ports/138772: emulators/qemu should build on SPARC

[Nathaniel Filardo]

>Number:         138772
>Category:       ports
>Synopsis:       emulators/qemu should build on SPARC
>Confidential:   no
>Severity:       non-critical
>Priority:       low
>Responsible:    freebsd-ports-bugs
>State:          open
>Quarter:        
>Keywords:       
>Date-Required:
>Class:          sw-bug
>Submitter-Id:   current-users
>Arrival-Date:   Sat Sep 12 19:40:04 UTC 2009
>Closed-Date:
>Last-Modified:
>Originator:     Nathaniel Filardo
>Release:        CURRENT
>Organization:
>Environment:
FreeBSD nwf 8.0-BETA2 FreeBSD 8.0-BETA2 #2: Sun Aug 16 13:00:36 EDT 2009     root at hydra.priv.oc.ietfng.org:/systank/obj/systank/src/sys/NWFKERN  sparc64

>Description:
emulators/qemu excludes sparc64 via ONLY_FOR_ARCH but there's relatively little that seems to need to change to fix that.
>How-To-Repeat:

>Fix:
Relaxen and watchen das blinkenlights for libmath; this hack for i386 and x68_64 is unneded on sparc64.

dyngen-exec.h needs to be told that we define {u,}int64_t already

cpu-exec.c needed to include sys/ucontext.h on FreeBSD too, not just Linux

cpu-exec.c:cpu_signal_handler needed to be told in the right spot :
#elif defined(__FreeBSD__)
    ucontext_t *uc = puc;
    void *sigmask = &uc->sigmask;
    unsigned long pc = uc->uc_mcontext.mc_tpc;

How much of this should be worked around in ports vs. how much should go upstream I am not sure.  I'm also not sufficiently familiar with the workings of ports to propose a patch, sorry. :)

>Release-Note:
>Audit-Trail:
>Unformatted: