Revert of 406cd0fee12ad99ba6d6f7d55f743cc4c6697955

Revert of 406cd0fee12ad99ba6d6f7d55f743cc4c6697955

[Koen Kooi]

Hi,

Since 406cd0fee12ad99ba6d6f7d55f743cc4c6697955 broke all 
glibc-internal-iconv builds and I'm too lazy/stupid to fix it properly 
for uclibc I reverted this commit.

Koen

Re: Revert of 406cd0fee12ad99ba6d6f7d55f743cc4c6697955

[Bernhard Reutner-Fischer]

On Wed, Nov 18, 2009 at 11:34:28AM +0100, Koen Kooi wrote:
>Hi,
>
>Since 406cd0fee12ad99ba6d6f7d55f743cc4c6697955 broke all
>glibc-internal-iconv builds and I'm too lazy/stupid to fix it
>properly for uclibc I reverted this commit.

for unpatched uClibc you could just
TARGET_CPPFLAGS += -D_Exit=_exit

Re: Revert of 406cd0fee12ad99ba6d6f7d55f743cc4c6697955

[Leon Woestenberg]

Hello Bernhard,

On Wed, Nov 18, 2009 at 12:41 PM, Bernhard Reutner-Fischer
<rep.dot.nop@gmail.com> wrote:
> On Wed, Nov 18, 2009 at 11:34:28AM +0100, Koen Kooi wrote:
>>Hi,
>>
>>Since 406cd0fee12ad99ba6d6f7d55f743cc4c6697955 broke all
>>glibc-internal-iconv builds and I'm too lazy/stupid to fix it
>>properly for uclibc I reverted this commit.
>
> for unpatched uClibc you could just
> TARGET_CPPFLAGS += -D_Exit=_exit
>
Thanks, and what's the patched uClibc alternative (and where's the
patch)? It defines _Exit()?

Maybe that's better for OE.

Regards,
-- 
Leon

Re: Revert of 406cd0fee12ad99ba6d6f7d55f743cc4c6697955

[Bernhard Reutner-Fischer]

On Wed, Nov 18, 2009 at 12:44:39PM +0100, Leon Woestenberg wrote:
>Hello Bernhard,
>
>On Wed, Nov 18, 2009 at 12:41 PM, Bernhard Reutner-Fischer
><rep.dot.nop@gmail.com> wrote:
>> On Wed, Nov 18, 2009 at 11:34:28AM +0100, Koen Kooi wrote:
>>>Hi,
>>>
>>>Since 406cd0fee12ad99ba6d6f7d55f743cc4c6697955 broke all
>>>glibc-internal-iconv builds and I'm too lazy/stupid to fix it
>>>properly for uclibc I reverted this commit.
>>
>> for unpatched uClibc you could just
>> TARGET_CPPFLAGS += -D_Exit=_exit
>>
>Thanks, and what's the patched uClibc alternative (and where's the
>patch)? It defines _Exit()?

I suggest to just #define it to _exit, yes.
(glibc has it as weak alias but i'd prefer to get away with a define),
like:

_Exit(): define to _exit()

Signed-off-by: Bernhard Reutner-Fischer <me>
---
 include/stdlib.h |    4 +++-
 1 files changed, 3 insertions(+), 1 deletions(-)

diff --git a/include/stdlib.h b/include/stdlib.h
index e462c1c..c993f8f 100644
--- a/include/stdlib.h
+++ b/include/stdlib.h
@@ -545,12 +545,14 @@ extern void exit (int __status) __THROW
__attribute__ ((__noreturn__));
 libc_hidden_proto(exit)
 __END_NAMESPACE_STD
 
-#ifdef __USE_ISOC99
+#if 0 /* def __USE_ISOC99 */
 __BEGIN_NAMESPACE_C99
 /* Terminate the program with STATUS without calling any of the
    functions registered with `atexit' or `on_exit'.  */
 extern void _Exit (int __status) __THROW __attribute__
((__noreturn__));
 __END_NAMESPACE_C99
+#else
+# define _Exit _exit
 #endif
 
 
-- 
1.6.5.2

Re: Revert of 406cd0fee12ad99ba6d6f7d55f743cc4c6697955

[Henning Heinold]

Hi Bernhard,

thanks for the patch, but _exit is not the problem. Problem is crappy iconv stuff.
But I will fix it now the right way.

Bye Henning

Re: Revert of 406cd0fee12ad99ba6d6f7d55f743cc4c6697955

[Bernhard Reutner-Fischer]

On Wed, Nov 18, 2009 at 01:30:12PM +0100, Bernhard Reutner-Fischer wrote:
>On Wed, Nov 18, 2009 at 12:44:39PM +0100, Leon Woestenberg wrote:
>>Hello Bernhard,
>>
>>On Wed, Nov 18, 2009 at 12:41 PM, Bernhard Reutner-Fischer
>><rep.dot.nop@gmail.com> wrote:
>>> On Wed, Nov 18, 2009 at 11:34:28AM +0100, Koen Kooi wrote:
>>>>Hi,
>>>>
>>>>Since 406cd0fee12ad99ba6d6f7d55f743cc4c6697955 broke all
>>>>glibc-internal-iconv builds and I'm too lazy/stupid to fix it
>>>>properly for uclibc I reverted this commit.
>>>
>>> for unpatched uClibc you could just
>>> TARGET_CPPFLAGS += -D_Exit=_exit
>>>
>>Thanks, and what's the patched uClibc alternative (and where's the
>>patch)? It defines _Exit()?
>
>I suggest to just #define it to _exit, yes.
>(glibc has it as weak alias but i'd prefer to get away with a define),
>like:
>
>_Exit(): define to _exit()

OTOH someone may want to override it locally, so to play safe let's
follow suite:
http://git.uclibc.org/uClibc/commit/?h=0_9_30&id=e4adfdc4128c5909adeefecdaf870e30172922a9

Re: Revert of 406cd0fee12ad99ba6d6f7d55f743cc4c6697955

[Holger Hans Peter Freyther]

On Wednesday 18 November 2009 13:30:12 Bernhard Reutner-Fischer wrote:
> I suggest to just #define it to _exit, yes.
> (glibc has it as weak alias but i'd prefer to get away with a define),
> like:

Two things..

1.) The patch can not be applied as someone's mailer has linewrapped it
2.) Does the issue still apply?