gcc-4.7.0-RC-20120302 fails to build for i686-pc-cygwin

gcc-4.7.0-RC-20120302 fails to build for i686-pc-cygwin

[Ryan Johnson]

Hi all,

I tried to bootstrap the gcc-4.7 RC and it fails because it expects to 
find <process.h> and the file actually lives in <cygwin/process.h> (see 
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52513).

I'm asking here because the gcc devs thought this would mean 4.6 is 
broken as well, but I have 4.6.2 running. Did process.h perhaps move 
between 1.7.10 and 1.7.11? I guess configure must be using linker rather 
than preprocessor tests for presence of spawnve, because it thinks 
(correctly) that the function exists.

Thanks!
Ryan


--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple

Re: gcc-4.7.0-RC-20120302 fails to build for i686-pc-cygwin

[marco atzeri]

On 3/7/2012 1:54 PM, Ryan Johnson wrote:
> Hi all,
>
> I tried to bootstrap the gcc-4.7 RC and it fails because it expects to
> find <process.h> and the file actually lives in <cygwin/process.h> (see
> http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52513).
>
> I'm asking here because the gcc devs thought this would mean 4.6 is
> broken as well, but I have 4.6.2 running. Did process.h perhaps move
> between 1.7.10 and 1.7.11? I guess configure must be using linker rather
> than preprocessor tests for presence of spawnve, because it thinks
> (correctly) that the function exists.
>
> Thanks!
> Ryan
>
>

it is in
/usr/include on 1.7.9
/usr/include/cygwin on 1.7.10

and was moved back on
/usr/include in 1.7.11

http://cygwin.com/cgi-bin2/package-cat.cgi?file=cygwin%2Fcygwin-1.7.9-1&grep=process.h
http://cygwin.com/cgi-bin2/package-cat.cgi?file=cygwin%2Fcygwin-1.7.11-1&grep=process.h


--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple

Re: gcc-4.7.0-RC-20120302 fails to build for i686-pc-cygwin

[Corinna Vinschen]

On Mar  7 07:54, Ryan Johnson wrote:
> Hi all,
> 
> I tried to bootstrap the gcc-4.7 RC and it fails because it expects
> to find <process.h> and the file actually lives in
> <cygwin/process.h> (see
> http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52513).
> 
> I'm asking here because the gcc devs thought this would mean 4.6 is
> broken as well, but I have 4.6.2 running. Did process.h perhaps move
> between 1.7.10 and 1.7.11? I guess configure must be using linker
> rather than preprocessor tests for presence of spawnve, because it
> thinks (correctly) that the function exists.

See http://cygwin.com/ml/cygwin-announce/2012-02/msg00041.html
We moved it back.  If you have it in cygwin/process.h, you didn't
update from 1.7.10 to 1.7.11.


Corinna

-- 
Corinna Vinschen                  Please, send mails regarding Cygwin to
Cygwin Project Co-Leader          cygwin AT cygwin DOT com
Red Hat

--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple

Re: gcc-4.7.0-RC-20120302 fails to build for i686-pc-cygwin

[Ryan Johnson]

On 07/03/2012 8:07 AM, Corinna Vinschen wrote:
> On Mar  7 07:54, Ryan Johnson wrote:
>> Hi all,
>>
>> I tried to bootstrap the gcc-4.7 RC and it fails because it expects
>> to find<process.h>  and the file actually lives in
>> <cygwin/process.h>  (see
>> http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52513).
>>
>> I'm asking here because the gcc devs thought this would mean 4.6 is
>> broken as well, but I have 4.6.2 running. Did process.h perhaps move
>> between 1.7.10 and 1.7.11? I guess configure must be using linker
>> rather than preprocessor tests for presence of spawnve, because it
>> thinks (correctly) that the function exists.
> See http://cygwin.com/ml/cygwin-announce/2012-02/msg00041.html
> We moved it back.  If you have it in cygwin/process.h, you didn't
> update from 1.7.10 to 1.7.11.
Ah, I do remember that, now that you mention, but I was running a 1.7.11 
snapshot and forgot to upgrade...

Thanks for the quick reply,
Ryan


--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple

Re: gcc-4.7.0-RC-20120302 fails to build for i686-pc-cygwin

[Dave Korn]

On 07/03/2012 13:34, Ryan Johnson wrote:
> On 07/03/2012 8:07 AM, Corinna Vinschen wrote:
>> On Mar  7 07:54, Ryan Johnson wrote:
>>> Hi all,
>>>
>>> I tried to bootstrap the gcc-4.7 RC and it fails because it expects
>>> to find<process.h>  and the file actually lives in
>>> <cygwin/process.h>  (see
>>> http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52513).
>>>
>>> I'm asking here because the gcc devs thought this would mean 4.6 is
>>> broken as well, but I have 4.6.2 running. Did process.h perhaps move
>>> between 1.7.10 and 1.7.11? I guess configure must be using linker
>>> rather than preprocessor tests for presence of spawnve, because it
>>> thinks (correctly) that the function exists.
>> See http://cygwin.com/ml/cygwin-announce/2012-02/msg00041.html
>> We moved it back.  If you have it in cygwin/process.h, you didn't
>> update from 1.7.10 to 1.7.11.
> Ah, I do remember that, now that you mention, but I was running a 1.7.11
> snapshot and forgot to upgrade...
> 
> Thanks for the quick reply,
> Ryan

