Re: [PATCH] rs6000: Fix up PCH on powerpc* [PR104323]

[Segher Boessenkool <segher@kernel.crashing.org>]

Hi!

On Tue, Feb 01, 2022 at 04:27:40PM +0100, Jakub Jelinek wrote:
> +/* PR target/104323 */
> +/* { dg-require-effective-target powerpc_altivec_ok } */
> +/* { dg-options "-maltivec" } */
> +
> +#include <altivec.h>
> testcase which I'm not including into testsuite because for some reason
> the test fails on non-powerpc* targets (is done even on those and fails
> because of missing altivec.h etc.),

powerpc_altivec_ok returns false if the target isn't Power, you can use
this in the testsuite fine?  Why does it still fail on other targets,
the test should be SKIPPED there?

Or wait, proc check_effective_target_powerpc_altivec_ok is broken, and
does not implement its intention or documentation.  Will fix.

> PCH is broken on powerpc*-*-* since the
> new builtin generator has been introduced.
> The generator contains or emits comments like:
>   /* #### Cannot mark this as a GC root because only pointer types can
>      be marked as GTY((user)) and be GC roots.  All trees in here are
>      kept alive by other globals, so not a big deal.  Alternatively,
>      we could change the enum fields to ints and cast them in and out
>      to avoid requiring a GTY((user)) designation, but that seems
>      unnecessarily gross.  */
> Having the fntypes stored in other GC roots can work fine for GC,
> ggc_collect will then always mark them and so they won't disappear from
> the tables, but it definitely doesn't work for PCH, which when the
> arrays with fntype members aren't GTY marked means on PCH write we create
> copies of those FUNCTION_TYPEs and store in *.gch that the GC roots should
> be updated, but don't store that rs6000_builtin_info[?].fntype etc. should
> be updated.  When PCH is read again, the blob is read at some other address,
> GC roots are updated, rs6000_builtin_info[?].fntype contains garbage
> pointers (GC freed pointers with random data, or random unrelated types or
> other trees).
> The following patch fixes that.  It stops any user markings because that
> is totally unnecessary, just skips fields we don't need to mark and adds
> GTY(()) to the 2 array variables.  We can get rid of all those global
> vars for the fn types, they can be now automatic vars.
> With the patch we get
>   {
>     &rs6000_instance_info[0].fntype,
>     1 * (RS6000_INST_MAX),
>     sizeof (rs6000_instance_info[0]),
>     &gt_ggc_mx_tree_node,
>     &gt_pch_nx_tree_node
>   },
>   {
>     &rs6000_builtin_info[0].fntype,
>     1 * (RS6000_BIF_MAX),
>     sizeof (rs6000_builtin_info[0]),
>     &gt_ggc_mx_tree_node,
>     &gt_pch_nx_tree_node
>   },
> as the new roots which is exactly what we want and significantly more
> compact than countless
>   {
>     &uv2di_ftype_pudi_usi,
>     1,
>     sizeof (uv2di_ftype_pudi_usi),
>     &gt_ggc_mx_tree_node,
>     &gt_pch_nx_tree_node
>   },
>   {
>     &uv2di_ftype_lg_puv2di,
>     1,
>     sizeof (uv2di_ftype_lg_puv2di),
>     &gt_ggc_mx_tree_node,
>     &gt_pch_nx_tree_node
>   },
>   {
>     &uv2di_ftype_lg_pudi,
>     1,
>     sizeof (uv2di_ftype_lg_pudi),
>     &gt_ggc_mx_tree_node,
>     &gt_pch_nx_tree_node
>   },
>   {
>     &uv2di_ftype_di_puv2di,
>     1,
>     sizeof (uv2di_ftype_di_puv2di),
>     &gt_ggc_mx_tree_node,
>     &gt_pch_nx_tree_node
>   },
> cases (822 of these instead of just those 4 shown).

Bill, can you review the builtin side of this?

> 	PR target/104323
> 	* config/rs6000/t-rs6000 (EXTRA_GTYPE_DEPS): Append rs6000-builtins.h
> 	rather than $(srcdir)/config/rs6000/rs6000-builtins.def.
> 	* config/rs6000/rs6000-gen-builtins.cc (write_decls): Don't use
> 	GTY((user)) for struct bifdata and struct ovlddata.  Instead add
> 	GTY((skip(""))) to members with pointer and enum types that don't need
> 	to be tracked.  Add GTY(()) to rs6000_builtin_info and rs6000_instance_info
> 	declarations.  Don't emit gt_ggc_mx and gt_pch_nx declarations.

Nice :-)

> 	(write_extern_fntype, write_fntype): Remove.
> 	(write_fntype_init): Emit the fntype vars as automatic vars instead
> 	of file scope ones.
> 	(write_header_file): Don't iterate with write_extern_fntype.
> 	(write_init_file): Don't iterate with write_fntype.  Don't emit
> 	gt_ggc_mx and gt_pch_nx definitions.

>    if (tf_found)
> -    fprintf (init_file, "  if (float128_type_node)\n  ");
> +    fprintf (init_file,
> +	     "  tree %s = NULL_TREE;\n  if (float128_type_node)\n    ",
> +	     buf);
>    else if (dfp_found)
> -    fprintf (init_file, "  if (dfloat64_type_node)\n  ");
> +    fprintf (init_file,
> +	     "  tree %s = NULL_TREE;\n  if (dfloat64_type_node)\n    ",
> +	     buf);

Things are much more readable if you break this into multiple print
statements.  I realise the existomg code is like that, but that could
be improved.

So, okay for trunk (modulo what Bill finds), and you can include the
testcase as well after I have fixed the selector.  Thanks Jakub!


Segher