2008-01-28 23:09:34

by Paolo Ciarrocchi

Subject: [PATCH 3/7] X86: Fix trailing statements should be on next line

Fix trailing statements should be on next line

Signed-off-by: Paolo Ciarrocchi <[email protected]>
---
arch/x86/math-emu/errors.c | 39 ++++++++++++++++++++++++++-------------
1 files changed, 26 insertions(+), 13 deletions(-)

diff --git a/arch/x86/math-emu/errors.c b/arch/x86/math-emu/errors.c
index 7f96db4..4419339 100644
--- a/arch/x86/math-emu/errors.c
+++ b/arch/x86/math-emu/errors.c
@@ -93,19 +93,32 @@ void FPU_printall(void)
partial_status = status_word();

#ifdef DEBUGGING
-if ( partial_status & SW_Backward ) printk("SW: backward compatibility\n");
-if ( partial_status & SW_C3 ) printk("SW: condition bit 3\n");
-if ( partial_status & SW_C2 ) printk("SW: condition bit 2\n");
-if ( partial_status & SW_C1 ) printk("SW: condition bit 1\n");
-if ( partial_status & SW_C0 ) printk("SW: condition bit 0\n");
-if ( partial_status & SW_Summary ) printk("SW: exception summary\n");
-if ( partial_status & SW_Stack_Fault ) printk("SW: stack fault\n");
-if ( partial_status & SW_Precision ) printk("SW: loss of precision\n");
-if ( partial_status & SW_Underflow ) printk("SW: underflow\n");
-if ( partial_status & SW_Overflow ) printk("SW: overflow\n");
-if ( partial_status & SW_Zero_Div ) printk("SW: divide by zero\n");
-if ( partial_status & SW_Denorm_Op ) printk("SW: denormalized operand\n");
-if ( partial_status & SW_Invalid ) printk("SW: invalid operation\n");
+if ( partial_status & SW_Backward )
+printk("SW: backward compatibility\n");
+if ( partial_status & SW_C3 )
+printk("SW: condition bit 3\n");
+if ( partial_status & SW_C2 )
+printk("SW: condition bit 2\n");
+if ( partial_status & SW_C1 )
+printk("SW: condition bit 1\n");
+if ( partial_status & SW_C0 )
+printk("SW: condition bit 0\n");
+if ( partial_status & SW_Summary )
+printk("SW: exception summary\n");
+if ( partial_status & SW_Stack_Fault )
+printk("SW: stack fault\n");
+if ( partial_status & SW_Precision )
+printk("SW: loss of precision\n");
+if ( partial_status & SW_Underflow )
+printk("SW: underflow\n");
+if ( partial_status & SW_Overflow )
+printk("SW: overflow\n");
+if ( partial_status & SW_Zero_Div )
+printk("SW: divide by zero\n");
+if ( partial_status & SW_Denorm_Op )
+printk("SW: denormalized operand\n");
+if ( partial_status & SW_Invalid )
+printk("SW: invalid operation\n");
#endif /* DEBUGGING */

printk(" SW: b=%d st=%ld es=%d sf=%d cc=%d%d%d%d ef=%d%d%d%d%d%d\n",
--
1.5.4.rc3

2008-01-28 23:45:13

by Willy Tarreau

Subject: Re: [PATCH 3/7] X86: Fix trailing statements should be on next line

Hi Paolo,

On Tue, Jan 29, 2008 at 12:07:44AM +0100, Paolo Ciarrocchi wrote:
> Fix trailing statements should be on next line

> -if ( partial_status & SW_C3 ) printk("SW: condition bit 3\n");
> -if ( partial_status & SW_C2 ) printk("SW: condition bit 2\n");
> -if ( partial_status & SW_C1 ) printk("SW: condition bit 1\n");
> -if ( partial_status & SW_C0 ) printk("SW: condition bit 0\n");
> -if ( partial_status & SW_Summary ) printk("SW: exception summary\n");
> -if ( partial_status & SW_Stack_Fault ) printk("SW: stack fault\n");
> -if ( partial_status & SW_Precision ) printk("SW: loss of precision\n");
> -if ( partial_status & SW_Underflow ) printk("SW: underflow\n");
> -if ( partial_status & SW_Overflow ) printk("SW: overflow\n");
> -if ( partial_status & SW_Zero_Div ) printk("SW: divide by zero\n");
> -if ( partial_status & SW_Denorm_Op ) printk("SW: denormalized operand\n");
> -if ( partial_status & SW_Invalid ) printk("SW: invalid operation\n");

