Philippe Virouleau authored 9 years ago and Philippe Virouleau committed 8 years ago

"affinity('depend | numa | core', 'expr1'[, 'expr2'])"

'expr1' is the affinity requested.
'expr2' defines if the affinity is "strict" (0 by default, anything but 0 means strict).

Using 'depend' will set the affinity of the task to the numa node where 'expr1' is located (therefore, a pointer is expected).

Philippe Virouleau authored 9 years ago and Philippe Virouleau committed 8 years ago

This specifies an algorithm for the initial ready tasks in the parallel section.

Philippe Virouleau authored 9 years ago and Philippe Virouleau committed 8 years ago

Example : one can use "#pragma omp task affinity(3)" to tell the runtime this task
has an affinity with NUMA node 3.

r283223 | davidsh | 2016-10-04 10:41:36 -0400 (Tue, 04 Oct 2016) | 1 line

[OpenMP] fix segfault when a variable referenced in reduction clause is a reference parameter
Differential Revision: http://reviews.llvm.org/D24524


git-svn-id: https://llvm.org/svn/llvm-project/cfe/branches/release_39@286972 91177308-0d34-0410-b5e6-96231b3b80d8

This patch adds the check for specifying both simdlen and safelen clauses on the 'distribute simd' or 'distribute parallel for simd' constructs.

Differential Revision: https://reviews.llvm.org/D22384


git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@275529 91177308-0d34-0410-b5e6-96231b3b80d8

This patch is to implement sema and parsing for 'target parallel for simd' pragma.

Differential Revision: http://reviews.llvm.org/D22096



git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@275365 91177308-0d34-0410-b5e6-96231b3b80d8

This patch is to remove duplicate code in ActOnOpenMPRegionStart. (NFC)

Differential Revision: http://reviews.llvm.org/D22177


git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@275323 91177308-0d34-0410-b5e6-96231b3b80d8

http://reviews.llvm.org/D22070




git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@275282 91177308-0d34-0410-b5e6-96231b3b80d8

http://reviews.llvm.org/D21904

This patch is similar to the implementation of 'private' clause: it adds a list of private pointers to be used within the target data region to store the device pointers returned by the runtime.
Please refer to the following document for a full description of what the runtime witll return in this case (page 10 and 11):
https://github.com/clang-omp/OffloadingDesign

I am happy to answer any question related to the runtime interface to help reviewing this patch.



git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@275271 91177308-0d34-0410-b5e6-96231b3b80d8

OpenMP relies on some helper expressions generated during semantic
analysis. But they are required only for codegen and not required in
dependent contexts. Patch removes generation of some of helper
expressions.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@274745 91177308-0d34-0410-b5e6-96231b3b80d8

Summary: This patch is an implementation of sema and parsing for the OpenMP composite pragma 'distribute simd'.

Differential Revision: http://reviews.llvm.org/D22007


git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@274604 91177308-0d34-0410-b5e6-96231b3b80d8

Summary: This patch is an implementation of sema and parsing for the OpenMP composite pragma 'distribute parallel for simd'.

Differential Revision: http://reviews.llvm.org/D21977



git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@274530 91177308-0d34-0410-b5e6-96231b3b80d8

construct


git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@274352 91177308-0d34-0410-b5e6-96231b3b80d8

Clang fails to diagnose cases such as
#pragma omp target
  while(0) {
    #pragma omp teams
    {}
  }

A patch by David Sheinkman.



git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@273908 91177308-0d34-0410-b5e6-96231b3b80d8

[OpenMP] Initial implementation of parse and sema for composite pragma 'distribute parallel for'

This patch is an initial implementation for #distribute parallel for.
The main differences that affect other pragmas are:

The implementation of 'distribute parallel for' requires blocking of the associated loop, where blocks are "distributed" to different teams and iterations within each block are scheduled to parallel threads within each team. To implement blocking, sema creates two additional worksharing directive fields that are used to pass the team assigned block lower and upper bounds through the outlined function resulting from 'parallel'. In this way, scheduling for 'for' to threads can use those bounds.
As a consequence of blocking, the stride of 'distribute' is not 1 but it is equal to the blocking size. This is returned by the runtime and sema prepares a DistIncrExpr variable to hold that value.
As a consequence of blocking, the global upper bound (EnsureUpperBound) expression of the 'for' is not the original loop upper bound (e.g. in for(i = 0 ; i < N; i++) this is 'N') but it is the team-assigned block upper bound. Sema creates a new expression holding the calculation of the actual upper bound for 'for' as UB = min(UB, PrevUB), where UB is the loop upper bound, and PrevUB is the team-assigned block upper bound.




