RFR(xxs): 8211845: A new switch to control verbosity of hs-err files

David Holmes david.holmes at oracle.com
Thu Nov 8 22:57:57 UTC 2018


Sorry Thomas, I delayed looking at this thinking it was more than just a 
flag definition.

I see it's pushed now, but the text indenting is inconsistent with 
existing code.

David

On 7/11/2018 9:07 PM, Thomas Stüfe wrote:
> Hi all,
> 
> may I please have reviews for this patch:
> 
> Bug: https://bugs.openjdk.java.net/browse/JDK-8211845
> 
> which is just the implementation of:
> 
> CSR: https://bugs.openjdk.java.net/browse/JDK-8211846
> 
> cr: http://cr.openjdk.java.net/~stuefe/webrevs/8211845-extensive-error-reports-switch/webrev.00/webrev/
> 
> We wanted a switch which would allow us to add types of output to the
> hs-err files which may under standard conditions considered to be
> either too lengthy or too security sensitive. This switch would be off
> under normal conditions but could be switched on in trusted
> environments or when one does not care about the size of hs-err files.
> For further information, please see the CSR.
> 
> Please note that I refrained from adding "enable security sensitive
> output" or similar text to the switch description. The reason is that
> this would implicitly promise that by default, when this switch is off
> (like, now), the hs-err file does not contain security sensitive
> information. And that we just cannot promise. After all, we already
> print out register contents etc.
> 
> Thanks, Thomas
> 


More information about the hotspot-runtime-dev mailing list