Jump to content

Mavericks kernel testing on AMD (formerly Mountain Lion kernel testing on AMD)


theconnactic
 Share

6,414 posts in this topic

Recommended Posts

 

I compiled a new kernel, using the latest sinetek source code. (added AnVSee3 and See42 opemu)

Use it to boot system normally.

But at compile time has been suggested that "xcrun: error: unable to find utility"ctfmerge", not a developer tool or inPATH"and "xcrun: error: unable to find utility "ctfconvert", not a developertool or in PATH".

It affects the compiler?

Thanks!

attachicon.gifmach_kernel V7 test.zip

 

 
tested the core, that's what is written at boot:
 
OPEMU: rep movsq
 
but the system is booted,
 
But that said DarwinDumper:
 

hw.packages: 1

hw.optional.floatingpoint: 1

hw.optional.mmx: 1

hw.optional.sse: 1

hw.optional.sse2: 1

hw.optional.sse3: 1

hw.optional.supplementalsse3: 0

hw.optional.sse4_1: 1

hw.optional.sse4_2: 0

hw.optional.x86_64: 1

hw.optional.aes: 1

hw.optional.avx1_0: 1

hw.optional.rdrand: 0

hw.optional.f16c: 1

hw.optional.enfstrg: 0

hw.optional.fma: 1

hw.optional.avx2_0: 0

hw.optional.bmi1: 0

hw.optional.bmi2: 0

hw.optional.rtm: 0

hw.optional.hle: 0

machdep.pmap.hashwalks: 1028734

Link to comment
Share on other sites

 

No. I'm a middle student now,and 14 years.

 

:hysterical: :hysterical: :drool::D

 

 

Edit:

 

I would just like to briefly point out some times, if someone has problems with its Nvidia graphics card, which should follow these instructions

 

http://www.insanelymac.com/forum/topic/285492-amd-issues-with-nvidia-cards-running-lion-64-bitsmountain-lion-dont-ask-help-here-the-help-topic-is-there-for-you/page-8

 

my post #159

Edited by spakk
Link to comment
Share on other sites

I compiled a new kernel, using the latest sinetek source code. (added AnVSee3 and See42 opemu)

Use it to boot system normally.

But at compile time has been suggested that "xcrun: error: unable to find utility"ctfmerge", not a developer tool or inPATH"and "xcrun: error: unable to find utility "ctfconvert", not a developertool or in PATH".

It affects the compiler?

Thanks!

Did you download all the dependencies & install them correctly ?? If yes then Try change you default Xcode path (even if you never have before), when I used previous versions of Xcode (Lion & ML) in Mavericks then went to compile a kernel in mavericks this would happen. If you have trouble with knowing what dependencies to download see the attached.

 

 

 

  • Like 2
Link to comment
Share on other sites

mach_sinetek

 

hw.packages: 1

hw.optional.floatingpoint: 1

hw.optional.mmx: 1

hw.optional.sse: 1

hw.optional.sse2: 1

hw.optional.sse3: 1

hw.optional.supplementalsse3: 0

hw.optional.sse4_1: 1

hw.optional.sse4_2: 0

hw.optional.x86_64: 1

hw.optional.aes: 1

hw.optional.avx1_0: 1

hw.optional.rdrand: 0

hw.optional.f16c: 1

hw.optional.enfstrg: 0

hw.optional.fma: 1

hw.optional.avx2_0: 0

hw.optional.bmi1: 0

hw.optional.bmi2: 0

hw.optional.rtm: 0

hw.optional.hle: 0

 

mach_kernel v6

 

 

hw.packages: 1

hw.optional.floatingpoint: 1

hw.optional.mmx: 1

hw.optional.sse: 1

hw.optional.sse2: 1

hw.optional.sse3: 1

hw.optional.supplementalsse3: 1

hw.optional.sse4_1: 1

hw.optional.sse4_2: 1

hw.optional.x86_64: 1

hw.optional.aes: 1

hw.optional.avx1_0: 1

hw.optional.rdrand: 0

hw.optional.f16c: 1

hw.optional.enfstrg: 0

hw.optional.fma: 1

hw.optional.avx2_0: 0

hw.optional.bmi1: 0

hw.optional.bmi2: 0

hw.optional.rtm: 0

hw.optional.hle: 0

 
Link to comment
Share on other sites

I read somewhat in Intel's specs and it seems that it works somewhat different.

The function for string compare is fine but the values always go to xmm0 and I implemented it to go to selected xmm reg.

Need to change this.

Also this makes me think that one of the arguments contains the count in 2 of the arguments (4 bits + 4 bits? meaning the count+1?)

I'm going to do a check on this this evening.

Also after fixing this I'm going to implement the 2 missing SSE4.2 instructions.

Keep you posted... first a delayed birthday celebration with my girlfriend (Cindy Claes) as I haven't done much on my birthday (3/1/2014) ;)

  • Like 13
Link to comment
Share on other sites

Everything working perfect with mach_sinetek kernel :) Only reboot is not working but it is no important because I always had a problems with reboot with Mavericks kernels ;) Great work spakk!! :thumbsup_anim:

 

EDIT: :sorcerer: :sorcerer:   HaPpY BiRtHdAy AnDy!!!!   :sorcerer: :sorcerer: To your girlfriend too :D

Edited by tragediana150
  • Like 2
Link to comment
Share on other sites

I read somewhat in Intel's specs and it seems that it works somewhat different.

