Crate baracuda_cusolver_sys 

Raw FFI + dynamic loader for NVIDIA cuSOLVER (Dense + Sparse + Refactor).

baracuda-cusolver wraps this with a safe, typed API. Use this crate directly only if you need a function that the safe layer hasn’t wrapped yet (in which case please file a bug).

Structs

Cusolver

Loaded cuSOLVER shared library plus a per-symbol OnceLock of function pointers.

CusolverMg

Loaded cuSOLVERMg shared library plus a per-symbol OnceLock of function pointers.

cuComplex

Single-precision complex number, ABI-compatible with cuBLAS / cuSOLVER cuComplex.

cuDoubleComplex

Double-precision complex number, ABI-compatible with cuBLAS / cuSOLVER cuDoubleComplex.

cusolverStatus_t

Status / error code returned by cuSOLVER FFI calls.

Enums

cublasDiagType_t

Diagonal-unit selector (matches the cuBLAS enum).

cublasFillMode_t

Triangular fill mode (matches the cuBLAS enum).

cublasOperation_t

Transpose selector — same values as cublasOperation_t (N=0, T=1, C=2).

cublasSideMode_t

Side selector for one-sided operators (apply from left or right).

cudaDataType

Element-dtype selector used by the generic-API (cusolverDnX*) routines.

cusolverEigMode_t

Eigenvalue-only vs eigenvalue-and-vector selector for eigensolvers.

cusolverEigRange_t

Subset-of-spectrum selector for partial eigensolvers.

cusolverEigType_t

Generalized eigenproblem variant for sygv* / hegv* routines.

Functions

cusolver

Lazy-load the cuSOLVER dense / sparse / refactor shared library and return its function-pointer table.

cusolver_mg

Lazy-load the cuSOLVERMg multi-GPU shared library and return its function-pointer table.

Type Aliases

PFN_cusolverDnCCgels

cuSOLVER: single-precision complex least-squares solver (A*X = B). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnCCgels_bufferSize

cuSOLVER: single-precision complex workspace-size query for least-squares solver (A*X = B). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnCgeqrf

cuSOLVER: single-precision complex QR factorization (Householder). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnCgeqrf_bufferSize

cuSOLVER: single-precision complex workspace-size query for QR factorization (Householder). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnCgesvd

cuSOLVER: single-precision complex singular value decomposition. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnCgesvd_bufferSize

cuSOLVER: single-precision complex workspace-size query for singular value decomposition. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnCgesvdj

cuSOLVER: single-precision complex Jacobi-method singular value decomposition. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnCgesvdjBatched

cuSOLVER: single-precision complex batched Jacobi-method singular value decomposition. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnCgesvdjBatched_bufferSize

cuSOLVER: single-precision complex workspace-size query for batched Jacobi-method singular value decomposition. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnCgesvdj_bufferSize

cuSOLVER: single-precision complex workspace-size query for Jacobi-method singular value decomposition. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnCgetrf

cuSOLVER: single-precision complex LU factorization with partial pivoting. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnCgetrf_bufferSize

cuSOLVER: single-precision complex workspace-size query for LU factorization with partial pivoting. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnCgetrs

cuSOLVER: single-precision complex solve linear system using LU factors. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnCheevd

cuSOLVER: single-precision complex Hermitian eigendecomposition (divide-and-conquer). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnCheevd_bufferSize

cuSOLVER: single-precision complex workspace-size query for Hermitian eigendecomposition (divide-and-conquer). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnCheevj

cuSOLVER: single-precision complex Hermitian eigendecomposition (Jacobi). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnCheevjBatched

cuSOLVER: single-precision complex batched Hermitian eigendecomposition (Jacobi). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnCheevjBatched_bufferSize

cuSOLVER: single-precision complex workspace-size query for batched Hermitian eigendecomposition (Jacobi). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnCheevj_bufferSize

cuSOLVER: single-precision complex workspace-size query for Hermitian eigendecomposition (Jacobi). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnCpotrf

cuSOLVER: single-precision complex Cholesky factorization. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnCpotrf_bufferSize

