[rocclr] Zero cache size on Raven Ridge #670

imyxh · 2021-08-29T05:52:50Z

imyxh
Aug 29, 2021

Hello, and thank you so much for putting these packages together!

I have absolutely no idea what this error means so I'm not even sure if I'm filing an issue in the right place. But maybe someone could help me diagnose this.

When I build rocm-opencl-runtime, and then run clinfo, or try to do something in blender, the program aborts due to the following:

clinfo: /home/imyxh/aur/rocm.d/rocclr/src/ROCclr-rocm-4.3.1/device/rocm/rocdevice.cpp:1106: bool roc::Device::populateOCLDeviceConstants(): Assertion `cachesize[0] > 0' failed.

The code in current master seems to be here.

I'm using the Raven Ridge 2400G. Any pointers would be appreciated!

tpkessler · 2021-08-29T12:08:41Z

tpkessler
Aug 29, 2021
Maintainer

Hi @imyxh, did you already check AMD's notes on Raven Ridge?

0 replies

imyxh · 2021-08-29T14:28:59Z

imyxh
Aug 29, 2021
Author

Hi @imyxh, did you already check AMD's notes on Raven Ridge?

Oh, yes, I understand that Raven Ridge is not supported by much of ROCm. I only filed this issue because rocm-opencl-runtime used to work decently (some bugs from blender, etc.) until I recently decided to upgrade my packages. I've never seen this zero cache size error before.

0 replies

SilverLuke · 2021-10-22T15:54:06Z

SilverLuke
Oct 22, 2021

I have the same error with this system:

Linux 5.10.75-1-lts
Ryzen 5 3400G
RX 580 8GB
There is any workaround ?

0 replies

tpkessler · 2021-11-21T17:52:06Z

tpkessler
Nov 21, 2021
Maintainer

ROCm was updated to 4.5.0. Could you please test with the newest version @imyxh @SilverLuke ?

0 replies

SilverLuke · 2021-11-21T17:56:15Z

SilverLuke
Nov 21, 2021

Yes I get the same error. I have tried just now with rocm 4.5.0.

Edit: Someone can explain to me how ROCR_VISIBLE_DEVICES works? Because changing this variable make clinfo do not crash but it says that I haven't any device "opencl ready"

0 replies

imyxh · 2021-11-22T05:15:02Z

imyxh
Nov 22, 2021
Author

ROCm was updated to 4.5.0. Could you please test with the newest version @imyxh @SilverLuke ?

Yes, I still see this bug in clinfo.

0 replies

tpkessler · 2021-11-22T18:30:48Z

tpkessler
Nov 22, 2021
Maintainer

Someone can explain to me how ROCR_VISIBLE_DEVICES works?

It's explained in the documentation.

Yes, I still see this bug in clinfo.

What about rocminfo?

0 replies

imyxh · 2021-11-22T23:00:04Z

imyxh
Nov 22, 2021
Author

What about rocminfo?

This is the output of my rocminfo (sorry for long paste):

ROCk module is loaded
=====================    
HSA System Attributes    
=====================    
Runtime Version:         1.1
System Timestamp Freq.:  1000.000000MHz
Sig. Max Wait Duration:  18446744073709551615 (0xFFFFFFFFFFFFFFFF) (timestamp count)
Machine Model:           LARGE                              
System Endianness:       LITTLE                             

