Re: [PATCH] Fix PR53708

2012-06-20 Thread Richard Guenther
On Wed, 20 Jun 2012, Richard Guenther wrote: > On Wed, 20 Jun 2012, Iain Sandoe wrote: > > > Hi, > > > > On 20 Jun 2012, at 09:23, Richard Guenther wrote: > > > > > On Tue, 19 Jun 2012, Iain Sandoe wrote: > > > > > >> > > >> On 19 Jun 2012, at 22:41, Mike Stump wrote: > > >> > > >>> On Jun 1

Re: [PATCH] Fix PR53708

2012-06-20 Thread Richard Guenther
On Wed, 20 Jun 2012, Iain Sandoe wrote: > Hi, > > On 20 Jun 2012, at 09:23, Richard Guenther wrote: > > > On Tue, 19 Jun 2012, Iain Sandoe wrote: > > > >> > >> On 19 Jun 2012, at 22:41, Mike Stump wrote: > >> > >>> On Jun 19, 2012, at 12:22 PM, Iain Sandoe wrote: > On 19 Jun 2012, at 13

Re: [PATCH] Fix PR53708

2012-06-20 Thread Iain Sandoe
Hi, On 20 Jun 2012, at 09:23, Richard Guenther wrote: > On Tue, 19 Jun 2012, Iain Sandoe wrote: > >> >> On 19 Jun 2012, at 22:41, Mike Stump wrote: >> >>> On Jun 19, 2012, at 12:22 PM, Iain Sandoe wrote: On 19 Jun 2012, at 13:53, Dominique Dhumieres wrote: > On Tue, 19 Jun 2012

Re: [PATCH] Fix PR53708

2012-06-20 Thread Richard Guenther
On Tue, 19 Jun 2012, Dominique Dhumieres wrote: > On Tue, 19 Jun 2012, Richard Guenther wrote: > > > > > Richard Guenther writes: > > > > We are too eager to bump alignment of some decls when vectorizing. > > > > The fix is to not bump alignment of decls the user explicitely > > > > aligned or t

Re: [PATCH] Fix PR53708

2012-06-20 Thread Richard Guenther
On Tue, 19 Jun 2012, Iain Sandoe wrote: > > On 19 Jun 2012, at 22:41, Mike Stump wrote: > > > On Jun 19, 2012, at 12:22 PM, Iain Sandoe wrote: > >> On 19 Jun 2012, at 13:53, Dominique Dhumieres wrote: > >> > >>> On Tue, 19 Jun 2012, Richard Guenther wrote: > > > Richard Guenther wri

Re: [PATCH] Fix PR53708

2012-06-19 Thread Iain Sandoe
On 19 Jun 2012, at 22:41, Mike Stump wrote: > On Jun 19, 2012, at 12:22 PM, Iain Sandoe wrote: >> On 19 Jun 2012, at 13:53, Dominique Dhumieres wrote: >> >>> On Tue, 19 Jun 2012, Richard Guenther wrote: > Richard Guenther writes: >> We are too eager to bump alignment of some decl

Re: [PATCH] Fix PR53708

2012-06-19 Thread Mike Stump
On Jun 19, 2012, at 5:53 AM, domi...@lps.ens.fr (Dominique Dhumieres) wrote: > On Tue, 19 Jun 2012, Richard Guenther wrote: >> >>> Richard Guenther writes: We are too eager to bump alignment of some decls when vectorizing. The fix is to not bump alignment of decls the user explicitely >

Re: [PATCH] Fix PR53708

2012-06-19 Thread Mike Stump
On Jun 19, 2012, at 12:22 PM, Iain Sandoe wrote: > On 19 Jun 2012, at 13:53, Dominique Dhumieres wrote: > >> On Tue, 19 Jun 2012, Richard Guenther wrote: >>> Richard Guenther writes: > We are too eager to bump alignment of some decls when vectorizing. > The fix is to not bump align

Re: [PATCH] Fix PR53708

2012-06-19 Thread Iain Sandoe
On 19 Jun 2012, at 13:53, Dominique Dhumieres wrote: > On Tue, 19 Jun 2012, Richard Guenther wrote: >> >>> Richard Guenther writes: We are too eager to bump alignment of some decls when vectorizing. The fix is to not bump alignment of decls the user explicitely aligned or that ar

Re: [PATCH] Fix PR53708

2012-06-19 Thread Dominique Dhumieres
On Tue, 19 Jun 2012, Richard Guenther wrote: > > > Richard Guenther writes: > > > We are too eager to bump alignment of some decls when vectorizing. > > > The fix is to not bump alignment of decls the user explicitely > > > aligned or that are used in an unknown way. > > > > I thought attribute(

Re: [PATCH] Fix PR53708

2012-06-19 Thread Richard Guenther
On Tue, 19 Jun 2012, Richard Sandiford wrote: > Richard Guenther writes: > > We are too eager to bump alignment of some decls when vectorizing. > > The fix is to not bump alignment of decls the user explicitely > > aligned or that are used in an unknown way. > > I thought attribute((__aligned__)

Re: [PATCH] Fix PR53708

2012-06-19 Thread Richard Sandiford
Richard Guenther writes: > We are too eager to bump alignment of some decls when vectorizing. > The fix is to not bump alignment of decls the user explicitely > aligned or that are used in an unknown way. I thought attribute((__aligned__)) only set a minimum alignment for variables? Most usees I