[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <87h6w711yc.ffs@tglx>
Date: Tue, 31 Jan 2023 10:35:23 +0100
From: Thomas Gleixner <tglx@...utronix.de>
To: Shanker Donthineni <sdonthineni@...dia.com>,
Marc Zyngier <maz@...nel.org>, Michael Walle <michael@...le.cc>
Cc: Sebastian Andrzej Siewior <bigeasy@...utronix.de>,
Hans de Goede <hdegoede@...hat.com>,
Wolfram Sang <wsa+renesas@...g-engineering.com>,
Shanker Donthineni <sdonthineni@...dia.com>,
linux-kernel@...r.kernel.org
Subject: Re: [PATCH 4/5] genirq: Use the common function irq_expand_nr_irqs()
On Sun, Jan 29 2023 at 18:57, Shanker Donthineni wrote:
> Subject: genirq: Use the common function ...
genirq: Unify irq_expand_nr_irqs()
irq_expand_nr_irqs() is implemented as a stub function for !SPARSEIRQ
builds. That's not necessary as the SPARSEIRQ version returns -ENOMEM
correctly even for the !SPARSEIRQ case as the ....
But this common function is non-obvious for the !SPARSEIRQ case. It at
least needs a comment
> +static int irq_expand_nr_irqs(unsigned int nr)
> +{
> + if (nr > MAX_SPARSE_IRQS)
> + return -ENOMEM;
> + nr_irqs = nr;
> + return 0;
> +}
or preferrably something like this:
if (!IS_ENABLED(CONFIG_SPARSEIRQ) || nr > MAX_SPARSE_IRQS)
return -ENOMEM;
which makes it entirely clear and also allows the compiler to optimize
is down to a 'return -ENOMEM'.
Thanks,
tglx
Powered by blists - more mailing lists