guarantee(false && "hipErrorNoBinaryForGpu: Coudn't find binary for current devices!")

[CyberShadow]

Hi and thank you for all your work in putting this together.

I'm running into an error when trying to run a simple Python program. After having built and installed tensorflow-rocm and dependencies, I'm trying the following simple Python script:

from tensorflow.python.client import device_lib
print(device_lib.list_local_devices())

It produces the following output:

2021-03-18 17:08:24.637639: I tensorflow/core/platform/cpu_feature_guard.cc:142] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  SSE3 SSE4.1 SSE4.2 AVX AVX2 FMA
To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
2021-03-18 17:08:24.639302: I tensorflow/compiler/jit/xla_gpu_device.cc:99] Not creating XLA devices, tf_xla_enable_xla_devices not set
2021-03-18 17:08:24.639375: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libamdhip64.so
/build/hip-rocclr/src/HIP-rocm-4.0.0/rocclr/hip_code_object.cpp:120: guarantee(false && "hipErrorNoBinaryForGpu: Coudn't find binary for current devices!")
[1]    2964311 abort (core dumped)  python test.py

I don't know which binary it is referring to that it cannot find. Grepping strace output for ENOENT does not produce any illuminating results.

rocminfo output

�[37mROCk module is loaded�[0m
�[37mAble to open /dev/kfd read-write�[0m
=====================    
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 Threadripper 3960X 24-Core Processor
  Uuid:                    CPU-XX                             
  Marketing Name:          AMD Ryzen Threadripper 3960X 24-Core Processor
  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:                      32768(0x8000) KB                   
  Chip ID:                 0(0x0)                             
  Cacheline Size:          64(0x40)                           
  Max Clock Freq. (MHz):   3800                               
  BDFID:                   0                                  
  Internal Node ID:        0                                  
  Compute Unit:            48                                 
  SIMDs per CU:            0                                  
  Shader Engines:          0                                  
  Shader Arrs. per Eng.:   0                                  
  WatchPts on Addr. Ranges:1                                  
  Features:                None
  Pool Info:               
    Pool 1                   
      Segment:                 GLOBAL; FLAGS: KERNARG, FINE GRAINED
      Size:                    131776176(0x7dabeb0) KB            
      Allocatable:             TRUE                               
      Alloc Granule:           4KB                                
      Alloc Alignment:         4KB                                
      Accessible by all:       TRUE                               
    Pool 2                   
      Segment:                 GLOBAL; FLAGS: COARSE GRAINED      
      Size:                    131776176(0x7dabeb0) KB            
      Allocatable:             TRUE                               
      Alloc Granule:           4KB                                
      Alloc Alignment:         4KB                                
      Accessible by all:       TRUE                               
  ISA Info:                
    N/A                      
*******                  
Agent 2                  
*******                  
  Name:                    gfx900                             
  Uuid:                    GPU-021502761a624124               
  Marketing Name:          Vega 10 XL/XT [Radeon RX Vega 56/64]
  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:                 26751(0x687f)                      
  Cacheline Size:          64(0x40)                           
  Max Clock Freq. (MHz):   1630                               
  BDFID:                   19712                              
  Internal Node ID:        1                                  
  Compute Unit:            64                                 
  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:                    8372224(0x7fc000) 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--gfx900          
      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 ***             

clinfo output

Number of platforms                               1
  Platform Name                                   AMD Accelerated Parallel Processing
  Platform Vendor                                 Advanced Micro Devices, Inc.
  Platform Version                                OpenCL 2.0 AMD-APP.dbg (3212.0)
  Platform Profile                                FULL_PROFILE
  Platform Extensions                             cl_khr_icd cl_amd_event_callback 
  Platform Extensions function suffix             AMD