git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@273884 91177308-0d34-0410-b5e6-96231b3b80d8

 [OpenMP] Initial implementation of parse and sema for composite pragma 'distribute parallel for'




git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@273709 91177308-0d34-0410-b5e6-96231b3b80d8

http://reviews.llvm.org/D21564

This patch is an initial implementation for #distribute parallel for.
The main differences that affect other pragmas are:

The implementation of 'distribute parallel for' requires blocking of the associated loop, where blocks are "distributed" to different teams and iterations within each block are scheduled to parallel threads within each team. To implement blocking, sema creates two additional worksharing directive fields that are used to pass the team assigned block lower and upper bounds through the outlined function resulting from 'parallel'. In this way, scheduling for 'for' to threads can use those bounds.
As a consequence of blocking, the stride of 'distribute' is not 1 but it is equal to the blocking size. This is returned by the runtime and sema prepares a DistIncrExpr variable to hold that value.
As a consequence of blocking, the global upper bound (EnsureUpperBound) expression of the 'for' is not the original loop upper bound (e.g. in for(i = 0 ; i < N; i++) this is 'N') but it is the team-assigned block upper bound. Sema creates a new expression holding the calculation of the actual upper bound for 'for' as UB = min(UB, PrevUB), where UB is the loop upper bound, and PrevUB is the team-assigned block upper bound.



git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@273705 91177308-0d34-0410-b5e6-96231b3b80d8

-Wfor-loop-analysis warnings for a for-loop with a condition variable. In such
a case, the loop condition variable is modified on each iteration of the loop
by definition.

Original commit message:

Rearrange condition handling so that semantic checks on a condition variable
are performed before the other substatements of the construct are parsed,
rather than deferring them until the end. This allows better error recovery
from semantic errors in the condition, improves diagnostic order, and is a
prerequisite for C++17 constexpr if.


git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@273600 91177308-0d34-0410-b5e6-96231b3b80d8

as it caused a regression in -Wfor-loop-analysis.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@273589 91177308-0d34-0410-b5e6-96231b3b80d8

are performed before the other substatements of the construct are parsed,
rather than deferring them until the end. This allows better error recovery
from semantic errors in the condition, improves diagnostic order, and is a
prerequisite for C++17 constexpr if.


git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@273548 91177308-0d34-0410-b5e6-96231b3b80d8

Since D21243 fixes relative clang-tidy tests.

This reverts commit a71d9fbd.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@273312 91177308-0d34-0410-b5e6-96231b3b80d8

instantiation.

Added checks for non-dependent context when trygin to capture
non-constant schedule chunk expression for proper codegen of outlined
functions.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@272775 91177308-0d34-0410-b5e6-96231b3b80d8

This reverts r272296, since there are clang-tidy failures that appear to
be caused by this change.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@272310 91177308-0d34-0410-b5e6-96231b3b80d8

These ExprWithCleanups are added for holding a RunCleanupsScope not
for destructor calls; rather, they are for lifetime marks. This requires
ExprWithCleanups to keep a bit to indicate whether it have cleanups with
side effects (e.g. dtor calls).

Differential Revision: http://reviews.llvm.org/D20498

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@272296 91177308-0d34-0410-b5e6-96231b3b80d8

Summary:
Target regions require globals to be captured. This patch fixes a bug exposed when that happens in a template function.


Reviewers: hfinkel, arpith-jacob, kkwli0, carlo.bertolli, ABataev

Subscribers: guansong, ABataev, cfe-commits, caomhin, fraggamuffin

Differential Revision: http://reviews.llvm.org/D18110

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@271001 91177308-0d34-0410-b5e6-96231b3b80d8

Summary:
The patch contains the parsing and sema support for the `from` clause. 

Patch based on the original post by Kelvin Li.

Reviewers: hfinkel, carlo.bertolli, kkwli0, arpith-jacob, ABataev

Subscribers: caomhin, cfe-commits

Differential Revision: http://reviews.llvm.org/D18488

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@270882 91177308-0d34-0410-b5e6-96231b3b80d8

Summary:
The patch contains the parsing and sema support for the `to` clause. 

Patch based on the original post by Kelvin Li.

Reviewers: carlo.bertolli, hfinkel, kkwli0, arpith-jacob, ABataev

Subscribers: caomhin, cfe-commits

Differential Revision: http://reviews.llvm.org/D18597

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@270880 91177308-0d34-0410-b5e6-96231b3b80d8

Summary:
This patch is to add parsing and sema support for `target update` directive. Support for the `to` and `from` clauses will be added by a different patch.  This patch also adds support for other clauses that are already implemented upstream and apply to `target update`, e.g. `device` and `if`.

