On Thu, 22 Dec 2022, Jakub Jelinek wrote:
> Hi!
>
> The following place in value_replacement is after proving that
> x == cst1 ? cst2 : x
> phi result is only used in a comparison with constant which doesn't
> care if it compares cst1 or cst2 and replaces it with x.
> The testcase is miscompiled because we have after the replacement
> incorrect range info for the phi result, we would need to
> effectively union the phi result range with cst1 (oarg in the code)
> because previously that constant might be missing in the range, but
> newly it can appear (we've just verified that the single use stmt
> of the phi result doesn't care about that value in particular).
>
> The following patch just resets the info, bootstrapped/regtested
> on x86_64-linux and i686-linux, ok for trunk?
OK.
> Aldy/Andrew, how would one instead union the SSA_NAME_RANGE_INFO
> with some INTEGER_CST and store it back into SSA_NAME_RANGE_INFO
> (including adjusting non-zero bits and the like)?
There's no get_range_info on SSA_NAMEs (anymore?) but you can
construct a value_range from the INTEGER_CST singleton and
union that into the SSA_NAMEs range and then do set_range_info
with the altered range I guess.
Richard.
> 2022-12-22 Jakub Jelinek <ja...@redhat.com>
>
> PR tree-optimization/108166
> * tree-ssa-phiopt.cc (value_replacement): For the maybe_equal_p
> case turned into equal_p reset SSA_NAME_RANGE_INFO of phi result.
>
> * g++.dg/torture/pr108166.C: New test.
>
> --- gcc/tree-ssa-phiopt.cc.jj 2022-10-28 11:00:53.970243821 +0200
> +++ gcc/tree-ssa-phiopt.cc 2022-12-21 14:27:58.118326548 +0100
> @@ -1491,6 +1491,12 @@ value_replacement (basic_block cond_bb,
> default:
> break;
> }
> + if (equal_p)
> + /* After the optimization PHI result can have value
> + which it couldn't have previously.
> + We could instead of resetting it union the range
> + info with oarg. */
> + reset_flow_sensitive_info (gimple_phi_result (phi));
> if (equal_p && MAY_HAVE_DEBUG_BIND_STMTS)
> {
> imm_use_iterator imm_iter;
> --- gcc/testsuite/g++.dg/torture/pr108166.C.jj 2022-12-21
> 14:31:02.638661322 +0100
> +++ gcc/testsuite/g++.dg/torture/pr108166.C 2022-12-21 14:30:45.441909725
> +0100
> @@ -0,0 +1,26 @@
> +// PR tree-optimization/108166
> +// { dg-do run }
> +
> +bool a, b;
> +int d, c;
> +
> +const int &
> +foo (const int &f, const int &g)
> +{
> + return !f ? f : g;
> +}
> +
> +__attribute__((noipa)) void
> +bar (int)
> +{
> +}
> +
> +int
> +main ()
> +{
> + c = foo (b, 0) > ((b ? d : b) ?: 8);
> + a = b ? d : b;
> + bar (a);
> + if (a != 0)
> + __builtin_abort ();
> +}
>
> Jakub
>
>
--
Richard Biener <rguent...@suse.de>
SUSE Software Solutions Germany GmbH, Frankenstrasse 146, 90461 Nuernberg,
Germany; GF: Ivo Totev, Andrew Myers, Andrew McDonald, Boudien Moerman;
HRB 36809 (AG Nuernberg)
