Re: Default compute dimensions (runtime)

2019-04-25 Thread Thomas Schwinge
tics to be too noisy. See also the several 'dg-bogus' with XFAIL that you're proposing to add? We should think about that some more. So, the following is relevant right now: > --- a/gcc/doc/invoke.texi > +++ b/gcc/doc/invoke.texi > @@ -2167,8 +2167,12 @@ have support for

Re: Default compute dimensions

2019-01-30 Thread Thomas Schwinge
w{@uref{http > implies @option{-pthread}, and thus is only supported on targets that > have support for @option{-pthread}. > > -[...] > +@item -fopenacc-dim=@var{geom} > +@opindex fopenacc-dim > +@cindex OpenACC accelerator programming > +Specify default compute dimensions [...]

Re: Default compute dimensions (runtime)

2018-10-05 Thread Julian Brown
* testsuite/libgomp.oacc-c-c++-common/loop-default-compile.c: New. commit a918a8739ae7652250c978b0ececa181a587b0c0 Author: Julian Brown Date: Fri Oct 5 11:11:47 2018 -0700 OpenACC default compute dimensions 20xx-xx-xx Nathan Sidwell Tom de Vries Thomas Schwinge

[gomp4] runtime default compute dimensions

2016-08-30 Thread Nathan Sidwell
This patch interrogates the target device to determine default gemotry at runtime. This has the greatest difference on gang partitioning, where there's a noticeable sawtooth in the relationship between number of gangs and execution time. Picking the number of gangs as an exact multiple of num

Re: Default compute dimensions (runtime)

2016-02-03 Thread Alexander Monakov
On Wed, 3 Feb 2016, Nathan Sidwell wrote: > You can only override at runtime those dimensions that you said you'd override > at runtime when you compiled your program. Ah, I see. That's not obvious to me, so perhaps added documentation can be expanded to explain that? (I now see that the plugin

Re: Default compute dimensions (runtime)

2016-02-03 Thread Nathan Sidwell
On 02/03/16 11:10, Alexander Monakov wrote: Hello, On Wed, 3 Feb 2016, Nathan Sidwell wrote: 1) extend the -fopenacc-dim=X:Y:Z syntax to allow '-' indicating a runtime choice. (0 also indicates that, but I thought best to have an explicit syntax as well). Does it work when the user specifies

Re: Default compute dimensions (runtime)

2016-02-03 Thread Alexander Monakov
Hello, On Wed, 3 Feb 2016, Nathan Sidwell wrote: > 1) extend the -fopenacc-dim=X:Y:Z syntax to allow '-' indicating a runtime > choice. (0 also indicates that, but I thought best to have an explicit syntax > as well). Does it work when the user specifies one of the dimensions, so that references

Default compute dimensions (runtime)

2016-02-03 Thread Nathan Sidwell
-dim-default-2.c: New. Index: gcc/doc/invoke.texi === --- gcc/doc/invoke.texi (revision 233084) +++ gcc/doc/invoke.texi (working copy) @@ -1969,7 +1969,12 @@ have support for @option{-pthread}. Specify default compute dimensions for parallel offload regions that do not explicitly specify. The @v

Re: Default compute dimensions

2016-02-01 Thread Nathan Sidwell
On 02/01/16 13:42, Jakub Jelinek wrote: Your patch broke bootstrap on ILP32 hosts, I'm testing following fix. Supporting unsigned values from 0x8000U to 0xU only on LP64 hosts and not on ILP64 hosts sounds really weird, I think it is better to only support 1 to 0x7fffU. yes, I

Re: Default compute dimensions

2016-02-01 Thread H.J. Lu
On Mon, Feb 1, 2016 at 8:15 AM, Nathan Sidwell wrote: > On 02/01/16 10:32, Jakub Jelinek wrote: >> >> On Mon, Feb 01, 2016 at 09:15:05AM -0500, Nathan Sidwell wrote: >>> >>> On 01/29/16 10:18, Jakub Jelinek wrote: On Thu, Jan 28, 2016 at 10:38:51AM -0500, Nathan Sidwell wrote: >

Re: Default compute dimensions

2016-02-01 Thread Jakub Jelinek
On Mon, Feb 01, 2016 at 11:15:13AM -0500, Nathan Sidwell wrote: > On 02/01/16 10:32, Jakub Jelinek wrote: > >On Mon, Feb 01, 2016 at 09:15:05AM -0500, Nathan Sidwell wrote: > >>On 01/29/16 10:18, Jakub Jelinek wrote: > >>>On Thu, Jan 28, 2016 at 10:38:51AM -0500, Nathan Sidwell wrote: > This pa

Re: Default compute dimensions

2016-02-01 Thread Nathan Sidwell
On 02/01/16 10:32, Jakub Jelinek wrote: On Mon, Feb 01, 2016 at 09:15:05AM -0500, Nathan Sidwell wrote: On 01/29/16 10:18, Jakub Jelinek wrote: On Thu, Jan 28, 2016 at 10:38:51AM -0500, Nathan Sidwell wrote: This patch adds default compute dimension handling. Users rarely specify compute dime

Re: Default compute dimensions

2016-02-01 Thread Jakub Jelinek
On Mon, Feb 01, 2016 at 09:15:05AM -0500, Nathan Sidwell wrote: > On 01/29/16 10:18, Jakub Jelinek wrote: > >On Thu, Jan 28, 2016 at 10:38:51AM -0500, Nathan Sidwell wrote: > >>This patch adds default compute dimension handling. Users rarely specify > >>compute dimensions, expecting the toolchain

Re: Default compute dimensions

2016-02-01 Thread Nathan Sidwell
On 01/29/16 10:18, Jakub Jelinek wrote: On Thu, Jan 28, 2016 at 10:38:51AM -0500, Nathan Sidwell wrote: This patch adds default compute dimension handling. Users rarely specify compute dimensions, expecting the toolchain to DTRT. More savvy users would like to specify global defaults. This pa

Re: Default compute dimensions

2016-01-29 Thread Nathan Sidwell
On 01/29/16 10:18, Jakub Jelinek wrote: On Thu, Jan 28, 2016 at 10:38:51AM -0500, Nathan Sidwell wrote: This patch adds default compute dimension handling. Users rarely specify compute dimensions, expecting the toolchain to DTRT. More savvy users would like to specify global defaults. This pa

Re: Default compute dimensions

2016-01-29 Thread Jakub Jelinek
On Thu, Jan 28, 2016 at 10:38:51AM -0500, Nathan Sidwell wrote: > This patch adds default compute dimension handling. Users rarely specify > compute dimensions, expecting the toolchain to DTRT. More savvy users would > like to specify global defaults. This patch permits both. Isn't it better to

Default compute dimensions

2016-01-28 Thread Nathan Sidwell
, incomplete, and subject to -change in future versions of GCC. See -@w{@uref{https://gcc.gnu.org/wiki/OpenACC}} for more information. +@item -fopenacc-dim=@var{geom} +@opindex fopenacc-dim +@cindex OpenACC accelerator programming +Specify default compute dimensions for parallel offload regions that do