chris1111 Posted March 2, 2015 Share Posted March 2, 2015 Sorry but you are totally wrong because I think you need and want decides what to do and where. (no????) Standard = install to target volume Install Chameleon in the ESP = install to the ESP of the parent disk of the target Volume (maybe only the title/description must be modified to something like "Install selected options in the ESP") If you want write something in the org.chameleon.Boot.plist (or install modules) in the target volume or in its ESP, without installing the bootloader, you have to select "Dont install the bootloader" = "ne pas installer le chargeur de ...", but always choosing where: ESP or target volume No?? EDIT same behavior for the Clover installer: clover.png and basically same identical description Yes, because there was only one target, now you can choose (anyway "Standard" = target volume is the default) EDIT II *if the disk of the selected Volume does not have any EFI partition (the case for MBR partition scheme), and you have selected "Install to ESP" the installer automatically redirect to the target volume. *@Ermac this behavior, if unwelcome, can be changed simply aborting the installation. oK thanks Micky1979 Now I understand 1 Link to comment Share on other sites More sharing options...
luxinhaus Posted March 3, 2015 Share Posted March 3, 2015 So the change I made in this TEST revision... failed So due to the big gap between the "x86" A.K.A. working version for 2248 and the actual 2588 is a lot of code to review... I need your help to know exactly witch was latest working version of Chameleon and witch version the problem starting from... Probably I do some error (I'm human after all ;-))..... After determining the "buggy" commit I/we can try to work on FIX it... This process need the ability to compile by yourself the source from the SVN A quick guide: Download and compile "old" revision (in this example 2248) Fireup your terminal and create a dir called 2248. . cds that folder: then svn co -r 2248 http://forge.voodooprojects.org/svn/chameleon/trunk/ cd trunk make distclean make your compiled binary are stored into sym/i386 subfolder install it manually or using Chameleon Wizard Let me know where the problem (revision) start for you. Cordially ErmaC Im on it ErmaC. have patience on me while i dig up the buggy commit. Thanks again 1 Link to comment Share on other sites More sharing options...
ErmaC Posted March 3, 2015 Author Share Posted March 3, 2015 Im on it ErmaC. have patience on me while i dig up the buggy commit. Thanks again Here a list of the "possible" commit (so don't need to download all the revision starting from 2248) 2248 OK 2253 (20 changes) 2254 (1 change) 2255 (35 changes but safe... po file update not chameleon core) 2257 (1 change safe) 2258 (1 change) 2259 (42 changes) 2260 (2 changes safe) 2261 (1 change safe) 2263 (13 changes) 2265 (1 change) 2266 (11 changes) 2268 (4 changes safe) 2269 (12 changes) 2281 (6 changes) 2282 (1 change safe) 2283 (1 change safe) 2284 (4 changes) 2285 (9 changes) 2286 (32 changes safe) 2288 (6 changes) 2289 (1 change) 2290 (15 changes safe) 2318 (27 changes) 2325 (5 changes) 2327 (53 changes) The "red" commit is where "probably" the problem is located The "green" are safe because are only indent translation file or typo the commit in the middle (example 2249, 2250, 2251, 2252, 2253) you can skip it because are not a trunk changes. ErmaC 1 Link to comment Share on other sites More sharing options...
luxinhaus Posted March 3, 2015 Share Posted March 3, 2015 @ErmaC Last working version was 2334 1 Link to comment Share on other sites More sharing options...
ErmaC Posted March 3, 2015 Author Share Posted March 3, 2015 @ErmaC Last working version was 2334 THX So one more test I compile the 2334 (just fix some compile errors due to old xcode version of that commit) and also the 2342 where probably the error "live" So: 2334 ->2334_i386.zip should be ok 2342 ->2342_i386.zip don't work...? Let me know ErmaC 1 Link to comment Share on other sites More sharing options...
Micky1979 Posted March 3, 2015 Share Posted March 3, 2015 Sorry Fabio , any test by me would be useless since all the revisions are working for me on all PCs I have (3) Link to comment Share on other sites More sharing options...
luxinhaus Posted March 3, 2015 Share Posted March 3, 2015 THX So one more test I compile the 2334 (just fix some compile errors due to old xcode version of that commit) and also the 2342 where probably the error "live" So: 2334 ->2334_i386.zip should be ok 2342 ->2342_i386.zip don't work...? Let me know ErmaC Your 2342 works, but cant verify coz I cant compile 2342 from the trunk due to error though, so I cant compare Il take a short break and be back tomorrow Thank You Link to comment Share on other sites More sharing options...
