Jump to content

[AMD] Yosemite Kernel Testing (for help use the Help Topic)


Duran Keeley
 Share

1,898 posts in this topic

Recommended Posts

:)

 

@spakk,

 
at the same time, I understand the reaction Bronya, Bronya kernel code for 10 years, it is wise to listen, so is Andy, the last kernel Andy works even better than Bronya but it works every two :)
 
Before posting your work quite honorable, try to see if the results are better, it is useless to go back.
 
I indulge with everyone, take the reproaches of Bronya as advice, no one but he and Andy are able to see your mistakes.
 
@spakk ,
 
en même temps , je comprends la réaction de Bronya , Bronya code du kernel depuis 10 ans  , il est sage de l'écouter , c'est pareil pour Andy , le dernier kernel d'Andy fonctionne encore mieux que celui de Bronya mais il fonctionne tout les deux :) 
 
Avant de poster votre travail tout à fait honorable , essayez le pour voir si les résultats sont meilleurs , il ne sert à rien de retourner en arrière . 
 
j'appréci tout le monde , prenez les reproches de Bronya comme un conseil , personne à part lui et Andy sont capable de voir vos erreurs .
Link to comment
Share on other sites

works on Sempron LE-1250 :)

 

moins de bug graphics , Safari bug , Chrome works , FireFox works , AppleStore no works , capture png icône red , capture png sélection works , Youtube HD works , QuickTime bug , FinalCutProx ? manque de mémoire . 

 

small vidéo Sempron :)

 

test_sempron.mp4.zip

Last login: Wed Dec 31 13:11:51 on console
iMac-de-gils:~ gils$ uname -a
Darwin iMac-de-gils.local 14.0.0 Darwin Kernel Version 14.0.0: di 30 dec 2014 22:54:58 CET; joriss:xnu-2782.1.97-AnV/BUILD/obj/RELEASE_X86_64 x86_64
iMac-de-gils:~ gils$ sysctl machdep
machdep.user_idle_level: 0
machdep.cpu.max_basic: 1
machdep.cpu.max_ext: 2147483672
machdep.cpu.vendor: AuthenticAMD
machdep.cpu.brand_string: AMD Sempron(tm) Processor LE-1250
machdep.cpu.family: 15
machdep.cpu.model: 127
machdep.cpu.extmodel: 7
machdep.cpu.extfamily: 0
machdep.cpu.stepping: 2
machdep.cpu.feature_bits: 35188793670655
machdep.cpu.extfeature_bits: 1210841299967
machdep.cpu.signature: 462834
machdep.cpu.brand: 0
machdep.cpu.features: FPU VME DE PSE TSC MSR PAE MCE CX8 APIC SEP MTRR PGE MCA CMOV PAT PSE36 CLFSH MMX FXSR SSE SSE2 SSE3 CX16
machdep.cpu.extfeatures: SYSCALL EM64T LAHF PREFETCHW RDTSCP TSCI
machdep.cpu.cores_per_package: 1
machdep.cpu.microcode_version: 21
machdep.cpu.processor_flag: 1
machdep.cpu.cache.linesize: 64
machdep.cpu.cache.L2_associativity: 16
machdep.cpu.cache.size: 524288
machdep.cpu.tlb.inst.small: 32
machdep.cpu.tlb.inst.large: 8
machdep.cpu.tlb.data.small: 32
machdep.cpu.tlb.data.small_level1: 512
machdep.cpu.tlb.data.large: 8
machdep.cpu.address_bits.physical: 40
machdep.cpu.address_bits.virtual: 48
machdep.cpu.core_count: 1
machdep.cpu.thread_count: 1
machdep.vectors.timer: 221
machdep.vectors.IPI: 222
machdep.pmap.hashwalks: 958298
machdep.pmap.hashcnts: 3254530
machdep.pmap.hashmax: 37
machdep.pmap.kernel_text_ps: 2097152
machdep.pmap.kern_pv_reserve: 3000
machdep.memmap.Conventional: 3220566016
machdep.memmap.RuntimeServices: 0
machdep.memmap.ACPIReclaim: 57344
machdep.memmap.ACPINVS: 139264
machdep.memmap.PalCode: 0
machdep.memmap.Reserved: 2273280
machdep.memmap.Unusable: 0
machdep.memmap.Other: 0
machdep.tsc.frequency: 2188823300
machdep.tsc.deep_idle_rebase: 1
machdep.tsc.nanotime.tsc_base: 126990181228
machdep.tsc.nanotime.ns_base: 0
machdep.tsc.nanotime.scale: 1962226597
machdep.tsc.nanotime.shift: 0
machdep.tsc.nanotime.generation: 2
machdep.misc.panic_restart_timeout: 1266874889
machdep.misc.interrupt_latency_max: 0x0 0x93 0x3b600
machdep.misc.machine_check_panic: 
machdep.misc.timer_queue_trace: 
machdep.misc.nmis: 0
machdep.eager_timer_evaluations: 338
machdep.eager_timer_evaluation_max: 265140
iMac-de-gils:~ gils$ 

