[Patch] New: CR16 port

Pompapathi V Gadad Pompapathi.V.Gadad@nsc.com
Mon Jul 16 13:01:00 GMT 2007


'Rask Ingemann Lambertsen' wrote:
> On Mon, Jul 16, 2007 at 02:26:06PM +0530, Pompapathi V Gadad wrote:
>
>   
>> Index: doc/extend.texi
>> ===================================================================
>> --- doc/extend.texi	(revision 126669)
>> +++ doc/extend.texi	(working copy)
>> @@ -2100,11 +2100,10 @@ This attribute is ignored for R8C target
>>  
>>  @item interrupt
>>  @cindex interrupt handler functions
>> -Use this attribute on the ARM, AVR, C4x, CRX, M32C, M32R/D, m68k, MS1,
>> -and Xstormy16 ports to indicate that the specified function is an
>> -interrupt handler.  The compiler will generate function entry and exit
>> -sequences suitable for use in an interrupt handler when this attribute
>> -is present.
>> +Use this attribute on the ARM, AVR, C4x, CR16, CRX, M32C, M32R/D, MS1, 
>> +and Xstormy16 ports to indicate that the specified function is an interrupt 
>> +handler. The compiler will generate function entry and exit sequences suitable
>> +for use in an interrupt handler when this attribute is present.
>>  
>>  Note, interrupt handlers for the Blackfin, H8/300, H8/300H, H8S, and
>>  SH processors can be specified via the @code{interrupt_handler} attribute.
>>     
>
>    You're still modifying the m68k documentation.
>   
I am extremely sorry, I do not know how m68k changes has gone in to this 
patch. I will be careful now on.
>   
>> Index: config/cr16/cr16.md
>>     
> [...]
>   
>> +(define_mode_attr tIsa      [(QI "b") (HI "w") (SI "d") (SF "d")])
>> +(define_mode_attr tIsaLong  [                  (SI "d") (SF "d")])
>> +(define_mode_attr tIsaShort [(QI "b") (HI "w")                  ])
>>     
>
>    You should be able to use tIsa instead of tIsaLong and tIsaShort
> everywhere.
>
>   
>> +(define_mode_attr lImmShort [(QI "4") (HI "4")                  ])
>> +(define_mode_attr lImmLong  [                  (SI "6") (SF "6")])
>> +(define_mode_attr lImmArith [(QI "4") (HI "4") (SI "6") (SF "6")])
>>     
>
>    And something similiar goes here.
>
>   
>> +;;  Negation Instructions
>> +; move 0,r sub x,r. More efficient than xor -1,x add 1,x
>> +(define_expand "neg<mode>2"
>> +  [(set (match_operand:CR16IM 0 "register_operand" "")
>> +        (const_int 0))
>> +   (parallel [(set (match_dup 0) 
>> +        (minus:CR16IM (match_dup 0)
>> +          (match_operand:CR16IM 1 "register_operand" "")))
>> +    (clobber (reg:CC CC_REGNUM))]) 
>> +  ]
>> +  ""
>> +  "")
>>     
>
>    You said you would delete this expander. Did it not work?
>
>   
I will make necessary changes and post the pacthes.
Thanks a lot,
Pompa




More information about the Gcc-patches mailing list