Chris Etheridge Posted November 2, 2013 Share Posted November 2, 2013 So should I leave that one out of the mix, and just do ATISupport.kext and ATIFramebuffer.kext ? Or maybe rename ATI5000 to AMD5000 ?edit: Just noticed now that everything 'ATI' related has been renamed to AMD. Should I rename? Link to comment Share on other sites More sharing options...
spakk Posted November 2, 2013 Share Posted November 2, 2013 I found this here on http://forge.voodooprojects.org/p/chameleon/source/tree/4/branches/rekursor/i386/libsaio/pci_root.c /* 2 * Copyright 2009 netkas 3 */ 4 5 #include "libsaio.h" 6 #include "bootstruct.h" 7 8 #ifndef DEBUG_PCIROOT 9 #define DEBUG_PCIROOT 0 10 #endif 11 12 #if DEBUG_PCIROOT==1 13 #define DBG(x...) printf(x) 14 #else 15 #define DBG(x...) 16 #endif 17 18 int rootuid = 10; //value means function wasnt ran yet 19 20 unsigned int findrootuid(unsigned char * dsdt) 21 { 22 int i; 23 for (i=0; i<64; i++) //not far than 64 symbols from pci root 24 { 25 if(dsdt[i] == '_' && dsdt[i+1] == 'U' && dsdt[i+2] == 'I' && dsdt[i+3] == 'D' && dsdt[i+5] == 0x08) 26 { 27 return dsdt[i+4]; 28 } 29 } 30 printf("pci root uid not found\n"); 31 return 11; 32 } 33 34 unsigned int findpciroot(unsigned char * dsdt,int size) 35 { 36 int i; 37 for (i=0; i<size; i++) 38 { 39 if(dsdt[i] == 'P' && dsdt[i+1] == 'C' && dsdt[i+2] == 'I' && (dsdt[i+3] == 0x08 || dsdt [i+4] == 0x08)) 40 { 41 return findrootuid(dsdt+i); 42 } 43 } 44 45 printf("pci root not found\n"); 46 return 10; 47 } 48 49 50 /* Setup ACPI. Replace DSDT if DSDT.aml is found */ 51 int getPciRootUID() 52 { 53 int fd, version; 54 void *new_dsdt; 55 const char *dsdt_filename; 56 const char *val; 57 int user_uid_value; 58 char dirspec[512]; 59 int len,fsize; 60 61 if(rootuid < 10) return rootuid; 62 63 if (!getValueForKey("DSDT", &dsdt_filename, &len, &bootInfo->bootConfig)) 64 dsdt_filename="DSDT.aml"; 65 66 if (getValueForKey("-pci1", &val, &len, &bootInfo->bootConfig)) //fallback 67 { 68 user_uid_value = 1; 69 rootuid = user_uid_value; 70 return rootuid; 71 } 72 else user_uid_value = 0; 73 74 75 // Check booting partition 76 sprintf(dirspec,"%s",dsdt_filename); 77 fd=open (dirspec,0); 78 if (fd<0) 79 {// Check Extra on booting partition 80 sprintf(dirspec,"/Extra/%s",dsdt_filename); 81 fd=open (dirspec,0); 82 if (fd<0) 83 {// Fall back to booter partition 84 sprintf(dirspec,"bt(0,0)/Extra/%s",dsdt_filename); 85 fd=open (dirspec,0); 86 if (fd<0) 87 { 88 verbose("No DSDT found, using 0 as uid value.\n"); 89 rootuid = user_uid_value; 90 return rootuid; 91 } 92 } 93 } 94 95 // Load replacement DSDT 96 new_dsdt=(void*)MALLOC(file_size (fd)); 97 if (!new_dsdt) 98 { 99 printf("Couldn't allocate memory for DSDT\n"); 100 rootuid = user_uid_value; 101 return rootuid; 102 } 103 fsize = file_size(fd); 104 if (read (fd, new_dsdt, file_size (fd))!=file_size (fd)) 105 { 106 printf("Couldn't read file\n"); 107 rootuid = user_uid_value; 108 return rootuid; 109 } 110 close (fd); 111 rootuid=findpciroot(new_dsdt, fsize); 112 if(rootuid == 11)rootuid=0; //usualy when _UID isnt present, it means uid is zero 113 if(rootuid == 10) //algo failed, PCI0 wasnt found; 114 { 115 printf("pci root uid value wasnt found, using zero, if you want it to be 1, use -pci1 flag"); 116 rootuid = user_uid_value; 117 } 118 free(new_dsdt); 119 return rootuid; 120 } who has an idea? @Shaneee, the cause is the lack of DSDT.aml, that's it, Link to comment Share on other sites More sharing options...
