quic: Use __builtin_clz if available

Different processors has specific instructions to count leading
zero bits. This includes: x86. x64, arm, ppc.
For portability reason the behaviour of __builtin_clz is not
defined if the value is zero so test for it.
Currently the function is not called with the value or 0.
This increase performance decoding of about 4-5% on a x64 machine
(code size decreases a little too, but about 0.1%).

Signed-off-by: Frediano Ziglio <fziglio@redhat.com>
Acked-by: Christophe Fergeau <cfergeau@redhat.com>
This commit is contained in:
Frediano Ziglio 2018-05-13 00:36:01 +01:00
parent b208389334
commit 1dcdefa8b3

View File

@ -280,6 +280,12 @@ static const BYTE lzeroes[256] = {
/* count leading zeroes */
static unsigned int cnt_l_zeroes(const unsigned int bits)
{
if (spice_extra_checks) {
spice_assert(bits != 0);
}
#if defined(__GNUC__) && __GNUC__ >= 4
return __builtin_clz(bits);
#else
if (bits & 0xff800000) {
return lzeroes[bits >> 24];
} else if (bits & 0xffff8000) {
@ -289,6 +295,7 @@ static unsigned int cnt_l_zeroes(const unsigned int bits)
} else {
return 24 + lzeroes[bits & 0x000000ff];
}
#endif
}
#define QUIC_FAMILY_8BPC