From mboxrd@z Thu Jan 1 00:00:00 1970 From: Arnd Bergmann Subject: Re: [PATCH v3 2/3] ata: ahci_platform: Manage SATA PHY Date: Thu, 9 Jan 2014 11:05:20 +0100 Message-ID: <201401091105.20727.arnd@arndb.de> References: <1389174428-31414-1-git-send-email-rogerq@ti.com> <5878159.3LjUeZKdYB@wuerfel> <52CE40FA.5070605@ti.com> Mime-Version: 1.0 Content-Type: Text/Plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Return-path: Received: from moutng.kundenserver.de ([212.227.17.9]:62102 "EHLO moutng.kundenserver.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750800AbaAIKFd (ORCPT ); Thu, 9 Jan 2014 05:05:33 -0500 In-Reply-To: <52CE40FA.5070605@ti.com> Sender: linux-ide-owner@vger.kernel.org List-Id: linux-ide@vger.kernel.org To: Roger Quadros Cc: Kishon Vijay Abraham I , tj@kernel.org, sergei.shtylyov@cogentembedded.com, b.zolnierkie@samsung.com, hdegoede@redhat.com, linux-ide@vger.kernel.org, linux-kernel@vger.kernel.org, devicetree@vger.kernel.org, Balaji T K On Thursday 09 January 2014, Roger Quadros wrote: > >> if (!IS_ERR(hpriv->phy)) { > >> rc = phy_init(hpriv->phy); > >> if (rc) > >> goto disable_unprepare_clk; > >> > >> rc = phy_power_on(hpriv->phy); > >> if (rc) { > >> phy_exit(hpriv->phy); > >> goto disable_unprepare_clk; > >> } > >> } > > > > As I said, I'd prefer to set hpriv->phy to NULL in case of -ENODEV, > > but functionally it seems right (with the fixup from your other mail). > > > > Why do you prefer setting hpriv->phy to NULL instead of using IS_ERR() check > before hpriv->phy is used? > The latter seems to be the norm at least among clock framework users. Two reasons: 1. It feels more natural to read "if (clk)" in driver code as a check for "a clock exists" than "if (!IS_ERR(clk))". 2. It clarifies that this code path is only there to check for the clk-not-present case, not for other errors. Obviously the first check after clk_get needs to be IS_ERR() because that is the documented interface, but then you should decide on the action based on the specific error. Arnd