Shaneee Posted November 2, 2013 Share Posted November 2, 2013 So should I leave that one out of the mix, and just do ATISupport.kext and ATIFramebuffer.kext ? Or maybe rename ATI5000 to AMD5000 ? Best bet is to back up all the AMD* and ATI* kexts on mavericks and try all the Lion ones. Link to comment Share on other sites More sharing options...
Diazz72 Posted November 2, 2013 Share Posted November 2, 2013 ok finaly i got in to OS Do only i have artifact when using web browser on HD 7870 or others too? Tnx evryone for help 1 Link to comment Share on other sites More sharing options...
ameris_cyning Posted November 2, 2013 Share Posted November 2, 2013 I must give credit where credit is due You guys are doing {censored} awesome We Intel brethren are happy to see our AMD brothers succeed 3 Link to comment Share on other sites More sharing options...
Chris Etheridge Posted November 2, 2013 Share Posted November 2, 2013 Getting stuck here now; http://s22.postimg.org/ts2thuwpd/image.jpg Link to comment Share on other sites More sharing options...
Shaneee Posted November 2, 2013 Share Posted November 2, 2013 Getting stuck here now; http://s22.postimg.org/ts2thuwpd/image.jpg That's because the graphics is not loading. Link to comment Share on other sites More sharing options...
Chris Etheridge Posted November 2, 2013 Share Posted November 2, 2013 That's because the graphics is not loading. Ah figures, hmm so kexts is a no go at all then Link to comment Share on other sites More sharing options...
tragediana150 Posted November 2, 2013 Share Posted November 2, 2013 For family 16 now fixed . Updated . mach_rc5_blacklist_fix.zip Ok, so now Launchpad is red again I do not see any difference between last rc5 kernel, for me it's running But!! Finally I can enter to the "system report", but maybe it's just a coincidence. Someone can tell me what's going on with the error in the "preview"? Because my "preview" works well (in terms of preview images). Hmm? Link to comment Share on other sites More sharing options...
Shaneee Posted November 2, 2013 Share Posted November 2, 2013 Ah figures, hmm so kexts is a no go at all then Didn't think it would work but was worth a try. Link to comment Share on other sites More sharing options...
Diazz72 Posted November 2, 2013 Share Posted November 2, 2013 essayes firefox Hmm, chrome is not working, but firefox is working, any idea??? Link to comment Share on other sites More sharing options...
Shaneee Posted November 2, 2013 Share Posted November 2, 2013 Hmm, chrome is not working, but firefox is working, any idea??? You need to disable hardware acceleration in chrome advanced settings. Link to comment Share on other sites More sharing options...
Diazz72 Posted November 2, 2013 Share Posted November 2, 2013 essayes AtiConfig=Hamashi GraphicsEnabler=No I am using clover. . .i'll try chameleon later Link to comment Share on other sites More sharing options...
Diazz72 Posted November 2, 2013 Share Posted November 2, 2013 FBConfig=Hamashi GraphicsInjector=No idem HD 7950 #3 nop...not working ;( EDIT: In chameleon i even can't boot, when i try my kext/drivers load. . .and afther that black screen. . .pc is not resetting, but black screen is here :& Link to comment Share on other sites More sharing options...
Diazz72 Posted November 3, 2013 Share Posted November 3, 2013 HD 4xxx good yepp...but i sold my hd 4890. . . ;( Link to comment Share on other sites More sharing options...
Sinetek Posted November 3, 2013 Share Posted November 3, 2013 what if we compiled XNU as an i386 binary and ran it with Lion's known-working kexts? also, Snow Leopard x86_64 worked without any graphics problems even on NVIDIA. Is this last statement correct? There might be a difference in some ACPI kext. 5 Link to comment Share on other sites More sharing options...
theconnactic Posted November 3, 2013 Author Share Posted November 3, 2013 Snow Leopard x86_64 worked without any graphics problems even on NVIDIA. Is this last statement correct? Yes, I used to run it with a G210. Quite unstable, but for reasons other than graphics. Link to comment Share on other sites More sharing options...
Sinetek Posted November 3, 2013 Share Posted November 3, 2013 So on AMD the blitting operations (ie anything with big images/pictures, and preview.app in general) seems to fail. Accelerated blitting operations. Would be cool to find a way to correctly replicate the bugs and program a small test app. 1 Link to comment Share on other sites More sharing options...
