This is the mail archive of the gcc@gcc.gnu.org mailing list for the GCC project.


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

[tree-ssa] mainline destabilizing tree-ssa?


Hi!

I trip over more and more problems coming possibly from mainline merges,
like PR14579 or the following ICE

tramp3d-v3.cpp:10547: error: unable to find a register to spill in class
`AREG'
tramp3d-v3.cpp:10547: error: this is the insn:
(insn:HI 38 37 39 2 tramp3d-v3.cpp:10534 (parallel [
            (set (reg:SI 1 dx [90])
                (truncate:SI (lshiftrt:DI (mult:DI (zero_extend:DI (reg:SI
3 bx [88]))
                            (zero_extend:DI (reg:SI 1 dx [91])))
                        (const_int 32 [0x20]))))
            (clobber (scratch:SI))
            (clobber (reg:CC 17 flags))
        ]) 179 {*umulsi3_highpart_insn} (insn_list 37 (nil))
    (expr_list:REG_UNUSED (reg:CC 17 flags)
        (expr_list:REG_UNUSED (scratch:SI)
            (expr_list:REG_DEAD (reg:SI 3 bx [88])
                (expr_list:REG_DEAD (reg:SI 1 dx [91])
                    (expr_list:REG_UNUSED (reg:CC 17 flags)
                        (expr_list:REG_UNUSED (scratch:SI)
                            (expr_list:REG_EQUAL (truncate:SI (lshiftrt:DI
(mult:DI (zero_extend:DI (reg:SI 3 bx [88]))
                                            (const_double -1431655765
[0xaaaaaaab] 0 [0x0] 0 [0x0] 0 [0x0] 0 [0x0] 0 [0x0]))
                                        (const_int 32 [0x20])))
                                (nil)))))))))
tramp3d-v3.cpp:10547: internal compiler error: in spill_failure, at
reload1.c:1887
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://gcc.gnu.org/bugs.html> for instructions.

triggered by -funroll-loops.

This makes it difficult to provide up-to-date performance comparisons.
Also it seems to require me tracking mainline as well (which I didn't do
for now, as tree-ssa and 3.4 seemed enough) - reporting these problems
against tree-ssa probably makes no sense.

Can we please have more selective mainline merges? I.e. merge only
bugfixes?

Thanks,

Richard.

--
Richard Guenther <richard dot guenther at uni-tuebingen dot de>
WWW: http://www.tat.physik.uni-tuebingen.de/~rguenth/


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