post-1093405-0-97380900-1420029848_thumb.png

post-1093405-0-42366900-1420029872_thumb.png

Edited by gils83
Link to comment
Share on other sites

Sure it does.

Every instruction till SSE 4.2.

Only problem is it's only part hasswell, it has thus CPUID model and extmodel issue.

Which needs the fix I provided above.

I just tested it, no opemu errors, all's good!

Build attached, patch based on BSA R3.

Trust me when I say: this is the only way... spent 2 days fishing out the differences reading through TONS of Intel manuals...

This works...

NOTE: If this patch gets in the way on another Intel CPU: -nopentiumgfix boot argument will disable it...

EDIT: Bronzovka, look here:

http://ark.intel.com/nl/products/80796/Intel-Pentium-Processor-G3240-3M-Cache-3_10-GHz

Thing to look at:

 

Extensies instructieset SSE4.1/4.2

Other CPU's matching this are the Pentium G3220 and G3258 in the Pentium series that I know of.

Also Celeron G needs this kind of fix or no luck.

Mobile Hasswell series Celeron U and Pentium U also match these problems btw.

Hence the generic fix ;)

My test results with current kernel: BSA_Yos_R3_PentiumG

AppStore works without problems

good graphic

Desktop Icons good

launchpad icons good

Safari with patch without WebKit crash

Boot time soon

  • Like 3
Link to comment
Share on other sites

@Andy 

 