==========               
HSA Agents               
==========               
*******                  
Agent 1                  
*******                  
  Name:                    AMD Ryzen 5 2400G with Radeon Vega Graphics
  Uuid:                    CPU-XX                             
  Marketing Name:          AMD Ryzen 5 2400G with Radeon Vega Graphics
  Vendor Name:             CPU                                
  Feature:                 None specified                     
  Profile:                 FULL_PROFILE                       
  Float Round Mode:        NEAR                               
  Max Queue Number:        0(0x0)                             
  Queue Min Size:          0(0x0)                             
  Queue Max Size:          0(0x0)                             
  Queue Type:              MULTI                              
  Node:                    0                                  
  Device Type:             CPU                                
  Cache Info:              
    L1:                      32(0x20) KB                        
  Chip ID:                 5597(0x15dd)                       
  Cacheline Size:          64(0x40)                           
  Max Clock Freq. (MHz):   3600                               
  BDFID:                   2048                               
  Internal Node ID:        0                                  
  Compute Unit:            8                                  
  SIMDs per CU:            4                                  
  Shader Engines:          1                                  
  Shader Arrs. per Eng.:   1                                  
  WatchPts on Addr. Ranges:4                                  
  Features:                None
  Pool Info:               
    Pool 1                   
      Segment:                 GLOBAL; FLAGS: FINE GRAINED        
      Size:                    16776832(0xfffe80) KB              
      Allocatable:             TRUE                               
      Alloc Granule:           4KB                                
      Alloc Alignment:         4KB                                
      Accessible by all:       TRUE                               
    Pool 2                   
      Segment:                 GLOBAL; FLAGS: KERNARG, FINE GRAINED
      Size:                    16776832(0xfffe80) KB              
      Allocatable:             TRUE                               
      Alloc Granule:           4KB                                
      Alloc Alignment:         4KB                                
      Accessible by all:       TRUE                               
  ISA Info:                
*******                  
Agent 2                  
*******                  
  Name:                    gfx902                             
  Uuid:                    GPU-XX                             
  Marketing Name:          AMD Ryzen 5 2400G with Radeon Vega Graphics
  Vendor Name:             AMD                                
  Feature:                 KERNEL_DISPATCH                    
  Profile:                 FULL_PROFILE                       
  Float Round Mode:        NEAR                               
  Max Queue Number:        128(0x80)                          
  Queue Min Size:          4096(0x1000)                       
  Queue Max Size:          131072(0x20000)                    
  Queue Type:              MULTI                              
  Node:                    0                                  
  Device Type:             GPU                                
  Cache Info:              
  Chip ID:                 5597(0x15dd)                       
  Cacheline Size:          64(0x40)                           
  Max Clock Freq. (MHz):   1250                               
  BDFID:                   2048                               
  Internal Node ID:        0                                  
  Compute Unit:            11                                 
  SIMDs per CU:            4                                  
  Shader Engines:          1                                  
  Shader Arrs. per Eng.:   1                                  
  WatchPts on Addr. Ranges:4                                  
  Features:                KERNEL_DISPATCH 
  Fast F16 Operation:      FALSE                              
  Wavefront Size:          64(0x40)                           
  Workgroup Max Size:      1024(0x400)                        
  Workgroup Max Size per Dimension:
    x                        1024(0x400)                        
    y                        1024(0x400)                        
    z                        1024(0x400)                        
  Max Waves Per CU:        160(0xa0)                          
  Max Work-item Per CU:    10240(0x2800)                      
  Grid Max Size:           4294967295(0xffffffff)             
  Grid Max Size per Dimension:
    x                        4294967295(0xffffffff)             
    y                        4294967295(0xffffffff)             
    z                        4294967295(0xffffffff)             
  Max fbarriers/Workgrp:   32                                 
  Pool Info:               
    Pool 1                   
      Segment:                 GROUP                              
      Size:                    64(0x40) KB                        
      Allocatable:             FALSE                              
      Alloc Granule:           0KB                                
      Alloc Alignment:         0KB                                
      Accessible by all:       FALSE                              
  ISA Info:                
    ISA 1                    
      Name:                    amdgcn-amd-amdhsa--gfx902:xnack+   
      Machine Models:          HSA_MACHINE_MODEL_LARGE            
      Profiles:                HSA_PROFILE_BASE                   
      Default Rounding Mode:   NEAR                               
      Default Rounding Mode:   NEAR                               
      Fast f16:                TRUE                               
      Workgroup Max Size:      1024(0x400)                        
      Workgroup Max Size per Dimension:
        x                        1024(0x400)                        
        y                        1024(0x400)                        
        z                        1024(0x400)                        
      Grid Max Size:           4294967295(0xffffffff)             
      Grid Max Size per Dimension:
        x                        4294967295(0xffffffff)             
        y                        4294967295(0xffffffff)             
        z                        4294967295(0xffffffff)             
      FBarrier Max Size:       32                                 
*** Done ***

This appears to be the same output as without rocm-opencl-runtime installed. Applications like blender still fail with the zero cache size error, though.

