[PATCH] rs6000: Add "cannot_copy" attribute, use it (PR67788, PR67789)

2015-09-30 Thread Segher Boessenkool
After the shrink-wrapping patches the prologue will often be pushed "deeper" into the function, which in turn means the software trace cache pass will more often want to duplicate the basic block containing the prologue. This caused failures for 32-bit SVR4 with -msecure-plt PIC. This configurati

Re: [PATCH] rs6000: Add "cannot_copy" attribute, use it (PR67788, PR67789)

2015-10-01 Thread Richard Biener
On Thu, Oct 1, 2015 at 8:08 AM, Segher Boessenkool wrote: > After the shrink-wrapping patches the prologue will often be pushed > "deeper" into the function, which in turn means the software trace cache > pass will more often want to duplicate the basic block containing the > prologue. This cause

Re: [PATCH] rs6000: Add "cannot_copy" attribute, use it (PR67788, PR67789)

2015-10-01 Thread Bernd Schmidt
Do we have other ports with local labels in define_insns? I see some in darwin.md as well which your patch doesn't handle btw., otherwise suspicious %0: also appears (only) in h8300.md. arc.md also has a suspicious case in its doloop_end_i pattern. It is reasonably common, and defining cannot

Re: [PATCH] rs6000: Add "cannot_copy" attribute, use it (PR67788, PR67789)

2015-10-01 Thread David Edelsohn
On Thu, Oct 1, 2015 at 2:08 AM, Segher Boessenkool wrote: > After the shrink-wrapping patches the prologue will often be pushed > "deeper" into the function, which in turn means the software trace cache > pass will more often want to duplicate the basic block containing the > prologue. This cause

Re: [PATCH] rs6000: Add "cannot_copy" attribute, use it (PR67788, PR67789)

2015-10-01 Thread David Edelsohn
On Thu, Oct 1, 2015 at 2:08 AM, Segher Boessenkool wrote: > After the shrink-wrapping patches the prologue will often be pushed > "deeper" into the function, which in turn means the software trace cache > pass will more often want to duplicate the basic block containing the > prologue. This cause

Re: [PATCH] rs6000: Add "cannot_copy" attribute, use it (PR67788, PR67789)

2015-10-01 Thread Segher Boessenkool
On Thu, Oct 01, 2015 at 12:14:44PM +0200, Richard Biener wrote: > On Thu, Oct 1, 2015 at 8:08 AM, Segher Boessenkool > wrote: > > After the shrink-wrapping patches the prologue will often be pushed > > "deeper" into the function, which in turn means the software trace cache > > pass will more ofte

Re: [PATCH] rs6000: Add "cannot_copy" attribute, use it (PR67788, PR67789)

2015-10-01 Thread Segher Boessenkool
On Thu, Oct 01, 2015 at 10:08:50AM -0400, David Edelsohn wrote: > Is this expensive enough that it is worth limiting the definition of > the hook to configurations that include 32-bit SVR4 support so that > not every configuration incurs the overhead? I don't think so. That won't save the call to

Re: [PATCH] rs6000: Add "cannot_copy" attribute, use it (PR67788, PR67789)

2015-10-01 Thread Alan Modra
On Thu, Oct 01, 2015 at 12:18:08PM -0500, Segher Boessenkool wrote: > On Thu, Oct 01, 2015 at 12:14:44PM +0200, Richard Biener wrote: > > So even if not "easy", can you try? > > I did, and after half a day had a big mess and lots of things failing, > no idea where this was headed, and in the meant

Re: [PATCH] rs6000: Add "cannot_copy" attribute, use it (PR67788, PR67789)

2015-10-01 Thread Segher Boessenkool
On Fri, Oct 02, 2015 at 10:24:07AM +0930, Alan Modra wrote: > On Thu, Oct 01, 2015 at 12:18:08PM -0500, Segher Boessenkool wrote: > > On Thu, Oct 01, 2015 at 12:14:44PM +0200, Richard Biener wrote: > > > So even if not "easy", can you try? > > > > I did, and after half a day had a big mess and lot

Re: [PATCH] rs6000: Add "cannot_copy" attribute, use it (PR67788, PR67789)

2015-10-02 Thread Richard Biener
On Fri, Oct 2, 2015 at 3:14 AM, Segher Boessenkool wrote: > On Fri, Oct 02, 2015 at 10:24:07AM +0930, Alan Modra wrote: >> On Thu, Oct 01, 2015 at 12:18:08PM -0500, Segher Boessenkool wrote: >> > On Thu, Oct 01, 2015 at 12:14:44PM +0200, Richard Biener wrote: >> > > So even if not "easy", can you