the Pentium D 915 works but bug graphics :(

Last login: Fri Jan  2 07:58:01 on console
Mac-Pro-de-gils:~ gils$ uname -a
Darwin Mac-Pro-de-gils.local 14.0.0 Darwin Kernel Version 14.0.0: di 30 dec 2014 22:54:58 CET; joriss:xnu-2782.1.97-AnV/BUILD/obj/RELEASE_X86_64 x86_64
Mac-Pro-de-gils:~ gils$ sysctl machdep
machdep.user_idle_level: 0
machdep.cpu.max_basic: 6
machdep.cpu.max_ext: 2147483656
machdep.cpu.vendor: GenuineIntel
machdep.cpu.brand_string: Intel(R) Pentium(R) D CPU 2.80GHz
machdep.cpu.family: 15
machdep.cpu.model: 6
machdep.cpu.extmodel: 0
machdep.cpu.extfamily: 0
machdep.cpu.stepping: 4
machdep.cpu.feature_bits: 251366180912127
machdep.cpu.extfeature_bits: 4832888832
machdep.cpu.signature: 3940
machdep.cpu.brand: 0
machdep.cpu.features: FPU VME DE PSE TSC MSR PAE MCE CX8 APIC SEP MTRR PGE MCA CMOV PAT PSE36 CLFSH DS ACPI MMX FXSR SSE SSE2 SS HTT TM PBE SSE3 DTES64 MON DSCPL EST CID CX16 TPR PDCM
machdep.cpu.extfeatures: SYSCALL XD EM64T LAHF
machdep.cpu.logical_per_package: 2
machdep.cpu.cores_per_package: 2
machdep.cpu.microcode_version: 4
machdep.cpu.processor_flag: 2
machdep.cpu.mwait.linesize_min: 64
machdep.cpu.mwait.linesize_max: 64
machdep.cpu.mwait.extensions: 0
machdep.cpu.mwait.sub_Cstates: 0
machdep.cpu.thermal.sensor: 0
machdep.cpu.thermal.dynamic_acceleration: 0
machdep.cpu.thermal.invariant_APIC_timer: 0
machdep.cpu.thermal.thresholds: 0
machdep.cpu.thermal.ACNT_MCNT: 0
machdep.cpu.thermal.core_power_limits: 0
machdep.cpu.thermal.fine_grain_clock_mod: 0
machdep.cpu.thermal.package_thermal_intr: 0
machdep.cpu.thermal.hardware_feedback: 0
machdep.cpu.thermal.energy_policy: 0
machdep.cpu.cache.linesize: 64
machdep.cpu.cache.L2_associativity: 8
machdep.cpu.cache.size: 2048
machdep.cpu.tlb.inst.large: 128
machdep.cpu.tlb.data.large: 64
machdep.cpu.address_bits.physical: 36
machdep.cpu.address_bits.virtual: 48
machdep.cpu.core_count: 2
machdep.cpu.thread_count: 2
machdep.vectors.timer: 221
machdep.vectors.IPI: 222
machdep.pmap.hashwalks: 431246
machdep.pmap.hashcnts: 1851225
machdep.pmap.hashmax: 65
machdep.pmap.kernel_text_ps: 2097152
machdep.pmap.kern_pv_reserve: 2000
machdep.memmap.Conventional: 2145841152
machdep.memmap.RuntimeServices: 0
machdep.memmap.ACPIReclaim: 57344
machdep.memmap.ACPINVS: 139264
machdep.memmap.PalCode: 0
machdep.memmap.Reserved: 9056256
machdep.memmap.Unusable: 0
machdep.memmap.Other: 0
machdep.tsc.frequency: 2792648180
machdep.tsc.deep_idle_rebase: 1
machdep.tsc.nanotime.tsc_base: 242102792148
machdep.tsc.nanotime.ns_base: 0
machdep.tsc.nanotime.scale: 1537955023
machdep.tsc.nanotime.shift: 0
machdep.tsc.nanotime.generation: 2
machdep.misc.panic_restart_timeout: 1266874889
machdep.misc.interrupt_latency_max: 0x0 0xde 0xb42adc
machdep.misc.machine_check_panic: 
machdep.misc.timer_queue_trace: 
machdep.misc.nmis: 0
machdep.eager_timer_evaluations: 70
machdep.eager_timer_evaluation_max: 370157
Mac-Pro-de-gils:~ gils$ 


 

Hi AnV

 

Your Source finishing (anv-r2 & Bronya-RC3)

 

 

And ......

 

 

Can you please help to fix it AppleNforceATA?
 
Let SATA DVD can works.
 
or give me some tips.
 
THX...
 

 

 

 

:)

 

the Nvidia and VIA chipset problems have nothing to do with the kernel !!  ;)

post-1093405-0-65120600-1420211771_thumb.png

Link to comment
Share on other sites

 

Hi AnV

 

Your Source finishing (anv-r2 & Bronya-RC3)

 

 

And ......

 

 

Can you please help to fix it AppleNforceATA?
 
Let SATA DVD can works.
 
or give me some tips.
 
THX...
 

 

It seems to misdetect partition type on DVD's.

Small override should fix this.

This driver is old however, and should be replaced by custom SuperNForceATA.

See the SuperVIAATA for Yosemite thread on upgrade diff ;)

  • Like 3
Link to comment
Share on other sites

Please test the WebKit patch installer now and let me know if the problem of non patching is resolved on this topic: AMD General Chat

You can find it here: WebKit Patch Installer

 

When reporting your results please check that in System ▸ Library ▸ Frameworks:

  • JavaScriptCore.framework
  • WebKit.framework

 

 

Hi Duran,

 

gives news to the webkit, I have still the Problem with Safari that web page from time to time crashes.

 

thanks

Link to comment
Share on other sites

I modified, compiled JmicronATA.kext for Maverick and Yosemite ! In JmicronATA i added dev_id 0x2362197b !

Support list devices  : 

0x2361197b

0x2362197b

0x2363197b

0x2365197b

0x2366197b

0x2368197b

 

You want this driver ? ))) 

  • Like 2
Link to comment
Share on other sites

I did a bit of development.

