RFR (XS) 8241545: Shenandoah: purge root work overwrites counters after JDK-8228818

Zhengyu Gu zgu at redhat.com
Tue Mar 24 19:07:27 UTC 2020


Looks good to me.

Thanks for fixing it.

-Zhengyu

On 3/24/20 3:03 PM, Aleksey Shipilev wrote:
> Bug:
>    https://bugs.openjdk.java.net/browse/JDK-8241545
> 
> Fix:
> 
> diff -r 97a3e6ce2652 src/hotspot/share/gc/shenandoah/shenandoahPhaseTimings.hpp
> --- a/src/hotspot/share/gc/shenandoah/shenandoahPhaseTimings.hpp        Tue Mar 24 18:46:48 2020 +0100
> +++ b/src/hotspot/share/gc/shenandoah/shenandoahPhaseTimings.hpp        Tue Mar 24 20:02:16 2020 +0100
> @@ -78,4 +78,5 @@
>     f(purge_class_unload,                             "    Unload Classes")              \
>     f(purge_par,                                      "    Parallel Cleanup")            \
> +  SHENANDOAH_GC_PAR_PHASE_DO(purge_par_roots,       "      PC: ", f)                   \
>     f(purge_cldg,                                     "    CLDG")                        \
>     f(complete_liveness,                              "  Complete Liveness")             \
> @@ -136,4 +137,5 @@
>     f(full_gc_purge_class_unload,                     "      Unload Classes")            \
>     f(full_gc_purge_par,                              "    Parallel Cleanup")            \
> +  SHENANDOAH_GC_PAR_PHASE_DO(full_gc_purge_roots,   "      PC: ", f)                   \
>     f(full_gc_purge_cldg,                             "    CLDG")                        \
>     f(full_gc_calculate_addresses,                    "  Calculate Addresses")           \
> 
> Testing: eyeballing gc+stats logs
> 



More information about the shenandoah-dev mailing list