Merge branch 'main' into cleanup

CHANGELOG.md

@@ -138,3 +138,20 @@ Features:
 
 Bug fixes:
  - Fixed a problem where warning messages would be displayed even though everything worked correctly.
+
+
+### 0.35.2
+
+Bug fixes:
+ - Fixed a bug where the CUDA setup failed due to a wrong function call.
+
+### 0.35.3
+
+Bug fixes:
+ - Fixed a bug in the CUDA Setup which led to an incomprehensible error if no GPU was detected.
+
+### 0.35.4
+
+Bug fixes:
+ - Fixed a bug in the CUDA Setup failed with the cuda runtime was found, but not the cuda library.
+ - Fixed a bug where not finding the cuda runtime led to an incomprehensible error.

bitsandbytes/cextension.py

@@ -116,7 +116,7 @@ try:
         CUDASetup.get_instance().generate_instructions()
         CUDASetup.get_instance().print_log_stack()
         raise RuntimeError('''
-        CUDA Setup failed despite GPU being available. Inspect the CUDA SETUP outputs to fix your environment!
+        CUDA Setup failed despite GPU being available. Inspect the CUDA SETUP outputs aboveto fix your environment!
         If you cannot find any issues and suspect a bug, please open an issue with detals about your environment:
         https://github.com/TimDettmers/bitsandbytes/issues''')
     lib.cadam32bit_g32
@@ -124,8 +124,6 @@ try:
     lib.get_cusparse.restype = ct.c_void_p
     COMPILED_WITH_CUDA = True
 except AttributeError:
-    warn(
+    warn("The installed version of bitsandbytes was compiled without GPU support. "
-        "The installed version of bitsandbytes was compiled without GPU support. "
+        "8-bit optimizers and GPU quantization are unavailable.")
-        "8-bit optimizers and GPU quantization are unavailable."
-    )
     COMPILED_WITH_CUDA = False

bitsandbytes/cuda_setup/main.py

@@ -17,6 +17,7 @@ evaluation:
 """
 
 import ctypes
+import torch
 
 from bitsandbytes.cextension import CUDASetup
 
@@ -28,14 +29,17 @@ def check_cuda_result(cuda, result_val):
     if result_val != 0:
         error_str = ctypes.c_char_p()
         cuda.cuGetErrorString(result_val, ctypes.byref(error_str))
-        CUDASetup.get_instance.add_log_entry(f"CUDA exception! Error code: {error_str.value.decode()}")
+        CUDASetup.get_instance().add_log_entry(f"CUDA exception! Error code: {error_str.value.decode()}")
+
 
-def get_cuda_version(cuda, cudart_path):
 # https://docs.nvidia.com/cuda/cuda-runtime-api/group__CUDART____VERSION.html#group__CUDART____VERSION
+def get_cuda_version(cuda, cudart_path):
+    if cuda is None: return None
+
     try:
         cudart = ctypes.CDLL(cudart_path)
     except OSError:
-        CUDASetup.get_instance.add_log_entry(f'ERROR: libcudart.so could not be read from path: {cudart_path}!')
+        CUDASetup.get_instance().add_log_entry(f'ERROR: libcudart.so could not be read from path: {cudart_path}!')
         return None
 
     version = ctypes.c_int()
@@ -55,7 +59,7 @@ def get_cuda_lib_handle():
     try:
         cuda = ctypes.CDLL("libcuda.so")
     except OSError:
-        CUDA_RUNTIME_LIB.get_instance().add_log_entry('CUDA SETUP: WARNING! libcuda.so not found! Do you have a CUDA driver installed? If you are on a cluster, make sure you are on a CUDA machine!')
+        CUDASetup.get_instance().add_log_entry('CUDA SETUP: WARNING! libcuda.so not found! Do you have a CUDA driver installed? If you are on a cluster, make sure you are on a CUDA machine!')
         return None
     check_cuda_result(cuda, cuda.cuInit(0))
 
@@ -73,7 +77,6 @@ def get_compute_capabilities(cuda):
     # bits taken from https://gist.github.com/f0k/63a664160d016a491b2cbea15913d549
     """
 
-
     nGpus = ctypes.c_int()
     cc_major = ctypes.c_int()
     cc_minor = ctypes.c_int()
@@ -87,9 +90,7 @@ def get_compute_capabilities(cuda):
         ref_major = ctypes.byref(cc_major)
         ref_minor = ctypes.byref(cc_minor)
         # 2. call extern C function to determine CC
-        check_cuda_result(
+        check_cuda_result(cuda, cuda.cuDeviceComputeCapability(ref_major, ref_minor, device))
-            cuda, cuda.cuDeviceComputeCapability(ref_major, ref_minor, device)
-        )
         ccs.append(f"{cc_major.value}.{cc_minor.value}")
 
     return ccs
@@ -102,11 +103,11 @@ def get_compute_capability(cuda):
     capabilities are downwards compatible. If no GPUs are detected, it returns
     None.
     """
-    ccs = get_compute_capabilities(cuda)
+    if cuda is None: return None
-    if ccs:
+
     # TODO: handle different compute capabilities; for now, take the max
-        return ccs[-1]
+    ccs = get_compute_capabilities(cuda)
-    return None
+    if ccs: return ccs[-1]
 
