rtl-ssa: Don't leave NOTE_INSN_DELETED around

Message ID ZTJLpun/ghfYhY2d@arm.com
State Accepted
Headers
Series rtl-ssa: Don't leave NOTE_INSN_DELETED around |

Checks

Context Check Description
snail/gcc-patch-check success Github commit url

Commit Message

Alex Coplan Oct. 20, 2023, 9:43 a.m. UTC
  Hi,

This patch tweaks change_insns to also call ::remove_insn to ensure the
underlying RTL insn gets removed from the insn chain in the case of a
deletion.

This avoids leaving NOTE_INSN_DELETED around after deleting insns.

For movement, the RTL insn chain is updated earlier in change_insns with
the call to move_insn.  For deletion, it seems reasonable to do it here.

Bootstrapped/regtested on aarch64-linux-gnu in combination with the upcoming
aarch64 load pair pass, i.e.:
https://gcc.gnu.org/pipermail/gcc-patches/2023-October/633601.html
and all uncommitted prerequisites except for 6/11 i.e.:
https://gcc.gnu.org/pipermail/gcc-patches/2023-October/633349.html
which shows that 6/11 is no longer required in combination with this patch (as
expected).

Thanks,
Alex

gcc/ChangeLog:

	* rtl-ssa/changes.cc (function_info::change_insns): Ensure we call
	::remove_insn on deleted insns.
  

Comments

Richard Sandiford Oct. 20, 2023, 10:36 a.m. UTC | #1
Alex Coplan <alex.coplan@arm.com> writes:
> Hi,
>
> This patch tweaks change_insns to also call ::remove_insn to ensure the
> underlying RTL insn gets removed from the insn chain in the case of a
> deletion.
>
> This avoids leaving NOTE_INSN_DELETED around after deleting insns.
>
> For movement, the RTL insn chain is updated earlier in change_insns with
> the call to move_insn.  For deletion, it seems reasonable to do it here.
>
> Bootstrapped/regtested on aarch64-linux-gnu in combination with the upcoming
> aarch64 load pair pass, i.e.:
> https://gcc.gnu.org/pipermail/gcc-patches/2023-October/633601.html
> and all uncommitted prerequisites except for 6/11 i.e.:
> https://gcc.gnu.org/pipermail/gcc-patches/2023-October/633349.html
> which shows that 6/11 is no longer required in combination with this patch (as
> expected).
>
> Thanks,
> Alex
>
> gcc/ChangeLog:
>
> 	* rtl-ssa/changes.cc (function_info::change_insns): Ensure we call
> 	::remove_insn on deleted insns.

OK, thanks.

Richard

> diff --git a/gcc/rtl-ssa/changes.cc b/gcc/rtl-ssa/changes.cc
> index b11a88e0919..d6874e5627c 100644
> --- a/gcc/rtl-ssa/changes.cc
> +++ b/gcc/rtl-ssa/changes.cc
> @@ -732,7 +732,11 @@ function_info::change_insns (array_slice<insn_change *> changes)
>        insn_change &change = *changes[i];
>        insn_info *insn = change.insn ();
>        if (change.is_deletion ())
> -	remove_insn (insn);
> +	{
> +	  if (rtx_insn *rtl = insn->rtl ())
> +	    ::remove_insn (rtl); // Remove the underlying RTL insn.
> +	  remove_insn (insn);
> +	}
>        else if (insn_info *placeholder = placeholders[i])
>  	{
>  	  // Check if earlier movements turned a move into a no-op.
  

Patch

diff --git a/gcc/rtl-ssa/changes.cc b/gcc/rtl-ssa/changes.cc
index b11a88e0919..d6874e5627c 100644
--- a/gcc/rtl-ssa/changes.cc
+++ b/gcc/rtl-ssa/changes.cc
@@ -732,7 +732,11 @@  function_info::change_insns (array_slice<insn_change *> changes)
       insn_change &change = *changes[i];
       insn_info *insn = change.insn ();
       if (change.is_deletion ())
-	remove_insn (insn);
+	{
+	  if (rtx_insn *rtl = insn->rtl ())
+	    ::remove_insn (rtl); // Remove the underlying RTL insn.
+	  remove_insn (insn);
+	}
       else if (insn_info *placeholder = placeholders[i])
 	{
 	  // Check if earlier movements turned a move into a no-op.