ErmaC Posted January 28, 2014 Author Share Posted January 28, 2014 Can we get some information to know what we need to focus for in this new test version?In the revision 2281 was introduced a boolean keyFlag to change the size of the table 17 See this closed issue http://forge.voodooprojects.org/p/chameleon/issues/361/ So one of the "focus" will be to know if this issue is also fixed w/o that key (of course who was affected) Other "focus": know if the merged code cause issue.. System Profiler show correct value? bdmesg still show empty field) ErmaC Link to comment Share on other sites More sharing options...
nyolc8 Posted January 28, 2014 Share Posted January 28, 2014 Processor table with r2346 test: Type: 4, Length: 35, Handle: 0x0004 Processor Information: Socket Designation: LGA775 Type: Central Processor Family: 0x1 Manufacturer: Intel® Corporation ID: 0x6FB Processor Version: Intel® Core2 Duo CPU External Clock: 396MHz Maximum Clock: 3168MHz Current Clock: 3168MHz Upgrade: Socket 478 Serial Number: ** PRIVATE ** Asset Tag: Intel® Core2Duo E6750 CPU @ 3.15GHz Part Number: Not Specified Processor table with r2354 test: Type: 4, Length: 38, Handle: 0x0004 Processor Information: Socket Designation: LGA775 Type: Central Processor Family: 0x1 Manufacturer: Intel® Corporation ID: 0x6FB Processor Version: Intel® Core2 Duo CPU External Clock: 396MHz Maximal Clock: 3168MHz Current Clock: 3168MHz Upgrade: Socket 478 Serial Number: ** PRIVATE ** Asset Tag: Intel® Core2Duo E6750 CPU @ 3.15GHz Part Number: Core Count: 0 Core Enabled: 0 Thread Count: 0 So with this latest test version, I'm getting these strange empty "core" lines and "part number" is empty again. Other tables are the same as with r2346. Is this because your version didn't get merged with Bungo's branch? Or another bug? Link to comment Share on other sites More sharing options...
ErmaC Posted January 28, 2014 Author Share Posted January 28, 2014 I'm getting these strange empty "core" lines and "part number" is empty again. Is this because your version didn't get merged with Bungo's branch? Or another bug? Try this… (In the previous was uncommented experimental field) I hope now is ok. (this is the reason why we need test and tester) 2354(TEST 2) --> also nyolc8 can you provide the info for your CPU (the first part of the bdmesg) from he previous test revision? I would like to link the "experimental empty field in the correct way to show the proper info. what I mean is --------------------------------------------- --------------- CPU INFO -------------------- --------------------------------------------- Brand String: Intel® Core i7 CPU 920 @ 2.67GHz Vendor: 0x756e6547 Family: 0x6 ExtFamily: 0x0 Signature: 106a5 Processor type: Intel Original OEM Processor Model: 0x1a ExtModel: 0x1 Stepping: 0x5 MaxCoef/CurrCoef: 0x0/0x14 MaxDiv/CurrDiv: 0x0/0x0 TSCFreq: 2672MHz FSBFreq: 134MHz CPUFreq: 2672MHz Cores: 4 Logical processor: 8 Features: 0x000002ff --------------------------------------------- Type: 4, Length: 38, Handle: 0x0004 Processor Information: Socket Designation: LGA1366 Type: Central Processor Family: 0xC6 Manufacturer: Intel® Corporation ID: 0x106A5 Processor Version: Intel® Core i7 CPU 920 @ 2.67GHz External Clock: 133MHz Maximal Clock: 2672MHz Current Clock: 2666MHz Upgrade: Other Serial Number: ** PRIVATE ** Asset Tag: To Be Filled By O.E.M. Part Number: To Be Filled By O.E.M. Core Count: 4 Core Enabled: 4 Thread Count: 8 ErmaC Link to comment Share on other sites More sharing options...
