linux-kernel - RE: [PATCH v4] x86/Hyper-V: Support for free page reporting

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]

Date:   Fri, 19 Mar 2021 21:30:50 +0000
From:   Michael Kelley <mikelley@...rosoft.com>
To:     Sunil Muthuswamy <sunilmut@...rosoft.com>,
        Matheus Castello <matheus@...tello.eng.br>,
        "linux-hyperv@...r.kernel.org" <linux-hyperv@...r.kernel.org>,
        Haiyang Zhang <haiyangz@...rosoft.com>,
        Stephen Hemminger <sthemmin@...rosoft.com>,
        Wei Liu <liuwe@...rosoft.com>,
        Tianyu Lan <Tianyu.Lan@...rosoft.com>,
        Wei Liu <wei.liu@...nel.org>, vkuznets <vkuznets@...hat.com>
CC:     KY Srinivasan <kys@...rosoft.com>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: RE: [PATCH v4] x86/Hyper-V: Support for free page reporting

From: Sunil Muthuswamy <sunilmut@...rosoft.com>  Sent: Friday, March 19, 2021 2:21 PM
> 
> > What's the strategy for this flag in the unlikely event that the hypercall fails?
> > It doesn't seem right to have hv_query_ext_cap() fail, but leave the
> > static flag set to true.  Just move that line down to after the status check
> > has succeeded?
> 
> That call should not fail in any normal circumstances. The current idea was to
> avoid repeating the same call on persistent failure. 

OK, I can see that as a valid strategy.  And the assumption is that a failed
hypercall would leave hv_extended_cap unmodified and hence all zeros.

I'm OK with this approach if you want to keep it.  But perhaps add a short
comment about the intent so it doesn't look like a bug. :-)

Michael

> But, since we don't expect
> the query capability to be called in any kind of hot path, I am ok moving this
> down.
> 
> >
> > Other than the above about the flag when the hypercall fails,
> > everything else looks good.
> Thanks for the review.