linux-kernel - Re: [RFC v2 1/2] cgroup: sev: Add misc cgroup controller

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:   Wed, 3 Mar 2021 22:12:32 -0800
From:   Vipin Sharma <vipinsh@...gle.com>
To:     Tejun Heo <tj@...nel.org>
Cc:     mkoutny@...e.com, rdunlap@...radead.org, thomas.lendacky@....com,
        brijesh.singh@....com, jon.grimm@....com, eric.vantassell@....com,
        pbonzini@...hat.com, hannes@...xchg.org, frankja@...ux.ibm.com,
        borntraeger@...ibm.com, corbet@....net, seanjc@...gle.com,
        vkuznets@...hat.com, wanpengli@...cent.com, jmattson@...gle.com,
        joro@...tes.org, tglx@...utronix.de, mingo@...hat.com,
        bp@...en8.de, hpa@...or.com, gingell@...gle.com,
        rientjes@...gle.com, dionnaglaze@...gle.com, kvm@...r.kernel.org,
        x86@...nel.org, cgroups@...r.kernel.org, linux-doc@...r.kernel.org,
        linux-kernel@...r.kernel.org
Subject: Re: [RFC v2 1/2] cgroup: sev: Add misc cgroup controller

On Wed, Mar 03, 2021 at 10:42:37AM -0500, Tejun Heo wrote:
> > +	atomic_t usage;
> > +};
> 
> Can we do 64bits so that something which counts memory can use this too?
> 
Sure.

> > +
> > +		if (usage > capacity)
> > +			return -EBUSY;
> 
> I'd rather go with allowing bringing down capacity below usage so that the
> users can set it to a lower value to drain existing usages while denying new
> ones. It's not like it's difficult to check the current total usage from the
> caller side, so I'm not sure it's very useful to shift the condition check
> here.
> 

Okay, I will change the code to set new capacity unconditionally.

Right now there is no API for the caller to know total usage, unless they
keep their own tally, I was thinking it will be useful to add one more API

unsigned long misc_cg_res_total_usage(enum misc_res_type type)

It will return root_cg usage for "type" resource.
Will it be fine?

> > +			pr_info("cgroup: charge rejected by misc controller for %s resource in ",
> > +				misc_res_name[type]);
> > +			pr_cont_cgroup_path(i->css.cgroup);
> > +			pr_cont("\n");
> 
> Should have commented on this in the priv thread but don't print something
> on every rejection. This often becomes a nuisance and can make an easy DoS
> vector at worst. If you wanna do it, print it once per cgroup or sth like
> that.

I didn't think in that way. Thanks, I will print it once per cgroup.

Thanks
Vipin