Re: webrev.01 of 8011402: Move blacklisting certificate logic from hard code to data

2013-09-18 Thread Sean Mullan
On 09/17/2013 07:07 AM, Weijun Wang wrote: Webrev updated to version 02 at http://cr.openjdk.java.net/~weijun/8011402/webrev.02/ Changes since webrev.01: 1. Makefiles: - new build logic outside "ifndef OPENJDK" - Add a sed check to make sure open and close list use same algorithm 2

Re: webrev.01 of 8011402: Move blacklisting certificate logic from hard code to data

2013-09-18 Thread Erik Joelsson
Build change looks ok. /Erik On 2013-09-17 13:07, Weijun Wang wrote: Webrev updated to version 02 at http://cr.openjdk.java.net/~weijun/8011402/webrev.02/ Changes since webrev.01: 1. Makefiles: - new build logic outside "ifndef OPENJDK" - Add a sed check to make sure open and close l

Re: webrev.01 of 8011402: Move blacklisting certificate logic from hard code to data

2013-09-17 Thread Weijun Wang
Webrev updated to version 02 at http://cr.openjdk.java.net/~weijun/8011402/webrev.02/ Changes since webrev.01: 1. Makefiles: - new build logic outside "ifndef OPENJDK" - Add a sed check to make sure open and close list use same algorithm 2. Fingerprint calculation moved into X509CertIm

Re: webrev.01 of 8011402: Move blacklisting certificate logic from hard code to data

2013-09-12 Thread Weijun Wang
On 9/13/13 5:15 AM, Sean Mullan wrote: Ok, I suggested you use a WeakHashMap but now I'm a little concerned this could become a bottleneck if every certificate check has to lock the map. Hmm. Maybe we should go back to the previous code, that also had some concurrency issues but it was only pe

Re: webrev.01 of 8011402: Move blacklisting certificate logic from hard code to data

2013-09-12 Thread Sean Mullan
Ok, I suggested you use a WeakHashMap but now I'm a little concerned this could become a bottleneck if every certificate check has to lock the map. Hmm. Maybe we should go back to the previous code, that also had some concurrency issues but it was only per certificate, and wasn't too bad sinc

Re: webrev.01 of 8011402: Move blacklisting certificate logic from hard code to data

2013-09-11 Thread Weijun Wang
Yes, you're right. I need to try run it again with OPENJDK set. Thanks Max On 9/11/13 9:45 PM, Erik Joelsson wrote: On 2013-09-11 15:32, Weijun Wang wrote: Slightly updated at the same location. Added different algorithm check in the 2 makefiles. I should have noticed this earlier, but in

Re: webrev.01 of 8011402: Move blacklisting certificate logic from hard code to data

2013-09-11 Thread Erik Joelsson
On 2013-09-11 15:32, Weijun Wang wrote: Slightly updated at the same location. Added different algorithm check in the 2 makefiles. I should have noticed this earlier, but in CopyFiles.gmk, the whole thing is enclosed in ifndef OPENJDK. I would expect lines 403 to 415 to be moved outside of t

Re: webrev.01 of 8011402: Move blacklisting certificate logic from hard code to data

2013-09-11 Thread Weijun Wang
Slightly updated at the same location. Added different algorithm check in the 2 makefiles. Thanks Max On 9/11/13 3:57 PM, Weijun Wang wrote: Hi Sean and Erik An updated webrev is at http://cr.openjdk.java.net/~weijun/8011402/webrev.01/ Changes since the last webrev: - Some makefile cha

webrev.01 of 8011402: Move blacklisting certificate logic from hard code to data

2013-09-11 Thread Weijun Wang
Hi Sean and Erik An updated webrev is at http://cr.openjdk.java.net/~weijun/8011402/webrev.01/ Changes since the last webrev: - Some makefile changes * wildcard on closed file * make sure the file's first line is always "Algorithm=" - Move fingerprint cache for cert from X509CertImpl to