cuSOLVER: single-precision complex workspace-size query for Cholesky factorization. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnCpotri

cuSOLVER: single-precision complex matrix inverse from Cholesky factors. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnCpotri_bufferSize

cuSOLVER: single-precision complex workspace-size query for matrix inverse from Cholesky factors. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnCpotrs

cuSOLVER: single-precision complex solve linear system using Cholesky factors. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnCreate

cuSOLVER: create a dense cuSOLVER handle. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnCreateGesvdjInfo

cuSOLVER: create a Jacobi-SVD control object. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnCreateParams

cuSOLVER: create a generic-API parameter object. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnCreateSyevjInfo

cuSOLVER: create a Jacobi-eigendecomposition control object. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnCungqr

cuSOLVER: single-precision complex generate the explicit unitary Q from a QR factorization. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnCungqr_bufferSize

cuSOLVER: single-precision complex workspace-size query for generate the explicit unitary Q from a QR factorization. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnCunmqr

cuSOLVER: single-precision complex apply Q (or Q^H) from a QR factorization to a matrix. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnCunmqr_bufferSize

cuSOLVER: single-precision complex workspace-size query for apply Q (or Q^H) from a QR factorization to a matrix. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnDDgels

cuSOLVER: double-precision least-squares solver (A*X = B). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnDDgels_bufferSize

cuSOLVER: double-precision workspace-size query for least-squares solver (A*X = B). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnDestroy

cuSOLVER: destroy a dense cuSOLVER handle. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnDestroyGesvdjInfo

cuSOLVER: destroy a Jacobi-SVD control object. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnDestroyParams

cuSOLVER: destroy a generic-API parameter object. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnDestroySyevjInfo

cuSOLVER: destroy a Jacobi-eigendecomposition control object. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnDgeqrf

cuSOLVER: double-precision QR factorization (Householder). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnDgeqrf_bufferSize

cuSOLVER: double-precision workspace-size query for QR factorization (Householder). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnDgesvd

cuSOLVER: double-precision singular value decomposition. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnDgesvd_bufferSize

cuSOLVER: double-precision workspace-size query for singular value decomposition. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnDgesvdj

cuSOLVER: double-precision Jacobi-method singular value decomposition. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnDgesvdjBatched

cuSOLVER: double-precision batched Jacobi-method singular value decomposition. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnDgesvdjBatched_bufferSize

cuSOLVER: double-precision workspace-size query for batched Jacobi-method singular value decomposition. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnDgesvdj_bufferSize

cuSOLVER: double-precision workspace-size query for Jacobi-method singular value decomposition. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnDgetrf

cuSOLVER: double-precision LU factorization with partial pivoting. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnDgetrf_bufferSize

cuSOLVER: double-precision workspace-size query for LU factorization with partial pivoting. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnDgetrs

cuSOLVER: double-precision solve linear system using LU factors. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnDorgqr

cuSOLVER: double-precision generate the explicit Q from a QR factorization. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnDorgqr_bufferSize

cuSOLVER: double-precision workspace-size query for generate the explicit Q from a QR factorization. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnDormqr

cuSOLVER: double-precision apply Q (or Q^T) from a QR factorization to a matrix. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnDormqr_bufferSize

cuSOLVER: double-precision workspace-size query for apply Q (or Q^T) from a QR factorization to a matrix. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnDpotrf

cuSOLVER: double-precision Cholesky factorization. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnDpotrf_bufferSize

cuSOLVER: double-precision workspace-size query for Cholesky factorization. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnDpotri

cuSOLVER: double-precision matrix inverse from Cholesky factors. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnDpotri_bufferSize

cuSOLVER: double-precision workspace-size query for matrix inverse from Cholesky factors. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnDpotrs

cuSOLVER: double-precision solve linear system using Cholesky factors. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnDsyevd

cuSOLVER: double-precision symmetric eigendecomposition (divide-and-conquer). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnDsyevd_bufferSize

cuSOLVER: double-precision workspace-size query for symmetric eigendecomposition (divide-and-conquer). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnDsyevj

