Bit Shoveler

Installed Enoch 2773 just now on my ancient Core 2 Quad Hack Pro, still running 10.10.5. No problems observed. CPU temps seem a degree or two higher, but if it's real it's not enough to worry about. Thanks to all the developers and contributors!

Thanks for your continued efforts! I'm looking forward to installing El Capitan on my old Core 2 Quad Hack Pro. Once all the source changes find their way into the trunk, I'll crank up the static analysis tools and report back.

Mr___G, I was getting that on the nVidia driver, til I switched back to the Apple driver. I wasn't seeing any performance benefit with the nVidia driver, so no great loss.

Very interesting. I worked on an open source Mac driver for these chips a few years ago. I'll give this a spin on my new Yosemite installation and see how it goes. Thank you Mieze!

I've updated my branch from the trunk (including all the updates ErmaC mentions) and checked in a bunch of changes, mostly little fixes to problems exposed by automated source code analysis. There are no new features, just fixes. I'm booted from it as I write. You can download the installer and MD5 signature from my own website. Use at your own risk, but please report any problems. Many thanks to the entire team. I hope my contributions are useful.

WIth Yosemite out, I'm going through the codebase again with the Clang static analyzer and Cppcheck, and I'll have patches for review in a couple of days. Most of what I've found so far is problems in error handling. So by the time the bug is triggered, something was already broken (usually memory allocation). But fixing them can mean the difference between an error message telling you what happened and a freeze. My changes will be in SVN under branches/chucko . Feel free to try them out, I have a pretty ancient Hack and many of the newer features aren't applicable to my machine.

Go into the source file and comment out the unused routine. You'll find a couple of them.

I finally found a workaround to my linker woes of a couple months ago: remove the '-x' option from the offending 'ld' command. Working under latest XCode on 10.9.4. 'svn diff' below shows the change. hackintosh_eth0:trunk chucko$ svn diff i386/modules/MakeInc.dir Index: i386/modules/MakeInc.dir =================================================================== --- i386/modules/MakeInc.dir (revision 2390) +++ i386/modules/MakeInc.dir (working copy) @@ -126,7 +126,7 @@ @ld -arch i386 \ -alias _$(MODULE_START) start \ -dylib -read_only_relocs suppress \ - -S -x -Z -dead_strip_dylibs \ + -S -Z -dead_strip_dylibs \ -no_uuid \ -current_version $(MODULE_VERSION) -compatibility_version $(MODULE_COMPAT_VERSION) \ -final_output $(MODULE_NAME) \ hackintosh_eth0:trunk chucko$

Well, grump. I hadn't tried to build Chameleon in a while and updated from the SVN repo. First the build barfed on unused functions. So I #ifdef'd them out. Then the linker blew up in uClibcxx. Turns out the problem is Apple's custom version of ld. If code is compiled with Link Time Optimization, and the linker is set to do dead code stripping, the linker fails. Apple is aware of the problem but doesn't seem to have a fix yet. If you have the latest XCode 5.1 installed on Mavericks 10.9.2, how do you build Chameleon?!

Latest version (6.0.1123, downloaded automatically by HWMonitor app) asks for password to decode. I canceled out of password dialog and tried to install anyway. Much trouble ensued. Finally was able to reinstall 6.0.1115.

Is this a new problem? I've been doing some work on the XML parser code. It's possible I introduced this. I will take a look when I have time, but it won't be for a couple of days. Can you post a sample file that provokes the problem?

Anybody have a hint at all?

I recently installed Mavericks on the old Core 2 Quad Hack in my signature (had been running Lion). When I did, I cleaned up a bunch of kexts that didn't seem to be needed any more. One of these was OpenHaltRestart.kext. Since then, sleep works again (yay!!), and restart is just fine, but shutting down is inconsistent. Sometimes it works right away, sometimes it takes about 8 seconds after the video goes off, sometimes 25 seconds… and sometimes it just hangs with the power on indefinitely. Usually if I hit the reset button and reboot all the way back to Mavericks, the second shutdown goes better, but this morning it took 2 resets. I already have a custom DSDT. I confess I'm using Chimera as a bootloader, but I'm now on the Chameleon dev team, mostly working cleanup items, so that will change in the next couple days. Is there a way to make shutdown work without using a hacky kext? (Is this in a FAQ or wiki somewhere?) I can provide more info, e.g. DSDT source and DarwinDumper info, later tonight when I get home from work. Thanks.

As the author of RealtekR1000SL.kext, I can assure you it does nothing at all that should affect the hard drives' icons. Mine are still orange.

I have been too busy to update it since the last binary release. Those sources should correspond exactly to the last kext. I do plan to keep working on it when I find time.