+ira Posted November 3, 2013 Share Posted November 3, 2013 what if we compiled XNU as an i386 binary and ran it with Lion's known-working kexts? also, Snow Leopard x86_64 worked without any graphics problems even on NVIDIA. Is this last statement correct? There might be a difference in some ACPI kext. My guess is that it depends on wich nVidia graphics card model is on target, in my case I only had QE/CI with i386 even on Snow Leopard. But i think we should try, I'll provide a SL installation for more detailed information... Cheers 1 Link to comment Share on other sites More sharing options...
Sinetek Posted November 3, 2013 Share Posted November 3, 2013 I'll Make a folder with images of the following types, and you guys tell which fail AI - Adobe Illustrator Artwork files BMP – Windows Bitmap files DNG – Digital Negative files DAE - Collada 3D files EPS – Encapsulated PostScript files (after an automatic conversion to PDF) FAX – faxes FPX – FlashPix files GIF – Graphics Interchange Format files HDR – High Dynamic Range Image files ICNS – Apple Icon Image files ICO – Windows icon files JPEG 2000 – JPEG 2000 files JPEG – Joint Photographic Experts Group files OpenEXR – OpenEXR files PS – Adobe PostScript files (after an automatic conversion to PDF) PSD – Adobe Photoshop files PICT – QuickDraw image files PDF – Portable Document Format files PNG – Portable Network Graphics files PNTG – MacPaint Bitmap Graphic files QTIF – QuickTime image files RAD – Radiance Scene Description files RAW – Raw image files SGI – Silicon Graphics Image files TGA – TARGA image files TIF, TIFF – Tagged Image File Format files XBM – X BitMap files PPT – PowerPoint files 1 Link to comment Share on other sites More sharing options...
Duran Keeley Posted November 3, 2013 Share Posted November 3, 2013 So on AMD the blitting operations (ie anything with big images/pictures, and preview.app in general) seems to fail. Accelerated blitting operations. Would be cool to find a way to correctly replicate the bugs and program a small test app. I can setup remote to my AMD if you need a test machine 1 Link to comment Share on other sites More sharing options...
Shaneee Posted November 3, 2013 Share Posted November 3, 2013 So on AMD the blitting operations (ie anything with big images/pictures, and preview.app in general) seems to fail. Accelerated blitting operations. Would be cool to find a way to correctly replicate the bugs and program a small test app. This is from a nvidia attempted boot if it helps any, http://pastebin.com/wDbstiM5 1 Link to comment Share on other sites More sharing options...
theconnactic Posted November 3, 2013 Author Share Posted November 3, 2013 The most baffling thing I can remember about the nVidia bug was the following behavior: whenever I started up Mountain Lion after previously restarting from i386 Lion, the screen would display a distorted image, or even quite clear pieces of, the Lion desktop and all open windows - like it was simply dumping onscreen the contents of VRAM. 1 Link to comment Share on other sites More sharing options...
theconnactic Posted November 3, 2013 Author Share Posted November 3, 2013 I'll Make a folder with images of the following types, and you guys tell which fail AI - Adobe Illustrator Artwork files BMP – Windows Bitmap files DNG – Digital Negative files DAE - Collada 3D files EPS – Encapsulated PostScript files (after an automatic conversion to PDF) FAX – faxes FPX – FlashPix files GIF – Graphics Interchange Format files HDR – High Dynamic Range Image files ICNS – Apple Icon Image files ICO – Windows icon files JPEG 2000 – JPEG 2000 files JPEG – Joint Photographic Experts Group files OpenEXR – OpenEXR files PS – Adobe PostScript files (after an automatic conversion to PDF) PSD – Adobe Photoshop files PICT – QuickDraw image files PDF – Portable Document Format files PNG – Portable Network Graphics files PNTG – MacPaint Bitmap Graphic files QTIF – QuickTime image files RAD – Radiance Scene Description files RAW – Raw image files SGI – Silicon Graphics Image files TGA – TARGA image files TIF, TIFF – Tagged Image File Format files XBM – X BitMap files PPT – PowerPoint files Okay. Will do. 2 Link to comment Share on other sites More sharing options...
Diazz72 Posted November 3, 2013 Share Posted November 3, 2013 I hope photoshop will work w/o acceleration Link to comment Share on other sites More sharing options...
Recommended Posts