[Diff since v3.3.0](v3.3.0...v3.3.1)

**Merged pull requests:**
- Update to v2.3.1 (#70) (@ludvigak)

**Closed issues:**
- Tests fail on MacOS aarch64 (#64)

[Diff since v3.2.0](v3.2.0...v3.3.0)

**Merged pull requests:**
- Fix Codecov action (#59) (@ludvigak)
- Fix docs to include CUDA (#60) (@ludvigak)
- Temporarily use safe FFTW planning (#66) (@jkrimmer)
- Update base library to v2.3.0 (#67) (@ludvigak)

**Closed issues:**
- Redundant argument checking for out-of-place nufftndm calls (#46)
- Segfault: Calling finufft and fftw in threaded loop (#62)

[Diff since v3.1.1](v3.1.1...v3.2.0)

**Merged pull requests:**
- Remove redundant checks whereever possible (#47) (@jkrimmer)
- Minor type stability and allocation improvements (#52) (@jipolanco)
- Add missing errors codes (#56) (@ludvigak)
- Fix error codes (again) (#57) (@ludvigak)
- cuFINUFFT interface (#58) (@ludvigak)

**Closed issues:**
- Tests crash process for Julia 1.8.5 (#48)
- Interface to GPU code cufinufft (#49)
- Possible sub-optimal flags in cmake  (#55)

[Diff since v3.1.0](v3.1.0...v3.1.1)

**Merged pull requests:**
- Update CI.yml (#53) (@ludvigak)
- Update base library to v2.2.0 (#54) (@ludvigak)

**Closed issues:**
- Having trouble understanding what nufft type 1 is doing (#50)
- nufft1d2 not working (#51)

[Diff since v3.0.2](v3.0.2...v3.1.0)

**Closed issues:**
- In-place versions nufftndm! require dtype argument for single-precision transforms (#42)
- Choosing tolerance too small throws an error (#43)

**Merged pull requests:**
- Fix error thrown in case of tolerance too low (#44) (@jkrimmer)
- Infer dtype for finufft_makeplan (#45) (@jkrimmer)

[Diff since v3.0.1](v3.0.1...v3.0.2)

**Merged pull requests:**
- Update base library to v2.1.0 (#41) (@ludvigak)

[Diff since v3.0.0](v3.0.0...v3.0.1)

**Closed issues:**
- FINUFFT.jl fails to run (yet) on Julia 1.7.1 / macOS arm64 (#36)
- Multiple plans simultaneously  (#39)

**Merged pull requests:**
- Keep references to input arrays in guru plan struct (#40) (@ludvigak)

[Diff since v2.1.0](v2.1.0...v3.0.0)

**Closed issues:**
- Thread-safety (#28)
- on recovering FFTW result? (#30)
- finufft_setpts should be finufft_setpts! (#34)
- finufft_destroy crashes Julia (#35)

**Merged pull requests:**
- Add guru interface and single/double precision calling interface (#33) (@lulibin)
- Release v3.0 and fix CI for newer Julia versions (#37) (@ludvigak)

[Diff since v2.0.0](v2.0.0...v2.1.0)

**Closed issues:**
- ComplexF32 data support (#22)

**Merged pull requests:**
- Implement interfaces to 32-bit floating point calls (#27) (@ludvigak)

[Diff since v1.0.0](v1.0.0...v2.0.0)

**Closed issues:**
- Provide binary dependency with JLL package (#19)
- Problem in macOS Catalina when Julia updated (#23)

**Merged pull requests:**
- Install the library using the corresponding JLL package (#21) (@giordano)
- Compatibility with finufft v2.0.2 stable (#24) (@jonas-kr)
- Switch CI to GitHub Actions (#25) (@ludvigak)