Coverage report: /home/ellis/comp/core/ffi/cuda/type.lisp
| Kind | Covered | All | % |
| expression | 0 | 14 | 0.0 |
| branch | 0 | 0 | nil |
Key
Not instrumented
Conditionalized out
Executed
Not executed
Both branches taken
One branch taken
Neither branch taken
1
;;; type.lisp --- CUDA Types8
(define-alien-type cu-device int)9
(define-alien-type cu-context (* t))10
(define-alien-type cu-module (* t))11
(define-alien-type cu-function (* t))12
(define-alien-type cu-stream (* t))13
(define-alien-type cu-event (* t))14
(define-alien-type cu-graphics-resource (* t))15
(define-alien-type cu-device-ptr unsigned-long-long)17
(define-alien-enum (cu-event-flag int)23
(define-alien-enum (cu-stream-flag int)27
(define-alien-enum (cu-memhostalloc-flag int)32
(define-alien-enum (cu-memhostregister-flag int)40
;; The API call returned with no errors. In the case of query calls, this41
;; also means that the operation being queried is complete (see42
;; ::cuEventQuery() and ::cuStreamQuery()).45
* This indicates that one or more of the parameters passed to the API call46
* is not within an acceptable range of values.48
:error-invalid-value 151
* The API call failed because it was unable to allocate enough memory or52
* other resources to perform the requested operation.54
:error-out-of-memory 257
* This indicates that the CUDA driver has not been initialized with58
* ::cuInit() or that initialization has failed.60
:error-not-initialized 363
* This indicates that the CUDA driver is in the process of shutting down.65
:error-deinitialized 468
* This indicates profiler is not initialized for this run. This can69
* happen when the application is running with external profiling tools70
* like visual profiler.72
:error-profiler-disabled 576
* This error return is deprecated as of CUDA 5.0. It is no longer an error77
* to attempt to enable/disable the profiling via ::cuProfilerStart or78
* ::cuProfilerStop without initialization.80
:error-profiler-not-initialized 684
* This error return is deprecated as of CUDA 5.0. It is no longer an error85
* to call cuProfilerStart() when profiling is already enabled.87
:error-profiler-already-started 791
* This error return is deprecated as of CUDA 5.0. It is no longer an error92
* to call cuProfilerStop() when profiling is already disabled.94
:error-profiler-already-stopped 897
* This indicates that the CUDA driver that the application has loaded is a98
* stub library. Applications that run with the stub rather than a real99
* driver loaded will result in CUDA API returning this error.101
:error-stub-library 34104
* This indicates that requested CUDA device is unavailable at the current105
* time. Devices are often unavailable due to use of106
* ::CU_COMPUTEMODE_EXCLUSIVE_PROCESS or ::CU_COMPUTEMODE_PROHIBITED.108
:error-device-unavailable 46111
* This indicates that no CUDA-capable devices were detected by the installed117
* This indicates that the device ordinal supplied by the user does not118
* correspond to a valid CUDA device or that the action requested is119
* invalid for the specified device.121
:error-invalid-device 101124
* This error indicates that the Grid license is not applied.126
:error-device-not-licensed 102129
* This indicates that the device kernel image is invalid. This can also130
* indicate an invalid CUDA module.132
:error-invalid-image 200135
* This most frequently indicates that there is no context bound to the136
* current thread. This can also be returned if the context passed to an137
* API call is not a valid handle (such as a context that has had138
* ::cuCtxDestroy() invoked on it). This can also be returned if a user139
* mixes different API versions (i.e. 3010 context with 3020 API calls).140
* See ::cuCtxGetApiVersion() for more details.141
* This can also be returned if the green context passed to an API call142
* was not converted to a ::CUcontext using ::cuCtxFromGreenCtx API.144
:error-invalid-context 201147
* This indicated that the context being supplied as a parameter to the148
* API call was already the active context.150
* This error return is deprecated as of CUDA 3.2. It is no longer an151
* error to attempt to push the active context via ::cuCtxPushCurrent().153
:error-context-already-current 202156
* This indicates that a map or register operation has failed.158
:error-map-failed 205161
* This indicates that an unmap or unregister operation has failed.163
:error-unmap-failed 206166
* This indicates that the specified array is currently mapped and thus167
* cannot be destroyed.169
:error-array-is-mapped 207172
* This indicates that the resource is already mapped.174
:error-already-mapped 208177
* This indicates that there is no kernel image available that is suitable178
* for the device. This can occur when a user specifies code generation179
* options for a particular CUDA source file that do not include the180
* corresponding device configuration.182