cuSOLVER: double-precision symmetric eigendecomposition (Jacobi). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnDsyevjBatched

cuSOLVER: double-precision batched symmetric eigendecomposition (Jacobi). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnDsyevjBatched_bufferSize

cuSOLVER: double-precision workspace-size query for batched symmetric eigendecomposition (Jacobi). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnDsyevj_bufferSize

cuSOLVER: double-precision workspace-size query for symmetric eigendecomposition (Jacobi). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnGetStream

cuSOLVER: query the CUDA stream bound to a dense cuSOLVER handle. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnSSgels

cuSOLVER: single-precision least-squares solver (A*X = B). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnSSgels_bufferSize

cuSOLVER: single-precision workspace-size query for least-squares solver (A*X = B). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnSetStream

cuSOLVER: bind a CUDA stream to a dense cuSOLVER handle. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnSgeqrf

cuSOLVER: single-precision QR factorization (Householder). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnSgeqrf_bufferSize

cuSOLVER: single-precision workspace-size query for QR factorization (Householder). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnSgesvd

cuSOLVER: single-precision singular value decomposition. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnSgesvd_bufferSize

cuSOLVER: single-precision workspace-size query for singular value decomposition. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnSgesvdj

cuSOLVER: single-precision Jacobi-method singular value decomposition. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnSgesvdjBatched

cuSOLVER: single-precision batched Jacobi-method singular value decomposition. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnSgesvdjBatched_bufferSize

cuSOLVER: single-precision workspace-size query for batched Jacobi-method singular value decomposition. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnSgesvdj_bufferSize

cuSOLVER: single-precision workspace-size query for Jacobi-method singular value decomposition. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnSgetrf

cuSOLVER: single-precision LU factorization with partial pivoting. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnSgetrf_bufferSize

cuSOLVER: single-precision workspace-size query for LU factorization with partial pivoting. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnSgetrs

cuSOLVER: single-precision solve linear system using LU factors. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnSorgqr

cuSOLVER: single-precision generate the explicit Q from a QR factorization. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnSorgqr_bufferSize

cuSOLVER: single-precision workspace-size query for generate the explicit Q from a QR factorization. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnSormqr

cuSOLVER: single-precision apply Q (or Q^T) from a QR factorization to a matrix. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnSormqr_bufferSize

cuSOLVER: single-precision workspace-size query for apply Q (or Q^T) from a QR factorization to a matrix. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnSpotrf

cuSOLVER: single-precision Cholesky factorization. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnSpotrf_bufferSize

cuSOLVER: single-precision workspace-size query for Cholesky factorization. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnSpotri

cuSOLVER: single-precision matrix inverse from Cholesky factors. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnSpotri_bufferSize

cuSOLVER: single-precision workspace-size query for matrix inverse from Cholesky factors. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnSpotrs

cuSOLVER: single-precision solve linear system using Cholesky factors. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnSsyevd

cuSOLVER: single-precision symmetric eigendecomposition (divide-and-conquer). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnSsyevd_bufferSize

cuSOLVER: single-precision workspace-size query for symmetric eigendecomposition (divide-and-conquer). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnSsyevj

cuSOLVER: single-precision symmetric eigendecomposition (Jacobi). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnSsyevjBatched

cuSOLVER: single-precision batched symmetric eigendecomposition (Jacobi). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnSsyevjBatched_bufferSize

cuSOLVER: single-precision workspace-size query for batched symmetric eigendecomposition (Jacobi). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnSsyevj_bufferSize

cuSOLVER: single-precision workspace-size query for symmetric eigendecomposition (Jacobi). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnXgeqrf

cuSOLVER: generic-API QR factorization (Householder) (dtype configurable via cudaDataType). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnXgeqrf_bufferSize

cuSOLVER: generic-API workspace-size query for QR factorization (Householder) (dtype configurable via cudaDataType). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnXgetrf

cuSOLVER: generic-API LU factorization with partial pivoting (dtype configurable via cudaDataType). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnXgetrf_bufferSize

