|WARNING| pw104330 [PATCH] [v5,1/1] gpu/cuda: introduce CUDA driver

|WARNING| pw104330 [PATCH] [v5,1/1] gpu/cuda: introduce CUDA driver

[dpdklab]

[-- Attachment #1: Type: text/plain, Size: 1916 bytes --]

Test-Label: iol-spell-check-testing
Test-Status: WARNING
http://dpdk.org/patch/104330

_Testing issues_

Submitter: Elena Agostini <eagostini@nvidia.com>
Date: Monday, November 15 2021 22:36:02 
DPDK git baseline: Repo:dpdk
  Branch: master
  CommitID:d4fb4eb08725a3e667042c1f5cd3ed72b94505a7

104330 --> testing fail

Test environment and result as below:

+--------------+-------------+
| Environment  | spell_check |
+==============+=============+
| Ubuntu 20.04 | FAIL        |
+--------------+-------------+

==== 20 line log output for Ubuntu 20.04 (spell_check): ====
dpdk/doc/guides/sample_app_ug/ipsec_secgw.rst:94 vector-size
dpdk/doc/guides/sample_app_ug/ipsec_secgw.rst:94 vector-tmo
dpdk/doc/guides/sample_app_ug/ipsec_secgw.rst:94 vector-tmo
dpdk/doc/guides/sample_app_ug/ipsec_secgw.rst:311 vector-size
dpdk/doc/guides/sample_app_ug/ipsec_secgw.rst:312 vector-tmo
dpdk/doc/guides/sample_app_ug/ipsec_secgw.rst:333 vector-size
dpdk/doc/guides/sample_app_ug/ipsec_secgw.rst:335 vector-tmo
dpdk/doc/guides/nics/mlx5.rst:619 WQEs
dpdk/doc/guides/nics/mlx5.rst:624 re-activeate
dpdk/doc/guides/nics/mlx5.rst:631 set-priv-flags
dpdk/doc/guides/nics/mlx5.rst:632 show-priv-flags
dpdk/doc/guides/vdpadevs/sfc.rst:22 libefx
dpdk/doc/guides/rel_notes/release_21_11.rst:114 CUDA
dpdk/doc/guides/rel_notes/release_21_11.rst:176 AFM
dpdk/doc/guides/rel_notes/release_21_11.rst:257 RxTx
dpdk/doc/guides/dmadevs/dpaa.rst:43 dpaa-based
dpdk/doc/guides/index.rst:25 gpus
dpdk/doc/guides/eventdevs/dsw.rst:48 MAINT
dpdk/doc/guides/eventdevs/dsw.rst:70 MAINT
Errors found: 172
==== End log output ====

Ubuntu 20.04
	Kernel: 4.18.0-240.10.1.el8_3.x86_64
	Compiler: gcc 9.3.0-17ubuntu1~20.04

To view detailed results, visit:
https://lab.dpdk.org/results/dashboard/patchsets/20208/

UNH-IOL DPDK Community Lab

To manage your email subscriptions, visit: 
https://lab.dpdk.org/results/dashboard/preferences/subscriptions/

|WARNING| pw104330 [PATCH v5 1/1] gpu/cuda: introduce CUDA driver

[checkpatch]

Test-Label: checkpatch
Test-Status: WARNING
http://dpdk.org/patch/104330

_coding style issues_


WARNING:LONG_LINE: line length of 109 exceeds 100 columns
#326: FILE: drivers/gpu/cuda/cuda.c:24:
+enum cuError (*sym_cuGetProcAddress)(const char *symbol, void **pfn, int cudaVersion, uint64_t flags) = NULL;

WARNING:LONG_LINE: line length of 114 exceeds 100 columns
#485: FILE: drivers/gpu/cuda/cuda.c:183:
+	res = sym_cuGetProcAddress("cuGetErrorString", (void **) (&pfn_cuGetErrorString), cuda_driver_version, 0);