:error-no-binary-for-gpu 209185
* This indicates that a resource has already been acquired.187
:error-already-acquired 210190
* This indicates that a resource is not mapped.192
:error-not-mapped 211195
* This indicates that a mapped resource is not available for access as an198
:error-not-mapped-as-array 212201
* This indicates that a mapped resource is not available for access as a204
:error-not-mapped-as-pointer 213207
* This indicates that an uncorrectable ECC error was detected during210
:error-ecc-uncorrectable 214213
* This indicates that the ::CUlimit passed to the API call is not214
* supported by the active device.216
:error-unsupported-limit 215219
* This indicates that the ::CUcontext passed to the API call can220
* only be bound to a single CPU thread at a time but is already221
* bound to a CPU thread.223
:error-context-already-in-use 216226
* This indicates that peer access is not supported across the given229
:error-peer-access-unsupported 217232
* This indicates that a PTX JIT compilation failed.234
:error-invalid-ptx 218237
* This indicates an error with OpenGL or DirectX context.239
:error-invalid-graphics-context 219242
* This indicates that an uncorrectable NVLink error was detected during the245
:error-nvlink-uncorrectable 220248
* This indicates that the PTX JIT compiler library was not found.250
:error-jit-compiler-not-found 221253
* This indicates that the provided PTX was compiled with an unsupported toolchain.255
:error-unsupported-ptx-version 222258
* This indicates that the PTX JIT compilation was disabled.260
:error-jit-compilation-disabled 223263
* This indicates that the ::CUexecAffinityType passed to the API call is not264
* supported by the active device.266
:error-unsupported-exec-affinity 224269
* This indicates that the code to be compiled by the PTX JIT contains270
* unsupported call to cudaDeviceSynchronize.272
:error-unsupported-devside-sync 225275
* This indicates that the device kernel source is invalid. This includes276
* compilation/linker errors encountered in device code or user error.278
:error-invalid-source 300281
* This indicates that the file specified was not found.283
:error-file-not-found 301286
* This indicates that a link to a shared object failed to resolve.288
:error-shared-object-symbol-not-found 302291
* This indicates that initialization of a shared object failed.293
:error-shared-object-init-failed 303296
* This indicates that an OS call failed.298
:error-operating-system 304301
* This indicates that a resource handle passed to the API call was not302
* valid. Resource handles are opaque types like ::CUstream and ::CUevent.304
:error-invalid-handle 400307
* This indicates that a resource required by the API call is not in a308
* valid state to perform the requested operation.310
:error-illegal-state 401313
* This indicates an attempt was made to introspect an object in a way that314
* would discard semantically important information. This is either due to315
* the object using funtionality newer than the API version used to316
* introspect it or omission of optional return arguments.318
:error-lossy-query 402321
* This indicates that a named symbol was not found. Examples of symbols322
* are global/constant variable names, driver function names, texture names,328
* This indicates that asynchronous operations issued previously have not329
* completed yet. This result is not actually an error, but must be indicated330
* differently than ::CUDA_SUCCESS (which indicates completion). Calls that331
* may return this value include ::cuEventQuery() and ::cuStreamQuery().336
* While executing a kernel, the device encountered a337
* load or store instruction on an invalid memory address.338
* This leaves the process in an inconsistent state and any further CUDA work339
* will return the same error. To continue using CUDA, the process must be terminated342
:error-illegal-address 700345
* This indicates that a launch did not occur because it did not have346
* appropriate resources. This error usually indicates that the user has347
* attempted to pass too many arguments to the device kernel, or the348
* kernel launch specifies too many threads for the kernel's register349
* count. Passing arguments of the wrong size (i.e. a 64-bit pointer350
* when a 32-bit int is expected) is equivalent to passing too many351
* arguments and can also result in this error.353
:error-launch-out-of-resources 701356
* This indicates that the device kernel took too long to execute. This can357
* only occur if timeouts are enabled - see the device attribute358
* ::CU_DEVICE_ATTRIBUTE_KERNEL_EXEC_TIMEOUT for more information.359
* This leaves the process in an inconsistent state and any further CUDA work360
* will return the same error. To continue using CUDA, the process must be terminated363
:error-launch-timeout 702366
* This error indicates a kernel launch that uses an incompatible texturing369
