This is the mail archive of the
binutils@sourceware.org
mailing list for the binutils project.
Re: Your dwarf3 change to gas/dwarf2dbg.c caused gas testsuite regressions for cris-elf
- From: Hans-Peter Nilsson <hans-peter dot nilsson at axis dot com>
- To: rth at redhat dot com
- Cc: hans-peter dot nilsson at axis dot com, binutils at sourceware dot org
- Date: Wed, 7 Sep 2005 19:50:08 +0200
- Subject: Re: Your dwarf3 change to gas/dwarf2dbg.c caused gas testsuite regressions for cris-elf
> Date: Wed, 7 Sep 2005 10:32:27 -0700
> From: Richard Henderson <rth@redhat.com>
> On Wed, Sep 07, 2005 at 05:41:38PM +0200, Hans-Peter Nilsson wrote:
> > FAIL: gas/cris/rd-dw2-1
>
> What are you attempting to test here? My guess is that you're
> actually interested in seeing that the line numbers themselves,
> so the most appropriate fix is to match "Special opcode .*:".
The purpose is to check that the dwarf2 target hooks
(dwarf2_emit_insn) is in the right place and time; that address
and line information is correct. So your suggestion would
probably match. Besides stuff added to the "directory table"
and a new "opcode base", that is.
Hmm, just spotted in gas.log. This looks wrong:
line " Set ISA to 134555844"
(gas/cris/rd-dw2-10). That number, 134555844=0x80528c4 seems
like a common virtual address on i386-linux. Uninitialized
thingy?
Oh, and another FAIL: I hope assume it's safe to change that "1"
to "[1-8]":
x/src/gas/testsuite/lib/run ../as-new -o dump.o /h/hp/binutils/cvs_latest/src/gas/testsuite/gas/lns/lns-common-1.s
x/gas/testsuite/../../binutils/readelf -wl dump.o > dump.out
regexp_diff match failure
regexp "^ Minimum Instruction Length: 1$"
line " Minimum Instruction Length: 2"
FAIL: lns-common-1
brgds, H-P