From mboxrd@z Thu Jan 1 00:00:00 1970 From: nsekhar@ti.com (Sekhar Nori) Date: Wed, 23 Nov 2016 17:43:09 +0530 Subject: [PATCH 1/3] of: base: add support to get machine compatible string In-Reply-To: References: <1479811311-3080-1-git-send-email-bgolaszewski@baylibre.com> <1479811311-3080-2-git-send-email-bgolaszewski@baylibre.com> <5ce9fb9f-459a-562b-2e9f-85d35f9ec035@arm.com> <67a3c2c7-0cb9-9764-2710-6ee66fc4dde4@ti.com> <2a644b8c-d91e-5ab1-200b-00f749a36307@arm.com> Message-ID: <11467504-c700-cbfa-a945-be9ec8776144@ti.com> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org On Wednesday 23 November 2016 05:37 PM, Sudeep Holla wrote: >> So, the if(!of_node_get()) is just an expensive NULL pointer check. I >> think >> it is better to be explicit about it by not using of_node_get/put() at >> all. >> How about: >> > > Are we planning to use this in any time sensitive paths? Anyways I am > fine removing them. Not worried about the time taken as much as it serving as a bad example and getting carried over to other places where the impact might actually be real. > >> +int of_machine_get_model_name(const char **model) >> +{ >> + int error; >> + >> + if (!of_root) >> + return -EINVAL; >> + >> + error = of_property_read_string(of_root, "model", model); >> + if (error) >> + error = of_property_read_string_index(of_root, >> "compatible", >> + 0, model); >> + return error; >> +} >> +EXPORT_SYMBOL(of_machine_get_model_name); >> >> I know the patch is already in -next so I guess it depends on how >> strongly >> Rob feels about this. > > Frank expressed his concerns and it may be reverted. Didn't notice that. I will check that thread. Thanks! Regards, Sekhar