Messages in this thread | | | Date | Fri, 28 Jul 2023 10:52:59 -0700 | Subject | Re: [PATCH] Add Silicom Platform Driver | From | Guenter Roeck <> |
| |
On 7/28/23 09:47, Christophe JAILLET wrote: > Le 28/07/2023 à 14:59, Huibin Shi a écrit : >> Christophe, >> >> Thanks for the comments. See my comments below. >> >> Updated patch will be sent out later after review comments from other reviewers are addressed. >> >> Henry >> -----Original Message----- >> From: Christophe JAILLET <christophe.jaillet@wanadoo.fr> >> Sent: Tuesday, July 25, 2023 5:03 PM >> To: Henry Shi <henryshi2018@gmail.com>; hbshi69@hotmail.com; tglx@linutronix.de; mingo@redhat.com; bp@alien8.de; dave.hansen@linux.intel.com; x86@kernel.org; hpa@zytor.com; hdegoede@redhat.com; markgross@kernel.org; jdelvare@suse.com; linux@roeck-us.net; linux-kernel@vger.kernel.org; platform-driver-x86@vger.kernel.org; linux-hwmon@vger.kernel.org >> Cc: hb_shi2003@yahoo.com; Huibin Shi <henrys@silicom-usa.com>; Wen Wang <wenw@silicom-usa.com> >> Subject: Re: [PATCH] Add Silicom Platform Driver >> >> Caution: This is an external email. Please take care when clicking links or opening attachments. >> >> >> Le 18/07/2023 à 18:01, Henry Shi a écrit : >>> The Silicom platform (silicom-platform) Linux driver for Swisscom >>> Business Box (Swisscom BB) as well as Cordoba family products is a >>> software solution designed to facilitate the efficient management and >>> control of devices through the integration of various Linux >>> frameworks. This platform driver provides seamless support for device >>> management via the Linux LED framework, GPIO framework, Hardware >>> Monitoring (HWMON), and device attributes. The Silicom platform >>> driver's compatibility with these Linux frameworks allows applications >>> to access and control Cordoba family devices using existing software >>> that is compatible with these frameworks. This compatibility >>> simplifies the development process, reduces dependencies on >>> proprietary solutions, and promotes interoperability with other >>> Linux-based systems and software. >>> >>> Signed-off-by: Henry Shi <henryshi2018@gmail.com> >>> --- >> >> [...] >> >>> +static int __init silicom_mc_leds_register(struct device *dev, >>> + const struct led_classdev_mc >>> +*mc_leds) { >>> + struct led_classdev_mc *led; >>> + int i, err; >>> + >>> + for (i = 0; mc_leds[i].led_cdev.name; i++) { >>> + /* allocate and copy data from the init constansts */ >>> + led = devm_kzalloc(dev, sizeof(struct led_classdev_mc), >>> + GFP_KERNEL); >> >> sizeof(*led) is shorter. >> Mostly a matter of taste. >> >> Maybe even devm_kmemdup()? >> >> Henry: thanks. Devm_kmemdup() API requires additional argument that is not necessary of this driver. I prefer devm_kzalloc for now. > > CJ: The only additionnal parameter I can think of are the one of memcpy() ... > >> >>> + if (IS_ERR_OR_NULL(led)) { >> >> if (!led) >> is enough. >> >> Henry: OK, changed >> >>> + dev_err(dev, "Failed to alloc >>> + led_classdev_mc[%d]: %ld\n", i, PTR_ERR(led)); >> >> This kind of message is useless and should be removed (checkpatch should warn about it) >> >> Henry: OK, removed. >> >>> + return -ENOMEM; >>> + } >>> + memcpy(led, &mc_leds[i], sizeof(*led)); > > ... here. > > devm_kzalloc() + this memcpy() could be done with only devm_kmemdup(). > > This is mostly a matter of taste. >
I don't think that using (or not using) available API functions should be a matter of taste, or the next step might be to re-implement memcpy().
Guenter
| |