Patch to get wvstreams recipe functional

Patch to get wvstreams recipe functional

[Kirtika Ruchandani]

[-- Attachment #1: Type: text/plain, Size: 395 bytes --]

Greetings,
Please find attached the quilt patches for the wvstreams 4.4.1 recipe.
The one in the OE tree currently doesn't build correctly.
It would be great if someone could test the above patches and check if the
build goes fine.
Regards,

-- 
Kirtika Ruchandani
Maemo Google Summer of Code  student
Sophomore, CS&E @ IIT-Madras
IRC: rkirti / oespirit
http://rkirti.wordpress.com

[-- Attachment #2: gcc4.3.patch --]
[-- Type: text/x-patch, Size: 618 bytes --]

Index: wvstreams-4.4.1/include/wvserialize.h
===================================================================
--- wvstreams-4.4.1.orig/include/wvserialize.h	2009-07-28 17:46:50.000000000 +0530
+++ wvstreams-4.4.1/include/wvserialize.h	2009-07-28 17:48:30.000000000 +0530
@@ -348,8 +348,9 @@
  * (zero) character.  Serialized WvStrings are guaranteed not to contain nul
  * except as the last character.
  */
+/*No storage class allowed here as per gcc 4.3 specs*/
 template <>
-extern WvString _wv_deserialize<WvString>(WvBuf &buf);
+WvString _wv_deserialize<WvString>(WvBuf &buf);
 
 
 /** Deserialize a WvBuf. */

[-- Attachment #3: dont-forget-header.patch --]
[-- Type: text/x-patch, Size: 470 bytes --]

Index: wvstreams-4.4.1/include/uniconfkey.h
===================================================================
--- wvstreams-4.4.1.orig/include/uniconfkey.h	2009-07-28 17:42:39.000000000 +0530
+++ wvstreams-4.4.1/include/uniconfkey.h	2009-07-28 17:42:55.000000000 +0530
@@ -9,7 +9,7 @@
 
 #include "wvstring.h"
 #include "wvlinklist.h"
-
+#include <limits.h>
 
 // The Python headers try to #define ANY as void.  If we're building Python
 // wrappers, get rid of that.

Re: Patch to get wvstreams recipe functional

[Khem Raj]

On (28/07/09 20:07), Kirtika Ruchandani wrote:
> Greetings,
> Please find attached the quilt patches for the wvstreams 4.4.1 recipe.
> The one in the OE tree currently doesn't build correctly.
> It would be great if someone could test the above patches and check if the
> build goes fine.
> Regards,
> 
> -- 
> Kirtika Ruchandani
> Maemo Google Summer of Code  student
> Sophomore, CS&E @ IIT-Madras
> IRC: rkirti / oespirit
> http://rkirti.wordpress.com

> Index: wvstreams-4.4.1/include/wvserialize.h
> ===================================================================
> --- wvstreams-4.4.1.orig/include/wvserialize.h	2009-07-28 17:46:50.000000000 +0530
> +++ wvstreams-4.4.1/include/wvserialize.h	2009-07-28 17:48:30.000000000 +0530
> @@ -348,8 +348,9 @@
>   * (zero) character.  Serialized WvStrings are guaranteed not to contain nul
>   * except as the last character.
>   */
> +/*No storage class allowed here as per gcc 4.3 specs*/
>  template <>
> -extern WvString _wv_deserialize<WvString>(WvBuf &buf);
> +WvString _wv_deserialize<WvString>(WvBuf &buf);
>  
>  
>  /** Deserialize a WvBuf. */

> Index: wvstreams-4.4.1/include/uniconfkey.h
> ===================================================================
> --- wvstreams-4.4.1.orig/include/uniconfkey.h	2009-07-28 17:42:39.000000000 +0530
> +++ wvstreams-4.4.1/include/uniconfkey.h	2009-07-28 17:42:55.000000000 +0530
> @@ -9,7 +9,7 @@
>  
>  #include "wvstring.h"
>  #include "wvlinklist.h"
> -
> +#include <limits.h>
>  
>  // The Python headers try to #define ANY as void.  If we're building Python
>  // wrappers, get rid of that.


FWIW patch looks ok. Can you send it with rest of recipe changes ?

Thx

-Khem

Re: Patch to get wvstreams recipe functional

[Khem Raj]

On (28/07/09 20:07), Kirtika Ruchandani wrote:
> Greetings,
> Please find attached the quilt patches for the wvstreams 4.4.1 recipe.
> The one in the OE tree currently doesn't build correctly.
> It would be great if someone could test the above patches and check if the
> build goes fine.
> Regards,
> 
> -- 
> Kirtika Ruchandani
> Maemo Google Summer of Code  student
> Sophomore, CS&E @ IIT-Madras
> IRC: rkirti / oespirit
> http://rkirti.wordpress.com

> Index: wvstreams-4.4.1/include/wvserialize.h
> ===================================================================
> --- wvstreams-4.4.1.orig/include/wvserialize.h	2009-07-28 17:46:50.000000000 +0530
> +++ wvstreams-4.4.1/include/wvserialize.h	2009-07-28 17:48:30.000000000 +0530
> @@ -348,8 +348,9 @@
>   * (zero) character.  Serialized WvStrings are guaranteed not to contain nul
>   * except as the last character.
>   */
> +/*No storage class allowed here as per gcc 4.3 specs*/
>  template <>
> -extern WvString _wv_deserialize<WvString>(WvBuf &buf);
> +WvString _wv_deserialize<WvString>(WvBuf &buf);
>  
>  
>  /** Deserialize a WvBuf. */

> Index: wvstreams-4.4.1/include/uniconfkey.h
> ===================================================================
> --- wvstreams-4.4.1.orig/include/uniconfkey.h	2009-07-28 17:42:39.000000000 +0530
> +++ wvstreams-4.4.1/include/uniconfkey.h	2009-07-28 17:42:55.000000000 +0530
> @@ -9,7 +9,7 @@
>  
>  #include "wvstring.h"
>  #include "wvlinklist.h"
> -
> +#include <limits.h>
>  
>  // The Python headers try to #define ANY as void.  If we're building Python
>  // wrappers, get rid of that.


FWIW patch looks ok. Can you send it with recipe changes ?
so it can be applied 
Thx

-Khem