cuSOLVER: generic-API workspace-size query for LU factorization with partial pivoting (dtype configurable via cudaDataType). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnXgetrs

cuSOLVER: generic-API solve linear system using LU factors (dtype configurable via cudaDataType). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnXpotrf

cuSOLVER: generic-API Cholesky factorization (dtype configurable via cudaDataType). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnXpotrf_bufferSize

cuSOLVER: generic-API workspace-size query for Cholesky factorization (dtype configurable via cudaDataType). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnXpotrs

cuSOLVER: generic-API solve linear system using Cholesky factors (dtype configurable via cudaDataType). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnXsyevd

cuSOLVER: generic-API symmetric eigendecomposition (divide-and-conquer) (dtype configurable via cudaDataType). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnXsyevd_bufferSize

cuSOLVER: generic-API workspace-size query for symmetric eigendecomposition (divide-and-conquer) (dtype configurable via cudaDataType). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnXsyevjSetMaxSweeps

cuSOLVER: set the maximum number of sweeps on a Jacobi-eigendecomposition control object. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnXsyevjSetTolerance

cuSOLVER: set convergence tolerance on a Jacobi-eigendecomposition control object. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnZZgels

cuSOLVER: double-precision complex least-squares solver (A*X = B). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnZZgels_bufferSize

cuSOLVER: double-precision complex workspace-size query for least-squares solver (A*X = B). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnZgeqrf

cuSOLVER: double-precision complex QR factorization (Householder). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnZgeqrf_bufferSize

cuSOLVER: double-precision complex workspace-size query for QR factorization (Householder). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnZgesvd

cuSOLVER: double-precision complex singular value decomposition. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnZgesvd_bufferSize

cuSOLVER: double-precision complex workspace-size query for singular value decomposition. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnZgesvdj

cuSOLVER: double-precision complex Jacobi-method singular value decomposition. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnZgesvdjBatched

cuSOLVER: double-precision complex batched Jacobi-method singular value decomposition. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnZgesvdjBatched_bufferSize

cuSOLVER: double-precision complex workspace-size query for batched Jacobi-method singular value decomposition. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnZgesvdj_bufferSize

cuSOLVER: double-precision complex workspace-size query for Jacobi-method singular value decomposition. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnZgetrf

cuSOLVER: double-precision complex LU factorization with partial pivoting. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnZgetrf_bufferSize

cuSOLVER: double-precision complex workspace-size query for LU factorization with partial pivoting. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnZgetrs

cuSOLVER: double-precision complex solve linear system using LU factors. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnZheevd

cuSOLVER: double-precision complex Hermitian eigendecomposition (divide-and-conquer). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnZheevd_bufferSize

cuSOLVER: double-precision complex workspace-size query for Hermitian eigendecomposition (divide-and-conquer). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnZheevj

cuSOLVER: double-precision complex Hermitian eigendecomposition (Jacobi). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnZheevjBatched

cuSOLVER: double-precision complex batched Hermitian eigendecomposition (Jacobi). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnZheevjBatched_bufferSize

cuSOLVER: double-precision complex workspace-size query for batched Hermitian eigendecomposition (Jacobi). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnZheevj_bufferSize

cuSOLVER: double-precision complex workspace-size query for Hermitian eigendecomposition (Jacobi). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnZpotrf

cuSOLVER: double-precision complex Cholesky factorization. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnZpotrf_bufferSize

cuSOLVER: double-precision complex workspace-size query for Cholesky factorization. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnZpotri

cuSOLVER: double-precision complex matrix inverse from Cholesky factors. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnZpotri_bufferSize

cuSOLVER: double-precision complex workspace-size query for matrix inverse from Cholesky factors. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnZpotrs

cuSOLVER: double-precision complex solve linear system using Cholesky factors. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnZungqr

cuSOLVER: double-precision complex generate the explicit unitary Q from a QR factorization. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnZungqr_bufferSize

cuSOLVER: double-precision complex workspace-size query for generate the explicit unitary Q from a QR factorization. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnZunmqr

