error in src/powerpc/ffi_darwin.c

error in src/powerpc/ffi_darwin.c

[chennam]

Hi Anthony,

I am testing libffi v 3.0.10 & 3.0.11 on PowerPC/AIX, I think there is an
error in src/powerpc/ffi_darwin.c  
1. It misses a "break" for case FFI_AIX:
2. It fails the testcases for FFI_BAD_ABI (e.g. err_bad_abi.c)


Here is the local modification I made:
Could you validate them ?

--- ffi_darwin.c        (revision 229)
+++ ffi_darwin.c        (working copy)
@@ -1065,11 +1065,12 @@
       closure->cif = cif;
       closure->fun = fun;
       closure->user_data = user_data;
+      break;

     default:

-      FFI_ASSERT(0);
-      break;
+//    FFI_ASSERT(0);
+      return FFI_BAD_ABI;
     }
   return FFI_OK;
 }
-- 
View this message in context: http://old.nabble.com/error-in-src-powerpc-ffi_darwin.c-tp33421057p33421057.html
Sent from the Sourceware - libffi-discuss mailing list archive at Nabble.com.

Re: error in src/powerpc/ffi_darwin.c

[Timothy Wall]

That looks fine (although I think in general much of the bad_abi handling in libffi is broken).

This doesn't affect your failing closure tests, does it?

On Mar 1, 2012, at 8:14 AM, chennam wrote:

> 
> Hi Anthony,
> 
> I am testing libffi v 3.0.10 & 3.0.11 on PowerPC/AIX, I think there is an
> error in src/powerpc/ffi_darwin.c  
> 1. It misses a "break" for case FFI_AIX:
> 2. It fails the testcases for FFI_BAD_ABI (e.g. err_bad_abi.c)
> 
> 
> Here is the local modification I made:
> Could you validate them ?
> 
> --- ffi_darwin.c        (revision 229)
> +++ ffi_darwin.c        (working copy)
> @@ -1065,11 +1065,12 @@
>       closure->cif = cif;
>       closure->fun = fun;
>       closure->user_data = user_data;
> +      break;
> 
>     default:
> 
> -      FFI_ASSERT(0);
> -      break;
> +//    FFI_ASSERT(0);
> +      return FFI_BAD_ABI;
>     }
>   return FFI_OK;
> }
> -- 
> View this message in context: http://old.nabble.com/error-in-src-powerpc-ffi_darwin.c-tp33421057p33421057.html
> Sent from the Sourceware - libffi-discuss mailing list archive at Nabble.com.
> 

Re: error in src/powerpc/ffi_darwin.c

[chennam]

No. unfortunately, I am still waiting for an answer for my previous post on
failing closure tests, or more precisely error in aix_closure.S after return
from ffi_closure_helper_DARWIN()



technomage wrote:
> 
> That looks fine (although I think in general much of the bad_abi handling
> in libffi is broken).
> 
> This doesn't affect your failing closure tests, does it?
> 
> On Mar 1, 2012, at 8:14 AM, chennam wrote:
> 
>> 
>> Hi Anthony,
>> 
>> I am testing libffi v 3.0.10 & 3.0.11 on PowerPC/AIX, I think there is an
>> error in src/powerpc/ffi_darwin.c  
>> 1. It misses a "break" for case FFI_AIX:
>> 2. It fails the testcases for FFI_BAD_ABI (e.g. err_bad_abi.c)
>> 
>> 
>> Here is the local modification I made:
>> Could you validate them ?
>> 
>> --- ffi_darwin.c        (revision 229)
>> +++ ffi_darwin.c        (working copy)
>> @@ -1065,11 +1065,12 @@
>>       closure->cif = cif;
>>       closure->fun = fun;
>>       closure->user_data = user_data;
>> +      break;
>> 
>>     default:
>> 
>> -      FFI_ASSERT(0);
>> -      break;
>> +//    FFI_ASSERT(0);
>> +      return FFI_BAD_ABI;
>>     }
>>   return FFI_OK;
>> }
>> -- 
>> View this message in context:
>> http://old.nabble.com/error-in-src-powerpc-ffi_darwin.c-tp33421057p33421057.html
>> Sent from the Sourceware - libffi-discuss mailing list archive at
>> Nabble.com.
>> 
> 
> 
> 

-- 
View this message in context: http://old.nabble.com/error-in-src-powerpc-ffi_darwin.c-tp33421057p33421292.html
Sent from the Sourceware - libffi-discuss mailing list archive at Nabble.com.