On Mon, Oct 4, 2021 at 12:31 PM Martin Liška <mli...@suse.cz> wrote:
>
> Hello.
>
> The patch sets -1 for x_flag_var_tracking and x_flag_var_tracking_assignments
> similarly
> to toplev.c.
>
> Patch can bootstrap on x86_64-linux-gnu and survives regression tests.
>
> Ready to be installed?
> Thanks,
> Martin
>
> PR debug/102585
>
> gcc/ChangeLog:
>
> * opts.c (finish_options): Similarly to toplev.c
> (process_options), set value -1.
>
> gcc/testsuite/ChangeLog:
>
> * gcc.dg/pr102585.c: New test.
> ---
> gcc/opts.c | 3 +++
> gcc/testsuite/gcc.dg/pr102585.c | 6 ++++++
> 2 files changed, 9 insertions(+)
> create mode 100644 gcc/testsuite/gcc.dg/pr102585.c
>
> diff --git a/gcc/opts.c b/gcc/opts.c
> index 6503980cd33..f2f058b3ece 100644
> --- a/gcc/opts.c
> +++ b/gcc/opts.c
> @@ -1343,6 +1343,9 @@ finish_options (struct gcc_options *opts, struct
> gcc_options *opts_set,
> opts->x_flag_complex_method = 1;
> else if (opts_set->x_flag_cx_fortran_rules)
> opts->x_flag_complex_method = opts->x_flag_default_complex_method;
> +
> + if (opts->x_flag_var_tracking_assignments && !opts->x_flag_var_tracking)
> + opts->x_flag_var_tracking = opts->x_flag_var_tracking_assignments = -1;
I can see how this affects later tests in process_options like
if (flag_var_tracking_uninit == AUTODETECT_VALUE)
flag_var_tracking_uninit = flag_var_tracking;
would now become -1 instead of zero?
So iff then probably this whole block would need to move but the option
handling code is a sophisticated mess between FE, target and middle-end ...
diff --git a/gcc/toplev.c b/gcc/toplev.c
index ec9f998a49b..fec72b1365a 100644
--- a/gcc/toplev.c
+++ b/gcc/toplev.c
@@ -2147,8 +2147,6 @@ standard_type_bitsize (int bitsize)
static void
do_compile ()
{
- process_options ();
-
/* Don't do any more if an error has already occurred. */
if (!seen_error ())
{
@@ -2368,6 +2366,8 @@ toplev::main (int argc, char **argv)
/* Exit early if we can (e.g. -help). */
if (!exit_after_options)
{
+ process_options ();
+
if (m_use_TV_TOTAL)
start_timevars ();
do_compile ();
would at least make the flow a bit more obvious ;) IIRC process_options
was/is supposed to be the place to emit diagnostics about options but
clearly it's doing even more. I guess outlining the call to
lang_hooks.post_options
would be another piece to make the flow clearer, likewise setting of
optimization_default_node ...
Richard.
> }
>
> #define LEFT_COLUMN 27
> diff --git a/gcc/testsuite/gcc.dg/pr102585.c b/gcc/testsuite/gcc.dg/pr102585.c
> new file mode 100644
> index 00000000000..efd066b4a4e
> --- /dev/null
> +++ b/gcc/testsuite/gcc.dg/pr102585.c
> @@ -0,0 +1,6 @@
> +/* PR debug/102585 */
> +/* { dg-do compile } */
> +/* { dg-options "-fvar-tracking-assignments -fno-var-tracking" } */
> +
> +#pragma GCC optimize 0
> +void d_demangle_callback_Og() { int c = 0; }
> --
> 2.33.0
>
