linux-kernel - Re: [PATCH v2 2/5] KVM: add KVM request variants without barrier

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Wed, 1 Mar 2017 17:58:02 +0100
From:   Radim Krčmář <rkrcmar@...hat.com>
To:     David Hildenbrand <david@...hat.com>
Cc:     linux-kernel@...r.kernel.org, kvm@...r.kernel.org,
        Paolo Bonzini <pbonzini@...hat.com>,
        Andrew Jones <drjones@...hat.com>,
        Marc Zyngier <marc.zyngier@....com>,
        Christian Borntraeger <borntraeger@...ibm.com>,
        Cornelia Huck <cornelia.huck@...ibm.com>,
        James Hogan <james.hogan@...tec.com>,
        Paul Mackerras <paulus@...abs.org>,
        Christoffer Dall <christoffer.dall@...aro.org>
Subject: Re: [PATCH v2 2/5] KVM: add KVM request variants without barrier

2017-02-27 11:18+0100, David Hildenbrand:
> Am 27.02.2017 um 11:02 schrieb David Hildenbrand:
>> Am 24.02.2017 um 20:49 schrieb Radim Krčmář:
>>> The leading underscores denote that the call is just a bitop wrapper.
>> 
>> Actually, the leading underscore is misleading
>> 
>> If we want to match the semantics of set/test/clear_bit, using a leading
>> underscore might feel like using the non-atomic variants like
>> __clear_bit and friends.
>> 
>> I'd prefer to simply drop the underscore.
>> 
> 
> Okay, this is not really possible for __kvm_request_set(). Hm.....

Yeah, requests are always atomic, but have some extra cruft on top of
bit operations and underscores are similar in the sense of doing less
that the non-underscored version.  Also, the underscores were something
to make its use look undesirable in the code.

kvm_request_set and kvm_request_test_and_clear use a barrier and
kvm_request_test could be expected to do so as well.

I think that a barrier makes no sense in kvm_request_clear, but called
it with underscores for consistency with others and also because I think
that some callers of could use a second thought.