about 2-5 times faster than gnu libcs qsort()
And should be 100% binary identical across platforms.
I will bump the version once the API is certainly stable and
everyone is happy with the API.
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
* cus/stable:
ffplay: fix -vismv 1
ffplay: rename buffer source instance from "src" to "ffplay_buffer"
ffplay: use key=val syntax for the buffersrc args
Merged-by: Michael Niedermayer <michaelni@gmx.at>
Add spaces around operators, fix brace placement and whitespace to
match K&R style, vertically align code, remove redundant != 0 and
convert x == 0 into !x, drop useless braces.
Signed-off-by: Martin Storsjö <martin@martin.st>
While here:
- add missing .version and .category,
- make .class_name consistent across filters,
- align declarations.
Signed-off-by: Paul B Mahol <onemda@gmail.com>
Add atempo audio filter for adjusting audio tempo without affecting
pitch. This filter implements WSOLA algorithm with fast cross
correlation calculation in frequency domain.
Signed-off-by: Pavel Koshevoy <pavel@homestead.aragog.com>
Signed-off-by: Stefano Sabatini <stefasab@gmail.com>
Defining restrict results - for some compilers - in changing other
uses of the restrict keyword also, e.g. __declspec(restrict) gets
changed to __declspec(__restrict) on MSVC. This causes compilation
failures. Therefore, using a private namespace macro instead is
more reliable and robust.
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
* qatar/master:
flacdec: read attached pictures.
lavf: don't segfault when a NULL filename is passed to avformat_open_input()
Merged-by: Michael Niedermayer <michaelni@gmx.at>
This can easily happen when the caller is using a custom AVIOContext.
Behave as if the filename was an empty string in this case.
CC: libav-stable@libav.org
Basically to make code clearer and adherent to the
standard. RFC 1321, on page 2 states
Let the symbol "+" denote addition of words (i.e., modulo-2^32
addition). Let X <<< s denote the 32-bit value obtained by circularly
shifting (rotating) X left by s bit positions.
on page 3, section 3.3 states:
A four-word buffer (A,B,C,D) is used to compute the message digest.
Here each of A, B, C, D is a 32-bit register.
so the algorithm needs to work with integers that are exactly 32bits
in length. And indeed in struct AVMD5 the MD buffer is declared as
"uint32_t ABCD[4];", while in the function that performs the block
transformation the state variables were "unsigned int"s. On
architectures where sizeof(unsigned int) != sizeof(uint32_t) this
could be a problem, although I can't name such an architecture from
the top of my head.
On a side note, both the reference implementation in RFC 1321 and the
gnulib implementation (used by md5sum program on GNU systems) use
uint32_t in the transform function.
Signed-off-by: Michael Niedermayer <michaelni@gmx.at>
Remove duplicated out-of-loop init code, and do color matrix
initialization at the beginning of each y iteration.
Allow factorization and avoid an useless matrix update in the last
iteration.