Gringo Vermelho Posted January 28, 2014 Share Posted January 28, 2014 It worked for me - bdmesg outputs from r2352 and r2354 test 1: bdmesg r2352 + r2354.zip I like how the output isn't crowded with empty tables anymore. Here's bdmesg from test2: r2354_2.txt I didn't look too closely but I think nothing changed. For reference, here's my current smbios.plist: Gringo_smbios.plist.zip I don't understand why I'm getting Core Enabled: 1 though, but this is not new. The CPU history window in Activity Monitor shows activity on all four cores. 1 Link to comment Share on other sites More sharing options...
nyolc8 Posted January 28, 2014 Share Posted January 28, 2014 Try this… (In the previous was uncommented experimental field) I hope now is ok. (this is the reason why we need test and tester) 2354(TEST 2) --> i386_v2354-TEST.zip also nyolc8 can you provide the info for your CPU (the first part of the bdmesg) from he previous test revision? I would like to link the "experimental empty field in the correct way to show the proper info. what I mean is ErmaC CPU + processor table from v2354 test1: Chameleon 2.2svn (svn-r2354) [2014-01-28 17:05:00] msr(485): ia32_perf_stat 0x0600082a max: 8 current: 8 --------------------------------------------- --------------- CPU INFO ------------------ --------------------------------------------- CPU: Brand String: Intel® Core2 Duo CPU E6750 @ 2.66GHz CPU: Vendor/Family/ExtFamily: 0x756e6547/0x6/0x0 CPU: Model/ExtModel/Stepping: 0xf/0x0/0xb CPU: MaxCoef/CurrCoef: 0x8/0x8 CPU: MaxDiv/CurrDiv: 0x0/0x0 CPU: TSCFreq: 3168MHz CPU: FSBFreq: 396MHz CPU: CPUFreq: 3168MHz CPU: NoCores/NoThreads: 2/2 CPU: Features: 0x0000024f --------------------------------------------- ... ... ... Type: 4, Length: 38, Handle: 0x0004 Processor Information: Socket Designation: LGA775 Type: Central Processor Family: 0x1 Manufacturer: Intel® Corporation ID: 0x6FB Processor Version: Intel® Core2 Duo CPU External Clock: 396MHz Maximal Clock: 3168MHz Current Clock: 3168MHz Upgrade: Socket 478 Serial Number: ** PRIVATE ** Asset Tag: Intel® Core2Duo E6750 CPU @ 3.15GHz Part Number: Core Count: 0 Core Enabled: 0 Thread Count: 0 CPU + processor table from v2354 test2: Chameleon 2.2svn (svn-r2354) [2014-01-28 17:05:00] msr(485): ia32_perf_stat 0x0600082a max: 8 current: 8 --------------------------------------------- --------------- CPU INFO ------------------ --------------------------------------------- CPU: Brand String: Intel® Core2 Duo CPU E6750 @ 2.66GHz CPU: Vendor/Family/ExtFamily: 0x756e6547/0x6/0x0 CPU: Model/ExtModel/Stepping: 0xf/0x0/0xb CPU: MaxCoef/CurrCoef: 0x8/0x8 CPU: MaxDiv/CurrDiv: 0x0/0x0 CPU: TSCFreq: 3168MHz CPU: FSBFreq: 396MHz CPU: CPUFreq: 3168MHz CPU: NoCores/NoThreads: 2/2 CPU: Features: 0x0000024f --------------------------------------------- ... ... ... Type: 4, Length: 38, Handle: 0x0004 Processor Information: Socket Designation: LGA775 Type: Central Processor Family: 0x1 Manufacturer: Intel® Corporation ID: 0x6FB Processor Version: Intel® Core2 Duo CPU External Clock: 396MHz Maximal Clock: 3168MHz Current Clock: 3168MHz Upgrade: Socket 478 Serial Number: ** PRIVATE ** Asset Tag: Intel® Core2Duo E6750 CPU @ 3.15GHz Part Number: Core Count: 0 Core Enabled: 0 Thread Count: 0 Same problem Also the "part number" still empty. (Btw how the date of the chameleon can be the same for both test versions? Is that normal?) Link to comment Share on other sites More sharing options...