:error-launch-incompatible-texturing 703372
* This error indicates that a call to ::cuCtxEnablePeerAccess() is373
* trying to re-enable peer access to a context which has already374
* had peer access to it enabled.376
:error-peer-access-already-enabled 704379
* This error indicates that ::cuCtxDisablePeerAccess() is380
* trying to disable peer access which has not been enabled yet381
* via ::cuCtxEnablePeerAccess().383
:error-peer-access-not-enabled 705386
* This error indicates that the primary context for the specified device387
* has already been initialized.389
:error-primary-context-active 708392
* This error indicates that the context current to the calling thread393
* has been destroyed using ::cuCtxDestroy, or is a primary context which394
* has not yet been initialized.396
:error-context-is-destroyed 709399
* A device-side assert triggered during kernel execution. The context400
* cannot be used anymore, and must be destroyed. All existing device401
* memory allocations from this context are invalid and must be402
* reconstructed if the program is to continue using CUDA.407
* This error indicates that the hardware resources required to enable408
* peer access have been exhausted for one or more of the devices409
* passed to ::cuCtxEnablePeerAccess().411
:error-too-many-peers 711414
* This error indicates that the memory range passed to ::cuMemHostRegister()415
* has already been registered.417
:error-host-memory-already-registered 712420
* This error indicates that the pointer passed to ::cuMemHostUnregister()421
* does not correspond to any currently registered memory region.423
:error-host-memory-not-registered 713426
* While executing a kernel, the device encountered a stack error.427
* This can be due to stack corruption or exceeding the stack size limit.428
* This leaves the process in an inconsistent state and any further CUDA work429
* will return the same error. To continue using CUDA, the process must be terminated432
:error-hardware-stack-error 714435
* While executing a kernel, the device encountered an illegal instruction.436
* This leaves the process in an inconsistent state and any further CUDA work437
* will return the same error. To continue using CUDA, the process must be terminated440
:error-illegal-instruction 715443
* While executing a kernel, the device encountered a load or store instruction444
* on a memory address which is not aligned.445
* This leaves the process in an inconsistent state and any further CUDA work446
* will return the same error. To continue using CUDA, the process must be terminated449
:error-misaligned-address 716452
* While executing a kernel, the device encountered an instruction453
* which can only operate on memory locations in certain address spaces454
* (global, shared, or local), but was supplied a memory address not455
* belonging to an allowed address space.456
* This leaves the process in an inconsistent state and any further CUDA work457
* will return the same error. To continue using CUDA, the process must be terminated460
:error-invalid-address-space 717463
* While executing a kernel, the device program counter wrapped its address space.464
* This leaves the process in an inconsistent state and any further CUDA work465
* will return the same error. To continue using CUDA, the process must be terminated468
:error-invalid-pc 718471
* An exception occurred on the device while executing a kernel. Common472
* causes include dereferencing an invalid device pointer and accessing473
* out of bounds shared memory. Less common cases can be system specific - more474
* information about these cases can be found in the system specific user guide.475
* This leaves the process in an inconsistent state and any further CUDA work476
* will return the same error. To continue using CUDA, the process must be terminated479
:error-launch-failed 719482
* This error indicates that the number of blocks launched per grid for a kernel that was483
* launched via either ::cuLaunchCooperativeKernel or ::cuLaunchCooperativeKernelMultiDevice484
* exceeds the maximum number of blocks as allowed by ::cuOccupancyMaxActiveBlocksPerMultiprocessor485
* or ::cuOccupancyMaxActiveBlocksPerMultiprocessorWithFlags times the number of multiprocessors486
* as specified by the device attribute ::CU_DEVICE_ATTRIBUTE_MULTIPROCESSOR_COUNT.488
:error-cooperative-launch-too-large 720491
* This error indicates that the attempted operation is not permitted.493
:error-not-permitted 800496
* This error indicates that the attempted operation is not supported497
* on the current system or device.499
:error-not-supported 801502
* This error indicates that the system is not yet ready to start any CUDA503
* work. To continue using CUDA, verify the system configuration is in a504
* valid state and all required driver daemons are actively running.505
