to work) under any unix system. # # It is based in large part on Sam Lantinga's script, which in turn was # based partly on Ray Kelm's script, which in turn was built on # Mo Dejong's script for doing the same, but with some added fixes. # # My changes: # 1. Adapted the script to the new packaging of MinGW GCC, which is # currently split into core and auxiliary components. # 2. The script now determines the GCC and BINUTILS directory name # directly from the tar file contents. This gets around common # problems due to the directory names not always following the # expected patterns. # 3. Grouped together and simplified the macros that users need to # define. # 4. Made optional components truly optional -- leave the # corresponding archive names blank and they will be ignored. # 5. Included an option to purge the installation directory before # installing the current cross-compiler. # # NOTE: If you choose a destination directory for the installation (set # in the macro PREFIX) for which you do not have write access, you will # need to run this script with root (or equivalent) privileges. # # # Updated by Igor Mikolic-Torreira <igormt(a)alumni.caltech.edu> # # Additional changes by MitchMurphy 2005/12/29 # # 2.1) binutils-2.16.91-20050827-1-src.tar.gz as per JoelDice # 2.2) MAKEFLAGS for parallel make # 2.3) built under Mac OS X 10.4 # function configure_binutils { cd "$BUILDDIR" rm -rf "binutils-$TARGET" mkdir "binutils-$TARGET" cd "binutils-$TARGET" echo "Configuring binutils" "$SRCDIR/$BINUTILS/configure" --prefix="$PREFIX" -- target="$TARGET" --disable-nls \ --with-gcc --with-gnu-as --with-gnu-ld &> configure.log #--with-gcc --with-gnu-as --with-gnu-ld --disable- shared &> configure.log function build_binutils { cd "$BUILDDIR/binutils-$TARGET" echo "Building binutils" make $MAKEFLAGS CFLAGS="-O2 -fno-exceptions" LDFLAGS="-s" &> make.log if test $? -ne 0; then echo "make of binutils failed - log available: binutils-$TARGET/make.log" exit 1 fi cd "$BUILDDIR" } function install_binutils { cd "$BUILDDIR/binutils-$TARGET" echo "Installing binutils" make $MAKEFLAGS install &> make-install.log if test $? -ne 0; then echo "install of binutils failed - log available: binutils-$TARGET/make-install.log" exit 1 fi cd "$BUILDDIR" } function configure_gcc { cd "$BUILDDIR" rm -rf "gcc-$TARGET" mkdir "gcc-$TARGET" cd "gcc-$TARGET" echo "Configuring gcc" "$SRCDIR/$GCC/configure" -v \ --prefix="$PREFIX" --target="$TARGET" \ --with-headers="$PREFIX/$TARGET/include" \ --with-gcc --with-gnu-ld --with-gnu-as \ --enable-threads --disable-nls --enable-languages= $GCC_LANGS \ --disable-win32-registry --enable-sjlj-exceptions -- enable-libgcj \ --disable-java-awt --without-x --enable-java- gc=boehm --disable-libgcj-debug \ --enable-interpreter --enable-hash-synchronization -- enable-libstdcxx-debug \ &> configure.log function build_gcc { cd "$BUILDDIR/gcc-$TARGET" echo "Building gcc" make $MAKEFLAGS CFLAGS="-O2" CXXFLAGS="-O2" GCJFLAGS="-O2" LDFLAGS="-s" DEBUG_FLAGS="-g0" &> make.log if test $? -ne 0; then echo "make of gcc failed - log available: gcc- $TARGET/make.log" exit 1 fi if [ "x$GCC_ADA" != "x" ]; then cd gcc make $MAKEFLAGS "CFLAGS=-O2" "LDFLAGS=-s" gnatlib_and_tools &> make-gnatlib.log if test $? -ne 0; then echo "make of gnatlib and tools failed - log available: gcc-$TARGET/make-gnatlib.log" exit 1 fi fi cd "$BUILDDIR" } function install_gcc { cd "$BUILDDIR/gcc-$TARGET" echo "Installing gcc" make $MAKEFLAGS install &> make-install.log if test $? -ne 0; then echo "install of gcc failed - log available: gcc- $TARGET/make-install.log" exit 1 fi cd "$BUILDDIR" } # Add extra binary links if [ ! -f "$PREFIX/$TARGET/bin/objdump" ]; then ln "$PREFIX/bin/$TARGET-objdump" "$PREFIX/$TARGET/ bin/objdump" fi # strip all the binaries ls "$PREFIX"/bin/* "$PREFIX/$TARGET"/bin/* | egrep -v '.dll $' | egrep -v 'gccbug$' | while read file; do strip "$file" done This email sent to site_archiver(a)lists.apple.com

