This is the mail archive of the
binutils@sources.redhat.com
mailing list for the binutils project.
Re: PATCH: Set XFAIL for non PE targets
- From: "H. J. Lu" <hjl at lucon dot org>
- To: "Aaron W. LaFramboise" <aaron98wiridge9 at aaronwl dot com>
- Cc: binutils at sources dot redhat dot com
- Date: Wed, 27 Oct 2004 07:51:41 -0700
- Subject: Re: PATCH: Set XFAIL for non PE targets
- References: <20041026232647.GA25570@lucon.org> <417EFC83.4060905@aaronwl.com>
On Tue, Oct 26, 2004 at 08:40:19PM -0500, Aaron W. LaFramboise wrote:
> H. J. Lu wrote:
>
> > +if { "$target_xfail" == "yes" } {
> > + setup_xfail *-*
> > +}
> > +
>
> Theres some code near the top of the file:
>
> > if {![istarget "i*86-*-*"]} {
> > return
> > }
> >
> > if {![istarget "i*86-*-*pe*"] \
> > && ![istarget "i*86-*-cygwin*"] \
> > && ![istarget "i*86-*-mingw32*"] } {
> > set target_xfail "yes"
> > } else {
> > set target_xfail "no"
> > }
>
> Does the variable target_xfail do anything by itself? Could `set
No, it is a local variable.
> target_xfail "yes"' just be changed to setup_xfail?
>
> Also, the first line just returns if the target isn't i386. Is this
> right? Should it call unsupported first?
>
> In fact, these tests really have nothing to do with i386 any more than
> anything else; they're tests for PE targets. In other words, i386-elf
> shouldn't be xfailed: it should just be unsupported, the same as
> arm-elf. And, for that matter, all of *-pe, such as arm-pe, should
> really be xfailed, because those targets really should be able to handle
> all dlltool tests.
>
If PE is enable, the binutils will support PE, except for assembler
which will only support the default target in most cases. If the test
doesn't call assembler, it should work.
H.J.