Re: [PATCH] shrink-wrap: New spread_components

2016-11-28 Thread Jeff Law
On 11/09/2016 02:46 PM, Segher Boessenkool wrote: This patch changes spread_components to use a simpler algorithm that puts prologue components as early as possible, and epilogue components as late as possible. This allows better scheduling, and also saves a bit of code size. The blocks that ru

ping x2: Re: [PATCH] shrink-wrap: New spread_components

2016-11-27 Thread Segher Boessenkool
Ping. On Sat, Nov 19, 2016 at 11:24:34AM -0600, Segher Boessenkool wrote: > Ping. > > On Wed, Nov 09, 2016 at 09:46:55PM +, Segher Boessenkool wrote: > > This patch changes spread_components to use a simpler algorithm that > > puts prologue components as early as possible, and epilogue compon

Re: [PATCH] shrink-wrap: New spread_components

2016-11-19 Thread Segher Boessenkool
Ping. On Wed, Nov 09, 2016 at 09:46:55PM +, Segher Boessenkool wrote: > This patch changes spread_components to use a simpler algorithm that > puts prologue components as early as possible, and epilogue components > as late as possible. This allows better scheduling, and also saves a > bit of

Re: [PATCH] shrink-wrap: New spread_components

2016-11-11 Thread Kyrill Tkachov
On 09/11/16 21:46, Segher Boessenkool wrote: This patch changes spread_components to use a simpler algorithm that puts prologue components as early as possible, and epilogue components as late as possible. This allows better scheduling, and also saves a bit of code size. The blocks that run wi

[PATCH] shrink-wrap: New spread_components

2016-11-09 Thread Segher Boessenkool
This patch changes spread_components to use a simpler algorithm that puts prologue components as early as possible, and epilogue components as late as possible. This allows better scheduling, and also saves a bit of code size. The blocks that run with some specific component enabled after this pa