ErmaC Posted March 3, 2015 Author Share Posted March 3, 2015 Your 2342 works, but cant verify coz I cant compile 2342 from the trunk due to error though, so I cant compare Il take a short break and be back tomorrow Thank You Ok tonight I will post the next 3 (precomplied) revision... I'm busy now... ErmaC Link to comment Share on other sites More sharing options...
luxinhaus Posted March 3, 2015 Share Posted March 3, 2015 @ErmaC An update: 2342 = Works 2343 - 2410 = I can't compile due to compiling error 2411 - Latest = FileEFINVRAM KP on x32 machine Therefore, the issue lives somewhere between 2343 - 2410 revision Error I'm getting during compiling using different version of Xcode, 4.2 / 4.6.3 / 5.1.1 How to fix this so i can help you narrow down the problem stringTable.c:43:1: error: unused function 'keyncmp' [-Werror,-Wunused-function] keyncmp(const char *str, const char *key, int n) ^ 1 error generated. make[3]: *** [/Users/Batman/Desktop/2343/obj/i386/libsaio/stringTable.o] Error 1 make[2]: *** [all-recursive] Error 1 make[1]: *** [all-recursive] Error 1 make: *** [all] Error 2 Link to comment Share on other sites More sharing options...
ochu Posted March 3, 2015 Share Posted March 3, 2015 (edited) "Can be the opposite??? this is the part of code in the script that install the stage2 (boot) in Chameleon/Enoch: # Writing stage 2 WRITE_STAGE2() { if [ $InstallBootloader = "1" ];then echo -e "${mainLine}\nWRITING STAGE 2:" cp -R "${i386Dir}/${stage2Loader}" "${choicedVolume}/" <-- copy and replace any existing "boot" file (see man cp) chflags hidden "${choicedVolume}/${stage2Loader}" <-- set the file hidden, but sure it can't fail (see man chflags) echo "Stage 2 (boot) writed to ${choicedVolume}." fi } chflags is a command that change the "file flags" and "hidden" option is defined in the man page as "hidden set the hidden flag [Hide item from GUI]". Probably Chameleon Wizard live the boot file visible, instead the official Installer hide it. Is not a bug, is wanted!" -------------- mm..very strange. I use the installers and the boot file appears at the root. Finishing a Chameleon Wizard install and the boot file reverts to hide state. (Darwin/x86 boot v5.0.132 - Chameleon v2.3svn r2538) Edited March 3, 2015 by ochu Link to comment Share on other sites More sharing options...
Micky1979 Posted March 3, 2015 Share Posted March 3, 2015 Just tested, but the boot file get hidden. Your are keeping the Volume opened at the Finder during the installation? If the Volume is opened in the Finder need to be refreshed. Have you checked it after a reboot? Anyway running this in the terminal: ls -lO /boot give me this result: -rw-r--r--@ 1 root wheel hidden 359936 Mar 3 17:38 /boot can you test this please after the installation? EDIT mm..very strange. I use the installers and the boot file appears at the root. Finishing a Chameleon Wizard install and the boot file reverts to hide state. (Darwin/x86 boot v5.0.132 - Chameleon v2.3svn r2538) Just to be sure only latest installer of Enoch/Chameleon hide the boot file since: [2569] Update Chameleon pkg Installer (Allow install on ESP). (Credits to Micky1979) running older installers this was not possible...if this is the case (the r2538 mentioned by you)...the case is solved, otherwise your chflags command have a bug (mmmh)! Anyway this is not a problem for the bootloader or its operation. Link to comment Share on other sites More sharing options...
ErmaC Posted March 4, 2015 Author Share Posted March 4, 2015 @ErmaC An update: 2342 = Works 2343 - 2410 = I can't compile due to compiling error 2411 - Latest = FileEFINVRAM KP on x32 machine Therefore, the issue lives somewhere between 2343 - 2410 revision How to fix this so i can help you narrow down the problem Here the diff for fix compile error (to use the patch/diff... cd the rev dir ex: ermac@Insanely:~/Documents/chammy/2399$ ls trunk ermac@Insanely:~/Documents/chammy/2399$ cd trunk/ ermac@Insanely:~/Documents/chammy/2399/trunk$ patch -p1 < fix-compile2399.txt patching file i386/libsaio/fake_efi.c patching file i386/libsaio/stringTable.c patching file i386/libsaio/vbe.c patching file i386/modules/AcpiCodec/acpi_codec.c patching file i386/modules/MakeInc.dir You can revert the changes with patch -p1 -R < fix-compile2399.txt ) 2343-> fix-compile2343.txt ErmaC 1 Link to comment Share on other sites More sharing options...
luxinhaus Posted March 4, 2015 Share Posted March 4, 2015 @ErmaC Just an update from above diff... 2343-2371 = works So we are down to 2381 - 2410 to pinpoint the issue Il be back. more compiling 1 Link to comment Share on other sites More sharing options...
