x86: Split AVX2 and AVX-512 mc asm into separate files
Fix compilation with nasm 2.15
meson: Workaround missing aarch64 normalisation
CLI: Fix help text
x86: Add minor mc 8-tap optimizations
CI: Fix PIC test on 32-bit x86
x86: Fix 32-bit build with PIC enabled
x86inc: Improve warnings for use of unsupported instructions
x86inc: Add missing 'blsr' BMI1 instruction
x86inc: Add template defines for EVEX broadcasts
Accumulate leb128 value using uint64_t as intermediate type
Simplify checks for leb128() and leb() output overflow
Simplify checks for dav1d_get_uleb128 out overflow
Assert sz is nonzero in dav1d_init_get_bits()
Update NEWS for 0.7.1
Extract y related operations out of warp_affine inner loop
x86: Branch before waiting on popcnt in ipred_z AVX2 functions
arm32: Add a NEON implementation of MSAC
arm64: msac: Add a special cased implementation of decode_hi_tok
arm64: msac: Use a narrower vector length for adapt4 in one place
arm64: msac: Clarify the register use in one macro
cli: Avoid large intermediates in the windows get_time_nanos
cli: Get the elapsed time if printing progress, regardless of the fps value
Update NEWS for 0.7.1
x86: Add put/prep_bilin_scaled AVX2 asm
x86: Add prep_8tap_scaled AVX2 asm
Add clamping back to mv projection
arm64: itx: Simplify and clarify the sub_sp macro a little bit
arm: itx: Add NEON implementation of itx for 8 bpc
meson: Use dav1d_src_root everywhere for consistency
Remove redundant memset in itx DSP initialization
meson: Make docs generation subproject-safe
x86: Adapt SSSE3 prep_8tap to SSE2
x86: Adapt SSSE3 prep_bilin to SSE2
arm64: itx16: Add a missed eob check in the 16x8 transform
arm64: itx16: Remove a leftover unused macro parameter
x86: Fix compiler warnings when using nasm 2.15
Avoid compiling logging functions when logging is disabled
CI: Enable coverage reports
Range of operating point is 0 - 31, not 0 - 32
arm: Add an export parameter to the const macro
x86: Add put_8tap_scaled AVX2 asm
meson: favor _aligned_malloc over posix_memalign
x86: Add minor looprestoration asm optimizations
dav1dplay: use new pl_chroma_location API
dav1dplay: allow resizing the window
dav1dplay: don't freeze on render errors
Update NEWS for 0.7.0
dav1dplay: support on-GPU film grain synthesis
dav1dplay: handle all supported csps/reprs/bitdepths
dav1dplay: move and simplify pl_image generation
dav1dplay: don't write directly to iparams.extensions
Fix swapped define guards in dav1dplay’s libplacebo renderer
Update NEWS for 0.7.0
checkasm: x86: Check for stack corruption
tools: add missing fopen error handling
Dav1dPlay: Split placebo renderer into two
Dav1dPlay: Remove redundant log message
Dav1dPlay: Remove unused renderer_info member
Dav1dPlay: Allow runtime renderer selection
Dav1dPlay: Fix renderer selection
Dav1dPlay: Split renderers into different files
Dav1dPlay: Add support for OpenGL with libplacebo
Dav1dPlay: Split FIFO to different files
checkasm: arm: Offset the location of the stack canary reference
checkasm: arm32: Take the number of stack arguments into account when checking for stack clobbering
checkasm: arm64: Take the number of stack arguments into account when checking for stack clobbering
checkasm: Cosmetics
checkasm: Skip printing the seed when using --list-functions
checkasm: arm64: Avoid overwriting the v0/q0/d0/s0 register
checkasm: arm: Don't use blx to call checkasm_fail_func
CI: Add 32 bit instruction set test
CI: Optimise multi-threading tests
CI: Optimise instruction set tests
CI: Add multi-threading to conformance tests
CI: Run conformance tests with different instruction sets
checkasm: filmgrain: Fix benchmarking in 16 bpc mode
cli: Reduce fps fraction in ivf parsing
x86: Use 'test' instead of 'or' to compare with zero
x86: Unconditionally compile msac_init.c
x86-64: Do msac refill before calling dav1d_msac_init_x86()
msac: Avoid attempting to refill after eob has already been reached
arm64: itx: Add NEON implementation of itx for 10 bpc
arm: Mark global symbols hidden
arm64: itx: Prepare for other bitdepths
itx: Add a bpc parameter to the itx dsp init function
arm64: itx: Share code for the three horz_16x8 functions
arm64: itx: Fix the eob checking for dct_dct_64x16
arm64: itx: Simplify inv_txfm_horz_dct_32x8
arm64: itx: Minor optimizations for the 8x32 functions
arm64: itx: Cosmetic fix up
arm64: itx: Remove an unused constant
arm64: itx: Remove a todo comment about more special cased functions
arm64: itx: Remove a now unused macro
arm64: Explicitly forbid using the x18 register
checkasm: arm32: Check for stack overflows
checkasm: arm64: Check for stack overflows
x86: Eliminate redundant 3-operand register syntax in itx