This patch is based on the original post by Kelvin Li.

Reviewers: hfinkel, carlo.bertolli, kkwli0, arpith-jacob, ABataev

Subscribers: caomhin, cfe-commits

Differential Revision: http://reviews.llvm.org/D15944

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@270878 91177308-0d34-0410-b5e6-96231b3b80d8

OpenMP 4.5 adds support for doacross loop synchronization. Patch
implements codegen for this construct.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@270690 91177308-0d34-0410-b5e6-96231b3b80d8

For better performance and to unify code with offloading part we pass
scalar firstprivate values by value, instead of by reference. It will
remove some extra copying operations.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@269751 91177308-0d34-0410-b5e6-96231b3b80d8

Support OpenMP version 4.5 syntax for #pragma omp declare target.

Syntax:
  #pragma omp declare target (extended-list) new-line
or
  #pragma omp declare target clause[ [,] clause ... ] new-line

Where clause is one of the following:
  to(extended-list)
  link(list)

Differential Revision: http://reviews.llvm.org/D20011

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@268925 91177308-0d34-0410-b5e6-96231b3b80d8

Fixes a bug with analysis of data-sharing attributes in templates.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@268020 91177308-0d34-0410-b5e6-96231b3b80d8

OpenMP 4.5 defines 'taskloop simd' directive, which is combined
directive for 'taskloop' and 'simd' directives. Patch adds initial
codegen support for this directive and its 2 basic clauses 'safelen' and
'simdlen'.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@267872 91177308-0d34-0410-b5e6-96231b3b80d8

Summary:
Implement codegen for the map clause. All the new list items in 4.5 specification are supported. 

Fix bug in the generation of array sections that was exposed by some of the map clause tests: for pointer types the offsets have to be calculated from the pointee not the pointer.

Reviewers: hfinkel, kkwli0, carlo.bertolli, arpith-jacob, ABataev

Subscribers: ABataev, cfe-commits, caomhin, fraggamuffin

Differential Revision: http://reviews.llvm.org/D16749

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@267808 91177308-0d34-0410-b5e6-96231b3b80d8

Summary:
This patch adds logic to save the components of mappable expressions in the clause that uses it, so that they don't have to be recomputed during codegen. Given that the mappable components are (will be) used in several clauses a new geneneric implementation `OMPMappableExprListClause` is used that extends the existing `OMPVarListClause`.

This patch does not add new tests. The goal is to preserve the existing functionality while storing more info in the clauses.

Reviewers: hfinkel, carlo.bertolli, arpith-jacob, kkwli0, ABataev

Subscribers: cfe-commits, caomhin

Differential Revision: http://reviews.llvm.org/D19382

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@267560 91177308-0d34-0410-b5e6-96231b3b80d8

The taskloop construct specifies that the iterations of one or more associated loops will be executed in parallel using OpenMP tasks. The iterations are distributed across tasks created by the construct and scheduled to be executed.
The next code will be generated for the taskloop directive:
    #pragma omp taskloop num_tasks(N) lastprivate(j)
        for( i=0; i<N*GRAIN*STRIDE-1; i+=STRIDE ) {
          int th = omp_get_thread_num();
          #pragma omp atomic
            counter++;
          #pragma omp atomic
            th_counter[th]++;
          j = i;
    }

Generated code:
task = __kmpc_omp_task_alloc(NULL,gtid,1,sizeof(struct
task),sizeof(struct shar),&task_entry);
psh = task->shareds;
psh->pth_counter = &th_counter;
psh->pcounter = &counter;
psh->pj = &j;
task->lb = 0;
task->ub = N*GRAIN*STRIDE-2;
task->st = STRIDE;
__kmpc_taskloop(
NULL,             // location
gtid,             // gtid
task,             // task structure
1,                // if clause value
&task->lb,        // lower bound
&task->ub,        // upper bound
STRIDE,           // loop increment
0,                // 1 if nogroup specified
2,                // schedule type: 0-none, 1-grainsize, 2-num_tasks
N,                // schedule value (ignored for type 0)
(void*)&__task_dup_entry // tasks duplication routine
);

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@267395 91177308-0d34-0410-b5e6-96231b3b80d8

If loop control variable for simd-based directives is explicitly marked
as linear/lastprivate in clauses, codegen for such construct would
crash. Patch fixes this problem.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@267101 91177308-0d34-0410-b5e6-96231b3b80d8

If the untied clause is present on a task construct, any thread in the
team can resume the task region after a suspension. Patch adds proper
codegen for untied tasks.

git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@266853 91177308-0d34-0410-b5e6-96231b3b80d8