  Thanks for spotting that Ryan.  FTR, I figure it's not worth delaying the
GCC release to add a fix to support .10, since there were other significant
problems with it, and anyone who has it should be moving to .11 anyway.

    cheers,
      DaveK


--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple

Re: gcc-4.7.0-RC-20120302 fails to build for i686-pc-cygwin

[Denis Excoffier]

On Wed, Mar 07, 2012 at 01:38:43PM +0000, Dave Korn wrote:
>> On 07/03/2012 13:34, Ryan Johnson wrote:
>> > On 07/03/2012 8:07 AM, Corinna Vinschen wrote:
>> >> On Mar  7 07:54, Ryan Johnson wrote:
>> >>> Hi all,
>> >>>
>> >>> I tried to bootstrap the gcc-4.7 RC and it fails because it expects
>> >>> to find<process.h>  and the file actually lives in
>> >>> <cygwin/process.h>  (see
>> >>> http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52513).
>> >>>
>> >>> I'm asking here because the gcc devs thought this would mean 4.6 is
>> >>> broken as well, but I have 4.6.2 running. Did process.h perhaps move
>> >>> between 1.7.10 and 1.7.11? I guess configure must be using linker
>> >>> rather than preprocessor tests for presence of spawnve, because it
>> >>> thinks (correctly) that the function exists.
>> >> See http://cygwin.com/ml/cygwin-announce/2012-02/msg00041.html
>> >> We moved it back.  If you have it in cygwin/process.h, you didn't
>> >> update from 1.7.10 to 1.7.11.
>> > Ah, I do remember that, now that you mention, but I was running a 1.7.11
>> > snapshot and forgot to upgrade...
>> > 
>> > Thanks for the quick reply,
>> > Ryan
>> 
>>   Thanks for spotting that Ryan.  FTR, I figure it's not worth delaying the
>> GCC release to add a fix to support .10, since there were other significant
>> problems with it, and anyone who has it should be moving to .11 anyway.
>> 

Perhaps we could nevertheless find a means to move <process.h>
into <cygwin/process.h>, while not breaking GCC compilation during
the mean time? A patch to GCC-4.8 to search first for <cygwin/process.h>
and <process.h> if not found? And, for Cygwin,
cp -p cygwin/process.h process.h? SHTDI.

Regards,

Denis Excoffier.

--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple

Re: gcc-4.7.0-RC-20120302 fails to build for i686-pc-cygwin

[Christopher Faylor]

On Thu, Mar 22, 2012 at 08:25:09AM +0100, Denis Excoffier wrote:
>On Wed, Mar 07, 2012 at 01:38:43PM +0000, Dave Korn wrote:
>>> On 07/03/2012 13:34, Ryan Johnson wrote:
>>> > On 07/03/2012 8:07 AM, Corinna Vinschen wrote:
>>> >> On Mar  7 07:54, Ryan Johnson wrote:
>>> >>> Hi all,
>>> >>>
>>> >>> I tried to bootstrap the gcc-4.7 RC and it fails because it expects
>>> >>> to find<process.h>  and the file actually lives in
>>> >>> <cygwin/process.h>  (see
>>> >>> http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52513).
>>> >>>
>>> >>> I'm asking here because the gcc devs thought this would mean 4.6 is
>>> >>> broken as well, but I have 4.6.2 running. Did process.h perhaps move
>>> >>> between 1.7.10 and 1.7.11? I guess configure must be using linker
>>> >>> rather than preprocessor tests for presence of spawnve, because it
>>> >>> thinks (correctly) that the function exists.
>>> >> See http://cygwin.com/ml/cygwin-announce/2012-02/msg00041.html
>>> >> We moved it back.  If you have it in cygwin/process.h, you didn't
>>> >> update from 1.7.10 to 1.7.11.
>>> > Ah, I do remember that, now that you mention, but I was running a 1.7.11
>>> > snapshot and forgot to upgrade...
>>> > 
>>> > Thanks for the quick reply,
>>> > Ryan
>>> 
>>>   Thanks for spotting that Ryan.  FTR, I figure it's not worth delaying the
>>> GCC release to add a fix to support .10, since there were other significant
>>> problems with it, and anyone who has it should be moving to .11 anyway.
>>> 
>
>Perhaps we could nevertheless find a means to move <process.h>
>into <cygwin/process.h>, while not breaking GCC compilation during
>the mean time? A patch to GCC-4.8 to search first for <cygwin/process.h>
>and <process.h> if not found? And, for Cygwin,
>cp -p cygwin/process.h process.h? SHTDI.

Why are we still talking about this?  We moved the header file back
where it has been for a decade or more.  The case is now closed.

--
Problem reports:       http://cygwin.com/problems.html
FAQ:                   http://cygwin.com/faq/
Documentation:         http://cygwin.com/docs.html
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple