linux-kernel - Re: [PATCH 3/3] HID: google: whiskers: mask out extra flags in EC event

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] [day] [month] [year] [list]

Date:   Mon, 7 Oct 2019 15:48:32 -0700
From:   Brian Norris <briannorris@...omium.org>
To:     Ikjoon Jang <ikjn@...omium.org>
Cc:     linux-input@...r.kernel.org,
        Linux Kernel <linux-kernel@...r.kernel.org>,
        Dmitry Torokhov <dmitry.torokhov@...il.com>,
        Benson Leung <bleung@...omium.org>,
        Enric Balletbo i Serra <enric.balletbo@...labora.com>,
        Guenter Roeck <groeck@...omium.org>,
        Enrico Granata <egranata@...gle.com>,
        Ting Shen <phoenixshen@...omium.org>,
        Nicolas Boichat <drinkcat@...omium.org>
Subject: Re: [PATCH 3/3] HID: google: whiskers: mask out extra flags in EC event_type

On Sat, Oct 5, 2019 at 3:16 AM Ikjoon Jang <ikjn@...omium.org> wrote:
>
> Whiskers needs to get notifications from EC for getting current base
> attached state. EC sends extra bits in event_type field that receiver
> should mask out.

Notably, this patch was never actually landed upstream:

https://lore.kernel.org/patchwork/patch/1019477/
[PATCH] mfd: cros_ec: Add support for MKBP more event flags

and therefore, this EC_CMD_GET_NEXT_EVENT v2 handling is not yet truly
relevant. (i.e., no upstream-proper users should hit this bug yet.)
But that's also a reminder that we need a patch like this for *every*
cros_ec client driver that's using the event_type field. Other
unpatched drivers include
drivers/media/platform/cros-ec-cec/cros-ec-cec.c,
drivers/platform/chrome/cros_ec_chardev.c, and possibly others.

So I wonder: why don't we
(a) *really* try to upstream the above patch and
(b) fix it so that event_data.event_type *always* masks out
EC_MKBP_EVENT_TYPE_MASK
?

(We could still handle the EC_MKBP_HAS_MORE_EVENTS bit within
cros_ec.c, but there's no need for every other driver to have to know
anything about it.)

Of course, this is another reminder that we should *really* try to get
our cros_ec patches landed properly in upstream, because otherwise we
have a different set of bugs and features landing in various
downstream and mostly-upstream kernels.

Brian

...
> --- a/drivers/hid/hid-google-hammer.c
> +++ b/drivers/hid/hid-google-hammer.c
> @@ -96,8 +96,9 @@ static int cbas_ec_notify(struct notifier_block *nb,
>         struct cros_ec_device *ec = _notify;
>         unsigned long flags;
>         bool base_present;
> +       const u8 event_type = ec->event_data.event_type & EC_MKBP_EVENT_TYPE_MASK;
>
> -       if (ec->event_data.event_type == EC_MKBP_EVENT_SWITCH) {
> +       if (event_type == EC_MKBP_EVENT_SWITCH) {
>                 base_present = cbas_parse_base_state(
>                                         &ec->event_data.data.switches);
>                 dev_dbg(cbas_ec.dev,