possible pkcs11 headers problem

Andreas Sterbenz Andreas.Sterbenz at Sun.COM
Wed Oct 17 17:57:32 UTC 2007


Kelly,

there seems to be something wrong here. j2se/make/sun/security/pkcs11 now runs

# Running javah:
/java/re/jdk/1.7.0/promoted/latest/binaries/solaris-i586/bin/javah 
-bootclasspath ../../../../build/solaris-i586/classes -d 
../../../../build/solaris-i586/tmp/sun/sun.security.pkcs11/j2pkcs11/CClassHeaders/ 
\
         sun.security.pkcs11.wrapper.PKCS11 sun.security.pkcs11.Secmod

But that is wrong since it sets CLASSDESTDIR. It used to be (in JDK 6 and 
I believe earlier JDK 7 builds)

# Running javah:
../../../../build/solaris-i586/bin/javah -J-XX:ThreadStackSize=768 
-J-client -J-Xmx512m -J-Xms256m  -bootclasspath 
../../../../build/solaris-i586/tmp/sun/sun.security.pkcs11/j2pkcs11/classes \
         -d 
../../../../build/solaris-i586/tmp/sun/sun.security.pkcs11/j2pkcs11/CClassHeaders/ 
\
         sun.security.pkcs11.wrapper.PKCS11 sun.security.pkcs11.Secmod

Note the bootclasspath.

There was an email exchange with Jon back on 6/25 where he suggested that 
the bootclasspath for javah should be set to 
$(CLASSDESTDIR):$(CLASSBINDIR), but make/common/shared/Defs-java.gmk just says
JAVAHFLAGS += -bootclasspath $(CLASSBINDIR)

Should this be changed in Defs-java.gmk?

Andreas.

Christian Thalinger wrote:
> Hi!
> 
> While hunting down a bug in CACAO, which prevented b22 from building, I
> noticed that the build system generates the headers for
> 
> sun.security.pkcs11.wrapper.PKCS11
> sun.security.pkcs11.Secmod
> 
> using the classes taken from sunpkcs11.jar from the bootstrapping JDK
> and not the newly built classes (somewhere in a tmp/ directory).
> 
> Is this intended?
> 
> - twisti




More information about the build-dev mailing list