InfiniteMac OSx86  


Reply
 
Thread tools Display modes
  #161  
Old 09-03-2011, 11:58 AM
andyvand's Avatar
andyvand andyvand is offline
 
Join Date: Apr 2009
Location: Tienen
Posts: 515
Quote:
Originally Posted by kocoman View Post
I tried it but it give me kernel panic, and does not give me a backtrace..

how to debug the problem..



Uploaded with ImageShack.us
Damn... the problem is in the new SSE3 emulator...
0xFFFF0001 is 2nd instruction of the SSE3 emulator...
Guess I'll also need to modify the trap handler to fix this.
Reply With Quote
  #162  
Old 09-12-2011, 03:28 AM
kocoman kocoman is offline
Jaguar
 
Join Date: Jan 2009
Posts: 73
Quote:
Originally Posted by andyvand View Post
Damn... the problem is in the new SSE3 emulator...
0xFFFF0001 is 2nd instruction of the SSE3 emulator...
Guess I'll also need to modify the trap handler to fix this.
Are there plans to fix this?
Reply With Quote
  #163  
Old 09-12-2011, 07:08 AM
andyvand's Avatar
andyvand andyvand is offline
 
Join Date: Apr 2009
Location: Tienen
Posts: 515
Quote:
Originally Posted by kocoman View Post
Are there plans to fix this?
Yes, I'll need to study any screenshots you can give me of this.
It may help me put an override on the trap handler.
Reply With Quote
  #164  
Old 09-13-2011, 07:36 AM
kocoman kocoman is offline
Jaguar
 
Join Date: Jan 2009
Posts: 73
Quote:
Originally Posted by andyvand View Post
Yes, I'll need to study any screenshots you can give me of this.
It may help me put an override on the trap handler.
running "-s" is fine (no panic), so I thought it was some .kext problem

"-x", panics

Also it always fsck (take long time) every time I hard reboot the computer

Sometimes it would keep repeat many times the last 2 lines of the panic message, then reboot. But it never says the what the offending program is.

So I think this sse3emu is completely different than the XNU one?

I used to look at an Pentium Pro EMU that the Pentium did not have in Linux.. I forgot what it was now.. but it similar to the sse3emu problem
Reply With Quote
  #165  
Old 09-13-2011, 12:15 PM
andyvand's Avatar
andyvand andyvand is offline
 
Join Date: Apr 2009
Location: Tienen
Posts: 515
Quote:
Originally Posted by kocoman View Post
running "-s" is fine (no panic), so I thought it was some .kext problem

"-x", panics

Also it always fsck (take long time) every time I hard reboot the computer

Sometimes it would keep repeat many times the last 2 lines of the panic message, then reboot. But it never says the what the offending program is.

So I think this sse3emu is completely different than the XNU one?

I used to look at an Pentium Pro EMU that the Pentium did not have in Linux.. I forgot what it was now.. but it similar to the sse3emu problem
Try my 10.6.8 kernel.
It shouldn't have this issue...
Reply With Quote
  #166  
Old 10-03-2011, 03:07 PM
kocoman kocoman is offline
Jaguar
 
Join Date: Jan 2009
Posts: 73
Quote:
Originally Posted by andyvand View Post
Try my 10.6.8 kernel.
It shouldn't have this issue...
I did use the 10.6.8 AnV v0.4.1 you posted at the other thread here already, and the screenshot in the few last message was the result..

did you make changes to the 0.4.1 ?
Reply With Quote
  #167  
Old 10-13-2011, 04:52 AM
kocoman kocoman is offline
Jaguar
 
Join Date: Jan 2009
Posts: 73
Slice made a comment about the SSE3 problem at InsanelyMac

. SSE3 emulator implementation is wrong. The new emulator occupies 2 pages while you allocated only one.

CODE
+ printf("Enabling SSE3 emulator...");
+ /* Install into commpage. Actual patching of master_idt happens in start.s */
+ /* ASSERT(sse3emu_size == PAGE_SIZE); */
+ commpage_stuff2(_COMM_PAGE_SSE3EMU, &sse3emu_data, sse3emu_size, TRUE);
+ printf("done.\n");
Should be like
CODE
if (!(_cpu_capabilities & kHasSSE3)) {
printf("enabling %s emulator...", "SSE3");
// Install into commpage. Actual patching of master_idt happens elsewhere

commpage_stuff2(_COMM_PAGE_SSE2EMU, &sse3emu_ffff4000, sizeof(sse3emu_ffff4000), legacy);
commpage_stuff2(_COMM_PAGE_SSE2EMU2, &sse3emu_ffff5000, sizeof(sse3emu_ffff5000), legacy);
printf("OK!\n");
}
Reply With Quote
  #168  
Old 10-13-2011, 02:09 PM
andyvand's Avatar
andyvand andyvand is offline
 
Join Date: Apr 2009
Location: Tienen
Posts: 515
Quote:
Originally Posted by kocoman View Post
Slice made a comment about the SSE3 problem at InsanelyMac

. SSE3 emulator implementation is wrong. The new emulator occupies 2 pages while you allocated only one.

CODE
+ printf("Enabling SSE3 emulator...");
+ /* Install into commpage. Actual patching of master_idt happens in start.s */
+ /* ASSERT(sse3emu_size == PAGE_SIZE); */
+ commpage_stuff2(_COMM_PAGE_SSE3EMU, &sse3emu_data, sse3emu_size, TRUE);
+ printf("done.\n");
Should be like
CODE
if (!(_cpu_capabilities & kHasSSE3)) {
printf("enabling %s emulator...", "SSE3");
// Install into commpage. Actual patching of master_idt happens elsewhere

commpage_stuff2(_COMM_PAGE_SSE2EMU, &sse3emu_ffff4000, sizeof(sse3emu_ffff4000), legacy);
commpage_stuff2(_COMM_PAGE_SSE2EMU2, &sse3emu_ffff5000, sizeof(sse3emu_ffff5000), legacy);
printf("OK!\n");
}
Tried that allready.
It panics on the following instruction at 0xFFFF4001 -> lock inc %eax.
Old kernel checks which SSE3 emulator is requested...
Reply With Quote
  #169  
Old 10-26-2011, 04:18 AM
chitchit chitchit is offline
Jaguar
 
Join Date: Jan 2011
Posts: 97
Has any progress been made on the emulator?

I'd be interested in getting SL on my lappy!



💡 Deploy cloud instances seamlessly on DigitalOcean. Free credits ($100) for InfMac readers.

Reply With Quote
Reply