  Platform Name                                   AMD Accelerated Parallel Processing
Number of devices                                 1
  Device Name                                     gfx900
  Device Vendor                                   Advanced Micro Devices, Inc.
  Device Vendor ID                                0x1002
  Device Version                                  OpenCL 2.0 
  Driver Version                                  3212.0 (HSA1.1,LC)
  Device OpenCL C Version                         OpenCL C 2.0 
  Device Type                                     GPU
  Device Board Name (AMD)                         Vega 10 XL/XT [Radeon RX Vega 56/64]
  Device PCI-e ID (AMD)                           0x687f
  Device Topology (AMD)                           PCI-E, 0000:4d:00.0
  Device Profile                                  FULL_PROFILE
  Device Available                                Yes
  Compiler Available                              Yes
  Linker Available                                Yes
  Max compute units                               64
  SIMD per compute unit (AMD)                     4
  SIMD width (AMD)                                16
  SIMD instruction width (AMD)                    1
  Max clock frequency                             1630MHz
  Graphics IP (AMD)                               9.0
  Device Partition                                (core)
    Max number of sub-devices                     64
    Supported partition types                     None
    Supported affinity domains                    (n/a)
  Max work item dimensions                        3
  Max work item sizes                             1024x1024x1024
  Max work group size                             256
  Preferred work group size (AMD)                 256
  Max work group size (AMD)                       1024
  Preferred work group size multiple (kernel)     64
  Wavefront width (AMD)                           64
  Preferred / native vector sizes                 
    char                                                 4 / 4       
    short                                                2 / 2       
    int                                                  1 / 1       
    long                                                 1 / 1       
    half                                                 1 / 1        (cl_khr_fp16)
    float                                                1 / 1       
    double                                               1 / 1        (cl_khr_fp64)
  Half-precision Floating-point support           (cl_khr_fp16)
    Denormals                                     No
    Infinity and NANs                             No
    Round to nearest                              No
    Round to zero                                 No
    Round to infinity                             No
    IEEE754-2008 fused multiply-add               No
    Support is emulated in software               No
  Single-precision Floating-point support         (core)
    Denormals                                     Yes
    Infinity and NANs                             Yes
    Round to nearest                              Yes
    Round to zero                                 Yes
    Round to infinity                             Yes
    IEEE754-2008 fused multiply-add               Yes
    Support is emulated in software               No
    Correctly-rounded divide and sqrt operations  Yes
  Double-precision Floating-point support         (cl_khr_fp64)
    Denormals                                     Yes
    Infinity and NANs                             Yes
    Round to nearest                              Yes
    Round to zero                                 Yes
    Round to infinity                             Yes
    IEEE754-2008 fused multiply-add               Yes
    Support is emulated in software               No
  Address bits                                    64, Little-Endian
  Global memory size                              8573157376 (7.984GiB)
  Global free memory (AMD)                        8372224 (7.984GiB) 8372224 (7.984GiB)
  Global memory channels (AMD)                    64
  Global memory banks per channel (AMD)           4
  Global memory bank width (AMD)                  256 bytes
  Error Correction support                        No
  Max memory allocation                           7287183768 (6.787GiB)
  Unified memory for Host and Device              No
  Shared Virtual Memory (SVM) capabilities        (core)
    Coarse-grained buffer sharing                 Yes
    Fine-grained buffer sharing                   Yes
    Fine-grained system sharing                   No
    Atomics                                       No
  Minimum alignment for any data type             128 bytes
  Alignment of base address                       1024 bits (128 bytes)
  Preferred alignment for atomics                 
    SVM                                           0 bytes
    Global                                        0 bytes
    Local                                         0 bytes
  Max size for global variable                    7287183768 (6.787GiB)
  Preferred total size of global vars             8573157376 (7.984GiB)
  Global Memory cache type                        Read/Write
  Global Memory cache size                        16384 (16KiB)
  Global Memory cache line size                   64 bytes
  Image support                                   Yes
    Max number of samplers per kernel             26751
    Max size for 1D images from buffer            4294967295 pixels
    Max 1D or 2D image array size                 8192 images
    Base address alignment for 2D image buffers   256 bytes
    Pitch alignment for 2D image buffers          256 pixels
    Max 2D image size                             16384x16384 pixels
    Max 3D image size                             16384x16384x8192 pixels
    Max number of read image args                 128
    Max number of write image args                8
    Max number of read/write image args           64
  Max number of pipe args                         16
  Max active pipe reservations                    16
  Max pipe packet size                            2992216472 (2.787GiB)
  Local memory type                               Local
  Local memory size                               65536 (64KiB)
  Local memory size per CU (AMD)                  65536 (64KiB)
  Local memory banks (AMD)                        32
  Max number of constant args                     8
  Max constant buffer size                        7287183768 (6.787GiB)
  Preferred constant buffer size (AMD)            16384 (16KiB)
  Max size of kernel argument                     1024
  Queue properties (on host)                      
    Out-of-order execution                        No
    Profiling                                     Yes
  Queue properties (on device)                    
    Out-of-order execution                        Yes
    Profiling                                     Yes
    Preferred size                                262144 (256KiB)
    Max size                                      8388608 (8MiB)
  Max queues on device                            1
  Max events on device                            1024
  Prefer user sync for interop                    Yes
  Number of P2P devices (AMD)                     0
  Profiling timer resolution                      1ns
  Profiling timer offset since Epoch (AMD)        0ns (Thu Jan  1 00:00:00 1970)
  Execution capabilities                          
    Run OpenCL kernels                            Yes
    Run native kernels                            No
    Thread trace supported (AMD)                  No
    Number of async queues (AMD)                  8
    Max real-time compute queues (AMD)            8
    Max real-time compute units (AMD)             64
  printf() buffer size                            4194304 (4MiB)
  Built-in kernels                                (n/a)
  Device Extensions                               cl_khr_fp64 cl_khr_global_int32_base_atomics cl_khr_global_int32_extended_atomics cl_khr_local_int32_base_atomics cl_khr_local_int32_extended_atomics cl_khr_int64_base_atomics cl_khr_int64_extended_atomics cl_khr_3d_image_writes cl_khr_byte_addressable_store cl_khr_fp16 cl_khr_gl_sharing cl_amd_device_attribute_query cl_amd_media_ops cl_amd_media_ops2 cl_khr_image2d_from_buffer cl_khr_subgroups cl_khr_depth_images cl_amd_copy_buffer_p2p cl_amd_assembly_program 

