public inbox for gcc@gcc.gnu.org
 help / color / mirror / Atom feed
* gnat regressions on sparc, mips and x86
@ 2009-09-14 18:52 Joel Sherrill
  0 siblings, 0 replies; only message in thread
From: Joel Sherrill @ 2009-09-14 18:52 UTC (permalink / raw)
  To: gcc

[-- Attachment #1: Type: text/plain, Size: 811 bytes --]

Hi,

I have re-run GNAT/RTEMS on a number of targets
for 4.3.4, 4.4.1 and the SVN head.  I kept the
binutils, gdb, newlib and RTEMS the same.  The SVN head
has a lot of new failures for SPARC, MIPS, and x86.
The SPARC and MIPS have the same number of failures
(319) so I assume the same underlying cause.

The x86 has gone from about 20 failures which
are mostly deficiencies in qemu FPU accuracy
to about 225 with many indicating hardware exceptions.

And FWIW I was unable to build ARM on the head. :(

I am attaching my summary notes but would appreciate
any feedback.

-- 
Joel Sherrill, Ph.D.             Director of Research & Development
joel.sherrill@OARcorp.com        On-Line Applications Research
Ask me about RTEMS: a free RTOS  Huntsville AL 35805
   Support Available             (256) 722-9985



[-- Attachment #2: acats-summary.txt --]
[-- Type: text/plain, Size: 1221 bytes --]

NOTE: c380004 requires large stack?

powerpc/psim
============
+ 4.3.4
  - 2 fails: c380004 c974013
+ 4.4.1
  - 2 fails: c380004 c974013
+ SVN (4.5.0 20090910 r151592)
  - 2 fails: c380004 c974013

sparc/sis
=========
+ 4.3.4
  - 2 fails: c380004 c974013
+ 4.4.1
  - 2 fails: c380004 c974013
+ SVN (4.5.0 20090910 r151592)
  - 319 fails: including c380004 c974013

mips/jmr3904
============
+ 4.3.4
  - 3 fails: c380004 cxf3a01 cxf3a02
     cxf3a01, cxf3a02 - read of unmapped address
+ 4.4.1
  - 1 fails: c380004
+ SVN (4.5.0 20090910 r151592)
  - 319 fails: many unhandled exception 9

arm/edb7312
===========
+ 4.3.4
  - ? did it build
+ 4.4.1
  - ? did it build
+ SVN (4.5.0 20090910 r151592)
  - bug box

i386/pc386
==========
+ 4.3.4
  - 54 fails: c380004 plus
    many hardware exceptions
+ 4.4.1
  - 20 fails: c380004 plus
    c953002 - ran too long (has output)
    cc1307b - ran too long (but no output)

    cxg2002 cxg2003 cxg2004 cxg2006 cxg2007 cxg2010
    cxg2011 cxg2012 cxg2013 cxg2014 cxg2015 cxg2016
    cxg2017 cxg2018 cxg2019 cxg2020 cxg2021 - Qemu
    has problems with FPU simulation and fails.
    The pass on real hardware.

+ SVN (4.5.0 20090910 r151592)
  - 225 fails, many hardware exceptions

^ permalink raw reply	[flat|nested] only message in thread

only message in thread, other threads:[~2009-09-14 18:52 UTC | newest]

Thread overview: (only message) (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2009-09-14 18:52 gnat regressions on sparc, mips and x86 Joel Sherrill

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).