From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([208.118.235.92]:58204) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1UGXDr-0001XI-GD for qemu-devel@nongnu.org; Fri, 15 Mar 2013 12:11:15 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1UGXDn-0006JE-7S for qemu-devel@nongnu.org; Fri, 15 Mar 2013 12:11:11 -0400 Received: from [2a02:248:0:30:223:aeff:fefe:7f1c] (port=48340 helo=dns.kamp-intra.net) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1UGXDn-0006HY-0J for qemu-devel@nongnu.org; Fri, 15 Mar 2013 12:11:07 -0400 From: Peter Lieven Date: Fri, 15 Mar 2013 16:50:13 +0100 Message-Id: <1363362619-3190-5-git-send-email-pl@kamp.de> In-Reply-To: <1363362619-3190-1-git-send-email-pl@kamp.de> References: <1363362619-3190-1-git-send-email-pl@kamp.de> Subject: [Qemu-devel] [PATCHv2 4/9] bitops: use vector algorithm to optimize find_next_bit() List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: qemu-devel@nongnu.org Cc: Peter Lieven this patch adds the usage of buffer_find_nonzero_offset() to skip large areas of zeroes. compared to loop unrolling presented in an earlier patch this adds another 50% performance benefit for skipping large areas of zeroes. loop unrolling alone added close to 100% speedup. Signed-off-by: Peter Lieven --- util/bitops.c | 26 +++++++++++++++++++++++--- 1 file changed, 23 insertions(+), 3 deletions(-) diff --git a/util/bitops.c b/util/bitops.c index e72237a..3c301fa 100644 --- a/util/bitops.c +++ b/util/bitops.c @@ -42,10 +42,30 @@ unsigned long find_next_bit(const unsigned long *addr, unsigned long size, size -= BITS_PER_LONG; result += BITS_PER_LONG; } - while (size & ~(BITS_PER_LONG-1)) { - if ((tmp = *(p++))) { - goto found_middle; + while (size >= BITS_PER_LONG) { + if ((tmp = *p)) { + goto found_middle; + } + if (((uintptr_t) p) % sizeof(VECTYPE) == 0 + && size >= BITS_PER_BYTE * sizeof(VECTYPE) + * BUFFER_FIND_NONZERO_OFFSET_UNROLL_FACTOR) { + unsigned long tmp2 = + buffer_find_nonzero_offset(p, ((size / BITS_PER_BYTE) & + ~(BUFFER_FIND_NONZERO_OFFSET_UNROLL_FACTOR * + sizeof(VECTYPE) - 1))); + result += tmp2 * BITS_PER_BYTE; + size -= tmp2 * BITS_PER_BYTE; + p += tmp2 / sizeof(unsigned long); + if (!size) { + return result; + } + if (tmp2) { + if ((tmp = *p)) { + goto found_middle; + } + } } + p++; result += BITS_PER_LONG; size -= BITS_PER_LONG; } -- 1.7.9.5