DPCT1026

Intel® DPC++ Compatibility Tool Developer Guide and Reference

Download PDF

ID 768918

Date 7/13/2023

Version

Public

A newer version of this document is available. Customers should click here to go to the newest version.

Visible to Intel only — GUID: GUID-BCDCA2CF-33D4-4067-AE62-EF2F63DABD04

View Details

DPCT1026

Message

The call to <API name> was removed because <reason>.

Detailed Help

API calls from the original application, which do not have functionally compatible SYCL* API calls are removed if Intel® DPC++ Compatibility Tool determines that it should not affect the program logic.

Possible reasons for removal:

SYCL currently does not support setting resource limits on devices.
SYCL currently does not support associating USM with a specific queue.
SYCL currently does not support query operations on queues.
SYCL currently does not support capture operations on queues.
SYCL currently does not support configuring shared memory on devices.
SYCL currently does not support setting cache config on devices.
SYCL currently does not support registering of existing host memory for use by device. Use USM to allocate memory for use by host and device.
SYCL currently does not support setting flags for devices.
SYCL currently does not support memory access across peer devices.
There is no corresponding API in SYCL.
The call is redundant in SYCL.

Suggestions to Fix

Verify the code correctness.

For example, this original CUDA* code:

__device__ void bar(int *a) {
  __ldg(a);
}

void foo() {
  cudaLimit limit;
  cudaStream_t stream;
  unsigned int flags;
  cudaSharedMemConfig config;

  cudaDeviceSetLimit(limit, 0);
  cudaStreamAttachMemAsync(stream, nullptr);
  cudaStreamQuery(stream);
  cudaDeviceSetSharedMemConfig(config);
  cudaDeviceSetCacheConfig(cudaFuncCachePreferNone);
  cudaSetDeviceFlags(flags);
  cuInit(0);
}

results in the following migrated SYCL code:

void bar(int *a) {
  /*
  DPCT1026:0: The call to __ldg was removed because there is no corresponding
  API in SYCL.
  */
  *a;
}

void foo() {
  cudaLimit limit;
  dpct::queue_ptr stream;
  unsigned int flags;
  int config;

  /*
  DPCT1026:1: The call to cudaDeviceSetLimit was removed because SYCL currently
  does not support setting resource limits on devices.
  */
  /*
  DPCT1026:2: The call to cudaStreamAttachMemAsync was removed because SYCL
  currently does not support associating USM with a specific queue.
  */
  /*
  DPCT1026:3: The call to cudaStreamQuery was removed because SYCL currently
  does not support query operations on queues.
  */
  /*
  DPCT1026:4: The call to cudaDeviceSetSharedMemConfig was removed because SYCL
  currently does not support configuring shared memory on devices.
  */
  /*
  DPCT1026:5: The call to cudaDeviceSetCacheConfig was removed because SYCL
  currently does not support setting cache config on devices.
  */
  /*
  DPCT1026:6: The call to cudaSetDeviceFlags was removed because SYCL currently
  does not support setting flags for devices.
  */
  /*
  DPCT1026:7: The call to cuInit was removed because this call is redundant in
  SYCL.
  */
}

which is rewritten to:

void bar(int *a) {
  *a;
}

void foo() {
  cudaLimit limit;
  dpct::queue_ptr stream;
  unsigned int flags;
  size_t count;
  float *ptr;

  int config;
}

Select Your Language

Using Intel.com Search

Quick Links

Recent Searches

Advanced Search

Only search in

Intel® DPC++ Compatibility Tool Developer Guide and Reference

DPCT1026

Message

Detailed Help

Suggestions to Fix