Richard Biener writes:
> On Tue, 18 May 2021, Richard Biener wrote:
>
>> On Tue, 18 May 2021, Richard Sandiford wrote:
>>
>> > Richard Biener writes:
>> > > @@ -6621,9 +6637,31 @@ pass_expand::execute (function *fun)
>> > > (int) param_ssp_buffer_size);
>> > > }
>> > >
>
er_nonconstant_array_refs_r - I'll fix that. I can also make
> the above fool-proof by recoding the decls I marked addressable
> in a vec and traverse that for unsetting - would that be prefered?
So like this? Testing TREE_ADDRESSABLE where it formerly wasn't
already testing use_re
On Tue, 18 May 2021, Richard Sandiford wrote:
> Richard Biener writes:
> > @@ -6621,9 +6637,31 @@ pass_expand::execute (function *fun)
> > (int) param_ssp_buffer_size);
> > }
> >
> > + /* Temporarily mark PARM_DECLs and RESULT_DECLs we need to expand to
> > + memory addre
Richard Biener writes:
> @@ -6621,9 +6637,31 @@ pass_expand::execute (function *fun)
>(int) param_ssp_buffer_size);
> }
>
> + /* Temporarily mark PARM_DECLs and RESULT_DECLs we need to expand to
> + memory addressable so expand_function_start can emit the required
> +
This avoids setting TREE_ADDRESSABLE on variables we want to force to
the stack. Instead track those in a temporary bitmap and force
stack expansion that way, leaving TREE_ADDRESSABLE alone, not
pessimizing future alias queries.
Bootstrap and regtest running on x86_64-unknown-linux-gnu, comments