New kernel ready with opemu fully based on SSEPlus.

For most I used the faster SSE2 emulation functions (wherever available) thus also implementing a full MM/XMM save and restore function.

It should be pretty fast ;)

Test please, let me know the results. :D

BSA_YOS_R4.zip

  • Like 7
Link to comment
Share on other sites

same, with and without system.kext.

 

"invalid user opcode 64" with a bunch of hexadecimal digits (xx xx xx xx xx etc.) on Phenom X4 9550

 

Don't have enough time to test with System.kext at the moment, but it usually doesn't change anything on my system.

Link to comment
Share on other sites

Try R4B.

I also compiled both 64bit args.

There are X86_64 and X86_64H kernels built inside the package.

Let me know the results...

With x86_64 invalid user opcode 64 error during boot

With x86_64H - kernel panic:

post-302440-14208919313518_thumb.jpg

These results without replacing system.kext.

 

Same results with replaced system.kext.

Link to comment
Share on other sites

Try R4B.

I also compiled both 64bit args.

There are X86_64 and X86_64H kernels built inside the package.

Let me know the results...

 

With x86_64 invalid user opcode 64 error during boot

With x86_64H - kernel panic:

 

the same result as the others

Link to comment
Share on other sites

I did a bit of development.

New kernel ready with opemu fully based on SSEPlus.

For most I used the faster SSE2 emulation functions (wherever available) thus also implementing a full MM/XMM save and restore function.

It should be pretty fast ;)

Test please, let me know the results. :D

@Andy good kernel 86_64b, good Safari -QuickTime (iCloud no work)

for error opcode 64 I used old kext 

Schermata 2015-01-10 alle 22.02.50.png

Schermata 2015-01-10 alle 22.00.16.png

replace yosemite.zip

  • Like 3
Link to comment
Share on other sites

Yes ,i checked and received this "opcode64 error" - because don't work instructions ssse3 in opemu !

I tried modified only pshufb128 and palignr128, function "fetchoperands" and work, i tried in terminal testssse3:

bash-3.2# /Users/iMac/Desktop/testssse3 
Test pshufb PASS!
Test pshufb128 PASS!
Test phaddw FAILED: 120 bad
Test phaddw128 FAILED: 124 bad
Test phaddd FAILED: 120 bad
Test phaddd128 FAILED: 124 bad
Test phaddsw FAILED: 120 bad
Test phaddsw128 FAILED: 124 bad
Test pmaddubsw FAILED: 115 bad
Test pmaddubsw128 FAILED: 115 bad
Test phsubw FAILED: 120 bad
Test phsubw128 FAILED: 124 bad
Test phsubd FAILED: 120 bad
Test phsubd128 FAILED: 124 bad
Test phsubsw FAILED: 120 bad
Test phsubsw128 FAILED: 124 bad
Test psignb FAILED: 116 bad
Test psignb128 FAILED: 116 bad
Test psignw FAILED: 116 bad
Test psignw128 FAILED: 116 bad
Test psignd FAILED: 116 bad
Test psignd128 FAILED: 117 bad
...
 
i can fix all , and i need test on yosemite … 
  • Like 9
Link to comment
Share on other sites

I did and checked on Yosemite . this works, but same problem flash player , wrong color picture , and etc...

In opemu.c problems :

1) if(kernel_trap) {

            if(size_128) ((ssp_m128*)src)->ui = (*(__uint128_t*)address); <-- i replaced type __m128i to __uint128_t and i to ui , and works. 

            else ((ssp_m64*)src)->u64 = (*(uint64_t*)address);

        } else {

            //printf("xnu: da = %llx, rsp=%llx,  rip=%llx\n", address, reg_sel[4], r64->isf.rip);

            if(size_128) copyin(address, (char*)& ((ssp_m128*)src)->i, 16);

            else copyin(address, (char*)& ((ssp_m64*)src)->u64, 8);

        }

 

2) storeallxmm and loadallxmm don't working. I removed and works to GUI .

3) replace new to old instructions and works ...

 

This odd problem ((( . 

 

I sleep . ))

 

See opemu .

opemu_ssePlus.zip

  • Like 1
Link to comment
Share on other sites

 Share

×
×
  • Create New...