[RFA] Code gen improvement for sim-decode.scm

[RFA] Code gen improvement for sim-decode.scm

[Ben Elliston]

The following patch improves the generated extraction code -- such that an
`insn' variable is not defined for instructions that have zero ifields.

Comments?  Is there a better way to do this?

Ben


@@ -433,10 +433,13 @@
    " extract_" (gen-sym sfmt) ":\n"
    "  {\n"
    "    const IDESC *idesc = &" IDESC-TABLE-VAR "[itype];\n"
-   "    CGEN_INSN_INT insn = "
-   (if (adata-integral-insn? CURRENT-ARCH)
-       "entire_insn;\n"
-       "base_insn;\n")
+   (if (> (length (sfmt-iflds sfmt)) 0)
+       (string-append
+       "    CGEN_INSN_INT insn = "
+       (if (adata-integral-insn? CURRENT-ARCH)
+           "entire_insn;\n"
+           "base_insn;\n"))
+       "")
    (gen-define-field-macro sfmt)
    (gen-define-ifields (sfmt-iflds sfmt) (sfmt-length sfmt) "    " #f)
    "\n"

Re: [RFA] Code gen improvement for sim-decode.scm

[Frank Ch. Eigler]

Hi -

On Wed, Dec 06, 2000 at 11:40:18PM +1100, Ben Elliston wrote:
: The following patch improves the generated extraction code -- such that an
: `insn' variable is not defined for instructions that have zero ifields.
: [...]

Are you convinced that nothing can possibly use that variable
except ifield extraction?  Is the improvement to generated code
merely the removal of a gcc warning?  (It should be able to
zap an unused variable, resulting in no performance difference.)

- FChE
-- 
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.4 (GNU/Linux)
Comment: For info see http://www.gnupg.org

iD8DBQE6LumRVZbdDOm/ZT0RAsT+AJ9jjaI6pCyUw7ASE3CppAUI6DmqHwCcCB2+
OeT+D/iC9Q1Xm1ePIvynAH4=
=06so
-----END PGP SIGNATURE-----

Re: [RFA] Code gen improvement for sim-decode.scm

[Ben Elliston]

   Are you convinced that nothing can possibly use that variable except
   ifield extraction?  Is the improvement to generated code merely the
   removal of a gcc warning?  (It should be able to zap an unused
   variable, resulting in no performance difference.)

Nothing else can use it because it's defined within block scope. Yes, the
improvement to generated code is to remove a GCC warning.

Ben