On Thu, Nov 05, 2015 at 11:31:33AM -0600, Evandro Menezes wrote:
> James,
>
> Since other members of the "tune_params" structure were signed
> integers, even though negative numbers would make no sense for most
> either, I followed the same pattern.
>
> Regardless, here's a patch with unsigned in
James,
Since other members of the "tune_params" structure were signed integers,
even though negative numbers would make no sense for most either, I
followed the same pattern.
Regardless, here's a patch with unsigned integers as you requested:
[AArch64] Add extra tuning parameters for targ
On Wed, Nov 04, 2015 at 05:21:03PM -0600, Evandro Menezes wrote:
> Please, ignore the previous patch. This is the intended patch.
>
> Sorry.
>
> --
> Evandro Menezes
>
> On 11/04/2015 05:18 PM, Evandro Menezes wrote:
> >This patch adds extra tuning information about AArch64 targets:
> >
> > *
Please, ignore the previous patch. This is the intended patch.
Sorry.
--
Evandro Menezes
On 11/04/2015 05:18 PM, Evandro Menezes wrote:
This patch adds extra tuning information about AArch64 targets:
* Maximum number of case values before resorting to a jump table
The default values assu
This patch adds extra tuning information about AArch64 targets:
* Maximum number of case values before resorting to a jump table
The default values assumed independently of the specific backends
may be rather low for modern processors, which sport quite efficient
direct branch predictio
Following the suggestions to add the support for the Exynos M1 models,
the following series of patches are broken down into:
* add more target specific tuning data
* add heuristics tuning
* add the Exynos M1 cost model
* add the Exynos M1 scheduling model
Thank you,
--
Evandro Menezes
Hi, James.
On 10/28/2015 05:36 AM, James Greenhalgh wrote:
On Tue, Oct 27, 2015 at 06:12:48PM -0500, Evandro Menezes wrote:
This patch adds the scheduling and cost models for Exynos M1.
Though it?s a rather large patch, much of it is the DFA model for the
pipeline.? Still, I?d appreciate any f
Hi, Kyrill.
True dat, I missed arm-tune.md.
And, yes, bootstrapped cross and natively on AArch64 and cross on ARM.
Thank you,
--
Evandro Menezes
On 10/28/2015 05:57 AM, Kyrill Tkachov wrote:
Hi Evandro,
On 27/10/15 23:12, Evandro Menezes wrote:
This patch adds the scheduling and cost model
Andrew,
I need to do more investigation WRT prefetching, especially Kyrill's
recent patch.
The change to the peeling limit also benefits A57. If James or Marcus
could confirm this, I'd be glad to add it to all AArch64 targets,
assuming that Xgene would be fine with it too.
Thank you,
--
Hi Evandro,
On 27/10/15 23:12, Evandro Menezes wrote:
This patch adds the scheduling and cost models for Exynos M1.
Though it’s a rather large patch, much of it is the DFA model for the
pipeline. Still, I’d appreciate any feedback.
Please, commit if it’s alright.
Thank you,
On top of Jame
On Wed, Oct 28, 2015 at 6:36 PM, James Greenhalgh
wrote:
> On Tue, Oct 27, 2015 at 06:12:48PM -0500, Evandro Menezes wrote:
>> This patch adds the scheduling and cost models for Exynos M1.
>>
>> Though it?s a rather large patch, much of it is the DFA model for the
>> pipeline.? Still, I?d apprecia
On Tue, Oct 27, 2015 at 06:12:48PM -0500, Evandro Menezes wrote:
> This patch adds the scheduling and cost models for Exynos M1.
>
> Though it?s a rather large patch, much of it is the DFA model for the
> pipeline.? Still, I?d appreciate any feedback.
>
> Please, commit if it?s alright.
Hi Evand
This patch adds the scheduling and cost models for Exynos M1.
Though its a rather large patch, much of it is the DFA model for the
pipeline. Still, Id appreciate any feedback.
Please, commit if its alright.
Thank you,
--
Evandro Menezes
0001-AArch64-Add-scheduling-and-cost-models-for
13 matches
Mail list logo
