JDK-8270824 : SPECjvm2008 compress regression on x86 when +PreserveFramePointer
Type:Bug
Component:hotspot
Sub-Component:compiler
Affected Version:18
Priority:P4
Status:Closed
Resolution:Not an Issue
OS:linux
CPU:x86
Submitted:2021-07-16
Updated:2021-07-22
Resolved:2021-07-22
The Version table provides details related to the release that this issue/RFE will be addressed.
Unresolved : Release in which this issue/RFE will be addressed. Resolved: Release in which this issue/RFE has been resolved. Fixed : Release in which this issue/RFE has been fixed. The release containing this fix may be available for download as an Early Access Release or a General Availability Release.
Sorry for the confusion. I wasn't asking about a +/-PreserveFramePointer comparison. I was asking about +PreserveFramePointer with INTPRESSURE set to 12 vs 13.
21-07-2021
I will make a clean new test with 18-b6 with +/-PreserveFramePointer and update you when it is done.
21-07-2021
I believe we are seeing a 5% regression. [~ecaspole], can you confirm, and help Joshua with the necessary configuration to reproduce?
16-07-2021
ILW = Performance regression in SPECjvm2008 compress on x86 with -XX:+PreserveFramePointer, only one platform and only one benchmark?, no workaround = MLH = P4
16-07-2021
Note from [~jzhu]:
This issue (JDK-8270824) is to track "compress" regression with -XX:+PreserveFramePointer after JDK-8268858. There is another issue (JDK-8270808) marked as "Enhancement" used to track performance degradation in Flink Nexmark benchmark with -XX:+PreserveFramePointer found before JDK-8268858.
16-07-2021
Just now I ran six rounds of "compress" on x64 machine with "-XX:+PreserveFramePointer".
Thread number of "compress" test is configured as 1 in my testing environment.
See result at http://cr.openjdk.java.net/~jzhu/8269598/Book2_x86.pdf
There is no regression in my testing.