ErmaC Posted January 28, 2014 Author Share Posted January 28, 2014 I didn't look too closely but I think nothing changed. Yep is ok... I think probably due to my setting (set as MacPro so the board type is different) the "experimental part work for me as you see in the previous post ( 1 CPU, 4 core, 8 thread) but looking at your log… some not usual data... CPU: CPUFreq: 3410MHz CPU: NoCores/NoThreads: 4/4 CPU: Features: 0x0000027f ---- Core Count: 4 Core Enabled: 1 Thread Count: 0 as I said those "data are experimental and don't mind nothing (ATM) TEST 3 i386_v2354-TEST3.zip @Gringo… I think also the CPU detection need a huge update... ErmaC Link to comment Share on other sites More sharing options...
Gringo Vermelho Posted January 28, 2014 Share Posted January 28, 2014 4/4 is accurate for an i5-3570K but only 1 core enabled is weird. I suppose it could also be a BIOS bug? bdmesg test 3 2354_3.txt core count/enabled now gone. (Btw how the date of the chameleon can be the same for both test versions? Is that normal?) The date is the same because they were built on the same day. Like two text documents that you saved the same day.. no difference. Link to comment Share on other sites More sharing options...
nyolc8 Posted January 28, 2014 Share Posted January 28, 2014 TEST3 result: Chameleon 2.2svn (svn-r2354_Test3) [2014-01-28 18:39:18] msr(485): ia32_perf_stat 0x0600082a max: 8 current: 8 CPU: Brand String: Intel® Core2 Duo CPU E6750 @ 2.66GHz CPU: Vendor/Family/ExtFamily: 0x756e6547/0x6/0x0 CPU: Model/ExtModel/Stepping: 0xf/0x0/0xb CPU: MaxCoef/CurrCoef: 0x8/0x8 CPU: MaxDiv/CurrDiv: 0x0/0x0 CPU: TSCFreq: 3168MHz CPU: FSBFreq: 396MHz CPU: CPUFreq: 3168MHz CPU: NoCores/NoThreads: 2/2 CPU: Features: 0x0000024f ... ... ... Type: 4, Length: 35, Handle: 0x0004 Processor Information: Socket Designation: LGA775 Type: Central Processor Family: 0x1 Manufacturer: Intel® Corporation ID: 0x6FB Processor Version: Intel® Core2 Duo CPU External Clock: 396MHz Maximal Clock: 3168MHz Current Clock: 3168MHz Upgrade: Socket 478 Serial Number: ** PRIVATE ** Asset Tag: Intel® Core2Duo E6750 CPU @ 3.15GHz Part Number: "Core" lines disappeared, "part number" still empty. Link to comment Share on other sites More sharing options...
ErmaC Posted January 28, 2014 Author Share Posted January 28, 2014 TEST3 result: "part number" still empty. Very strange... This is the Bungo's Branch compiled (commited today) --> Bungo-Branch_v2353.zip (I don't understand what is missed from the code I merged)) Fabio Link to comment Share on other sites More sharing options...
nyolc8 Posted January 28, 2014 Share Posted January 28, 2014 Bungo branch v2353: Chameleon 2.2svn (Bungo branch) r2353 [2014-01-28 19:03:51] msr(497): ia32_perf_stat 0x0600082a max: 8 current: 8 CPU: Brand String: Intel® Core2 Duo CPU E6750 @ 2.66GHz CPU: Vendor/Family/ExtFamily: 0x756e6547/0x6/0x0 CPU: Model/ExtModel/Stepping: 0xf/0x0/0xb CPU: MaxCoef/CurrCoef: 0x8/0x8 CPU: MaxDiv/CurrDiv: 0x0/0x0 CPU: TSCFreq: 3168MHz CPU: FSBFreq: 396MHz CPU: CPUFreq: 3168MHz CPU: NoCores/NoThreads: 2/2 CPU: Features: 0x0000024f ... ... ... Type: 4, Length: 35, Handle: 0x0004 Processor Information: Socket Designation: LGA775 Type: Central Processor Family: 0x1 Manufacturer: Intel® Corporation ID: 0x6FB Processor Version: Intel® Core2 Duo CPU External Clock: 396MHz Maximal Clock: 3168MHz Current Clock: 3168MHz Upgrade: Socket 478 Serial Number: ** PRIVATE ** Asset Tag: Intel® Core2Duo E6750 CPU @ 3.15GHz Part Number: "Part Number" filled with "Not specified" when booting with Bungo's v2346. So something got broke since then. Link to comment Share on other sites More sharing options...
