This is the mail archive of the
binutils@sources.redhat.com
mailing list for the binutils project.
Re: sh-linux is now broken
- To: "H . J . Lu" <hjl at lucon dot org>
- Subject: Re: sh-linux is now broken
- From: NIIBE Yutaka <gniibe at m17n dot org>
- Date: Fri, 2 Nov 2001 23:35:43 +0900 (JST)
- Cc: binutils at sourceware dot cygnus dot com
- References: <20011031131037.A2203@lucon.org><20011031131238.A2284@lucon.org><200111010130.fA11UZr10017@mule.m17n.org><20011031214452.C9712@lucon.org><200111010556.fA15uNk12013@mule.m17n.org>
Conclusion: Those are not regressions by my change. Those tests
depend on GCC. I'm using (modified) GCC 3.0.2.
NIIBE Yutaka wrote:
> Yes, sure. I've fix some of them. I'm currently investigating following
> four cases remained:
>
> > -------- Unexpected failures were:
> > Running /home/gniibe/gniibe/c/src/ld/testsuite/ld-selective/selective.exp ...
> > FAIL: selective4
> > FAIL: selective5
This test depends on C++. I'll check Linux/x86 case, and if it will
emit same symbols, I'll change the test cases as expected failure.
> > -------- Unexpected successes were:
> > Running /home/gniibe/gniibe/c/src/ld/testsuite/ld-undefined/undefined.exp ...
> > XPASS: undefined function
> > XPASS: undefined line
It correctry emit function and the line number. I don't know the
reason why it's not expected to pass. I'll check another architecture
which uses DWARF2. Possibly, I'll change the test case as expected to
pass.
Later,
--