[Qemu-devel] MIPS exception number limits?

[Qemu-devel] MIPS exception number limits?

[陳韋任 (Wei-Ren Chen)]

Hi all,

  Wondering why MIPS limits exception number less then 0x100,
you can see such example in function do_raise_exception_err
(target-mips/op_helper.c). See below,

static inline void QEMU_NORETURN do_raise_exception_err(CPUMIPSState *env,
                                                        ...
{
  ...

#if 1
    if (exception < 0x100)
        qemu_log("%s: %d %d\n", __func__, exception, error_code);
#endif

  ...
}

  Anyone can help confirm this, stated in what spec? Also, do we
need "#if 1 .. #endif"? Thanks for help. =]  

Regards,
chenwj

-- 
Wei-Ren Chen (陳韋任)
Computer Systems Lab, Institute of Information Science,
Academia Sinica, Taiwan (R.O.C.)
Tel:886-2-2788-3799 #1667
Homepage: http://people.cs.nctu.edu.tw/~chenwj

Re: [Qemu-devel] MIPS exception number limits?

[Johnson, Eric]

> -----Original Message-----
> From: qemu-devel-bounces+ericj=mips.com@nongnu.org [mailto:qemu-devel-
> bounces+ericj=mips.com@nongnu.org] On Behalf Of ??? (Wei-Ren Chen)
> Sent: Thursday, November 22, 2012 11:34 AM
> To: qemu-devel@nongnu.org
> Subject: [Qemu-devel] MIPS exception number limits?
> 
> Hi all,
> 
>   Wondering why MIPS limits exception number less then 0x100,
> you can see such example in function do_raise_exception_err
> (target-mips/op_helper.c). See below,
> 
> static inline void QEMU_NORETURN do_raise_exception_err(CPUMIPSState *env,
>                                                         ...
> {
>   ...
> 
> #if 1
>     if (exception < 0x100)
>         qemu_log("%s: %d %d\n", __func__, exception, error_code);
> #endif
> 
>   ...
> }
> 
>   Anyone can help confirm this, stated in what spec? Also, do we
> need "#if 1 .. #endif"? Thanks for help. =]
> 
> Regards,
> chenwj
> 
> --
> Wei-Ren Chen (陳韋任)
> Computer Systems Lab, Institute of Information Science,
> Academia Sinica, Taiwan (R.O.C.)
> Tel:886-2-2788-3799 #1667
> Homepage: http://people.cs.nctu.edu.tw/~chenwj

The 0x100 is an arbitrary value known to be larger than the maximum possible CP0.Cause.ExcCode value.

From target-mips/cpu.h:
/* Dummy exception for conditional stores.  */
#define EXCP_SC 0x100

From target-mips/translate.c:
    gen_helper_0e0i(raise_exception, EXCP_SC); \

It can be seen the only use of EXCP_SC is for the macro OP_ST_ATOMIC when CONFIG_USER_ONLY is defined.

So basically the SC and SCD instructions for user-mode QEMU are implemented with a "dummy" exception.  Since it is not a real exception, it is not reported in the QEMU log file as an exception.

-Eric

Re: [Qemu-devel] MIPS exception number limits?

[Richard Henderson]

On 11/26/2012 09:25 PM, Johnson, Eric wrote:
> So basically the SC and SCD instructions for user-mode QEMU are implemented with a "dummy" exception.  Since it is not a real exception, it is not reported in the QEMU log file as an exception.

I'm not certain that's a helpful distinction to make when it comes to debugging qemu.


r~

Re: [Qemu-devel] MIPS exception number limits?

[Peter Maydell]

On 27 November 2012 21:18, Richard Henderson <rth@twiddle.net> wrote:
> On 11/26/2012 09:25 PM, Johnson, Eric wrote:
>> So basically the SC and SCD instructions for user-mode QEMU are implemented with a "dummy" exception.  Since it is not a real exception, it is not reported in the QEMU log file as an exception.
>
> I'm not certain that's a helpful distinction to make when it comes to debugging qemu.

And if it is, using a symbolic constant in the comparison
here might help point out what we're doing.

This is all bikeshedding over a fairly minor bit of code though.

-- PMM

Re: [Qemu-devel] MIPS exception number limits?

[Johnson, Eric]

> -----Original Message-----
> From: Richard Henderson [mailto:rth7680@gmail.com] On Behalf Of Richard
> Henderson
> Sent: Tuesday, November 27, 2012 1:19 PM
> To: Johnson, Eric
> Cc: "陳韋任 (Wei-Ren Chen)"; qemu-devel@nongnu.org
> Subject: Re: [Qemu-devel] MIPS exception number limits?
> 
> On 11/26/2012 09:25 PM, Johnson, Eric wrote:
> > So basically the SC and SCD instructions for user-mode QEMU are
> implemented with a "dummy" exception.  Since it is not a real exception,
> it is not reported in the QEMU log file as an exception.
> 
> I'm not certain that's a helpful distinction to make when it comes to
> debugging qemu.
> 
> 
> r~

It may not be a helpful distinction.

The under 0x100 check has been there since 2005 which is long before I looked at any QEMU code.  I was describing the code as it exists based on deductions from looking at the code.  I should have made that clear that my summary.

My first draft of a response I described the steps I did (I.E. vim op_helper.c, grep EXCP *.h, gitk op_helper.c, grep EXCP_SC *, vim translate.c) to get to the conclusions I made.  When I read it seemed like it might be interpreted as condescending which was not my intent.

In the original reply I also meant to say "removing the '#if 1' is fine with me."

/If/ the under 0x100 check remains, I would advocate using EXCP_SC instead of a magic constant.

-Eric