ErmaC Posted March 4, 2015 Author Share Posted March 4, 2015 @ErmaC Just an update from above diff... 2343-2371 = works So we are down to 2381 - 2410 to pinpoint the issue Il be back. more compiling OK let me know... Now I'm tired... sleep time... ErmaC 1 Link to comment Share on other sites More sharing options...
luxinhaus Posted March 4, 2015 Share Posted March 4, 2015 (edited) @ErmaC Lo and Behold…. 2382 was the culprit To make sure, i used fake_efi.c from 2381 and that fixed the issue. Can you show me where to edit the diff please. there are only 13 differences but cant nail it lol diff.zip Edited March 4, 2015 by luxinhaus 1 Link to comment Share on other sites More sharing options...
ErmaC Posted March 4, 2015 Author Share Posted March 4, 2015 @ErmaC Lo and Behold…. 2382 was the culprit To make sure, i used fake_efi.c from 2381 and that fixed the issue. Can you show me where to edit the diff please. there are only 13 differences but cant nail it lol Revision? 2382? Link to comment Share on other sites More sharing options...
luxinhaus Posted March 4, 2015 Share Posted March 4, 2015 @ErmaC Yes, it was revision 2382 Update.. finally solved it!!!! got the correct spot where the issue is coming from. this the the diff after I applied your fix-compile2382.txt. FixedDiff.zip 1 Link to comment Share on other sites More sharing options...
ErmaC Posted March 4, 2015 Author Share Posted March 4, 2015 @ErmaC Yes, it was revision 2382 Update.. finally solved it!!!! got the correct spot where the issue is coming from. this the the diff after I applied your fix-compile2382.txt. Good! Now from what version are you able to compile by yourself from the SVN with out fix compiling error? 2410? 2411? I'm not remember when we have fix the compiling error (I mean when the commit was made). try to apply this tiny diff (fake_efi.c) to a more recent revision (ex:2410) ErmaC 1 Link to comment Share on other sites More sharing options...
luxinhaus Posted March 4, 2015 Share Posted March 4, 2015 @ErmaC I need your YOUR patch starting from revision 2343 up to 2410 else compiling error. I applied MY patch starting from revision 2382 up to latest (2584) else KP on x32 machine. Next is to help you fix "AllocateKernelMemory error" that happened "after" revision 2510 Link to comment Share on other sites More sharing options...
ErmaC Posted March 4, 2015 Author Share Posted March 4, 2015 @ErmaC I need your YOUR patch starting from revision 2343 up to 2410 else compiling error. I applied MY patch starting from revision 2382 up to latest (2584) else KP on x32 machine. Next is to help you fix "AllocateKernelMemory error" that happened "after" revision 2510 I not completly understand what you mean with after 2510... You mean 2509 downloaded and compiled from the SVN is not affected by this issue??? I re-made the fix compiling error and add also the changes you propose for the fake_efi.c 2412-> fix-compile2412.txt ErmaC Link to comment Share on other sites More sharing options...
luxinhaus Posted March 4, 2015 Share Posted March 4, 2015 No. Everything is Ok now. What i mean by "after 2510" was another issue "AllocateKernelMemory error" by a user posted here Link to comment Share on other sites More sharing options...
luxinhaus Posted March 4, 2015 Share Posted March 4, 2015 2412 work? No KP on x32 machine? 2412 works. no more KP on x32 machine after applying the new fix-compile2412.txt Even 2588 now works after applying fake_efi.c fix. So, i can say that the issue on x32 has been solved 1 Link to comment Share on other sites More sharing options...
ErmaC Posted March 4, 2015 Author Share Posted March 4, 2015 So, i can say that the issue on x32 has been solved Just to be absolutly sure... 2584-> fix-compile2584.txt 2584 patched binaries -> Patched_i386_2584.zip ErmaC 1 Link to comment Share on other sites More sharing options...
luxinhaus Posted March 4, 2015 Share Posted March 4, 2015 @ErmaC I got KP on x32 with your Patched_i386_2584.zip. Your fix-compile2584.txt is wrong. compare yours to mine and youl see what went wrong with your diff. Or unless you really intended that to make sure if i will have the KP 1 Link to comment Share on other sites More sharing options...
ErmaC Posted March 4, 2015 Author Share Posted March 4, 2015 @ErmaC I got KP on x32 with your Patched_i386_2584.zip. Your fix-compile2584.txt is wrong. compare yours to mine and youl see what went wrong with your diff. Or unless you really intended that to make sure if i will have the KP Sorry I mix up the changes... And no is not intentional... This... I'm 100% are the good one... feedback pls! (also post your bdmesg) 2584-> fix-compile2584_V2.txt 2584 patched binaries -> Patched_i386_2584_V2.zip ErmaC 1 Link to comment Share on other sites More sharing options...
Recommended Posts