From: p.fedin@samsung.com (Pavel Fedin) Date: Thu, 29 Oct 2015 11:17:21 +0300 Subject: [PATCH v3 3/4] drivers: exynos-srom: Add support for bank configuration In-Reply-To: <5631C3D3.1050309@samsung.com> References: <048021141f295b0e70c852134d4ac9cc6f7f1da4.1446018918.git.p.fedin@samsung.com> <56318175.6000006@samsung.com> <006b01d11216$9831dc40$c89594c0$@samsung.com> <5631C3D3.1050309@samsung.com> Message-ID: <001201d11222$3c69a6c0$b53cf440$@samsung.com> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org Hello! > But the system won't get any information about the failure. Device using > these banks should not probe in such case. In order to achieve this effect, i believe, it's not enough to fail SROMc probe. Peripherials (such as smsc) should then be declared as subnodes. There's one problem with this, however. I could not manage to do it. If i place device's node inside of SROMc node, it does not get probed at all, and i don't really understand why. I tried to read the code and i suggest that only specific devices (like compatible="simple-bus") are searched for sub-devices. OTOH, this patch (http://www.spinics.net/lists/arm-kernel/msg449703.html) perfectly worked, and generic timer node placed inside of MCT node was perfectly probed by the kernel. So i still don't understand why i could not place smsc description inside of SROMc. May be you can help me with this? Kind regards, Pavel Fedin Expert Engineer Samsung Electronics Research center Russia