Missing instructions on Cyrix 686 processors

  • Thread starter Thread starter Mark Hobley
  • Start date Start date
M

Mark Hobley

Does anybody know where I can get a definitive list of IA32 instructions are
supported by a traditional Intel Pentium processor, but are not supported by a
Cyrix 686 processor?

Thanks in advance,

Mark.
 
Mark said:
Does anybody know where I can get a definitive list of IA32 instructions are
supported by a traditional Intel Pentium processor, but are not supported by a
Cyrix 686 processor?

Thanks in advance,

Mark.

TestCPU - processor detection and benchmark
http://testcpu.webz.cz/man.htm

The earlier 6x86 processor was missing the CPUID instruction which was
introduced with the Pentium MMX and K6 processors, but that was added
with the 6x86MX processor. So without the CPUID instruction, the 6x86
was basically the same as the 486 processor in capabilities. But the
CPUID was later backgrafted onto the 486 processors too.

Yousuf Khan
 
Yousuf Khan said:
The earlier 6x86 processor was missing the CPUID instruction which was
introduced with the Pentium MMX and K6 processors, but that was added
with the 6x86MX processor.

I have some Intel Pentium based computers without the MMX extensions,
and some AMD K5 processors.

I know that the Pentium based computers do not have CPUID. I was just
wondering if there is anything else missing on the 6x86.

I am getting "Invalid Instruction" crashes on a Linux kernel, and I am
just trying to get an idea as to what may be wrong, before I decompile
it.

The kernel is supposed to be built for a 486 machine, so I am not
expecting CPUID instructions, but there may be a bug in the compiler.

Cheers,

Mark.
 
Mark said:
I have some Intel Pentium based computers without the MMX extensions,
and some AMD K5 processors.

I know that the Pentium based computers do not have CPUID. I was just
wondering if there is anything else missing on the 6x86.

I am getting "Invalid Instruction" crashes on a Linux kernel, and I am
just trying to get an idea as to what may be wrong, before I decompile
it.

The kernel is supposed to be built for a 486 machine, so I am not
expecting CPUID instructions, but there may be a bug in the compiler.

"Invalid Instruction" errors often tended to be catchall errors for
programming errors (e.g. pointers pointing to wrong places), or bad memory.

If it's happening to the same program over and over again, then it's
possibly a bad program. If it happens at the same memory address but to
different programs, then it's possibly bad memory.

Yousuf Khan
 
Back
Top