[Bug binutils/15869] New: nm --size-sort test fails for alpha-unknown-linux-gnuecoff

[hp at sourceware dot org]

http://sourceware.org/bugzilla/show_bug.cgi?id=15869

            Bug ID: 15869
           Summary: nm --size-sort test fails for
                    alpha-unknown-linux-gnuecoff
           Product: binutils
           Version: 2.24 (HEAD)
            Status: NEW
          Severity: normal
          Priority: P2
         Component: binutils
          Assignee: unassigned at sourceware dot org
          Reporter: hp at sourceware dot org
              Host: i686-pc-linux-gnu
            Target: alpha-unknown-linux-gnuecoff

Build HEAD for alpha-unknown-linux-gnuecoff (*not* alpha-linux-ecoff); just an
ECOFF (or non-ELF) target. "make check-binutils" shows lots of failures,
including:

Running
/home/hp/binutils/fixmmix0820/src/binutils/testsuite/binutils-all/nm.exp ...
Version /home/hp/binutils/fixmmix0820/obE/binutils/nm-new 2.23.52.20130820
FAIL: nm (no arguments)
FAIL: nm -P
FAIL: nm --size-sort

The log for the --size-sort change shows:
/home/hp/binutils/fixmmix0820/obE/binutils/nm-new  --size-sort tmpdir/nm-1.o
Executing on host: /home/hp/binutils/fixmmix0820/obE/binutils/nm-new 
--size-sort tmpdir/nm-1.o   (timeout = 300)
00000008 T text_symbol2

0000000c T text_symbol1

0000000c T text_symbol3

00000008 T text_symbol2
0000000c T text_symbol1
0000000c T text_symbol3

FAIL: nm --size-sort

Note the unexpected symbol values; a plain "./nm-new tmpdir/nm-1.o" shows:
00000000 T text_symbol1
0000000c T text_symbol2
00000014 T text_symbol3

A guess is that the symbol values are fouled up in the generic nm size-sort
handler.

-- 
You are receiving this mail because:
You are on the CC list for the bug.