Gringo Vermelho Posted January 28, 2014 Share Posted January 28, 2014 It works for me, a blank string value injects "not specified". SMcpupartnumber Bungo r2353.txt Link to comment Share on other sites More sharing options...
ErmaC Posted January 28, 2014 Author Share Posted January 28, 2014 Bungo branch v2353: "Part Number" filled with "Not specified" when booting with Bungo's v2346. So something got broke since then. ? I don't see it… (I mean in your quote) @Gringo yep could be a "bug" in the BIOS... I try to explain... this data smbios.h /* ============================ Processor Information (Type 4) ============================== */ typedef struct SMBProcessorInformation { // 2.0+ spec (26 bytes) SMB_STRUCT_HEADER // Type 4 SMBString socketDesignation; SMBByte processorType; // CPU = 3 SMBByte processorFamily; // processor family enum SMBString manufacturer; SMBQWord processorID; // based on CPUID SMBString processorVersion; SMBByte voltage; // bit7 cleared indicate legacy mode SMBWord externalClock; // external clock in MHz SMBWord maximumClock; // max internal clock in MHz SMBWord currentClock; // current internal clock in MHz SMBByte status; SMBByte processorUpgrade; // processor upgrade enum // 2.1+ spec (32 bytes) SMBWord L1CacheHandle; SMBWord L2CacheHandle; SMBWord L3CacheHandle; // 2.3+ spec (35 bytes) SMBString serialNumber; SMBString assetTag; SMBString partNumber; // 2.5+ spec (40 bytes) // SMBByte coreCount; // SMBByte coreEnabled; // SMBByte threadCount; // SMBWord processorFuncSupport; // 2.6+ spec (42 bytes) // SMBWord processorFamily2; } __attribute__((packed)) SMBProcessorInformation; provide info for the dmi table n4 and coming from the bios the other info CPU: Brand String: Intel® Core i5-3570K CPU @ 3.40GHz CPU: Vendor/Family/ExtFamily: 0x756e6547/0x6/0x0 CPU: Model/ExtModel/Stepping: 0x3a/0x3/0x9 CPU: MaxCoef/CurrCoef: 0x0/0x22 CPU: MaxDiv/CurrDiv: 0x0/0x0 CPU: TSCFreq: 3410MHz CPU: FSBFreq: 100MHz CPU: CPUFreq: 3410MHz CPU: NoCores/NoThreads: 4/4 CPU: Features: 0x0000027f are catch from a cpu scan (so basically 2 way to provide the same info) Fabio 1 Link to comment Share on other sites More sharing options...
nyolc8 Posted January 28, 2014 Share Posted January 28, 2014 ? I don't see it… (I mean in your quote) ... Fabio You can see that in this post I made before: http://www.insanelymac.com/forum/topic/231075-chameleon-22svn-official-pkg-installer-binaries/?p=1989468 Link to comment Share on other sites More sharing options...
Gringo Vermelho Posted January 28, 2014 Share Posted January 28, 2014 @Gringo yep could be a "bug" in the BIOS... It is, look: Gringo_vanilla_smbios.txt I did that just the other day and I already forgot lol..... (latest dmidump, Linux Mint Live DVD) Link to comment Share on other sites More sharing options...
ErmaC Posted January 28, 2014 Author Share Posted January 28, 2014 You can see that in this post I made before: http://www.insanelymac.com/forum/topic/231075-chameleon-22svn-official-pkg-installer-binaries/?p=1989468 OK I see it now… I continue investigating… I'm off now It is, look: Gringo_vanilla_smbios.txt I did that just the other day and I already forgot lol..... (latest dmidump, Linux Mint Live DVD) yepx2 Serial Number: To Be Filled By O.E.M. Asset Tag: To Be Filled By O.E.M. Part Number: To Be Filled By O.E.M. Core Count: 4 Core Enabled: 1 Characteristics: 64-bit capable Handle 0x0005, DMI type 7, 19 bytes So this confirm is some "strange" things coming from your bios ErmaC 1 Link to comment Share on other sites More sharing options...
