]> git.ipfire.org Git - thirdparty/gcc.git/commitdiff
-finline-stringops: drop obsolete comment [PR112778]
authorAlexandre Oliva <oliva@adacore.com>
Thu, 21 Dec 2023 06:21:23 +0000 (03:21 -0300)
committerAlexandre Oliva <oliva@gnu.org>
Thu, 21 Dec 2023 06:21:23 +0000 (03:21 -0300)
When fixing the PR, I failed to remove the comment that raised the
very concern that the PR confirmed, and that the earlier patch for the
PR fixed.

for  gcc/ChangeLog

PR target/112778
* builtins.cc (try_store_by_multiple_pieces): Drop obsolete
comment.

gcc/builtins.cc

index 0f64feeedbad6df0685831292b5294eef0c1ee0e..125ea158ebfada8cbe57e9ec4cf4239c1c091ebf 100644 (file)
@@ -4491,10 +4491,6 @@ try_store_by_multiple_pieces (rtx to, rtx len, unsigned int ctz_len,
       if (max_len >> max_bits > min_len >> max_bits)
        tst_bits = max_bits;
     }
-  /* ??? Do we have to check that all powers of two lengths from
-     max_bits down to ctz_len pass can_store_by_pieces?  As in, could
-     it possibly be that xlenest passes while smaller power-of-two
-     sizes don't?  */
 
   by_pieces_constfn constfun;
   void *constfundata;