[PULL] Xserver build fixes for Windows MinGW target.

Jon TURNEY jon.turney at dronecode.org.uk
Wed Nov 28 06:35:53 PST 2012

On 27/11/2012 23:08, Keith Packard wrote:
> Jon TURNEY writes:
>>       os/utils.c: Fix compilation of OsBlockSIGIO when SIGIO isn't defined
> This should probably look like this instead, to avoid compilers
> complaining about unreachable code:
> diff --git a/os/utils.c b/os/utils.c
> index 3c520ad..0d77b84 100644
> --- a/os/utils.c
> +++ b/os/utils.c
> @@ -1227,10 +1227,10 @@ OsBlockSIGIO(void)
>          sigprocmask(SIG_BLOCK, &set, &PreviousSigIOMask);
>          ret = sigismember(&PreviousSigIOMask, SIGIO);
>          return ret;
> -    } else
> -        return 1;
> +    }
>  #endif
>  #endif
> +    return 1;
>  }

Good point.

>> Yaakov Selkowitz (2):
>>       hw/xwin: use raise instead of signal
> This should be fine (it's in POSIX.1-2001 according to the Linux
> raise(3) man page). If we find a system this breaks on, it will be easy
> to fix (testing for raise in the c library and using it only if present).

I'm not quite sure what you are telling me here.

Since this code is in hw/xwin, it's only ever going to be compiled for Windows
targets, anyhow.

Since raise() is in C89, I hope systems that it isn't available on are rare :-)

(Also, it looks like I made a mistake when writing the comment, as it should
say "use raise() instead of kill()" to match the code)

> The other patches were also Reviewed-by: Keith Packard <keithp at keithp.com>

Thanks for taking the time to review these.

More information about the xorg-devel mailing list