Bug 94012 - Tonga OpenCL clpeak vm faults
Summary: Tonga OpenCL clpeak vm faults
Status: RESOLVED FIXED
Alias: None
Product: DRI
Classification: Unclassified
Component: DRM/AMDgpu (show other bugs)
Version: DRI git
Hardware: Other All
: medium normal
Assignee: Default DRI bug account
QA Contact:
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2016-02-05 18:29 UTC by Andy Furniss
Modified: 2016-03-31 15:14 UTC (History)
0 users

See Also:
i915 platform:
i915 features:


Attachments
vm debug (24.95 KB, text/plain)
2016-02-05 18:29 UTC, Andy Furniss
no flags Details
compute shaders .bz2 (1.68 MB, application/octet-stream)
2016-02-05 18:30 UTC, Andy Furniss
no flags Details

Description Andy Furniss 2016-02-05 18:29:44 UTC
Created attachment 121540 [details]
vm debug

R9285 llvm,mesa,libclc yesterdays gits.

No ideea if this is a regression as I've only just noticed tonga got added to libclc so this is first time testing OpenCL.

clpeak - https://github.com/krrishnarraj/clpeak.git

produces vmfaults. It will run to completion.

R600_DEBUG=check_vm ./clpeak 

Platform: Clover
  Device: AMD TONGA (DRM 3.1.0, LLVM 3.9.0)
    Driver version  : 11.2.0-devel (Linux x64)
    Compute units   : 28
    Clock frequency : 0 MHz

    Global memory bandwidth (GBPS)
      float   : 1.45
      float2  : 1.44
      float4  : 1.45
      float8  : 1.45
      float16 : Detected a VM fault, exiting...

Attaching output and vm debug file from 

R600_DEBUG=check_vm,cs ./clpeak &>~/clpeak-cs
Comment 1 Andy Furniss 2016-02-05 18:30:59 UTC
Created attachment 121541 [details]
compute shaders .bz2
Comment 2 Andy Furniss 2016-03-31 15:14:31 UTC
fixed by llvm commit

commit d3adac51fcce66e8c79b77299fef9e5f6c4c646e
Author: Tom Stellard <thomas.stellard@amd.com>
Date:   Wed Mar 30 16:35:09 2016 +0000

    AMDGPU/SI: Enable lanemask tracking in misched
    
    Summary:
    This results in higher register usage, but should make it easier for
    the compiler to hide latency.
    
    This pass is a prerequisite for some more scheduler improvements, and I
    think the increase register usage with this patch is acceptable, because
    when combined with the scheduler improvements, the total register usage
    will decrease.
    
    shader-db stats:
    
    2382 shaders in 478 tests
    Totals:
    SGPRS: 48672 -> 49088 (0.85 %)
    VGPRS: 34148 -> 34847 (2.05 %)
    Code Size: 1285816 -> 1289128 (0.26 %) bytes
    LDS: 28 -> 28 (0.00 %) blocks
    Scratch: 492544 -> 573440 (16.42 %) bytes per wave
    Max Waves: 6856 -> 6846 (-0.15 %)
    Wait states: 0 -> 0 (0.00 %)
    
    Depends on D18451
    
    Reviewers: nhaehnle, arsenm
    
    Subscribers: arsenm, llvm-commits
    
    Differential Revision: http://reviews.llvm.org/D18452
    
    git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@264876 91177308-0d34-0410-b5e6-96231b3b80d8


Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.