This is the mail archive of the crossgcc@sources.redhat.com mailing list for the crossgcc project.

See the CrossGCC FAQ for lots more information.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

glibc-2.2.5 regression test results with sh4 and gcc-3.3


Hi all,
I can build x86->sh4 crosstoolchains with gcc-3.3/glibc-2.2.5 from scratch
without any trouble, and the resulting toolchains pass most of the
gcc regression tests.  However, there is trouble with many glibc tests.
(I know, glibc-2.2.5 is old, and I should be testing glibc-2.3.2.
I will try that again soon; last time I had trouble getting started with it.)

For what it's worth, the patches I'm using are
http://www.kegel.com/crosstool/crosstool-0.14/glibc-2.2.5-patches/
and
http://www.kegel.com/crosstool/crosstool-0.14/gcc-ss-3_3-20030714-patches/
(These are the minimal patches to get gcc/glibc to compile on all
the platforms glibc supports, and to get the test suites to work remotely.)

The trouble breaks down into four basic symptoms; see the summary below.
If you know of patches that address any of these, please let me know.
I've looked briefly at the patches in
http://www.sh-linux.org/rpm-2003/SRPMS/glibc-2.2.5-8.src.rpm
but didn't see anything obviously related.
Thanks!
- Dan

------------------

One test, test-fpucw, terminates with the message
  control word is 0x80004 but should be 0x80000.

The code for that test is simple:
  _FPU_GETCW (cw);

cw &= ~_FPU_RESERVED;

  if (cw != (_FPU_DEFAULT & ~_FPU_RESERVED))
    printf ("control word is 0x%lx but should be 0x%lx.\n",
            (long int) cw, (long int) (_FPU_DEFAULT & ~_FPU_RESERVED));

------------------

The following tests fail with:
  Floating point exception
  Error 136

test-fenv
test-misc
test-float
test-double
test-ifloat
test-idouble

------------------

The following tests fail with:
  Didn't expect signal from child: got `File size limit exceeded'
  Error 1

tst-preadwrite
tst-preadwrite64

(Daniel Jacobowitz posted a patch that may fix pread/pwrite; it's
at http://www.kegel.com/crosstool/crosstool-0.14/glibc-2.3.2-patches/glibc-drow-sh.patch
but I haven't really looked at it, nor have I tried backporting the
pread/pwrite part of the patch to glibc-2.2.5 yet.)

------------------

The following tests fail with errors like:
  Operation failed : Invalid argument
  aio_write: wrong size: 100, should be 1000
  comparison failed for aio_read test
  lio_listio (write): wrong size: 0, should be 1000
  aio_fsync (aio_write): wrong size: 0, should be 1000
  Error 1

tst-aio
tst-aio64

------------------

--
Dan Kegel
http://www.kegel.com
http://counter.li.org/cgi-bin/runscript/display-person.cgi?user=78045


------ Want more information? See the CrossGCC FAQ, http://www.objsw.com/CrossGCC/ Want to unsubscribe? Send a note to crossgcc-unsubscribe@sources.redhat.com


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]