Change from v2:
Correct the logic when calculating icount.
Change from v1:
According to Jan and Stefano''s comments, drop the previous logic which
setting the value in an arbitrary way and use Stefano''s suggestion
instead.
Currently, the max expiration time is 2147483647ns(INT32_MAX ns). This is enough
when guest is busy, but when guest is idle, the next timer will be later than
INT32_MAX ns. And those meaningless alarm will harm the pkg C-state.
PS: Since the overflow will not happen with the expression((delta / 1000) +
(delta % 1000 > 0 ? 1 : 0)), so i also removed the comments" To avoid
problems with overflow limit this to 2^32."
Signed-off-by: Yang Zhang <yang.z.zhang@Intel.com>
---
vl.c | 15 ++++++++-------
1 files changed, 8 insertions(+), 7 deletions(-)
diff --git a/vl.c b/vl.c
index be8587a..d30cb2c 100644
--- a/vl.c
+++ b/vl.c
@@ -1410,8 +1410,7 @@ static int64_t qemu_next_deadline(void)
delta = active_timers[QEMU_TIMER_VIRTUAL]->expire_time -
qemu_get_clock(vm_clock);
} else {
- /* To avoid problems with overflow limit this to 2^32. */
- delta = INT32_MAX;
+ delta = INT64_MAX;
}
if (delta < 0)
@@ -1427,9 +1426,11 @@ static uint64_t qemu_next_deadline_dyntick(void)
int64_t rtdelta;
if (use_icount)
- delta = INT32_MAX;
- else
- delta = (qemu_next_deadline() + 999) / 1000;
+ delta = INT64_MAX;
+ else {
+ delta = qemu_next_deadline();
+ delta = (delta / 1000) + (delta % 1000 > 0 ? 1 : 0);
+ }
if (active_timers[QEMU_TIMER_REALTIME]) {
rtdelta = (active_timers[QEMU_TIMER_REALTIME]->expire_time -
@@ -3872,8 +3873,8 @@ int main_loop(void)
env->icount_decr.u16.low = 0;
env->icount_extra = 0;
count = qemu_next_deadline();
- count = (count + (1 << icount_time_shift) - 1)
- >> icount_time_shift;
+ count = (count >> icount_time_shift) +
+ ((count & ((1 << icount_time_shift) - 1))
> 0 ? 1 : 0);
qemu_icount += count;
decr = (count > 0xffff) ? 0xffff : count;
count -= decr;
--
1.7.1