[okl4-developer] [OKL2.1]spurious exception as soon as OKlinux forks init.

[Geoffrey Lee]

On Tue, Sep 30, 2008 at 02:55:28AM -0700, Amit Daniel wrote:
> 
> 
> 

Hi Amit

I suggest modifying syscall_loop() to print a system call trace
and a exception trace (e.g. unaligned load/store) and work
out what went wrong.

	-gl

> Geoffrey Lee wrote:
> > 
> > On Mon, Sep 29, 2008 at 05:16:26AM -0700, Amit Daniel wrote:
> >> 
> >> Dear All,
> >> 
> >> I have tested some test examples with okl microkernel and iguana. They
> >> works
> >> fine.
> >> But when i try executing OKLinux as an application, spurious exception
> >> starts coming. I am building the image file as,
> >> tools/build.py machine=gta01 project=iguana wombat=True 
> >> enable_fastpaths=False
> >> This exception is coming as soon kernel tries to fork the /sbin/init
> >> process. I tried to put some print in main interrupt handler but those
> >> prints are not coming. so, the exception source is not clear.
> >> 
> > 
> > Hi Amit
> > 
> > The exception handler for Linux processes is the syscall_loop(), not 
> > interrupt_loop(), which I think may be what is confusing you.
> > 
> > Also, have you made any modifications to either the OKL4 or the OK Linux
> > distribution in any way?  If so what have you changed?
> > 
> > 	-gl
> > 
> > Hi Geoffrey,
> > 
> > I have just made 1 change which is adding the solution for clzeq given in
> > one of the previous mails in the mailing list. This clzeq instruction was
> > causing exception and enable_fastpaths=false only disables 2 clz
> > instruction. I am testing for s3c2440 board which is ARM920T based core.
> > Modified file is : arch/arm/pistachio/v5/src/traps.spp
> > The diff with the original is shown below,
> > <         ldreq   r1,     [r0, #OFS_TCB_POST_SYSCALL_CALLBACK]
> > <         clzeq   r1,     r1 /* r1 contains callback function, 0xFXXXXXXX
> > */
> > ---
> >>        // ldreq   r1,     [r0, #OFS_TCB_POST_SYSCALL_CALLBACK]
> >>       //  clzeq   r1,     r1 /* r1 contains callback function, 0xFXXXXXXX
> >> */
> >>       ldr     r1,     [r0, #OFS_TCB_POST_SYSCALL_CALLBACK]
> > 224a226,228
> >>       bne kernel_fixup
> >>       normal_return:
> > 233c237,241
> > <         subnes  pc,     lr,     #1
> > ---
> >>    //     subnes  pc,     lr,     #1
> >>       subs  pc,     lr,     #1
> >>       kernel_fixup:
> >>       cmp     r1,     #0
> >>       beq     normal_return
> > 
> > Thanks,
> > Amit Daniel
> > 
> > 
> >> View this message in context:
> >> http://n2.nabble.com/-OKL2.1-spurious-exception-as-soon-as-OKlinux-forks-init.-tp1125668p1125668.html
> >> Sent from the OKL4 Community Forum mailing list archive at Nabble.com.
> >> 
> >> 
> >> _______________________________________________
> >> Developer mailing list
> >> Developer at okl4.org
> >> https://lists.okl4.org/mailman/listinfo/developer
> >> 
> > 
> > -- 
> > 
> > 
> > _______________________________________________
> > Developer mailing list
> > Developer at okl4.org
> > https://lists.okl4.org/mailman/listinfo/developer
> > 
> > 
> 
> -- 
> View this message in context: http://n2.nabble.com/-OKL2.1-spurious-exception-as-soon-as-OKlinux-forks-init.-tp1125668p1128483.html
> Sent from the OKL4 Community Forum mailing list archive at Nabble.com.
> 
> 
> _______________________________________________
> Developer mailing list
> Developer at okl4.org
> https://lists.okl4.org/mailman/listinfo/developer
> 

--