RFR(s): 8218543: ThreadsList handling during error reporting can crash

Daniel D. Daugherty daniel.daugherty at oracle.com
Fri Oct 11 14:44:38 UTC 2019


On 10/11/19 7:08 AM, Robbin Ehn wrote:
> Hi all, please review.
>
> Issue:
> https://bugs.openjdk.java.net/browse/JDK-8218543
> Changeset:
> http://cr.openjdk.java.net/~rehn/8218543/v1/webrev/index.html

src/hotspot/share/utilities/vmError.hpp
     No comments (thanks for removing the friends).

src/hotspot/share/utilities/vmError.cpp
     No comments.

src/hotspot/share/runtime/threadSMR.cpp
     No comments.

src/hotspot/share/utilities/decoder.cpp
     No comments.

Thumbs up!

Dan


>
> We avoid the assert and let the error reporting use the ThreadsList in 
> an unsafe way. More discussion in issue.
>
> Note that there was an get_first_error_tid(), but for some reason 
> direct access to that field with friend was used. I change that to use 
> the get method and renamed the field following private _ naming 
> convention.
>
> Passes t1-2
>
> Thanks Robbin



More information about the hotspot-runtime-dev mailing list