This is the mail archive of the
binutils@sourceware.org
mailing list for the binutils project.
Re: Yet another fill-1 test fix
- From: Hans-Peter Nilsson <hp at bitrange dot com>
- To: Alan Modra <amodra at gmail dot com>
- Cc: binutils at sourceware dot org, Dave Brolley <brolley at redhat dot com>
- Date: Wed, 25 Oct 2017 11:47:06 -0400 (EDT)
- Subject: Re: Yet another fill-1 test fix
- Authentication-results: sourceware.org; auth=none
- References: <20171025075343.GJ27132@bubble.grove.modra.org>
On Wed, 25 Oct 2017, Alan Modra wrote:
> tic4x fails due to being a 4 octets per byte target, while tic54x is 2
> octets per byte.
>
> mmix still fails with
> fill-1.s:4: Error: unknown pseudo-op: `.l1:'
> fill-1.s:6: Error: unknown pseudo-op: `.l2:'
> fill-1.s:3: Error: .space specifies non-absolute value
But the .l1 error was gone with s/.L1/L1/, right?
> and if the labels are changed to L1 and L2 to avoid the mmix fail then
> mep-elf fails with
> fill-1.s:3: Error: .space specifies non-absolute value
> due to generating a fancy RELC reloc that isn't resolved.
>
> Since both of those look like they ought to be investigated by the
> target maintainers, I'm tweaking the test to fail on both targets.
I'm happy just knowing that the test passes for MMIX at this
time if .L1 and .L2 are both temporarily changed to L1 and L2
(".L1" is not a valid symbol), but thought that edit that may
not be best for all targets and as you noticed, another target
then fails, so I'm happy with the not-target. That's what I
intended to commit anyway, and I'll do that later if not sooner.
> That's not because I think the bugs ought to be fixed, but just that I
> wasn't comfortable sweeping this under the rug myself by adding mmix
> or mep to not-target.
I think it's reasonably investigated. The ".space specifies
non-absolute value" reasonably comes from the labels regarded as
non-local when the syntax trying to define them was invalid.
Thanks for having a look. Are there other suspicions?
brgds, H-P