From mboxrd@z Thu Jan 1 00:00:00 1970 From: Pavel Fedin Subject: RE: [PATCH 1/3] KVM: arm64: Implement vGICv3 distributor and redistributor access from userspace Date: Mon, 31 Aug 2015 10:35:05 +0300 Message-ID: <013f01d0e3bf$8ed1fed0$ac75fc70$@samsung.com> References: <20150830164237.GD24113@cbox> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Return-path: In-reply-to: <20150830164237.GD24113@cbox> Content-language: ru Sender: kvm-owner@vger.kernel.org To: 'Christoffer Dall' Cc: kvmarm@lists.cs.columbia.edu, kvm@vger.kernel.org, 'Marc Zyngier' List-Id: kvmarm@lists.cs.columbia.edu Hello! > > + len = vgic_v3_get_reg_size(attr); > > + if (len < 0) > > + return len; > > > > - return -ENXIO; > > + ret = vgic_v3_attr_regs_access(dev, attr, (len == 8) ? (void *)®64 : > > + (void *)®, len, false); > > this use of the ternary operator is terrible, but it should be solved if > you always use a u64 for the reg parameter. I also dislike this, but this is the best thing i could invent. This is dictated by put_user() and get_user(), which rely on typeof() of their arguments. Well, i could do some castings, but they are no less ugly, and would give more headache to bigendian systems. However, what about doing the same thing as GET/SET_ONE_REG does by just assuming that everything is 64-bit wide? This would automatically resolve two other issues you have commented on. By the way, handling it in userspace would also be simpler. Kind regards, Pavel Fedin Expert Engineer Samsung Electronics Research center Russia