未验证 提交 1e3b3647 编写于 作者: D dapan1121 提交者: GitHub

Merge pull request #8112 from taosdata/hotfix/td-10488-dev

[TD-10488]<hotfix>:fix error caused by double conversion when time is big enough
......@@ -380,10 +380,40 @@ int64_t convertTimePrecision(int64_t time, int32_t fromPrecision, int32_t toPrec
assert(toPrecision == TSDB_TIME_PRECISION_MILLI ||
toPrecision == TSDB_TIME_PRECISION_MICRO ||
toPrecision == TSDB_TIME_PRECISION_NANO);
static double factors[3][3] = { {1., 1000., 1000000.},
{1.0 / 1000, 1., 1000.},
{1.0 / 1000000, 1.0 / 1000, 1.} };
return (int64_t)((double)time * factors[fromPrecision][toPrecision]);
switch(fromPrecision) {
case TSDB_TIME_PRECISION_MILLI: {
switch (toPrecision) {
case TSDB_TIME_PRECISION_MILLI:
return time;
case TSDB_TIME_PRECISION_MICRO:
return time * 1000;
case TSDB_TIME_PRECISION_NANO:
return time * 1000000;
}
} // end from milli
case TSDB_TIME_PRECISION_MICRO: {
switch (toPrecision) {
case TSDB_TIME_PRECISION_MILLI:
return time / 1000;
case TSDB_TIME_PRECISION_MICRO:
return time;
case TSDB_TIME_PRECISION_NANO:
return time * 1000;
}
} //end from micro
case TSDB_TIME_PRECISION_NANO: {
switch (toPrecision) {
case TSDB_TIME_PRECISION_MILLI:
return time / 1000000;
case TSDB_TIME_PRECISION_MICRO:
return time / 1000;
case TSDB_TIME_PRECISION_NANO:
return time;
}
} //end from nano
default:
assert(0);
} //end switch fromPrecision
}
static int32_t getDuration(int64_t val, char unit, int64_t* result, int32_t timePrecision) {
......
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册