From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from smtp.codeaurora.org ([198.145.29.96]:50444 "EHLO smtp.codeaurora.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751350AbeBWGxb (ORCPT ); Fri, 23 Feb 2018 01:53:31 -0500 MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII; format=flowed Date: Fri, 23 Feb 2018 12:23:30 +0530 From: poza@codeaurora.org To: Christoph Hellwig Cc: Bjorn Helgaas , Philippe Ombredanne , Thomas Gleixner , Greg Kroah-Hartman , Kate Stewart , linux-pci@vger.kernel.org, linux-kernel@vger.kernel.org, Dongdong Liu , Keith Busch , Wei Zhang , Sinan Kaya , Timur Tabi Subject: Re: [PATCH v10 6/7] PCI: Unify wait for link active into generic pci In-Reply-To: <20180222184348.GE6267@infradead.org> References: <1519315332-26852-1-git-send-email-poza@codeaurora.org> <1519315332-26852-7-git-send-email-poza@codeaurora.org> <20180222184348.GE6267@infradead.org> Message-ID: Sender: linux-pci-owner@vger.kernel.org List-ID: On 2018-02-23 00:13, Christoph Hellwig wrote: >> >> +/** >> + * pci__wait_for_link - Wait for link till its active/inactive > > typo - just wants a single underscore. > >> + pcie_capability_read_word(pdev, PCI_EXP_LNKSTA, &lnk_status); >> + ret = !!(lnk_status & PCI_EXP_LNKSTA_DLLLA); > > no need for the !! when assigning to a boolean. bool is 8 bit, while #define PCI_EXP_LNKSTA_DLLLA 0x2000 /* Data Link Layer Link Active */ so I think we need it. otherwise it will treat lnk_status & PCI_EXP_LNKSTA_DLLLA as 0 even if 13th bit is set. > >> + >> + while ((ret != active) && (timeout > 0)) { > > No need for either pair of inner braces. > >> + msleep(10); >> + timeout -= 10; >> + pcie_capability_read_word(pdev, PCI_EXP_LNKSTA, &lnk_status); >> + ret = !!(lnk_status & PCI_EXP_LNKSTA_DLLLA); > > Same as above. > >> + } >> + >> + if (ret == active) >> + return true; > > Seems like the structure is a bit odd. Why not: > > for (;;) { > pcie_capability_read_word(pdev, PCI_EXP_LNKSTA, &lnk_status); > if ((lnk_status & PCI_EXP_LNKSTA_DLLLA) == active) > return true; > if (timeout <= 0) > break; > timeout -= 10; > } Thanks for suggestion, will do this way. thanks. Regards, Oza.