WARNING:LONG_LINE: line length of 110 exceeds 100 columns
#491: FILE: drivers/gpu/cuda/cuda.c:189:
+	res = sym_cuGetProcAddress("cuGetErrorName", (void **) (&pfn_cuGetErrorName), cuda_driver_version, 0);

WARNING:LONG_LINE: line length of 124 exceeds 100 columns
#497: FILE: drivers/gpu/cuda/cuda.c:195:
+	res = sym_cuGetProcAddress("cuPointerSetAttribute", (void **) (&pfn_cuPointerSetAttribute), cuda_driver_version, 0);

WARNING:LONG_LINE: line length of 122 exceeds 100 columns
#503: FILE: drivers/gpu/cuda/cuda.c:201:
+	res = sym_cuGetProcAddress("cuDeviceGetAttribute", (void **) (&pfn_cuDeviceGetAttribute), cuda_driver_version, 0);

WARNING:LONG_LINE: line length of 124 exceeds 100 columns
#509: FILE: drivers/gpu/cuda/cuda.c:207:
+	res = sym_cuGetProcAddress("cuDeviceGetByPCIBusId", (void **) (&pfn_cuDeviceGetByPCIBusId), cuda_driver_version, 0);

WARNING:LONG_LINE: line length of 112 exceeds 100 columns
#515: FILE: drivers/gpu/cuda/cuda.c:213:
+	res = sym_cuGetProcAddress("cuDeviceGetName", (void **) (&pfn_cuDeviceGetName), cuda_driver_version, 0);

WARNING:LONG_LINE: line length of 130 exceeds 100 columns
#521: FILE: drivers/gpu/cuda/cuda.c:219:
+	res = sym_cuGetProcAddress("cuDevicePrimaryCtxRetain", (void **) (&pfn_cuDevicePrimaryCtxRetain), cuda_driver_version, 0);

WARNING:LONG_LINE: line length of 101 exceeds 100 columns
#523: FILE: drivers/gpu/cuda/cuda.c:221:
+		rte_gpu_cuda_log(ERR, "Retrieve pfn_cuDevicePrimaryCtxRetain failed with %d
", res);

WARNING:LONG_LINE: line length of 132 exceeds 100 columns
#527: FILE: drivers/gpu/cuda/cuda.c:225:
+	res = sym_cuGetProcAddress("cuDevicePrimaryCtxRelease", (void **) (&pfn_cuDevicePrimaryCtxRelease), cuda_driver_version, 0);

WARNING:LONG_LINE: line length of 102 exceeds 100 columns
#529: FILE: drivers/gpu/cuda/cuda.c:227:
+		rte_gpu_cuda_log(ERR, "Retrieve pfn_cuDevicePrimaryCtxRelease failed with %d
", res);

WARNING:LONG_LINE: line length of 114 exceeds 100 columns
#533: FILE: drivers/gpu/cuda/cuda.c:231:
+	res = sym_cuGetProcAddress("cuDeviceTotalMem", (void **) (&pfn_cuDeviceTotalMem), cuda_driver_version, 0);

WARNING:LONG_LINE: line length of 118 exceeds 100 columns
#539: FILE: drivers/gpu/cuda/cuda.c:237:
+	res = sym_cuGetProcAddress("cuCtxGetApiVersion", (void **) (&pfn_cuCtxGetApiVersion), cuda_driver_version, 0);

WARNING:LONG_LINE: line length of 110 exceeds 100 columns
#545: FILE: drivers/gpu/cuda/cuda.c:243:
+	res = sym_cuGetProcAddress("cuCtxGetDevice", (void **) (&pfn_cuCtxGetDevice), cuda_driver_version, 0);

WARNING:LONG_LINE: line length of 112 exceeds 100 columns
#551: FILE: drivers/gpu/cuda/cuda.c:249:
+	res = sym_cuGetProcAddress("cuCtxSetCurrent", (void **) (&pfn_cuCtxSetCurrent), cuda_driver_version, 0);