NULL platform behavior
  clGetPlatformInfo(NULL, CL_PLATFORM_NAME, ...)  No platform
  clGetDeviceIDs(NULL, CL_DEVICE_TYPE_ALL, ...)   No platform
  clCreateContext(NULL, ...) [default]            No platform
  clCreateContext(NULL, ...) [other]              Success [AMD]
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_DEFAULT)  Success (1)
    Platform Name                                 AMD Accelerated Parallel Processing
    Device Name                                   gfx900
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_CPU)  No devices found in platform
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_GPU)  Success (1)
    Platform Name                                 AMD Accelerated Parallel Processing
    Device Name                                   gfx900
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_ACCELERATOR)  No devices found in platform
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_CUSTOM)  No devices found in platform
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_ALL)  Success (1)
    Platform Name                                 AMD Accelerated Parallel Processing
    Device Name                                   gfx900

Any advice on how to narrow down this problem would be appreciated.

Also, I don't know if this is related or not, but I'm slightly confused about the subject of kernel modules. Information such as here seems to say that I should expect to have amdkfd loaded. For me, it is not loaded, which may or may not be related to the problem above. However, despite this, /dev/kfd exists and rocminfo says "ROCk module is loaded", and I see that this repository's README.md doesn't mention kernel modules. Regardless if I try to install rock-dkms-bin, the DKMS build fails due to not being able to find kcl/backport/kcl_reservation_backport.h et al.

[CyberShadow]

With ROCm-Developer-Tools/HIP#2166 (comment) as a starting point, I traced what was causing the error:

• The problem occurs because the GPU binary blob passed to libamdhip64.so doesn't have the target we want.
  • The blob contains:
    • host-x86_64-unknown-linux
    • triple=hip-amdgcn-amd-amdhsa-gfx803�ELF@
  • But we want something with gfx900.
• This binary is sent to libamdhip64.so (via __hipRegisterFatBinary) from librccl.so (via __hip_module_ctor) in its module constructor.
• This module constructor is apparently generated by the compiler: https://github.com/llvm-mirror/clang/blob/master/lib/CodeGen/CGCUDANV.cpp#L477
• Checking the strings of the tensorflow package does seem to confirm that its GPU blobs are built only for the gfx803 target:

$ pacman -Qql python-tensorflow-opt-rocm | grep -v /$ | xargs strings | grep hip-amdgcn-amd-amdhsa 
strings: Warning: '/usr/lib/python3.9/site-packages/tensorflow/include/Eigen' is a directory
strings: Warning: '/usr/lib/python3.9/site-packages/tensorflow/include/absl' is a directory
strings: Warning: '/usr/lib/python3.9/site-packages/tensorflow/include/external' is a directory
strings: Warning: '/usr/lib/python3.9/site-packages/tensorflow/include/google' is a directory
strings: Warning: '/usr/lib/python3.9/site-packages/tensorflow/include/include' is a directory
strings: Warning: '/usr/lib/python3.9/site-packages/tensorflow/include/tensorflow' is a directory
strings: Warning: '/usr/lib/python3.9/site-packages/tensorflow/include/third_party' is a directory
strings: Warning: '/usr/lib/python3.9/site-packages/tensorflow/include/unsupported' is a directory
hip-amdgcn-amd-amdhsa-gfx803
hip-amdgcn-amd-amdhsa-gfx803
hip-amdgcn-amd-amdhsa-gfx803
hip-amdgcn-amd-amdhsa-gfx803
hip-amdgcn-amd-amdhsa-gfx803
hip-amdgcn-amd-amdhsa-gfx803
hip-amdgcn-amd-amdhsa-gfx803
hip-amdgcn-amd-amdhsa-gfx803
hip-amdgcn-amd-amdhsa-gfx803
hip-amdgcn-amd-amdhsa-gfx803
hip-amdgcn-amd-amdhsa-gfx803
hip-amdgcn-amd-amdhsa-gfx803
hip-amdgcn-amd-amdhsa-gfx803
hip-amdgcn-amd-amdhsa-gfx803
hip-amdgcn-amd-amdhsa-gfx803
hip-amdgcn-amd-amdhsa-gfx803
hip-amdgcn-amd-amdhsa-gfx803
... (etc.) ...

[CyberShadow]

Rebuilding tensorflow-rocm outside of a chroot now produces binary files which have hip-amdgcn-amd-amdhsa-gfx900 instead of -803. As expected, the newly built package works.

So, like others who encountered this problem before me, I stumbled my way to a solution, but the root cause remains undetermined. I suspect it may have something to do with the build order or chroot isolation.