0 replies

SilverLuke · 2021-11-23T00:08:01Z

SilverLuke
Nov 23, 2021

I get this from rocminfo :

ROCk module is loaded
=====================    
HSA System Attributes    
=====================    
Runtime Version:         1.1
System Timestamp Freq.:  1000.000000MHz
Sig. Max Wait Duration:  18446744073709551615 (0xFFFFFFFFFFFFFFFF) (timestamp count)
Machine Model:           LARGE                              
System Endianness:       LITTLE                             

==========               
HSA Agents               
==========               
*******                  
Agent 1                  
*******                  
Name:                    AMD Ryzen 5 3400G with Radeon Vega Graphics
Uuid:                    CPU-XX                             
Marketing Name:          AMD Ryzen 5 3400G with Radeon Vega Graphics
Vendor Name:             CPU                                
Feature:                 None specified                     
Profile:                 FULL_PROFILE                       
Float Round Mode:        NEAR                               
Max Queue Number:        0(0x0)                             
Queue Min Size:          0(0x0)                             
Queue Max Size:          0(0x0)                             
Queue Type:              MULTI                              
Node:                    0                                  
Device Type:             CPU                                
Cache Info:              
  L1:                      32(0x20) KB                        
Chip ID:                 5592(0x15d8)                       
Cacheline Size:          64(0x40)                           
Max Clock Freq. (MHz):   3700                               
BDFID:                   11776                              
Internal Node ID:        0                                  
Compute Unit:            8                                  
SIMDs per CU:            4                                  
Shader Engines:          1                                  
Shader Arrs. per Eng.:   1                                  
WatchPts on Addr. Ranges:4                                  
Features:                None
Pool Info:               
  Pool 1                   
    Segment:                 GLOBAL; FLAGS: FINE GRAINED        
    Size:                    33554048(0x1fffe80) KB             
    Allocatable:             TRUE                               
    Alloc Granule:           4KB                                
    Alloc Alignment:         4KB                                
    Accessible by all:       TRUE                               
  Pool 2                   
    Segment:                 GLOBAL; FLAGS: KERNARG, FINE GRAINED
    Size:                    33554048(0x1fffe80) KB             
    Allocatable:             TRUE                               
    Alloc Granule:           4KB                                
    Alloc Alignment:         4KB                                
    Accessible by all:       TRUE                               
ISA Info:                
*******                  
Agent 2                  
*******                  
Name:                    gfx902                             
Uuid:                    GPU-XX                             
Marketing Name:          AMD Ryzen 5 3400G with Radeon Vega Graphics
Vendor Name:             AMD                                
Feature:                 KERNEL_DISPATCH                    
Profile:                 FULL_PROFILE                       
Float Round Mode:        NEAR                               
Max Queue Number:        128(0x80)                          
Queue Min Size:          4096(0x1000)                       
Queue Max Size:          131072(0x20000)                    
Queue Type:              MULTI                              
Node:                    0                                  
Device Type:             GPU                                
Cache Info:              
Chip ID:                 5592(0x15d8)                       
Cacheline Size:          64(0x40)                           
Max Clock Freq. (MHz):   1400                               
BDFID:                   11776                              
Internal Node ID:        0                                  
Compute Unit:            11                                 
SIMDs per CU:            4                                  
Shader Engines:          1                                  
Shader Arrs. per Eng.:   1                                  
WatchPts on Addr. Ranges:4                                  
Features:                KERNEL_DISPATCH 
Fast F16 Operation:      FALSE                              
Wavefront Size:          64(0x40)                           
Workgroup Max Size:      1024(0x400)                        
Workgroup Max Size per Dimension:
  x                        1024(0x400)                        
  y                        1024(0x400)                        
  z                        1024(0x400)                        
Max Waves Per CU:        160(0xa0)                          
Max Work-item Per CU:    10240(0x2800)                      
Grid Max Size:           4294967295(0xffffffff)             
Grid Max Size per Dimension:
  x                        4294967295(0xffffffff)             
  y                        4294967295(0xffffffff)             
  z                        4294967295(0xffffffff)             
Max fbarriers/Workgrp:   32                                 
Pool Info:               
  Pool 1                   
    Segment:                 GROUP                              
    Size:                    64(0x40) KB                        
    Allocatable:             FALSE                              
