unused return values causing build failures in HEAD

unused return values causing build failures in HEAD

[Julian Graham]

Hey everyone,

So I'm trying to build Guile from HEAD after a hard drive wipe, and it
looks like the version of gcc that ships with Ubuntu 8.10, 4.3.2, now
warns when you don't check function return values, which of course
breaks Guile's warnings-as-errors build.  There are several offending
instances of this, and the functions they use are write, read, pipe,
and nice.  I'm happy to submit a patch that raises SCM_SYSERROR or
whatever in the appropriate places -- unless there's some other way
around this.


Regards,
Julian

Re: unused return values causing build failures in HEAD

[Ludovic Courtès]

Hi Julian,

"Julian Graham" <joolean@gmail.com> writes:

> So I'm trying to build Guile from HEAD after a hard drive wipe, and it
> looks like the version of gcc that ships with Ubuntu 8.10, 4.3.2, now
> warns when you don't check function return values, which of course
> breaks Guile's warnings-as-errors build.  There are several offending
> instances of this, and the functions they use are write, read, pipe,
> and nice.  I'm happy to submit a patch that raises SCM_SYSERROR or
> whatever in the appropriate places -- unless there's some other way
> around this.

You report appears to be similar to #24009:

  https://savannah.gnu.org/bugs/?24009

I'll see what can be done.

Thanks,
Ludo'.

Re: unused return values causing build failures in HEAD

[Julian Graham]

Oops, should've checked Savannah first.  Thanks for fixing that!


On Sun, Nov 30, 2008 at 11:35 AM, Ludovic Courtès <ludo@gnu.org> wrote:
> Hi Julian,
>
> "Julian Graham" <joolean@gmail.com> writes:
>
>> So I'm trying to build Guile from HEAD after a hard drive wipe, and it
>> looks like the version of gcc that ships with Ubuntu 8.10, 4.3.2, now
>> warns when you don't check function return values, which of course
>> breaks Guile's warnings-as-errors build.  There are several offending
>> instances of this, and the functions they use are write, read, pipe,
>> and nice.  I'm happy to submit a patch that raises SCM_SYSERROR or
>> whatever in the appropriate places -- unless there's some other way
>> around this.
>
> You report appears to be similar to #24009:
>
>  https://savannah.gnu.org/bugs/?24009
>
> I'll see what can be done.