* Re: [PATCH v3 01/11] driver: platform: Add helper for safer setting of driver_override
[not found] <20220227135214.145599-2-krzysztof.kozlowski@canonical.com>
@ 2022-02-28 20:03 ` Bjorn Helgaas
0 siblings, 0 replies; only message in thread
From: Bjorn Helgaas @ 2022-02-28 20:03 UTC (permalink / raw)
To: Krzysztof Kozlowski
Cc: linux-hyperv, Stuart Yoder, Rafael J. Wysocki, linux-pci,
linux-remoteproc, alsa-devel, Bjorn Andersson,
Srinivas Kandagatla, Vineeth Vijayan, Alexander Gordeev,
Fabio Estevam, linux-clk, linux-s390, Wei Liu, Stephen Hemminger,
Abel Vesa, Michael S. Tsirkin, Dexuan Cui, Linus Torvalds,
Andy Gross, NXP Linux Team, Heiko Carstens, Vasily Gorbik,
linux-arm-msm, Sascha Hauer, linux-spi, Mark Brown,
Rasmus Villemoes, Bjorn Helgaas, virtualization, linux-arm-kernel,
Laurentiu Tudor, Mathieu Poirier, Greg Kroah-Hartman,
Haiyang Zhang, Peter Oberparleiter, linux-kernel, Sven Schnelle,
Shawn Guo
On Sun, Feb 27, 2022 at 02:52:04PM +0100, Krzysztof Kozlowski wrote:
> Several core drivers and buses expect that driver_override is a
> dynamically allocated memory thus later they can kfree() it.
> +int driver_set_override(struct device *dev, const char **override,
> + const char *s, size_t len)
> +{
> + const char *new, *old;
> + char *cp;
> +
> + if (!dev || !override || !s)
> + return -EINVAL;
> +
> + /* We need to keep extra room for a newline */
It would help a lot to extend this comment with a hint about where the
room for a newline is needed. It was confusing even before, but it's
much more so now that the check is in a completely different file than
the "show" functions that need the space.
> + if (len >= (PAGE_SIZE - 1))
> + return -EINVAL;
_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/virtualization
^ permalink raw reply [flat|nested] only message in thread