    Alloc Granule:           0KB                                
    Alloc Alignment:         0KB                                
    Accessible by all:       FALSE                              
ISA Info:                
  ISA 1                    
    Name:                    amdgcn-amd-amdhsa--gfx902:xnack-   
    Machine Models:          HSA_MACHINE_MODEL_LARGE            
    Profiles:                HSA_PROFILE_BASE                   
    Default Rounding Mode:   NEAR                               
    Default Rounding Mode:   NEAR                               
    Fast f16:                TRUE                               
    Workgroup Max Size:      1024(0x400)                        
    Workgroup Max Size per Dimension:
      x                        1024(0x400)                        
      y                        1024(0x400)                        
      z                        1024(0x400)                        
    Grid Max Size:           4294967295(0xffffffff)             
    Grid Max Size per Dimension:
      x                        4294967295(0xffffffff)             
      y                        4294967295(0xffffffff)             
      z                        4294967295(0xffffffff)             
    FBarrier Max Size:       32                                 
*******                  
Agent 3                  
*******                  
Name:                    gfx803                             
Uuid:                    GPU-XX                             
Marketing Name:          AMD Radeon RX 580 Series           
Vendor Name:             AMD                                
Feature:                 KERNEL_DISPATCH                    
Profile:                 BASE_PROFILE                       
Float Round Mode:        NEAR                               
Max Queue Number:        128(0x80)                          
Queue Min Size:          4096(0x1000)                       
Queue Max Size:          131072(0x20000)                    
Queue Type:              MULTI                              
Node:                    1                                  
Device Type:             GPU                                
Cache Info:              
  L1:                      16(0x10) KB                        
Chip ID:                 26591(0x67df)                      
Cacheline Size:          64(0x40)                           
Max Clock Freq. (MHz):   1350                               
BDFID:                   4096                               
Internal Node ID:        1                                  
Compute Unit:            36                                 
SIMDs per CU:            4                                  
Shader Engines:          4                                  
Shader Arrs. per Eng.:   1                                  
WatchPts on Addr. Ranges:4                                  
Features:                KERNEL_DISPATCH 
Fast F16 Operation:      FALSE                              
Wavefront Size:          64(0x40)                           
Workgroup Max Size:      1024(0x400)                        
Workgroup Max Size per Dimension:
  x                        1024(0x400)                        
  y                        1024(0x400)                        
  z                        1024(0x400)                        
Max Waves Per CU:        40(0x28)                           
Max Work-item Per CU:    2560(0xa00)                        
Grid Max Size:           4294967295(0xffffffff)             
Grid Max Size per Dimension:
  x                        4294967295(0xffffffff)             
  y                        4294967295(0xffffffff)             
  z                        4294967295(0xffffffff)             
Max fbarriers/Workgrp:   32                                 
Pool Info:               
  Pool 1                   
    Segment:                 GLOBAL; FLAGS: COARSE GRAINED      
    Size:                    8388608(0x800000) KB               
    Allocatable:             TRUE                               
    Alloc Granule:           4KB                                
    Alloc Alignment:         4KB                                
    Accessible by all:       FALSE                              
  Pool 2                   
    Segment:                 GROUP                              
    Size:                    64(0x40) KB                        
    Allocatable:             FALSE                              
    Alloc Granule:           0KB                                
    Alloc Alignment:         0KB                                
    Accessible by all:       FALSE                              
ISA Info:                
  ISA 1                    
    Name:                    amdgcn-amd-amdhsa--gfx803          
    Machine Models:          HSA_MACHINE_MODEL_LARGE            
    Profiles:                HSA_PROFILE_BASE                   
    Default Rounding Mode:   NEAR                               
    Default Rounding Mode:   NEAR                               
    Fast f16:                TRUE                               
    Workgroup Max Size:      1024(0x400)                        
    Workgroup Max Size per Dimension:
      x                        1024(0x400)                        
      y                        1024(0x400)                        
      z                        1024(0x400)                        
    Grid Max Size:           4294967295(0xffffffff)             
    Grid Max Size per Dimension:
      x                        4294967295(0xffffffff)             
      y                        4294967295(0xffffffff)             
      z                        4294967295(0xffffffff)             
    FBarrier Max Size:       32                                 
*** Done ***