cuSOLVER: double-precision complex apply Q (or Q^H) from a QR factorization to a matrix. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverDnZunmqr_bufferSize

cuSOLVER: double-precision complex workspace-size query for apply Q (or Q^H) from a QR factorization to a matrix. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverGetVersion

cuSOLVER: return the cuSOLVER library version. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverMgCreate

cuSOLVER: create a multi-GPU dense cuSOLVERMg handle. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverMgCreateDeviceGrid

cuSOLVER: create a multi-GPU device grid. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverMgCreateMatrixDesc

cuSOLVER: create a multi-GPU matrix descriptor. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverMgDestroy

cuSOLVER: destroy a multi-GPU dense cuSOLVERMg handle. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverMgDestroyGrid

cuSOLVER: destroy a multi-GPU device grid. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverMgDestroyMatrixDesc

cuSOLVER: destroy a multi-GPU matrix descriptor. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverMgDeviceSelect

cuSOLVER: select GPU devices for a cuSOLVERMg handle. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverMgGetrf

cuSOLVER: multi-GPU LU factorization with partial pivoting. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverMgGetrf_bufferSize

cuSOLVER: multi-GPU workspace-size query for LU factorization with partial pivoting. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverMgPotrf

cuSOLVER: multi-GPU Cholesky factorization. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverMgPotrf_bufferSize

cuSOLVER: multi-GPU workspace-size query for Cholesky factorization. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverMgSyevd

cuSOLVER: multi-GPU symmetric eigendecomposition (divide-and-conquer). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverMgSyevd_bufferSize

cuSOLVER: multi-GPU workspace-size query for symmetric eigendecomposition (divide-and-conquer). See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverRfAnalyze

cuSOLVER: analyze the sparsity pattern of the refactor handle. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverRfCreate

cuSOLVER: create a sparse-refactor cuSOLVER handle. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverRfDestroy

cuSOLVER: destroy a sparse-refactor cuSOLVER handle. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverRfRefactor

cuSOLVER: refactor with new numerical values reusing the analyzed pattern. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverRfSetupDevice

cuSOLVER: supply sparse triangular factors and pivot vectors to the refactor handle. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverRfSolve

cuSOLVER: solve linear systems using the refactor handle. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverSpCreate

cuSOLVER: create a sparse cuSOLVER handle. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverSpDcsrlsvchol

cuSOLVER: double-precision sparse linear solve via Cholesky on a CSR matrix. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverSpDcsrlsvqr

cuSOLVER: double-precision sparse linear solve via QR on a CSR matrix. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverSpDestroy

cuSOLVER: destroy a sparse cuSOLVER handle. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverSpScsrlsvchol

cuSOLVER: single-precision sparse linear solve via Cholesky on a CSR matrix. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverSpScsrlsvqr

cuSOLVER: single-precision sparse linear solve via QR on a CSR matrix. See https://docs.nvidia.com/cuda/cusolver/index.html.

PFN_cusolverSpSetStream

cuSOLVER: bind a CUDA stream to a sparse cuSOLVER handle. See https://docs.nvidia.com/cuda/cusolver/index.html.

cudaLibMgGrid_t

Opaque multi-GPU device grid.

cudaLibMgMatrixDesc_t

Opaque multi-GPU matrix descriptor.

cusolverDnHandle_t

Opaque dense cuSOLVER handle.

cusolverDnIRSInfos_t

Opaque info object for iterative-refinement solvers.

cusolverDnIRSParams_t

Opaque parameter object for iterative-refinement solvers.

cusolverDnParams_t

Opaque generic-API parameter object for cusolverDnX* routines.

cusolverMgHandle_t

Opaque multi-GPU dense cuSOLVERMg handle.

cusolverRfHandle_t

Opaque sparse-refactor cuSOLVER handle.

cusolverSpHandle_t

Opaque sparse cuSOLVER handle.

gesvdjInfo_t

Opaque control object for Jacobi-based SVD.

syevjInfo_t

Opaque control object for Jacobi-based symmetric eigendecomposition.