This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [PATCH]: Fix PR rtl-optimization/31344
- From: "Uros Bizjak" <ubizjak at gmail dot com>
- To: "Ian Lance Taylor" <iant at google dot com>
- Cc: gcc-patches <gcc-patches at gcc dot gnu dot org>, "Jan Hubicka" <jh at suse dot cz>, "H. J. Lu" <hjl at lucon dot org>, "François-Xavier Coudert" <fxcoudert at gmail dot com>
- Date: Wed, 9 May 2007 10:00:47 +0200
- Subject: Re: [PATCH]: Fix PR rtl-optimization/31344
- References: <5787cf470705070309q61c3b8efp1d0ec85c617eeac1@mail.gmail.com> <m37irjztzs.fsf@localhost.localdomain>
On 07 May 2007 23:53:27 -0700, Ian Lance Taylor <iant@google.com> wrote:
I think something more like H.J.'s patch is likely to be right,
although I'm not very happy with his patch either. There is a lot of
special handling in expr.c for push instructions, and this is another
place which needs special handling.
The solution to this problem now looks to be stalled by two opinions:
a) i386 maintainer (Jan) thinks that this is a middle-end problem [1]
b) middle-end maintainer thinks this is a target problem
Two "competing" patches are:
a) target patch [2] that avoids code-path that produces invalid insn.
However, it is said in [2], that many tests will fail [4] with
proposed patch for i386-pc-linux-gnu.
b) middle-end patch [3] that avoids generation of invalid insn. There
are _zero_ new failures [5].
I don't mind that I'm an author of [3], I have no preference on either
fix, but I would like to fix this problem, as it is known for almost 3
months now.
Could I kindly ask for a final decision?
[1]: http://gcc.gnu.org/ml/gcc-patches/2007-04/msg01987.html
[2]: http://gcc.gnu.org/ml/gcc-patches/2007-03/msg01764.html
[3]: http://gcc.gnu.org/ml/gcc-patches/2007-05/msg00525.html
[4]: http://gcc.gnu.org/ml/gcc-testresults/2007-03/msg01311.html
[5]:
Native configuration is i386-pc-linux-gnu
=== gcc tests ===
Running target unix
FAIL: gcc.c-torture/execute/990127-2.c execution, -Os
FAIL: gcc.c-torture/execute/mayalias-2.c compilation, -O3 -g
(internal compiler error)
UNRESOLVED: gcc.c-torture/execute/mayalias-2.c execution, -O3 -g
FAIL: gcc.c-torture/execute/mayalias-3.c compilation, -O3 -g
(internal compiler error)
UNRESOLVED: gcc.c-torture/execute/mayalias-3.c execution, -O3 -g
XPASS: gcc.dg/tree-ssa/20040204-1.c scan-tree-dump-times link_error 0
XPASS: gcc.dg/tree-ssa/update-threading.c scan-tree-dump-times Invalid sum 0
FAIL: gcc.target/i386/pr21291.c (test for excess errors)
FAIL: gcc.target/i386/sse-vect-types.c (test for excess errors)
FAIL: gcc.target/i386/sse4a-extract.c (test for excess errors)
WARNING: gcc.target/i386/sse4a-extract.c compilation failed to produce
executable
FAIL: gcc.target/i386/sse4a-insert.c (test for excess errors)
WARNING: gcc.target/i386/sse4a-insert.c compilation failed to produce executable
FAIL: gcc.target/i386/sse4a-montsd.c (test for excess errors)
WARNING: gcc.target/i386/sse4a-montsd.c compilation failed to produce executable
FAIL: gcc.target/i386/sse4a-montss.c (test for excess errors)
WARNING: gcc.target/i386/sse4a-montss.c compilation failed to produce executable
=== gcc Summary ===
# of expected passes 45039
# of unexpected failures 9
# of unexpected successes 2
# of expected failures 120
# of unresolved testcases 2
# of untested testcases 28
# of unsupported tests 301
/home/uros/gcc-sandbox/gcc/xgcc version 4.3.0 20070508 (experimental)
Uros.