linux-kernel - Re: [RESEND PATCH v10 06/10] vmx: spp: Set up SPP paging table at vmentry/vmexit

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: Windows password security audit tool. GUI, reports in PDF.

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

Message-ID: <20200113065012.GE12253@local-michael-cet-test.sh.intel.com>
Date:   Mon, 13 Jan 2020 14:50:13 +0800
From:   Yang Weijiang <weijiang.yang@...el.com>
To:     Sean Christopherson <sean.j.christopherson@...el.com>
Cc:     Yang Weijiang <weijiang.yang@...el.com>, kvm@...r.kernel.org,
        linux-kernel@...r.kernel.org, pbonzini@...hat.com,
        jmattson@...gle.com, yu.c.zhang@...ux.intel.com,
        alazar@...defender.com, edwin.zhai@...el.com
Subject: Re: [RESEND PATCH v10 06/10] vmx: spp: Set up SPP paging table at
 vmentry/vmexit

On Fri, Jan 10, 2020 at 09:55:37AM -0800, Sean Christopherson wrote:
> On Thu, Jan 02, 2020 at 02:13:15PM +0800, Yang Weijiang wrote:
> > If write to subpage is not allowed, EPT violation generates
> > and it's handled in fast_page_fault().
> > 
> > In current implementation, SPPT setup is only handled in handle_spp()
> > vmexit handler, it's triggered when SPP bit is set in EPT leaf
> > entry while SPPT entries are not ready.
> > 
> > A SPP specific bit(11) is added to exit_qualification and a new
> > exit reason(66) is introduced for SPP.
> 
> ...
> 
> > diff --git a/arch/x86/kvm/mmu/mmu.c b/arch/x86/kvm/mmu/mmu.c
> > index 6f92b40d798c..c41791ebee65 100644
> > --- a/arch/x86/kvm/mmu/mmu.c
> > +++ b/arch/x86/kvm/mmu/mmu.c
> > @@ -6372,6 +6427,8 @@ unsigned long kvm_mmu_calculate_default_mmu_pages(struct kvm *kvm)
> >  	return nr_mmu_pages;
> >  }
> >  
> > +#include "spp.c"
> > +
> 
> Unless there is a *very* good reason for these shenanigans, spp.c needs
> to built via the Makefile like any other source.  If this is justified
> for whatever reason, then that justification needs to be very clearly
> stated in the changelog.

Yes, it looks odd. When extracted the SPP code from mmu.c, I found a lot
of functions in mmu.c should be exposed so that spp.c can see them, I
took them as unnecessary modification to mmu.c, so just add the spp.c file back
to mmu.c, if you suggest change it with a seperate object file, I'll do it.

> 
> In general, the code organization of this entire series likely needs to
> be overhauled.  There are gobs exports which are either completely
> unnecessary or completely backswards.
> 
> E.g. exporting VMX-only functions from spp.c, which presumably are only
> callbed by VMX.
> 
> 	EXPORT_SYMBOL_GPL(vmx_spp_flush_sppt);
> 	EXPORT_SYMBOL_GPL(vmx_spp_init);
> 
> Exporting ioctl helpers from the same file, which are presumably called
> only from x86.c.
> 
> 	EXPORT_SYMBOL_GPL(kvm_vm_ioctl_get_subpages);
> 	EXPORT_SYMBOL_GPL(kvm_vm_ioctl_set_subpages);

Thanks for the suggestion, I'll go over the patches and or-organize them.