AND I get this 2 lines in dmesg :

[43063.566067] Alloc host visible vram on small bar is not allowed
[43063.566636] Alloc host visible vram on small bar is not allowed

even when I run import tensorflow as tf; tf.config.list_physical_devices() (installed with pip install tensorflow-rocm) I get those lines in dmesg.
I have tried to change the ROCR_VISIBLE_DEVICES and the HIP_VISIBLE_DEVICES with the value "1,3" to hide the "Agent 2" (the integrated video card) now the clinfo output is this:

Number of platforms:				 1
  Platform Profile:				 FULL_PROFILE
  Platform Version:				 OpenCL 2.2 AMD-APP.dbg (3361.0)
  Platform Name:				 AMD Accelerated Parallel Processing
  Platform Vendor:				 Advanced Micro Devices, Inc.
  Platform Extensions:				 cl_khr_icd cl_amd_event_callback 


  Platform Name:				 AMD Accelerated Parallel Processing
Number of devices:				 0

0 replies

petm5 · 2021-11-23T18:09:17Z

petm5
Nov 23, 2021

I have the same bug on my system with the packages from arch4edu. Any fix yet?

0 replies

tpkessler · 2021-11-25T18:09:35Z

tpkessler
Nov 25, 2021
Maintainer

Not all packages on arch4edu have been updated to 4.5.0 yet. You can either wait or build them from source as described in the README of this repo.

0 replies

SilverLuke · 2021-12-14T09:33:55Z

SilverLuke
Dec 14, 2021

Are there any possible fixes in 4.5.2? Compile rocm require a lot of time for me.

0 replies

imyxh · 2021-12-18T03:18:23Z

imyxh
Dec 18, 2021
Author

Are there any possible fixes in 4.5.2? Compile rocm require a lot of time for me.

Nope, still broken.

$ pacman -Qs rocm
local/hsa-rocr 4.5.2-1
    ROCm Platform Runtime: ROCr a HPC market enhanced HSA based runtime
local/rocm-cmake 4.5.2-1
    CMake modules for common build tasks needed for the ROCm software stack
local/rocm-device-libs 4.5.2-1
    Radeon Open Compute - device libs
local/rocm-llvm 4.5.2-1
    Radeon Open Compute - LLVM toolchain (llvm, clang, lld)
local/rocm-opencl-runtime 4.5.2-1
    Radeon Open Compute - OpenCL runtime
local/rocminfo 4.5.2-1
    ROCm info tools - rocm_agent_enumerator
                                                                                                                                                                                                                                                                           
$ clinfo
clinfo: /home/imyxh/aur/rocm.d/rocm-opencl-runtime/src/ROCclr-rocm-4.5.2/device/rocm/rocdevice.cpp:1111: bool roc::Device::populateOCLDeviceConstants(): Assertion `cachesize[0] > 0' failed.
[1]    3380204 abort (core dumped)  clinfo

0 replies

acxz · 2022-04-10T15:23:32Z

acxz
Apr 10, 2022
Maintainer

@imyxh want to just give another bump to try this on the latest release

maybe the best way to solve this issue might be to do a bisection. i.e. find the last commit/version that worked for you and keep on bisecting the code commits to figure out which commit introduced the unwanted behavior.