 
 def evaluate_cuda_setup():
@@ -116,28 +117,31 @@ def evaluate_cuda_setup():
     #print('Welcome to bitsandbytes. For bug reports, please submit your error trace to: https://github.com/TimDettmers/bitsandbytes/issues')
     #print('For effortless bug reporting copy-paste your error into this form: https://docs.google.com/forms/d/e/1FAIpQLScPB8emS3Thkp66nvqwmjTEgxp8Y9ufuWTzFyr9kJ5AoI47dQ/viewform?usp=sf_link')
     #print('='*80)
-    #if not torch.cuda.is_available():
+    if not torch.cuda.is_available(): return 'libsbitsandbytes_cpu.so', None, None, None, None
-        #print('No GPU detected. Loading CPU library...')
-        #return binary_name
-
-    binary_name = "libbitsandbytes_cpu.so"
 
     cuda_setup = CUDASetup.get_instance()
     cudart_path = determine_cuda_runtime_lib_path()
-    if cudart_path is None:
-        cuda_setup.add_log_entry("WARNING: No libcudart.so found! Install CUDA or the cudatoolkit package (anaconda)!", is_warning=True)
-        return binary_name
-
-    cuda_setup.add_log_entry(f"CUDA SETUP: CUDA runtime path found: {cudart_path}")
     cuda = get_cuda_lib_handle()
     cc = get_compute_capability(cuda)
-    cuda_setup.add_log_entry(f"CUDA SETUP: Highest compute capability among GPUs detected: {cc}")
     cuda_version_string = get_cuda_version(cuda, cudart_path)
 
+    failure = False
+    if cudart_path is None:
+        failure = True
+        cuda_setup.add_log_entry("WARNING: No libcudart.so found! Install CUDA or the cudatoolkit package (anaconda)!", is_warning=True)
+    else:
+        cuda_setup.add_log_entry(f"CUDA SETUP: CUDA runtime path found: {cudart_path}")
 
-    if cc == '':
+    if cc == '' or cc is None:
-        cuda_setup.add_log_entry("WARNING: No GPU detected! Check your CUDA paths. Processing to load CPU-only library...", is_warning=True)
+        failure = True
-        return binary_name
+        cuda_setup.add_log_entry("WARNING: No GPU detected! Check your CUDA paths. Proceeding to load CPU-only library...", is_warning=True)
+    else:
+        cuda_setup.add_log_entry(f"CUDA SETUP: Highest compute capability among GPUs detected: {cc}")
+
+    if cuda is None:
+        failure = True
+    else:
+        cuda_setup.add_log_entry(f'CUDA SETUP: Detected CUDA version {cuda_version_string}')
 
     # 7.5 is the minimum CC vor cublaslt
     has_cublaslt = cc in ["7.5", "8.0", "8.6"]
@@ -148,16 +152,13 @@ def evaluate_cuda_setup():
 
     # we use ls -l instead of nvcc to determine the cuda version
     # since most installations will have the libcudart.so installed, but not the compiler
-    cuda_setup.add_log_entry(f'CUDA SETUP: Detected CUDA version {cuda_version_string}')
 
-    def get_binary_name():
+    if failure:
-        "if not has_cublaslt (CC < 7.5), then we have to choose  _nocublaslt.so"
+        binary_name = "libbitsandbytes_cpu.so"
-        bin_base_name = "libbitsandbytes_cuda"
+    elif has_cublaslt:
-        if has_cublaslt:
+        binary_name = f"libbitsandbytes_cuda{cuda_version_string}.so"
-            return f"{bin_base_name}{cuda_version_string}.so"
     else:
-            return f"{bin_base_name}{cuda_version_string}_nocublaslt.so"
+        "if not has_cublaslt (CC < 7.5), then we have to choose  _nocublaslt.so"
-
+        binary_name = f"libbitsandbytes_cuda{cuda_version_string}_nocublaslt.so"
-    binary_name = get_binary_name()
 
     return binary_name, cudart_path, cuda, cc, cuda_version_string

bitsandbytes/cuda_setup/paths.py

@@ -62,7 +62,7 @@ def warn_in_case_of_duplicates(results_paths: Set[Path]) -> None:
             "If you get `CUDA error: invalid device function` errors, the above "
             "might be the cause and the solution is to make sure only one "
             f"{CUDA_RUNTIME_LIB} in the paths that we search based on your env.")
-        CUDASetup.get_instance.add_log_entry(warning_msg, is_warning=True)
+        CUDASetup.get_instance().add_log_entry(warning_msg, is_warning=True)
 
 
 def determine_cuda_runtime_lib_path() -> Union[Path, None]:
@@ -88,7 +88,7 @@ def determine_cuda_runtime_lib_path() -> Union[Path, None]:
         if conda_cuda_libs:
             return next(iter(conda_cuda_libs))
 
-        CUDASetup.get_instance.add_log_entry(f'{candidate_env_vars["CONDA_PREFIX"]} did not contain '
+        CUDASetup.get_instance().add_log_entry(f'{candidate_env_vars["CONDA_PREFIX"]} did not contain '
             f'{CUDA_RUNTIME_LIB} as expected! Searching further paths...', is_warning=True)
 
     if "LD_LIBRARY_PATH" in candidate_env_vars:

setup.py

@@ -18,7 +18,7 @@ def read(fname):
 
 setup(
     name=f"bitsandbytes",
-    version=f"0.35.1",
+    version=f"0.35.4",
     author="Tim Dettmers",
     author_email="dettmers@cs.washington.edu",
     description="8-bit optimizers and matrix multiplication routines.",