linux-kernel - Re: [PATCH] perf/core: fast breakpoint modification via _IOC_MODIFY

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:   Tue, 7 Nov 2017 09:09:13 -0800
From:   Andi Kleen <ak@...ux.intel.com>
To:     Peter Zijlstra <peterz@...radead.org>
Cc:     Milind Chabbi <chabbi.milind@...il.com>, jolsa@...hat.com,
        Ingo Molnar <mingo@...hat.com>,
        Arnaldo Carvalho de Melo <acme@...nel.org>,
        Alexander Shishkin <alexander.shishkin@...ux.intel.com>,
        Namhyung Kim <namhyung@...nel.org>,
        Michael Ellerman <mpe@...erman.id.au>,
        Hari Bathini <hbathini@...ux.vnet.ibm.com>,
        Jin Yao <yao.jin@...ux.intel.com>,
        Kan Liang <kan.liang@...el.com>,
        Sukadev Bhattiprolu <sukadev@...ux.vnet.ibm.com>,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH] perf/core: fast breakpoint modification via
 _IOC_MODIFY_BREAKPOINT.

On Tue, Nov 07, 2017 at 09:15:41AM +0100, Peter Zijlstra wrote:
> On Mon, Nov 06, 2017 at 03:16:58PM -0800, Andi Kleen wrote:
> > > +static int _perf_event_modify_breakpoint(struct perf_event *bp,
> > > +					 struct perf_event_attr *attr)
> > > +{
> > > +	u64 old_addr = bp->attr.bp_addr;
> > > +	u64 old_len = bp->attr.bp_len;
> > > +	int old_type = bp->attr.bp_type;
> > > +	int err = 0;
> > > +
> > > +	_perf_event_disable(bp);
> > > +
> > > +	bp->attr.bp_addr = attr->bp_addr;
> > > +	bp->attr.bp_type = attr->bp_type;
> > > +	bp->attr.bp_len = attr->bp_len;
> > 
> > You don't check any of the other fields, so user space is free
> > to fill in junk. That means they can never be used for anything.
> > It would be better to check at least some of them for being
> > zero, and also that the type matches the break point.
> 
> Yes, the values should at the very least get the exact same validation
> they would get on creating an event with those values.

In this case the ioctl could be also generalized. Not call it _BREAKPOINT,
just _MODIFY. Just for now it would be only implemented for break points,
but that could be potentially extended later.

-Andi