Bungo Posted January 28, 2014 Share Posted January 28, 2014 Bungo branch v2353: "Part Number" filled with "Not specified" when booting with Bungo's v2346. So something got broke since then. I need your original and patched smbios to compare Link to comment Share on other sites More sharing options...
ErmaC Posted January 29, 2014 Author Share Posted January 29, 2014 I complete and merge into main trunk code a new flag for nvidia users (commit 2359) since the revision 1991 we have added the ability to assign a specific name at our Nvidia card Example: Asus GeForce GT 240 this is also displayed in the system profiler Now If there is some "nostalgic" or simply prefer the "classic" name showed in System Profiler add NvidiaGeneric=Yes in org.chameleon.Boot.plist Result: GeForce GT 240 ErmaC Link to comment Share on other sites More sharing options...
Gringo Vermelho Posted January 29, 2014 Share Posted January 29, 2014 If you can add this for ATI cards too I know someone who will be very happy. Also who is the guy with the amazing sig,, "awesome CPU, million dollar video card, tons of RAM..etc"? Now he can finally have About This Mac say "million dollar video card" Link to comment Share on other sites More sharing options...
Bungo Posted January 29, 2014 Share Posted January 29, 2014 Bungo branch v2353: "Part Number" filled with "Not specified" when booting with Bungo's v2346. So something got broke since then. Check spaces (" ") if "blank" entry (as described before). As reference use dmidecode because it's independent tool. Link to comment Share on other sites More sharing options...
nyolc8 Posted January 29, 2014 Share Posted January 29, 2014 Check spaces (" ") if "blank" entry (as described before). As reference use dmidecode because it's independent tool. Where should I check for spaces? In smbios.plist? Because I don't have (and never had but I got the "not specified") any SMcpupartnumber line there. Should I make one there? Also how can I get my original smbios? I can't find any app for Windows to do an smbios dump properly. edit: I just saw Gringo's #2026 post, I'll put that line into smbios.plist and report back. edit 2: adding SMcpupartnumber into smbios.plist fixed the empty part number line problem. 1 Link to comment Share on other sites More sharing options...
ErmaC Posted January 29, 2014 Author Share Posted January 29, 2014 edit 2: adding SMcpupartnumber into smbios.plist fixed the empty part number line problem. Finally! So we can confirm the correct implementation? PLS TEST: Chameleon rev 2360_TEST1 --> i386_v2360-TEST1.zip This will be merged into main code. Is the latest trunk 2359 + Bungo's smbios patch + some minor changes for the output (like dmidecode… inspired on Gringo's dump from linux provided yesterday). ErmaC Link to comment Share on other sites More sharing options...
nyolc8 Posted January 29, 2014 Share Posted January 29, 2014 v2360 test1: Handle: 0x0004, DMI type: 4, 38 bytes[/size] Processor Information Socket Designation: LGA775 Type: Central Processor Family: 0x1 Manufacturer: Intel® Corporation ID: 0x6FB Version: Intel® Core2 Duo CPU External Clock: 396 MHz Max Speed: 3168 MHz Current Speed: 3168 MHz Upgrade: Socket 478 Serial Number: ** PRIVATE ** Asset Tag: Intel® Core2Duo E6750 CPU @ 3.15GHz Part Number: Not Specified Everything looks OK here. 1 Link to comment Share on other sites More sharing options...
ErmaC Posted January 29, 2014 Author Share Posted January 29, 2014 v2360 test1: Everything looks OK here. Thx Could be possible a dmi dump from live linux distro (or if you have a linux distro on your system)? and also a complete bdmesg from OS X? ErmaC Link to comment Share on other sites More sharing options...
nyolc8 Posted January 29, 2014 Share Posted January 29, 2014 Full bdmesg:dmesg.txt I'll try to make a dmi dump from a linux live distro (if I can start it) Link to comment Share on other sites More sharing options...
ErmaC Posted January 29, 2014 Author Share Posted January 29, 2014 I'll try to make a dmi dump from a linux live distro (if I can start it) Thx. I w8 until you catch it. Any other feedback? Link to comment Share on other sites More sharing options...
Recommended Posts