0 replies

darkbasic · 2022-09-30T07:41:44Z

darkbasic
Sep 30, 2022

It still doesn't work with 5.2.3:

$ clinfo
clinfo: /home/niko/.cache/yay/rocm-opencl-runtime/src/ROCclr-rocm-5.2.3/device/rocm/rocdevice.cpp:1132: bool roc::Device::populateOCLDeviceConstants(): Assertion `cachesize[0] > 0' failed.
Aborted (core dumped)

Also, why do I have two agents?

$ /opt/rocm/bin/rocminfo
ROCk module is loaded
=====================    
HSA System Attributes    
=====================    
Runtime Version:         1.1
System Timestamp Freq.:  1000.000000MHz
Sig. Max Wait Duration:  18446744073709551615 (0xFFFFFFFFFFFFFFFF) (timestamp count)
Machine Model:           LARGE                              
System Endianness:       LITTLE                             

==========               
HSA Agents               
==========               
*******                  
Agent 1                  
*******                  
  Name:                    AMD Ryzen 5 2400G with Radeon Vega Graphics
  Uuid:                    CPU-XX                             
  Marketing Name:          AMD Ryzen 5 2400G with Radeon Vega Graphics
  Vendor Name:             CPU                                
  Feature:                 None specified                     
  Profile:                 FULL_PROFILE                       
  Float Round Mode:        NEAR                               
  Max Queue Number:        0(0x0)                             
  Queue Min Size:          0(0x0)                             
  Queue Max Size:          0(0x0)                             
  Queue Type:              MULTI                              
  Node:                    0                                  
  Device Type:             CPU                                
  Cache Info:              
    L1:                      32(0x20) KB                        
  Chip ID:                 5597(0x15dd)                       
  Cacheline Size:          64(0x40)                           
  Max Clock Freq. (MHz):   3600                               
  BDFID:                   12288                              
  Internal Node ID:        0                                  
  Compute Unit:            8                                  
  SIMDs per CU:            4                                  
  Shader Engines:          1                                  
  Shader Arrs. per Eng.:   1                                  
  WatchPts on Addr. Ranges:4                                  
  Features:                None
  Pool Info:               
    Pool 1                   
      Segment:                 GLOBAL; FLAGS: FINE GRAINED        
      Size:                    67108480(0x3fffe80) KB             
      Allocatable:             TRUE                               
      Alloc Granule:           4KB                                
      Alloc Alignment:         4KB                                
      Accessible by all:       TRUE                               
    Pool 2                   
      Segment:                 GLOBAL; FLAGS: KERNARG, FINE GRAINED
      Size:                    67108480(0x3fffe80) KB             
      Allocatable:             TRUE                               
      Alloc Granule:           4KB                                
      Alloc Alignment:         4KB                                
      Accessible by all:       TRUE                               
  ISA Info:                
*******                  
Agent 2                  
*******                  
  Name:                    gfx902                             
  Uuid:                    GPU-XX                             
  Marketing Name:          AMD Ryzen 5 2400G with Radeon Vega Graphics
  Vendor Name:             AMD                                
  Feature:                 KERNEL_DISPATCH                    
  Profile:                 FULL_PROFILE                       
  Float Round Mode:        NEAR                               
  Max Queue Number:        128(0x80)                          
  Queue Min Size:          64(0x40)                           
  Queue Max Size:          131072(0x20000)                    
  Queue Type:              MULTI                              
  Node:                    0                                  
  Device Type:             GPU                                
  Cache Info:              
  Chip ID:                 5597(0x15dd)                       
  Cacheline Size:          64(0x40)                           
  Max Clock Freq. (MHz):   1250                               
  BDFID:                   12288                              
  Internal Node ID:        0                                  
  Compute Unit:            11                                 
  SIMDs per CU:            4                                  
  Shader Engines:          1                                  
  Shader Arrs. per Eng.:   1                                  
  WatchPts on Addr. Ranges:4                                  
  Features:                KERNEL_DISPATCH 
  Fast F16 Operation:      TRUE                               
  Wavefront Size:          64(0x40)                           
  Workgroup Max Size:      1024(0x400)                        
  Workgroup Max Size per Dimension:
    x                        1024(0x400)                        
    y                        1024(0x400)                        
    z                        1024(0x400)                        
  Max Waves Per CU:        160(0xa0)                          
  Max Work-item Per CU:    10240(0x2800)                      
  Grid Max Size:           4294967295(0xffffffff)             
  Grid Max Size per Dimension:
    x                        4294967295(0xffffffff)             
    y                        4294967295(0xffffffff)             
    z                        4294967295(0xffffffff)             
  Max fbarriers/Workgrp:   32                                 
  Pool Info:               
    Pool 1                   
      Segment:                 GROUP                              
      Size:                    64(0x40) KB                        
      Allocatable:             FALSE                              
      Alloc Granule:           0KB                                
      Alloc Alignment:         0KB                                
      Accessible by all:       FALSE                              
  ISA Info:                
    ISA 1                    
      Name:                    amdgcn-amd-amdhsa--gfx902:xnack+   
      Machine Models:          HSA_MACHINE_MODEL_LARGE            
      Profiles:                HSA_PROFILE_BASE                   
      Default Rounding Mode:   NEAR                               
      Default Rounding Mode:   NEAR                               
      Fast f16:                TRUE                               
      Workgroup Max Size:      1024(0x400)                        
      Workgroup Max Size per Dimension:
        x                        1024(0x400)                        
        y                        1024(0x400)                        
        z                        1024(0x400)                        
      Grid Max Size:           4294967295(0xffffffff)             
      Grid Max Size per Dimension:
        x                        4294967295(0xffffffff)             
        y                        4294967295(0xffffffff)             
        z                        4294967295(0xffffffff)             
      FBarrier Max Size:       32                                 
*** Done ***

