This is the mail archive of the
libffi-discuss@sourceware.org
mailing list for the libffi project.
Re: s390x ffi_closure_helper_SYSV
- From: Tom Tromey <tom at tromey dot com>
- To: libffi-discuss at sourceware dot org
- Date: Fri, 18 Dec 2015 08:09:09 -0700
- Subject: Re: s390x ffi_closure_helper_SYSV
- Authentication-results: sourceware.org; auth=none
- References: <5672CF2D dot 2030401 at gmail dot com> <20151217153018 dot CDD291513 at oc7340732750 dot ibm dot com>
>>>>> "Ulrich" == Ulrich Weigand <uweigand-tA70FqPdS9bQT0dZR+AlfA@public.gmane.org> writes:
Ulrich> While this is not explicitly mentioned in the closure docs, the
Ulrich> code makes the assumption that this is handled likewise for closure
Ulrich> return values. This is also explicitly checked for in the libffi
Ulrich> test suite, e.g. testsuite/libffi.call/cls_sshort.c
Actually the closure docs are actively wrong right now.
I didn't realize this same widening behavior applied to closures (in
fact I thought I asked about this and got the answer I documented, but I
can't find the email, so it was probably just my error).
I sent a github PR to correct the closure documentation.
Tom