Floating point emulation problem

Floating point emulation problem

[Ivo Zivkov]

[-- Attachment #1: Type: text/plain, Size: 813 bytes --]


Dear List,

I am running kernel 2.2.5-15. I am trying to calculate sin(0.9), and it crashes on a 386 board with no f/p hardware. The message I get is:

"Unable to handle kernel paging request at virtual address 7f3c0070......"

The interesting thing is sin(0.8) works fine. On a Pentium the program executes fine for all values.

I tried in 2 different 386 boards, and I get the same problem. The program was compiled on R.H.6.0, and "libm" was present on the 386. I even linked the program statically to eliminate any library dependencies. 

This seems like a common problem, and easy to reproduce. Anybody had the same experience?

please reply to mailto:izivkov@yahoo.com

Regards,

Ivo

 



---------------------------------
Do You Yahoo!?
>From homework help to love advice, Yahoo! Experts has your answer.

[-- Attachment #2: Type: text/html, Size: 969 bytes --]

Re: Floating point emulation problem

[Alan Cox]

> I am running kernel 2.2.5-15. I am trying to calculate sin(0.9), and it crashes on a 386 board with no f/p hardware. The message I get is:
> 
> "Unable to handle kernel paging request at virtual address 7f3c0070......"
> 
> The interesting thing is sin(0.8) works fine. On a Pentium the program executes fine for all values.
> 
> I tried in 2 different 386 boards, and I get the same problem. The program was compiled on R.H.6.0, and "libm" was present on the 386. I even linked the program statically to eliminate any library dependencies. 
> 
> This seems like a common problem, and easy to reproduce. Anybody had the same experience?

Either upgrade to a later 2.2 kernel or rebuild the kernel with gcc 2.7.2.3. Its
a bug in the FPU code when compiled with later compilers

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
Please read the FAQ at http://www.tux.org/lkml/