* Claudiu Zissulescu <claudiu.zissule...@synopsys.com> [2017-07-24 10:42:54
+0200]:
> From: claziss <claz...@synopsys.com>
>
> Add support for -G option to ARC backend.
>
> gcc/
> 2017-04-24 Claudiu Zissulescu <claz...@synopsys.com>
>
> * config.gcc: Use g.opt for arc.
> * config/arc/arc.c (LEGITIMATE_SCALED_ADDRESS_P): Deleted,
> functionality moved to ...
> (legitimate_scaled_address_p): New function, ...here.
> (LEGITIMATE_SMALL_DATA_OFFSET_P): New define.
> (LEGITIMATE_SMALL_DATA_ADDRESS_P): Use the above define.
> (legitimate_offset_address_p): Delete TARGET_NO_SDATA_SET
> condition.
> (arc_override_options): Handle G option.
> (arc_output_pic_addr_const): Correct function definition.
> (arc_legitimate_address_p): Use legitimate_scaled_address_p.
> (arc_decl_anon_ns_mem_p): Delete.
> (arc_in_small_data_p): Overhaul this function to take into
> consideration the value given via G option.
> (arc_rewrite_small_data_1): Renamed and corrected old
> arc_rewrite_small_data function.
> (arc_rewrite_small_data): New function.
> (small_data_pattern): Don't use pic_offset_table_rtx.
> * config/arc/arc.h (CC1_SPEC): Recognize G option.
> * config/arc/simdext.md (movmisalignv2hi): Use
> prepare_move_operands function.
> (mov*): Likewise.
> (movmisalign*): Likewise.
>
> gcc/testsuite/
> 2017-04-24 Claudiu Zissulescu <claz...@synopsys.com>
>
> * gcc.target/arc/sdata-5.c: New test.
> * gcc.target/arc/arc700-stld-hazard.c: Update test options.
>From looking at other targets, I think that we need to add
documentation for -G into the ARC Options section of of
gcc/doc/invoke.texi.
> diff --git a/gcc/testsuite/gcc.target/arc/sdata-5.c
> b/gcc/testsuite/gcc.target/arc/sdata-5.c
> new file mode 100755
> index 0000000..0fdd52d
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/arc/sdata-5.c
> @@ -0,0 +1,8 @@
> +/* { dg-do compile } */
> +/* { dg-options "-w -Os" } */
> +
> +const a[1] = {};
> +static short b[] = {};
> +
> +c;
> +fn1() { return a + b[c]; }
Is it critical that we rely on default types here? I know it's
legitimate, but it just makes me sad to see new code using default
types.
Also, I guess that this test expects some things to be placed into the
small data section? Is there no pattern we can 'scan-assembler' for?
Or is this testing some other feature / bug?
Otherwise seems fine.
Thanks,
Andrew