The function for string compare is fine but the values always go to xmm0 and I implemented it to go to selected xmm reg.

Need to change this.

Also this makes me think that one of the arguments contains the count in 2 of the arguments (4 bits + 4 bits? meaning the count+1?)

I'm going to do a check on this this evening.

Also after fixing this I'm going to implement the 2 missing SSE4.2 instructions.

Keep you posted... first a delayed birthday celebration with my girlfriend (Cindy Claes) as I haven't done much on my birthday (3/1/2014) ;)

yeah it was right in my implementation i believe, tho the rest was incomplete (i only bothered with one func mode).

I think the code calling that instruction is just a strlen(), so only the flags register's result is used.

 but i agree it would be nice to have the whole 4.2 set.

 

happy age counter increment! cheers to the both of ya

  • Like 3
Link to comment
Share on other sites

yeah it was right in my implementation i believe, tho the rest was incomplete (i only bothered with one func mode).

I think the code calling that instruction is just a strlen(), so only the flags register's result is used.

but i agree it would be nice to have the whole 4.2 set.

 

happy age counter increment! cheers to the both of ya

yes, the functions with strlen style are right (ecx/rcx + flags reg) but the one with xmm0 and the ones with length specified need some adjusting.

I'll do that when I get home.

++age;

:D

Link to comment
Share on other sites

Happy birthday, Andy! :D

Sorry, I've totally ignored, even by me a Congratulations subsequently andy for your birthday :rofl:

  • Like 1
Link to comment
Share on other sites

I'm finally at home.

Maybe now I can correct the opemu and add the 2 missing instructions...

Keep you posted :D

EDIT: Ok, I ran a little test...

movl $4, %eax

movl $5, %edx

pcmpestri $0, %xmm1, %xmm0

= param a 4 bytes compare, param b 5 bytes compare...

Is what it compiles into...
2 params ... one in eax, one in edx
Ok, fixed this part...
Now change the code to store it in XMM0 and this is all fixed...
 
Ok, added this in the functions which need to store to XMM0:
    /* Set target to XMM0 */
    this->udo_dst->base = UD_R_XMM0;

Now try to compile...

 

Ok... had to modify...

New code:

    enum ud_type XMMDST = UD_R_XMM0;
    /* Set target to XMM0 */
    memcpy((void *)&(this->udo_dst->base), (void *)&XMMDST, sizeof(this->udo_dst->base));

I hope this is allowed...

New opemu attached...

Should work right for all 4 string compare functions... (need to test...)

OPEMU.zip

  • Like 5
Link to comment
Share on other sites

the kernel should be in Build folder.

if you build with:

make ARCH_CONFIGS=X86_64 KERNEL_CONFIGS=RELEASE install

it should be under build/dst

Link to comment
Share on other sites

 

Last login: Sun Jan  5 23:42:14 on ttys000
localhost:~ spakk$ uname -a
Darwin localhost 13.0.0 Darwin Kernel Version 13.0.0: So  5 Jan 2014 23:15:58 CET; *patched-sinetek/BUILD/obj/RELEASE_X86_64 x86_64
localhost:~ spakk$ 
I make just a few tests with the new kernel, thx Andy
Nice.

Do you use the new opemu with it?

the 4 sse4.2 cmp functions should be tested...

EDIT: that's not the new opemu I posted above...

I fixed that sprintf deprecation to use snprintf....

Link to comment
Share on other sites

really odd...

when I compile my kernel: not a single warning...

and I have all warn flags set...

Link to comment
Share on other sites

Excuse me for my clumsy work :rofl:

Last login: Mon Jan  6 02:16:41 on console
localhost:~ spakk$ uname -a
Darwin localhost 13.0.0 Darwin Kernel Version 13.0.0: Mo  6 Jan 2014 02:09:31 CET; *patched-sinetek/BUILD/obj/RELEASE_X86_64 x86_64
localhost:~ spakk$ 

BSA_kernel_V8.zip

Edited by spakk
  • Like 4
Link to comment
Share on other sites

  BSA_kernel_V8

reboot - ok 

shutdown - ok 

Phenom X4 945   -  SSE4.2  probably not

machdep.cpu.brand_string: AMD Phenom II X4 945 Processor

machdep.cpu.family: 16

machdep.cpu.model: 4

machdep.cpu.extmodel: 0

machdep.cpu.extfamily: 1

machdep.cpu.stepping: 3

machdep.cpu.feature_bits: 394984447 8396809

machdep.cpu.extfeature_bits: 4022598655 14335

machdep.cpu.signature: 1052483

machdep.cpu.brand: 0

machdep.cpu.features: FPU VME DE PSE TSC MSR PAE MCE CX8 APIC SEP MTRR PGE MCA CMOV PSE36 CLFSH MMX FXSR SSE SSE2 HTT SSE3 MON CX16 POPCNT

 

 

hw.optional.sse: 1

hw.optional.sse2: 1

hw.optional.sse3: 1

hw.optional.supplementalsse3: 0

hw.optional.sse4_1: 0

hw.optional.sse4_2: 0

Link to comment
Share on other sites

 Share

×
×
  • Create New...