> +if ( partial_status & SW_Backward )
> +printk("SW: backward compatibility\n");
> +if ( partial_status & SW_C3 )
> +printk("SW: condition bit 3\n");
> +if ( partial_status & SW_C2 )
> +printk("SW: condition bit 2\n");
> +if ( partial_status & SW_C1 )
> +printk("SW: condition bit 1\n");
> +if ( partial_status & SW_C0 )
> +printk("SW: condition bit 0\n");
> +if ( partial_status & SW_Summary )
> +printk("SW: exception summary\n");
> +if ( partial_status & SW_Stack_Fault )
> +printk("SW: stack fault\n");
> +if ( partial_status & SW_Precision )
> +printk("SW: loss of precision\n");
> +if ( partial_status & SW_Underflow )
> +printk("SW: underflow\n");
> +if ( partial_status & SW_Overflow )
> +printk("SW: overflow\n");
> +if ( partial_status & SW_Zero_Div )
> +printk("SW: divide by zero\n");
> +if ( partial_status & SW_Denorm_Op )
> +printk("SW: denormalized operand\n");
> +if ( partial_status & SW_Invalid )
> +printk("SW: invalid operation\n");
> #endif /* DEBUGGING */

Well, IMHO, the code was more readable and checkable in the initial
version. I think this is one example of exceptions where code appearance
is more important than style correctness.

Regards,
Willy

2008-01-29 12:43:33

by Stefan Richter

Subject: Re: [PATCH 3/7] X86: Fix trailing statements should be on next line

Willy Tarreau wrote:
> Hi Paolo,
>
> On Tue, Jan 29, 2008 at 12:07:44AM +0100, Paolo Ciarrocchi wrote:
>> Fix trailing statements should be on next line
>
>> -if ( partial_status & SW_C3 ) printk("SW: condition bit 3\n");
...
>> +if ( partial_status & SW_Backward )
>> +printk("SW: backward compatibility\n");
...
>
> Well, IMHO, the code was more readable and checkable in the initial
> version. I think this is one example of exceptions where code appearance
> is more important than style correctness.

Besides, the proposed style "fix" itself violates CodingStyle in at
least two obvious ways.
--
Stefan Richter
-=====-==--- ---= ===-=
http://arcgraph.de/sr/

2008-02-01 16:22:13

by Ingo Molnar

Subject: Re: [PATCH 3/7] X86: Fix trailing statements should be on next line

* Willy Tarreau <[email protected]> wrote:

> Hi Paolo,
>
> On Tue, Jan 29, 2008 at 12:07:44AM +0100, Paolo Ciarrocchi wrote:
> > Fix trailing statements should be on next line
>
> > -if ( partial_status & SW_C3 ) printk("SW: condition bit 3\n");
> > -if ( partial_status & SW_C2 ) printk("SW: condition bit 2\n");
> > -if ( partial_status & SW_C1 ) printk("SW: condition bit 1\n");
> > -if ( partial_status & SW_C0 ) printk("SW: condition bit 0\n");
> > -if ( partial_status & SW_Summary ) printk("SW: exception summary\n");
> > -if ( partial_status & SW_Stack_Fault ) printk("SW: stack fault\n");
> > -if ( partial_status & SW_Precision ) printk("SW: loss of precision\n");
> > -if ( partial_status & SW_Underflow ) printk("SW: underflow\n");
> > -if ( partial_status & SW_Overflow ) printk("SW: overflow\n");
> > -if ( partial_status & SW_Zero_Div ) printk("SW: divide by zero\n");
> > -if ( partial_status & SW_Denorm_Op ) printk("SW: denormalized operand\n");
> > -if ( partial_status & SW_Invalid ) printk("SW: invalid operation\n");
>
> > +if ( partial_status & SW_Backward )
> > +printk("SW: backward compatibility\n");
> > +if ( partial_status & SW_C3 )
> > +printk("SW: condition bit 3\n");
> > +if ( partial_status & SW_C2 )
> > +printk("SW: condition bit 2\n");
> > +if ( partial_status & SW_C1 )
> > +printk("SW: condition bit 1\n");
> > +if ( partial_status & SW_C0 )
> > +printk("SW: condition bit 0\n");
> > +if ( partial_status & SW_Summary )
> > +printk("SW: exception summary\n");
> > +if ( partial_status & SW_Stack_Fault )
> > +printk("SW: stack fault\n");
> > +if ( partial_status & SW_Precision )
> > +printk("SW: loss of precision\n");
> > +if ( partial_status & SW_Underflow )
> > +printk("SW: underflow\n");
> > +if ( partial_status & SW_Overflow )
> > +printk("SW: overflow\n");
> > +if ( partial_status & SW_Zero_Div )
> > +printk("SW: divide by zero\n");
> > +if ( partial_status & SW_Denorm_Op )
> > +printk("SW: denormalized operand\n");
> > +if ( partial_status & SW_Invalid )
> > +printk("SW: invalid operation\n");
> > #endif /* DEBUGGING */
>
> Well, IMHO, the code was more readable and checkable in the initial
> version. I think this is one example of exceptions where code
> appearance is more important than style correctness.

definitely so. The rule of thumb is: if in doubt, use your human taste.
Does it 'look' better? If not, dont do the change.

Ingo