This is an automated email from the git hooks/post-receive script.
Git pushed a change to branch master
in repository ffmpeg.
from 3c1aad218f avcodec/av1dec: use sw_pix_fmt for film grain subsampling
new da6007179c swscale/filters: allocate padding past end of weights
new 8e96400e8d Revert "swscale/uops: pad the horizontal filter weights to
whole blocks"
new 35c1620e7c swscale/internal: add ff_refstruct_calloc() helper
new da82a784f6 swscale/filters: use ff_refstruct_calloc()
new cc0c782634 swscale/format: use ff_refstruct_calloc()
new 810459c0ba swscale/uops: use ff_refstruct_calloc()
new 85fe84e5e3 swscale/ops: allocate external kernel data as refstructs
new c744f68541 swscale/ops_chain: track references instead of arbitrary
free callbacks
new 8a336146a7 swscale/ops_chain: drop unused helpers
new 29fb586192 swscale/ops_chain: drop explicit alignment from SwsOpPriv
new 174f470813 swscale/ops_chain: re-order SwsOpImpl to store func with
its data
new 8abe273736 swscale/uops_backend: drop unnecessary pointer indirection
new d094f3b348 swscale/x86/ops_common: call directly into [implq]
new 682ea4829e swscale/x86/ops_common: simplify stack layout
new 62234b7832 swscale/x86/ops_common: let x86inc set up the stack
new 80e3425ba2 swscale/ops_chain: use abstract packed data buffer
new 65caf98001 swscale/ops_chain: directly pack data into SwsOpPriv
new 3992bab06b swscale/uops_tmpl: pack pointer private data
new a76a235c58 swscale/uops_tmpl: pack the filter weights and size
new ae7e2412b7 swscale/x86/ops: pack the filter private data
new 82487bf39f swscale/x86/ops: pack SCALE constant as dword
new 3c76ec031b swscale/x86/ops: pack the remaining private data
new bf7c698bb2 swscale/aarch64/ops: pack kernel private data
new f2a88ef228 swscale/uops_tmpl: parametrize CONTINUE() by size of priv
data
new 5c6d7ebb17 swscale/x86/ops: add PRIV_SIZE to kernel setup
new b265036a85 swscale/x86: pre-load CLEAR continuation
new 6a82b9e8b9 swscale/aarch64/ops: add alignof_impl and FFALIGN
new 314b034a34 swscale/aarch64/ops: add helper to determine the priv size
per op
new 1d33874a82 swscale/ops: only skip over consumed data in SwsOpImpl
new e488f9caec swscale/ops_chain: replace SwsImplResult.priv by abstract
buffer
new 02dbbf671f swscale/ops_chain: make SwsOpImpl.priv generic
new 9d1ef0a23a swscale/uops_tmpl: keep largest element of block_t first
new 3eed75fa1a swscale/uops_tmpl: store the linear coefficients in the op
chain
new a2224fda66 swscale/x86/ops: store the f32 linear coefficients in the
op chain
new 2bf291526e swscale/aarch64/ops: store the linear coefficients in the
op chain
The 35 revisions listed above as "new" are entirely new to this
repository and will be described in separate emails. The revisions
listed as "add" were already present in the repository and have only
been added to this reference.
Summary of changes:
libswscale/aarch64/ops.c | 66 +++++++++-------
libswscale/aarch64/ops_asmgen.h | 2 +-
libswscale/aarch64/ops_impl.h | 31 +++++++-
libswscale/aarch64/ops_static.c | 18 +++--
libswscale/filters.c | 7 +-
libswscale/filters.h | 1 +
libswscale/format.c | 3 +-
libswscale/ops_chain.c | 57 ++++++++------
libswscale/ops_chain.h | 75 +++++++++---------
libswscale/swscale_internal.h | 8 ++
libswscale/uops.c | 3 +-
libswscale/uops_backend.c | 27 ++++---
libswscale/uops_tmpl.c | 170 ++++++++++++++++++++--------------------
libswscale/uops_tmpl.h | 38 +++++++--
libswscale/x86/ops.c | 81 ++++++++++++-------
libswscale/x86/ops_common.asm | 32 +++-----
libswscale/x86/ops_float.asm | 40 ++++++----
libswscale/x86/ops_include.asm | 16 ++--
libswscale/x86/ops_int.asm | 11 ++-
19 files changed, 402 insertions(+), 284 deletions(-)
--
To stop receiving notification emails like this one, please contact
[email protected].
_______________________________________________
ffmpeg-cvslog mailing list -- [email protected]
To unsubscribe send an email to [email protected]
