Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
MIME-Version: 1.0
References: <20190211122606.8662-1-brgl@bgdev.pl> <20190211122606.8662-29-brgl@bgdev.pl>
 <7b494951-7b37-3ee5-5ece-f486b7a566fe@arm.com>
In-Reply-To: <7b494951-7b37-3ee5-5ece-f486b7a566fe@arm.com>
From:   Bartosz Golaszewski <bgolaszewski@baylibre.com>
Date:   Mon, 11 Feb 2019 14:10:29 +0100
Message-ID: <CAMpxmJVeQe7sxfharvA8REsOdmeaQUr2f1ByJwHCc5NnaFYEsw@mail.gmail.com>
Subject: Re: [RESEND PATCH v2 28/33] ARM: davinci: cp-intc: unify error handling
To:     Marc Zyngier <marc.zyngier@arm.com>
Cc:     Bartosz Golaszewski <brgl@bgdev.pl>, Sekhar Nori <nsekhar@ti.com>,
        Kevin Hilman <khilman@kernel.org>,
        Thomas Gleixner <tglx@linutronix.de>,
        Jason Cooper <jason@lakedaemon.net>,
        David Lechner <david@lechnology.com>,
        arm-soc <linux-arm-kernel@lists.infradead.org>,
        LKML <linux-kernel@vger.kernel.org>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

pon., 11 lut 2019 o 14:08 Marc Zyngier <marc.zyngier@arm.com> napisa=C5=82(=
a):
>
> On 11/02/2019 12:26, Bartosz Golaszewski wrote:
> > From: Bartosz Golaszewski <bgolaszewski@baylibre.com>
> >
> > Use WARN_ON() on eny error in cp-intc initialization and drop all
> > custom error messages.
> >
> > Signed-off-by: Bartosz Golaszewski <bgolaszewski@baylibre.com>
> > Reviewed-by: David Lechner <david@lechnology.com>
> > ---
> >  arch/arm/mach-davinci/cp_intc.c | 10 +++-------
> >  1 file changed, 3 insertions(+), 7 deletions(-)
> >
> > diff --git a/arch/arm/mach-davinci/cp_intc.c b/arch/arm/mach-davinci/cp=
_intc.c
> > index f3787ae4cdbd..c1efb9390655 100644
> > --- a/arch/arm/mach-davinci/cp_intc.c
> > +++ b/arch/arm/mach-davinci/cp_intc.c
> > @@ -200,20 +200,16 @@ davinci_cp_intc_do_init(const struct davinci_cp_i=
ntc_config *config,
> >                       DAVINCI_CP_INTC_CHAN_MAP(offset));
> >
> >       irq_base =3D irq_alloc_descs(-1, 0, config->num_irqs, 0);
> > -     if (irq_base < 0) {
> > -             pr_warn("Couldn't allocate IRQ numbers\n");
> > -             irq_base =3D 0;
> > -     }
> > +     if (WARN_ON(irq_base < 0))
> > +             return irq_base;
> >
> >       /* create a legacy host */
> >       davinci_cp_intc_irq_domain =3D irq_domain_add_legacy(
> >                                       node, config->num_irqs, irq_base,=
 0,
> >                                       &davinci_cp_intc_irq_domain_ops, =
NULL);
> >
> > -     if (!davinci_cp_intc_irq_domain) {
> > -             pr_err("cp_intc: failed to allocate irq host!\n");
> > +     if (WARN_ON(!davinci_cp_intc_irq_domain))
> >               return -EINVAL;
> > -     }
> >
> >       set_handle_irq(davinci_cp_intc_handle_irq);
> >
> >
>
> I'm sorry, but how is turning an explicit message into a long stack
> trace really useful?
>

If any of these calls fails, the system is fried. I assumed that a
stack trace will point users straight to the offending line.

Bart