linux-kernel - Re: [PATCH v2 0/4] Add trace events for SD registers.

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]

Message-ID: <CAPDyKFqerMeQey0wJW3Ctohh77CJs6-PxSaR=T6ive_Rx-oMMA@mail.gmail.com>
Date:   Mon, 29 Apr 2019 11:57:02 +0200
From:   Ulf Hansson <ulf.hansson@...aro.org>
To:     Raul Rangel <rrangel@...omium.org>
Cc:     "linux-mmc@...r.kernel.org" <linux-mmc@...r.kernel.org>,
        linux-trace-devel@...r.kernel.org,
        Daniel Kurtz <djkurtz@...omium.org>, zwisler@...omium.org,
        Steven Rostedt <rostedt@...dmis.org>,
        hongjiefang <hongjiefang@...micro.com>,
        Jennifer Dahm <jennifer.dahm@...com>,
        Ingo Molnar <mingo@...hat.com>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
        Shawn Lin <shawn.lin@...k-chips.com>,
        Kyle Roeschley <kyle.roeschley@...com>,
        Avri Altman <avri.altman@....com>,
        Simon Horman <horms+renesas@...ge.net.au>
Subject: Re: [PATCH v2 0/4] Add trace events for SD registers.

On Wed, 24 Apr 2019 at 17:31, Raul Rangel <rrangel@...omium.org> wrote:
>
> On Tue, Apr 23, 2019 at 08:29:15AM +0200, Ulf Hansson wrote:
> > On Tue, 16 Apr 2019 at 20:33, Raul E Rangel <rrangel@...omium.org> wrote:
> > >
> > > I am not able to make a single event class for all these registers. They
> > > all have different struct sizes and different printf formats.
> > >
> > > Thanks for the reviews!
> > >
> > > Changes in v2:
> > > - Made trace_sd_scr print out flags.
> > > - Add BUILD_BUG_ON to make sure tracing stays in sync with structs.
> > > - memcpy using sizeof(__entry->raw)
> > >
> > > Raul E Rangel (4):
> > >   mmc: core: Add trace event for SD SCR response
> > >   mmc: core: Add trace event for SD SSR response
> > >   mmc: core: Add trace event for SD OCR response
> > >   mmc: core: Add trace event for CSD response
> > >
> > >  drivers/mmc/core/mmc.c     |   4 +
> > >  drivers/mmc/core/sd.c      |  10 ++
> > >  drivers/mmc/core/sd_ops.c  |   6 ++
> > >  include/trace/events/mmc.h | 204 +++++++++++++++++++++++++++++++++++++
> > >  4 files changed, 224 insertions(+)
> > >
> > > --
> > > 2.21.0.392.gf8f6787159e-goog
> > >
> >
> > Why do you need this? We already have these card registers reflected
> > though sysfs files, isn't that sufficient?
> >
> I was not actually aware that the registers were exposed via sysfs. I
> was debugging a problem where the host controller was returning all
> zeros when reading from the card. I wasn't aware that it was returning
> all zeros until I added tracing. It made it quite easy to diagnose the
> problem by just diffing the two traces.

This sounds like a quite an unusual problem, and I don't think having
the buffers printed via tracing is worth it.

Moreover, we already have tracing per command/request (but don't print
the buffers), that should cover most of error cases during init, don't
you think?

[...]

Kind regards
Uffe