From mboxrd@z Thu Jan 1 00:00:00 1970 From: timur@codeaurora.org (Timur Tabi) Date: Thu, 19 Oct 2017 17:44:59 -0500 Subject: [PATCH 1/2] [v5] pinctrl: qcom: disable GPIO groups with no pins In-Reply-To: References: <1504798409-32041-1-git-send-email-timur@codeaurora.org> <1504798409-32041-2-git-send-email-timur@codeaurora.org> <20171002174414.GL1165@minitux> <9f28c9dd-6277-6c75-1186-a834e15c5346@codeaurora.org> Message-ID: <819a0c67-3494-030d-3166-748081509c83@codeaurora.org> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org On 10/07/2017 06:07 AM, Linus Walleij wrote: > I mean that you add > unsigned long *line_valid_mask; > to struct gpio_chip using the same type of logic > as .irq_valid_mask. The mask should be optional. > > Then the operation gpiod_get[_index]() will FAIL if the > line is not valid. > > There is no need to check on every operation, there should > be no way to get a handle on the pin any other way. > > All new code should be using descriptors at this point so we > only need to design for that case. I think I'm missing something, because I implemented what you're asking for, but whenever I try to expose a GPIO from sysfs (e.g. echo 37 > /sys/class/gpio/export), gpiod_get_index() is not called. The only thing preventing the GPIO from being exported is the ->request callback: static int msm_gpio_request(struct gpio_chip *chip, unsigned int offset) { struct msm_pinctrl *pctrl = gpiochip_get_data(chip); const struct msm_pingroup *g = &pctrl->soc->groups[offset]; if (!g->npins) return -ENODEV; What operation is supposed to trigger a call to gpiod_get_index? -- Qualcomm Datacenter Technologies, Inc. as an affiliate of Qualcomm Technologies, Inc. Qualcomm Technologies, Inc. is a member of the Code Aurora Forum, a Linux Foundation Collaborative Project.