RFR: JDK-8079556: BACKOUT - Determining the desired PLAB size adjusts to the the number of threads at the wrong place

Thomas Schatzl thomas.schatzl at oracle.com
Thu May 7 12:28:12 UTC 2015


Hi,

On Thu, 2015-05-07 at 10:34 +0200, Jesper Wilhelmsson wrote:
> Hi,
> 
> Please review this backout of JDK-8073204 - Determining the desired PLAB size 
> adjusts to the the number of threads at the wrong place
> 
> After the change the size returned by PLABStats::desired_plab_sz() can get below 
> the minimum PLAB size due to truncating integer division in that function, and 
> actually missing alignment. This results in several places getting the wrong 
> results because of this broken invariant.
> 
> The fix will take more than a day and this change is now blocking a push of 
> hs-gc to main.
> 
> Thomas prepared this backout, I have reviewed it (looks good), so we need 
> another Reviewer to push it.
> 
> Backout bug: https://bugs.openjdk.java.net/browse/JDK-8079556
> Webrev: http://cr.openjdk.java.net/~jwilhelm/8073204/webrev_backout.00/
> Original change: http://hg.openjdk.java.net/jdk9/hs-gc/hotspot/rev/d57d15fb074b
> 
> The backout applies cleanly to latest hs-gc.
> 

  thanks everyone for handling this so quickly.

Thomas




More information about the hotspot-gc-dev mailing list