> On Tue, Feb 23, 2016 at 12:32 PM, Richard Biener wrote:
> > On Tue, 23 Feb 2016, Jan Hubicka wrote:
> >
> >> >
> >> > Ok, so maybe a better question to symtab would be if there is an
> >> > actual definition for what __builtin_FOO will call. Not really
> >> > whether that definition is cfun. O
On Tue, Feb 23, 2016 at 12:32 PM, Richard Biener wrote:
> On Tue, 23 Feb 2016, Jan Hubicka wrote:
>
>> >
>> > Ok, so maybe a better question to symtab would be if there is an
>> > actual definition for what __builtin_FOO will call. Not really
>> > whether that definition is cfun. Of course all t
On Tue, 23 Feb 2016, Jan Hubicka wrote:
> >
> > Ok, so maybe a better question to symtab would be if there is an
> > actual definition for what __builtin_FOO will call. Not really
> > whether that definition is cfun. Of course all the fortify
> > always-inline wrappers should not count as such
>
> Ok, so maybe a better question to symtab would be if there is an
> actual definition for what __builtin_FOO will call. Not really
> whether that definition is cfun. Of course all the fortify
> always-inline wrappers should not count as such (just in case
> the symtab code is confused about t
> On Mon, 22 Feb 2016, Jakub Jelinek wrote:
>
> > On Mon, Feb 22, 2016 at 01:44:09PM +0100, Richard Biener wrote:
> > > --- 1079,1086
> > > || !dominated_by_p (CDI_DOMINATORS,
> > > loop->latch, gimple_bb (stmt)))
> > > return;
> > > +
On Mon, 22 Feb 2016, Richard Biener wrote:
> On Mon, 22 Feb 2016, Jakub Jelinek wrote:
>
> > On Mon, Feb 22, 2016 at 01:44:09PM +0100, Richard Biener wrote:
> > > --- 1079,1086
> > > || !dominated_by_p (CDI_DOMINATORS,
> > > loop->latch, gimple_bb
On Mon, 22 Feb 2016, Jakub Jelinek wrote:
> On Mon, Feb 22, 2016 at 01:44:09PM +0100, Richard Biener wrote:
> > --- 1079,1086
> > || !dominated_by_p (CDI_DOMINATORS,
> > loop->latch, gimple_bb (stmt)))
> > return;
> > + if (cgraph_node::get (cfun->dec
On Mon, Feb 22, 2016 at 01:44:09PM +0100, Richard Biener wrote:
> --- 1079,1086
> || !dominated_by_p (CDI_DOMINATORS,
> loop->latch, gimple_bb (stmt)))
> return;
> + if (cgraph_node::get (cfun->decl)->aliases (BUILT_IN_MEMSET))
> + return;
P
The following fixes us to not transform loops into memcpy or memset
or not transform malloc + memset into calloc if we implement that
functionality itself (and thus we "alias" the symbol we'd call when
emitting a call to the builtin).
Bootstrap and regtest pending on x86_64-unknown-linux-gnu.
Th
