[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Qemu-devel] [PATCH 10/18] qemu_img: is_not_zero() optimization
From: |
Kevin Wolf |
Subject: |
[Qemu-devel] [PATCH 10/18] qemu_img: is_not_zero() optimization |
Date: |
Thu, 19 May 2011 14:33:24 +0200 |
From: Dmitry Konishchev <address@hidden>
I run qemu-img under profiler and realized, that most of CPU time is
consumed by is_not_zero() function. I had made a couple of optimizations
on it and got the following output for `time qemu-img convert -O qcow2
volume.qcow2 snapshot.qcow2`:
Original qemu-img:
real 0m56.159s
user 0m34.670s
sys 0m12.079s
Patched qemu-img:
real 0m34.805s
user 0m18.445s
sys 0m12.552s
Signed-off-by: Dmitry Konishchev <address@hidden>
Signed-off-by: Kevin Wolf <address@hidden>
---
qemu-img.c | 29 ++++++++++++++++++++++++++---
1 files changed, 26 insertions(+), 3 deletions(-)
diff --git a/qemu-img.c b/qemu-img.c
index 1da5484..4f162d1 100644
--- a/qemu-img.c
+++ b/qemu-img.c
@@ -496,14 +496,37 @@ static int img_commit(int argc, char **argv)
return 0;
}
+/*
+ * Checks whether the sector is not a zero sector.
+ *
+ * Attention! The len must be a multiple of 4 * sizeof(long) due to
+ * restriction of optimizations in this function.
+ */
static int is_not_zero(const uint8_t *sector, int len)
{
+ /*
+ * Use long as the biggest available internal data type that fits into the
+ * CPU register and unroll the loop to smooth out the effect of memory
+ * latency.
+ */
+
int i;
- len >>= 2;
- for(i = 0;i < len; i++) {
- if (((uint32_t *)sector)[i] != 0)
+ long d0, d1, d2, d3;
+ const long * const data = (const long *) sector;
+
+ len /= sizeof(long);
+
+ for(i = 0; i < len; i += 4) {
+ d0 = data[i + 0];
+ d1 = data[i + 1];
+ d2 = data[i + 2];
+ d3 = data[i + 3];
+
+ if (d0 || d1 || d2 || d3) {
return 1;
+ }
}
+
return 0;
}
--
1.7.2.3
- Re: [Qemu-devel] [PATCH 05/18] ide: Turn debug messages into assertions, (continued)
[Qemu-devel] [PATCH 04/18] hw/xen_disk: Remove unused local variable, Kevin Wolf, 2011/05/19
[Qemu-devel] [PATCH 06/18] Add documentation for qemu_progress_{init, print}(), Kevin Wolf, 2011/05/19
[Qemu-devel] [PATCH 08/18] qemu-tool: Stub out qemu-timer functions, Kevin Wolf, 2011/05/19
[Qemu-devel] [PATCH 07/18] ahci: Fix crashes on duplicate BH registration, Kevin Wolf, 2011/05/19
[Qemu-devel] [PATCH 09/18] qed: Periodically flush and clear need check bit, Kevin Wolf, 2011/05/19
[Qemu-devel] [PATCH 11/18] qed: support for growing images, Kevin Wolf, 2011/05/19
[Qemu-devel] [PATCH 12/18] ide: Split qdev "ide-drive" into "ide-hd" and "ide-cd", Kevin Wolf, 2011/05/19
[Qemu-devel] [PATCH 10/18] qemu_img: is_not_zero() optimization,
Kevin Wolf <=
[Qemu-devel] [PATCH 14/18] defaults: ide-cd, ide-hd and scsi-cd devices suppress default CD-ROM, Kevin Wolf, 2011/05/19
[Qemu-devel] [PATCH 13/18] scsi: Split qdev "scsi-disk" into "scsi-hd" and "scsi-cd", Kevin Wolf, 2011/05/19
[Qemu-devel] [PATCH 15/18] block QMP: Deprecate query-block's "type", drop info block's "type=", Kevin Wolf, 2011/05/19
[Qemu-devel] [PATCH 16/18] blockdev: Store -drive option media in DriveInfo, Kevin Wolf, 2011/05/19
[Qemu-devel] [PATCH 17/18] block: Remove type hint, it's guest matter, doesn't belong here, Kevin Wolf, 2011/05/19
[Qemu-devel] [PATCH 18/18] ahci: Fix non-NCQ accesses for LBA > 16bits, Kevin Wolf, 2011/05/19
Re: [Qemu-devel] [PULL 00/18] Block patches, Anthony Liguori, 2011/05/19