bug-binutils
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Bug ld/25677] Changes to --*magic options for pdp11-aout target


From: nickc at redhat dot com
Subject: [Bug ld/25677] Changes to --*magic options for pdp11-aout target
Date: Mon, 30 Mar 2020 12:12:50 +0000

https://sourceware.org/bugzilla/show_bug.cgi?id=25677

--- Comment #5 from Nick Clifton <nickc at redhat dot com> ---
Hi Stephen,

  Judging by the attached log, you seem to have gotten the hang of running the
linker testsuite.  Do not worry that there are quite a few failures for the
pdp11-aout target.  This is fairly normal with older targets that do not have
active maintainers.

  The important issue is the extract symbols failure in the script.exp tests:

    Running ../../ld/testsuite/ld-scripts/script.exp ...
    [...]
    FAIL: extract symbols

According to my logs this failure is new.  Ie it only happens after applying
your patch.  To find out more about the failure, have a look in the ld.log file
which should be in the ld directory.  When the test passes the log should
contain lines like this:

     [...]
    /home/nickc/work/builds/binutils/current/pdp11-dec-aout/ld/../binutils/size
 tmpdir/extract
    Executing on host: sh -c
{/home/nickc/work/builds/binutils/current/pdp11-dec-aout/ld/../binutils/size 
tmpdir/extract 2>&1}  /dev/null ld.tmp (timeout = 300)
    spawn [open ...]
       text        data     bss     dec     hex filename
          0           0       0       0       0 tmpdir/extract
       text        data     bss     dec     hex filename
          0           0       0       0       0 tmpdir/extract
    PASS: extract symbols

Note how the data field is zero.  What I found with your patch applied is that
this field is now 4.  So presumably some new data has been added to the
"extract" binary.

It may be that the new behaviour is correct, the data should be there.  Note
however that the extract binary has been created by running "objcopy
--extract-symbol" on another file, (see earlier in ld.log), so normally you
would only expect symbols to be present in the output file, not data.  So you
need to work out if the presence of the data represents a bug or not.

If the data should be there you will need to update the
ld/testsuite/ld-scripts/script.exp file.  The extract_symbol_test procedure
(starting at line 110) is where you need to look, and the specific test starts
at line 174.  There are some examples above of adding "expected failure"
results for other targets (and for earlier parts of the test), so you can use
those as a guide.

Cheers
  Nick

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


reply via email to

[Prev in Thread] Current Thread [Next in Thread]