From: Rui Wang Subject: Re: [PATCH v2 4/4] crypto: testmgr - Add a test case for import()/export() Date: Tue, 2 Feb 2016 22:16:34 +0800 Message-ID: <1454422594-19990-1-git-send-email-rui.y.wang@intel.com> Cc: rui.y.wang@intel.com, tim.c.chen@linux.intel.com, linux-crypto@vger.kernel.org, linux-kernel@vger.kernel.org To: herbert@gondor.apana.org.au Return-path: Received: from mga09.intel.com ([134.134.136.24]:46292 "EHLO mga09.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754451AbcBBOge (ORCPT ); Tue, 2 Feb 2016 09:36:34 -0500 Sender: linux-crypto-owner@vger.kernel.org List-ID: On Mon, Feb 1, 2016 4:22 PM Herbert Xu wrote: > > On Wed, Jan 27, 2016 at 05:08:38PM +0800, Rui Wang wrote: > > > > diff --git a/crypto/testmgr.h b/crypto/testmgr.h index > > da0a8fd..451e7eb 100644 > > --- a/crypto/testmgr.h > > +++ b/crypto/testmgr.h > > @@ -44,6 +44,7 @@ struct hash_testvec { > > unsigned short psize; > > unsigned char np; > > unsigned char ksize; > > + unsigned char partial; > > Why not make it unconditional? > I initially made it unconditional, but then I found that it can easily hang the machine during boot due to any import/export bug in any of the hash drivers. So I used this .partial flag to guard against this risk. Only when an author is confident that his driver can do this test, should he add this flag. What do you think? Thanks Rui