1 0

Re: Win32 cross-compiler for OSX?
by Finlay Dobbie 06 Jan '06

06 Jan '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Domainkey-signature: a=rsa-sha1; q=dns; c=nofws; s=beta; d=gmail.com; h=received:message-id:date:from:reply-to:to:subject:cc:in-reply-to:mime-version:content-type:content-transfer-encoding:content-disposition:references; b=Cv/MNgZOoAgJjmtGcZmaQ/GvpDkPrgKi2XKnFNKFFd2lfjmAOgdbJ65im5CP4dLrjWlczmErsNIoC+/YPx9W7jkkVcImlc2+XuWbQ/tGM9qjLrs/zCdIb4Z4J5An3Kw0ske1Rn6fgi3VN4NiV9gwsfGL/Nghnt4Lx+u8OROwh2w= On 06/01/06, jmzorko(a)mac.com <jmzorko(a)mac.com> wrote: > > Hello, all ... > > I would like to make Win32 executables from my PowerBook running > Panther (Tiger eventually). I know I can do this from a Debian box > using GNU gcc, but I was wondering if I can do it using Apple's gcc? <http://www.zonic.co.uk/products/mingw/> details how to set up a MinGW cross-compilation environment on Tiger. -- Finlay _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… This email sent to site_archiver(a)lists.apple.com

1 0

Re: Correct API to implement a timer in a real-time thread
by Stéphane Letz 06 Jan '06

06 Jan '06

site_archiver(a)lists.apple.com Delivered-To: Darwin-dev(a)lists.apple.com On Jan 5, 2006, at 2:31 PM, Herbie Robinson wrote: The HAL uses the CAGuard class from the Public Utilities in the CoreAudio SDK to handle all it's scheduling needs. CAGuard simply wraps up a pthread mutex and cond var as a unit. CAGuard uses the function pthread_cond_timedwait_relative_np() to handle blocking the thread for a specified amount of time. If I remember right, the difference between this function and the pthread_cond_timedwait () is that in the former you pass the interval to wait and the latter you pass the time to wake up. The reason for this is purely because of the implementation of CAGuard. In fact, if you look at the implementation of pthreads in Darwin, you'll find that pthread_cond_timedwait() is implemented pretty much the same way that CAGuard::WaitUntil() is. Didn't one of your co-workers post that the pthreads implementation uses a global lock that can cause priority inversion; so, at least some pthreads calls aren't safe to use in real-time threads? Yes. There is a spin lock in the pthread implementation that can cause some trouble under some conditions. It isn't clear how often this bug comes up, but it seems to be fairly rare (and almost non- existent on MP machines). It has a particular signature in kernel traces, so it's easy to spot and I have only ever seen one trace from a developer that showed this bug happening (which is what led me to it in the first place). It has been my experience that more often than not any problems folks have in their IOProcs are their own doing. The libc folks are looking at it as far as I know. In the mean time, the HAL will continue to use the pthread calls for all it's scheduling needs. If this particular bug is hurting you, you can switch to using some other synchronization API like mach semaphores or whatever you like. At any rate, this bug does not invalidate any of the advice about the algorithms to use to solve the problem Mr. Letz was looking at. Jeff Moore Core Audio Apple Here is the code I use now in the "time-constraint" thread: static void* TimerTask(void* unused) { // Initialization mach_timebase_info_data_t tbi; mach_timebase_info(&tbi); double invRatio = ((double)tbi.denom) / ((double)tbi.numer); double timeResNanos = gTimeResMilli * 1000000; // In nanosecond double nextDateNanos, curDateNanos = mach_absolute_time() / invRatio; while (1) { pthread_testcancel(); nextDateNanos = mach_absolute_time() / invRatio; while (curDateNanos < nextDateNanos) { long call = gTimeResMilli; while (call--) ClockHandler(gMem); curDateNanos += timeResNanos; } mach_wait_until(curDateNanos * invRatio); } return 0; } static void* TimerTask(void* unused) { // Initialization mach_timebase_info_data_t tbi; mach_timebase_info(&tbi); double invRatio = ((double)tbi.denom) / ((double)tbi.numer); double timeResNanos = gTimeResMilli * 1000000; // In nanosecond double nextDateNanos, curDateNanos = mach_absolute_time() / invRatio; while (1) { nextDateNanos = mach_absolute_time() / invRatio; while (curDateNanos < nextDateNanos) { long call = gTimeResMilli; while (call--) EMPTYClockHandler(gMem); curDateNanos += timeResNanos; } mach_wait_until(curDateNanos * invRatio); } return 0; } I tried to "capture/profile" the CPU spike without real success.... Any idea? Stephane Letz _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… At 12:08 PM -0800 1/5/06, Jeff Moore wrote: Has this been fixed and/or is there some safe set of pthreads calls? -- The strange thing is that even with a simpler version (without pthread_testcancel and with an empty ClockHandler) : I get strange CPU spike on Panther: average CPU is 0.5% then every 2 or 3 minutes i get a CPU spike up yo 80% during several seconds.... but the exact same code works without problems on Tiger (on the same machine) This email sent to site_archiver(a)lists.apple.com

