Details
-
Bug
-
Status: Closed (View Workflow)
-
Major
-
Resolution: Fixed
-
10.1(EOL)
-
None
Description
The point-select sysbench benchmark I ran recently on 10.2 shows rather high time spent in
high resolution timer syscalls (clock_gettime on Linux, and QueryPerformanceCounter on Windows)
The inspection reveals that clock() call introduced inside trx_start_low by this patch https://github.com/MariaDB/server/commit/74961760a4837d2deb33336329c28cf9ad9b4e9e is responsible for the most calls (also, in 10.2, clock() was erroneously done twice).
The call can be replaced by the value of THD::start_utime.
Attachments
Activity
| Field | Original Value | New Value |
|---|---|---|
| Attachment | 10.2.callgraph.G.txt.zip [ 43474 ] |
| Status | Open [ 1 ] | Confirmed [ 10101 ] |
| Fix Version/s | 10.1.23 [ 22503 ] | |
| Fix Version/s | 10.2 [ 14601 ] |
| Description |
The point-select sysbench benchmark I ran recently on 10.2 shows rather high time spent in
high resolution timer syscalls (clock_gettime on Linux, and QueryPerformanceCounter on Windows) - 2-3% of the CPU samples are spent inside this function. The inspection reveals that clock() call introduced inside trx_start_low by this patch https://github.com/MariaDB/server/commit/74961760a4837d2deb33336329c28cf9ad9b4e9e is responsible for the most calls (also, in 10.2, clock() was erroneously done twice). The call can be replaced by the value of THD::start_utime. |
The point-select sysbench benchmark I ran recently on 10.2 shows rather high time spent in
high resolution timer syscalls (clock_gettime on Linux, and QueryPerformanceCounter on Windows) The inspection reveals that clock() call introduced inside trx_start_low by this patch https://github.com/MariaDB/server/commit/74961760a4837d2deb33336329c28cf9ad9b4e9e is responsible for the most calls (also, in 10.2, clock() was erroneously done twice). The call can be replaced by the value of THD::start_utime. |
| Attachment | remove_clock.txt [ 43480 ] | |
| Attachment | with_clock.txt [ 43481 ] |
| Fix Version/s | 10.2 [ 14601 ] | |
| Resolution | Fixed [ 1 ] | |
| Status | Confirmed [ 10101 ] | Closed [ 6 ] |
| Workflow | MariaDB v3 [ 80096 ] | MariaDB v4 [ 151845 ] |
| Link | This issue causes MDEV-27486 [ MDEV-27486 ] |
| Link | This issue causes MDEV-27486 [ MDEV-27486 ] |