WARNING:LONG_LINE: line length of 112 exceeds 100 columns
#557: FILE: drivers/gpu/cuda/cuda.c:255:
+	res = sym_cuGetProcAddress("cuCtxGetCurrent", (void **) (&pfn_cuCtxGetCurrent), cuda_driver_version, 0);

WARNING:LONG_LINE: line length of 122 exceeds 100 columns
#563: FILE: drivers/gpu/cuda/cuda.c:261:
+	res = sym_cuGetProcAddress("cuCtxGetExecAffinity", (void **) (&pfn_cuCtxGetExecAffinity), cuda_driver_version, 0);

WARNING:LONG_LINE: line length of 102 exceeds 100 columns
#569: FILE: drivers/gpu/cuda/cuda.c:267:
+	res = sym_cuGetProcAddress("cuMemAlloc", (void **) (&pfn_cuMemAlloc), cuda_driver_version, 0);

WARNING:LONG_LINE: line length of 116 exceeds 100 columns
#581: FILE: drivers/gpu/cuda/cuda.c:279:
+	res = sym_cuGetProcAddress("cuMemHostRegister", (void **) (&pfn_cuMemHostRegister), cuda_driver_version, 0);

WARNING:LONG_LINE: line length of 120 exceeds 100 columns
#587: FILE: drivers/gpu/cuda/cuda.c:285:
+	res = sym_cuGetProcAddress("cuMemHostUnregister", (void **) (&pfn_cuMemHostUnregister), cuda_driver_version, 0);

WARNING:LONG_LINE: line length of 132 exceeds 100 columns
#593: FILE: drivers/gpu/cuda/cuda.c:291:
+	res = sym_cuGetProcAddress("cuMemHostGetDevicePointer", (void **) (&pfn_cuMemHostGetDevicePointer), cuda_driver_version, 0);

WARNING:LONG_LINE: line length of 102 exceeds 100 columns
#595: FILE: drivers/gpu/cuda/cuda.c:293:
+		rte_gpu_cuda_log(ERR, "Retrieve pfn_cuMemHostGetDevicePointer failed with %d
", res);

WARNING:LONG_LINE: line length of 134 exceeds 100 columns
#599: FILE: drivers/gpu/cuda/cuda.c:297:
+	res = sym_cuGetProcAddress("cuFlushGPUDirectRDMAWrites", (void **) (&pfn_cuFlushGPUDirectRDMAWrites), cuda_driver_version, 0);

WARNING:LONG_LINE: line length of 103 exceeds 100 columns
#974: FILE: drivers/gpu/cuda/cuda.c:672:
+		if ((uintptr_t) mem_alloc_list_tail->ptr_d != (uintptr_t) mem_alloc_list_tail->ptr_h) {

WARNING:LONG_LINE: line length of 102 exceeds 100 columns
#1228: FILE: drivers/gpu/cuda/cuda.c:926:
+							cuda_driver_version, CUDA_DRIVER_MIN_VERSION);

WARNING:LONG_LINE: line length of 105 exceeds 100 columns
#1391: FILE: drivers/gpu/cuda/cuda.c:1089:
+	rte_gpu_cuda_log_debug("dev id = %u name = %s
", dev->mpshared->info.dev_id, private->gpu_name);

WARNING:TYPO_SPELLING: 'tipically' may be misspelled - perhaps 'typically'?
#1447: FILE: drivers/gpu/cuda/cuda_loader.h:7:
+ * tipically found in /usr/local/cuda/include

WARNING:LONG_LINE_COMMENT: line length of 143 exceeds 100 columns
#1595: FILE: drivers/gpu/cuda/cuda_loader.h:155:
+	/* The returned attribute shall be interpreted as a bitmask, where the individual bits are described by the cuFlushGDRWriteOpts enum */

WARNING:LONG_LINE_COMMENT: line length of 206 exceeds 100 columns
#1597: FILE: drivers/gpu/cuda/cuda_loader.h:157:
+	/* GPUDirect RDMA writes to the device do not need to be flushed for consumers within the scope indicated by the returned attribute. See cuGDRWriteOrdering for the numerical values returned here. */