1 reply

biergaizi Jul 18, 2023

Also, why do I have two agents?

The Agent 1 is CPU-only, Agent 2 is the GPU. Because the GPU is integrated on the CPU, they have the same name, but their UUIDs are different - one begins with CPU- and another begins with GPU-. On systems with dedicated GPUs, the first agent is always CPU-only with the name of your CPU model (e.g. Xeon, EPYC).

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

rocm-arch

[rocclr] Zero cache size on Raven Ridge #670

{{title}}

Replies: 15 comments 1 reply

{{title}}

{{title}}

{{title}}

{{editor}}'s edit

{{editor}}'s edit

{{title}}

{{title}}

{{editor}}'s edit

{{editor}}'s edit

{{title}}

{{title}}

{{title}}

{{editor}}'s edit

{{editor}}'s edit

{{title}}

{{editor}}'s edit

{{editor}}'s edit

{{title}}

{{title}}

{{editor}}'s edit

{{editor}}'s edit

{{title}}

{{title}}

{{title}}

{{editor}}'s edit

{{editor}}'s edit

{{title}}

{{title}}

{{editor}}'s edit

{{editor}}'s edit

Select a reply

[rocclr] Zero cache size on Raven Ridge #670

Replies: 15 comments · 1 reply

tpkessler Aug 29, 2021 Maintainer

imyxh Aug 29, 2021 Author

tpkessler Nov 21, 2021 Maintainer

imyxh Nov 22, 2021 Author

tpkessler Nov 22, 2021 Maintainer

imyxh Nov 22, 2021 Author

tpkessler Nov 25, 2021 Maintainer

imyxh Dec 18, 2021 Author

acxz Apr 10, 2022 Maintainer

Replies: 15 comments 1 reply

tpkessler
Aug 29, 2021
Maintainer

imyxh
Aug 29, 2021
Author

tpkessler
Nov 21, 2021
Maintainer

imyxh
Nov 22, 2021
Author

tpkessler
Nov 22, 2021
Maintainer

imyxh
Nov 22, 2021
Author

tpkessler
Nov 25, 2021
Maintainer

imyxh
Dec 18, 2021
Author

acxz
Apr 10, 2022
Maintainer