RFR - 7163471: Licensee source bundle failed around 7u4

Bengt Rutisson bengt.rutisson at oracle.com
Tue Jun 5 01:27:59 PDT 2012


Nils,

This looks good to me. One minor thing if you have time:

Could we move the JFR defintion up to between JFR_SRCDIRS_EXIST and 
UCRYPTO? That way we would get a more natural order in the file: 
JFR_SRCDIRS_EXIST, JFR, UCRYPTO, SUBDIRS.

Ship it!
Bengt

On 2012-06-04 13:11, Nils Loodin wrote:
> In the scenario which the bugs talk about, i.e. licensee builds.
> They don´t have, and don't want, a define in the makefiles.
> Their only mechanism is to remove certain files based on a  blacklist.
>
> This is currently causing a build failure from the licensee sources.
>
> /Nisse
>
>
> On Jun 4, 2012, at 12:01 , Erik Joelsson wrote:
>
>> In what scenario would the directory not exist when OPENJDK is not defined?
>>
>> /Erik
>>
>> On 2012-06-04 11:10, Nils Loodin wrote:
>>> http://cr.openjdk.java.net/~nloodin/71663471/webrev.00/<http://cr.openjdk.java.net/%7Enloodin/71663471/webrev.00/>
>>>
>>> Simple fix, check if the directory exist or not, and if not don't build that subdirectory.
>>>
>>> Modelled the mechanic after another makefile in the jdk.
>>>
>>>



More information about the hotspot-runtime-dev mailing list