1 0

Win32 cross-compiler for OSX?
by jmzorko＠mac.com 06 Jan '06

06 Jan '06

site_archiver(a)lists.apple.com Delivered-To: darwin-dev(a)lists.apple.com Hello, all ... Regards, John Falling You - exploring the beauty of voice and sound http://www.fallingyou.com _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… I would like to make Win32 executables from my PowerBook running Panther (Tiger eventually). I know I can do this from a Debian box using GNU gcc, but I was wondering if I can do it using Apple's gcc? This email sent to site_archiver(a)lists.apple.com

1 0

Re: Correct API to implement a timer in a real-time thread
by Steve Sisak 05 Jan '06

05 Jan '06

site_archiver(a)lists.apple.com Delivered-To: Darwin-dev(a)lists.apple.com Not in Darwin, but MPDelayUntil() should be fairly accurate. -Steve _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… We try also low-level function like mach_wait_until but had problems on Panther. What is the recommended way (API...) to implement a timer? (The other relevant calls being UpTime(), DurationToAbsolute() and AddAbsoluteToAbsolute() -- typing these from memory, YMMV) This email sent to site_archiver(a)lists.apple.com

1 0

Re: Correct API to implement a timer in a real-time thread
by Stéphane Letz 05 Jan '06

05 Jan '06

site_archiver(a)lists.apple.com Delivered-To: Darwin-dev(a)lists.apple.com Le 5 janv. 06 à 15:35, Christian Brunschen a écrit : On Thu, 5 Jan 2006, Stéphane Letz wrote: But that's easy - here's some pseudocode: time_t start = now(); time_t next = start; time_t interval = // ... while(true) { time_t now = now(); while (next < now) { next += interval; } sleep(next - now); // do something } time_t start = now(); time_t next = start; time_t interval = // ... while(true) { time_t now = now(); next += interval; time_t ahead = next - now; if (ahead > 0) { sleep(ahead); } // do something } Thanks Stephane _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: http://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.app… [ ... ] The thing is that when using this kind of functions, i have to deal with cases where an abnormal latency occur in the system, then the wait function returns too late, I have to compensate for that for the next interrupt and so on. This will attempt to 'do something' at fixed intervals. Even if there is latency or an unexpected delay at some point, the next time to wake up will still be "start + (n * interval)" for soem value of n. If the unexpected latency exceeds one interval, then an interval will be skipped. If you need to avoid skipping any invocations - in other words, if such a latency burst should not result in any invocations being skipped, but instead in perhaps severl invocations happening as quickly as possible until the system has caught up - then you can modify the pseudocode above: Yes, this is the case I'll need... and that is what I was more or less already doing... This email sent to site_archiver(a)lists.apple.com

1 0