[Bug inline-asm/31705] New: inline assembler causes ICE: extract_constrain_insn_cached, at recog.c:2002

[Bug inline-asm/31705] New: inline assembler causes ICE: extract_constrain_insn_cached, at recog.c:2002

[fultonm at ca dot ibm dot com]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 2618 bytes --]

The source file:

double helperDoubleTan(double * a, double * b)
{
        double tmpA, tmpB;

        tmpA = *a;
        {
                double y[3] = { 0.0, 0.0, 0.0 };
                double r;
                int lq;

                lq = __ieee754_rem_pio2(tmpA, y);
                r = y[0];
                if (lq & 0x1) {
                  __asm__ __volatile__("fptan; fdivp;fchs;":
"=&t"(tmpB):"f"(r));
                  tmpB = __kernel_tan(y[0], y[1], 1 - ((lq & 1) << 1));
                } else {
                        tmpB = tan(r);
                }
        }
        *b = tmpB;
        return 0;
}

compiled with gcc -O -c testcase.c

results in:

testcase.c: In function ‘helperDoubleTan’:
testcase.c:17: warning: incompatible implicit declaration of built-in function
‘tan’
testcase.c:22: error: unrecognizable insn:
(insn 36 34 96 (parallel [
            (set (reg/v:DF 8 st [orig:58 tmpB.26 ] [58])
                (asm_operands/v:DF ("fptan; fdivp;fchs;") ("=&t") 0 [
                        (reg:DF 8 st)
                    ]
                     [
                        (asm_input:DF ("f"))
                    ] ("testcase.c") 14))
            (clobber (reg:QI 19 dirflag))
            (clobber (reg:QI 18 fpsr))
            (clobber (reg:QI 17 flags))
        ]) -1 (nil)
    (expr_list:REG_UNUSED (reg:QI 17 flags)
        (expr_list:REG_UNUSED (reg:QI 18 fpsr)
            (expr_list:REG_UNUSED (reg:QI 19 dirflag)
                (expr_list:REG_UNUSED (reg/v:DF 8 st [orig:58 tmpB.26 ] [58])
                    (nil))))))

Output from gcc -v:

gcc -v
Using built-in specs.
Target: i486-linux-gnu
Configured with: ../src/configure -v
--enable-languages=c,c++,fortran,objc,obj-c++,treelang --prefix=/usr
--enable-shared --with-system-zlib --libexecdir=/usr/lib
--without-included-gettext --enable-threads=posix --enable-nls
--program-suffix=-4.1 --enable-__cxa_atexit --enable-clocale=gnu
--enable-libstdcxx-debug --enable-mpfr --enable-checking=release i486-linux-gnu
Thread model: posix
gcc version 4.1.2 (Ubuntu 4.1.2-0ubuntu4)


-- 
           Summary: inline assembler causes ICE:
                    extract_constrain_insn_cached, at recog.c:2002
           Product: gcc
           Version: 4.1.2
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: inline-asm
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: fultonm at ca dot ibm dot com
 GCC build triplet:  i486-linux-gnu
  GCC host triplet:  i486-linux-gnu
GCC target triplet:  i486-linux-gnu


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=31705

[Bug target/31705] inline assembler causes ICE: extract_constrain_insn_cached, at recog.c:2002

[ubizjak at gmail dot com]

------- Comment #1 from ubizjak at gmail dot com  2007-04-26 08:22 -------
(In reply to comment #0)

> __asm__ __volatile__("fptan; fdivp;fchs;": "=&t"(tmpB):"f"(r));

These constraints are wrong. You need:

  __asm__ __volatile__ ("fptan; fdivp;fchs;" : "=t" (tmpB) : "0"(r) : "st(1)");


Also note that tmpB is unused, as you rewrite it in next statement.


-- 

ubizjak at gmail dot com changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
         Resolution|                            |INVALID


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=31705

[Bug target/31705] inline assembler causes ICE: extract_constrain_insn_cached, at recog.c:2002

[fultonm at ca dot ibm dot com]

------- Comment #2 from fultonm at ca dot ibm dot com  2007-04-26 16:17 -------
Thank you for the information. I guess the follow-on question is if generating
an internal compiler error is appropriate? I realize it's inline assembler, but
are you not able to recover from an invalid entry? Also - I should have
mentioned that at no-opt, there is no message or error. Perhaps no-opt is
generating invalid code?


-- 

fultonm at ca dot ibm dot com changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|RESOLVED                    |UNCONFIRMED
         Resolution|INVALID                     |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=31705