dbg feature: PrintMallocStatistics still wanted?

[Doerr, Martin]

Hi Coleen,

thanks for your feedback. I've created JDK-8227692: "Remove develop feature PrintMallocStatistics" and I'll send an RFR, soon.

Best regards,
Martin


> -----Original Message-----
> From: hotspot-dev <hotspot-dev-bounces at openjdk.java.net> On Behalf Of
> coleen.phillimore at oracle.com
> Sent: Montag, 15. Juli 2019 18:37
> To: hotspot-dev at openjdk.java.net
> Subject: Re: dbg feature: PrintMallocStatistics still wanted?
> 
> 
> I didn't realize this was still in the sources.  I think you should
> remove it.
> Coleen
> 
> On 7/15/19 9:06 AM, Doerr, Martin wrote:
> > Hi,
> >
> > I recently noticed that the implementation for PrintMallocStatistics slows
> down the VM in fastdbg builds even if the feature is not active:
> > https://bugs.openjdk.java.net/browse/JDK-8227597
> >
> > My current proposal just improves the performance impact:
> >
> http://cr.openjdk.java.net/~mdoerr/8227597_DBG_Inline_inc_bytes_allocat
> ed/webrev.01/
> >
> > But now, the question has come up, if PrintMallocStatistics is still needed
> since we have NMT. Note that PrintMallocStatistics is only available in dbg
> builds.
> > Does anybody still want to use it?
> > Would anybody vote for removing this feature?
> >
> > Best regards,
> > Martin
> >