RFR(XXXS): 7092245: G1: Wrong format specifier in G1PrintRegionLivenessInfo header output

Tony Printezis tony.printezis at oracle.com
Wed Sep 21 13:48:18 PDT 2011


Ramki,

Ramki Ramakrishna wrote:
>
> OK, in context, this seems fine. Below, a related comment ...
>
> As regards, the sizes in bytes, and the efficiency, would it be a good 
> idea to provide the units?
> (Is efficiency in bytes/ms?). Perhaps they are for internal 
> consumption and left without units
> explicitly stated (with the understanding that if you don't know the 
> units you should not be
> reading this log), but if this is for general user consumption, a 
> header line somewhere with
> the unit would not hurt :-)

Well, depends what you mean by internal or general use consumption. Of 
course, the parameter is available for everyone to use. But, 
realistically, who else is going to use it?

Anyway, IIRC the reason I had left out the units when I first introduced 
the G1PrintRegionLivenessInfo was so that the columns do not get too 
large. Would it work if we add a new header line with the units, like this?

### PHASE Post-Sorting @ 5.303
### HEAP  committed: 0xf2400000-0xf6400000  reserved: 
0xf2400000-0xf6400000  region-size: 1048576
###
###   type         address-range       used  prev-live  
next-live          gc-eff
###                                 (bytes)    (bytes)    (bytes)      
(bytes/ms)
###   OLD  0xf4d00000-0xf4e00000    1048576        680          0       
1036338.1
###   OLD  0xf4800000-0xf4900000    1048576        824          0       
1033205.5

Tony
> But this of course does not have to be done now.


More information about the hotspot-gc-dev mailing list