[PATCH] Clang: Reject OpenCL workgroup size requirement of 0

Joey Gouly joey.gouly at arm.com
Mon May 19 04:03:31 PDT 2014


Pedro, do you have commit access? If not, I can commit it for you.

 

Joey

 

From: cfe-commits-bounces at cs.uiuc.edu
[mailto:cfe-commits-bounces at cs.uiuc.edu] On Behalf Of Neil Henning
Sent: 19 May 2014 11:44
To: cfe-commits at cs.uiuc.edu
Subject: Re: [PATCH] Clang: Reject OpenCL workgroup size requirement of 0

 

Any reason why the additional test in invalid-kernel-attrs.cl only tests
whether the Z dimension is 0? Would be useful to ensure no regressions occur
that we test each dimension (Perhaps you could just submit additional
testing in a further review though, no need to hold up this patch any longer
- the patch LGTM otherwise!)

Cheers,
-Neil.

On 19/05/2014 11:30, Pedro Ferreira wrote:

Any more comments on this?

On 13/05/14 17:09, Pedro Ferreira wrote:

Your wish is my command. 

On 13/05/14 16:18, Joey Gouly wrote: 



LGTM too, but please add a test case to invalid-kernel-attrs.cl. 

-----Original Message----- 
From: cfe-commits-bounces at cs.uiuc.edu 
[mailto:cfe-commits-bounces at cs.uiuc.edu] On Behalf Of David Tweed 
Sent: 13 May 2014 11:50 
To: 'Pedro Ferreira'; cfe-commits at cs.uiuc.edu 
Subject: RE: [PATCH] Clang: Reject OpenCL workgroup size requirement of 0 

The patch LGTM. 
-----Original Message----- 
From: cfe-commits-bounces at cs.uiuc.edu 
[mailto:cfe-commits-bounces at cs.uiuc.edu] On Behalf Of Pedro Ferreira 
Sent: 13 May 2014 10:03 
To: cfe-commits at cs.uiuc.edu 
Subject: [PATCH] Clang: Reject OpenCL workgroup size requirement of 0 

OpenCL allows us to define the workgroup size used in a kernel in order 
to further optimise the code, specifically to that size. 
The syntax is 
__attribute__((reqd_work_group_size(X, Y, Z))) 

where X, Y and Z are compile-time known integers. 
Currently clang allows any of those to be 0, which would be illegal in 
OpenCL. 

The attached patch adds this validation; it's currently a draft and any 
comments would be welcome. 

(I reported this on bug 19699) 




_______________________________________________ 
cfe-commits mailing list 
cfe-commits at cs.uiuc.edu 
http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits 











_______________________________________________
cfe-commits mailing list
cfe-commits at cs.uiuc.edu
http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits







_______________________________________________
cfe-commits mailing list
cfe-commits at cs.uiuc.edu
http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits

 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/cfe-commits/attachments/20140519/446217b8/attachment.html>


More information about the cfe-commits mailing list