WARNING:LONG_LINE_COMMENT: line length of 108 exceeds 100 columns
#1621: FILE: drivers/gpu/cuda/cuda_loader.h:181:
+	/* 1 if this pointer maps to an allocation that is suitable for cudaIpcGetMemHandle, 0 otherwise **/

WARNING:LONG_LINE_COMMENT: line length of 112 exceeds 100 columns
#1627: FILE: drivers/gpu/cuda/cuda_loader.h:187:
+	/* 1 if this pointer is in a valid address range that is mapped to a backing allocation, 0 otherwise **/

WARNING:LONG_LINE_COMMENT: line length of 149 exceeds 100 columns
#1633: FILE: drivers/gpu/cuda/cuda_loader.h:193:
+	/* Returns the access flags the device associated with the current context has on the corresponding memory referenced by the pointer given */

WARNING:LONG_LINE_COMMENT: line length of 120 exceeds 100 columns
#1635: FILE: drivers/gpu/cuda/cuda_loader.h:195:
+	/* Returns the mempool handle for the allocation if it was allocated from a mempool. Otherwise returns NULL. **/

WARNING:LONG_LINE_COMMENT: line length of 121 exceeds 100 columns
#1642: FILE: drivers/gpu/cuda/cuda_loader.h:202:
+/* The CU_STREAM_WAIT_VALUE_FLUSH flag and the CU_STREAM_MEM_OP_FLUSH_REMOTE_WRITES MemOp are supported on the device. */

WARNING:LONG_LINE_COMMENT: line length of 129 exceeds 100 columns
#1646: FILE: drivers/gpu/cuda/cuda_loader.h:206:
+/* The device does not natively support ordering of remote writes. cuFlushGPUDirectRDMAWrites() can be leveraged if supported. */

WARNING:LONG_LINE_COMMENT: line length of 103 exceeds 100 columns
#1648: FILE: drivers/gpu/cuda/cuda_loader.h:208:
+/* Natively, the device can consistently consume remote writes, although other CUDA devices may not. */

WARNING:LONG_LINE: line length of 109 exceeds 100 columns
#1673: FILE: drivers/gpu/cuda/cuda_loader.h:233:
+extern enum cuError (*sym_cuGetProcAddress)(const char *symbol, void **pfn, int cudaVersion, uint64_t flags);

WARNING:LONG_LINE: line length of 118 exceeds 100 columns
#1712: FILE: drivers/gpu/cuda/cuda_loader.h:272:
+typedef enum cuError (*PFN_cuPointerSetAttribute_v6000)(const void *value, enum cuPtrAttr attribute, cuDevPtr_v2 ptr);

WARNING:LONG_LINE: line length of 101 exceeds 100 columns
#1713: FILE: drivers/gpu/cuda/cuda_loader.h:273:
+typedef enum cuError (*PFN_cuDeviceGetAttribute_v2000)(int *pi, enum cuDevAttr attrib, cuDev_v1 dev);

WARNING:LONG_LINE: line length of 135 exceeds 100 columns
#1727: FILE: drivers/gpu/cuda/cuda_loader.h:287:
+typedef enum cuError (*PFN_cuCtxGetExecAffinity_v11040)(struct cuExecAffinityParams *pExecAffinity, enum cuExecAffinityParamType type);

WARNING:LONG_LINE: line length of 109 exceeds 100 columns
#1734: FILE: drivers/gpu/cuda/cuda_loader.h:294:
+typedef enum cuError (*PFN_cuMemHostGetDevicePointer_v3020)(cuDevPtr_v2 *pdptr, void *p, unsigned int Flags);

WARNING:LONG_LINE: line length of 120 exceeds 100 columns
#1735: FILE: drivers/gpu/cuda/cuda_loader.h:295:
+typedef enum cuError (*PFN_cuFlushGPUDirectRDMAWrites_v11030)(enum cuFlushGDRTarget target, enum cuFlushGDRScope scope);

total: 0 errors, 42 warnings, 1590 lines checked