* More information about this error can be found in the system specific508
:error-system-not-ready 802511
* This error indicates that there is a mismatch between the versions of512
* the display driver and the CUDA driver. Refer to the compatibility documentation513
* for supported versions.515
:error-system-driver-mismatch 803518
* This error indicates that the system was upgraded to run with forward compatibility519
* but the visible hardware detected by CUDA does not support this configuration.520
* Refer to the compatibility documentation for the supported hardware matrix or ensure521
* that only supported hardware is visible during initialization via the CUDA_VISIBLE_DEVICES522
* environment variable.524
:error-compat-not-supported-on-device 804527
* This error indicates that the MPS client failed to connect to the MPS control daemon or the MPS server.529
:error-mps-connection-failed 805532
* This error indicates that the remote procedural call between the MPS server and the MPS client failed.534
:error-mps-rpc-failure 806537
* This error indicates that the MPS server is not ready to accept new MPS client requests.538
* This error can be returned when the MPS server is in the process of recovering from a fatal failure.540
:error-mps-server-not-ready 807543
* This error indicates that the hardware resources required to create MPS client have been exhausted.545
:error-mps-max-clients-reached 808548
* This error indicates the the hardware resources required to support device connections have been exhausted.550
:error-mps-max-connections-reached 809553
* This error indicates that the MPS client has been terminated by the server. To continue using CUDA, the process must be terminated and relaunched.555
:error-mps-client-terminated 810558
* This error indicates that the module is using CUDA Dynamic Parallelism, but the current configuration, like MPS, does not support it.560
:error-cdp-not-supported 811563
* This error indicates that a module contains an unsupported interaction between different versions of CUDA Dynamic Parallelism.565
:error-cdp-version-mismatch 812568
* This error indicates that the operation is not permitted when569
* the stream is capturing.571
:error-stream-capture-unsupported 900574
* This error indicates that the current capture sequence on the stream575
* has been invalidated due to a previous error.577
:error-stream-capture-invalidated 901580
* This error indicates that the operation would have resulted in a merge581
* of two independent capture sequences.583
:error-stream-capture-merge 902586
* This error indicates that the capture was not initiated in this stream.588
:error-stream-capture-unmatched 903591
* This error indicates that the capture sequence contains a fork that was592
* not joined to the primary stream.594
:error-stream-capture-unjoined 904597
* This error indicates that a dependency would have been created which598
* crosses the capture sequence boundary. Only implicit in-stream ordering599
* dependencies are allowed to cross the boundary.601
:error-stream-capture-isolation 905604
* This error indicates a disallowed implicit dependency on a current capture605
* sequence from cudaStreamLegacy.607
:error-stream-capture-implicit 906610
* This error indicates that the operation is not permitted on an event which611
* was last recorded in a capturing stream.613
:error-captured-event 907616
* A stream capture sequence not initiated with the ::CU_STREAM_CAPTURE_MODE_RELAXED617
* argument to ::cuStreamBeginCapture was passed to ::cuStreamEndCapture in a620
:error-stream-capture-wrong-thread 908623
* This error indicates that the timeout specified for the wait operation has lapsed.628
* This error indicates that the graph update was not performed because it included 629
* changes which violated constraints specific to instantiated graph update.631
:error-graph-exec-update-failure 910634
* This indicates that an async error has occurred in a device outside of CUDA.635
* If CUDA was waiting for an external device's signal before consuming shared data,636
* the external device signaled an error indicating that the data is not valid for637
* consumption. This leaves the process in an inconsistent state and any further CUDA638
* work will return the same error. To continue using CUDA, the process must be639
* terminated and relaunched.641
:error-external-device 911644
* Indicates a kernel launch error due to cluster misconfiguration.646
:error-invalid-cluster-size 912649
* Indiciates a function handle is not loaded when calling an API that requires652
:error-function-not-loaded 913655
* This error indicates one or more resources passed in are not valid resource656
* types for the operation.658
:error-invalid-resource-type 914661
* This error indicates one or more resources are insufficient or non-applicable for664
:error-invalid-resource-configuration 915667
* This indicates that an unknown internal error has occurred.