From mboxrd@z Thu Jan 1 00:00:00 1970 Date: Mon, 26 Feb 2018 13:00:36 +1100 From: "Tobin C. Harding" Subject: Re: [PATCH 2/4] leaking_addresses: simplify path skipping Message-ID: <20180226020036.GE2808@eros> References: <1519008649-15782-1-git-send-email-me@tobin.cc> <1519008649-15782-3-git-send-email-me@tobin.cc> <20180226012631.qff25qjngubxy4ia@cisco> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20180226012631.qff25qjngubxy4ia@cisco> To: Tycho Andersen Cc: Kernel Hardening , LKML List-ID: On Sun, Feb 25, 2018 at 06:26:31PM -0700, Tycho Andersen wrote: > Hi Tobin, > > On Mon, Feb 19, 2018 at 01:50:47PM +1100, Tobin C. Harding wrote: > > -# Do not parse these files under any subdirectory. > > -my @skip_parse_files_any = ('0', > > - '1', > > - '2', > > - 'pagemap', > > - 'events', > > - 'access', > > - 'registers', > > - 'snapshot_raw', > > - 'trace_pipe_raw', > > - 'ptmx', > > - 'trace_pipe'); > > It might be worth adding 'syscall' here; the pointers listed are user > pointers, and negative syscall args will show up like kernel pointers, > e.g. I get this output, which is spurious: > > /proc/31808/syscall: 0 0x3 0x55b107a38180 0x2000 0xffffffffffffffb0 0x55b107a302d0 0x55b107a38180 0x7fffa313b8e8 0x7ff098560d11 Nice. Will add. thanks, Tobin.