Common

#include "max/c/common.h"

Functions

M_version()

const char *M_version()

Gets the MAX Engine version.

• Returns:

  A string containing the semantic version of the MAX Engine.

M_newStatus()

M_Status *M_newStatus()

Creates a new status object.

This is required as an argument for several functions, such as M_newRuntimeContext() and M_compileModel(). They will update the status object and you can check for errors with M_isError() and get the status message with M_getError(). For example:

M_Status *status = M_newStatus();
M_RuntimeConfig *runtimeConfig = M_newRuntimeConfig();
M_RuntimeContext *context = M_newRuntimeContext(runtimeConfig, status);
if (M_isError(status)) {
  logError(M_getError(status));
  return EXIT_FAILURE;
}

• Returns:

  A pointer to the new status object. You are responsible for the memory associated with the pointer returned. You can deallocate the memory by calling M_freeStatus().

M_getError()

const char *M_getError(const M_Status *status)

Gets an error message from the M_Status parameter.

You should call this only if M_isError() is true.

• Parameters:

  status – The status object for reporting errors and other messages.

• Returns:

  A pointer to a null-terminated string containing the error message.

M_isError()

int M_isError(const M_Status *status)

Checks if status holds an error value.

• Parameters:

  status – The status object for reporting errors and other messages.

• Returns:

  0 if there is no error, 1 otherwise.

M_freeStatus()

void M_freeStatus(M_Status *status)

Deallocates the memory for the status object. No-op if status is NULL.

• Parameters:

  status – The status object for reporting errors and other messages.

M_sizeOf()

size_t M_sizeOf(M_Dtype type)

Gets the size (in bytes) of a data type.

• Parameters:

  type – The data type.

• Returns:

  Size in bytes of the given data type. If the data type is M_UNKNOWN, then 0.

M_getDynamicDimensionValue()

int64_t M_getDynamicDimensionValue()

Gets the value representing dynamic dimension.

Use this to indicate a dynamic tensor shape when declaring TorchScript input specs with M_TorchInputSpec.

• Returns:

  Value representing dynamic dimension.

M_getDynamicRankValue()

int64_t M_getDynamicRankValue()

Gets the value representing dynamic rank.

• Returns:

  Value representing dynamic rank.