From mboxrd@z Thu Jan 1 00:00:00 1970 From: Tejun Heo Subject: Re: [PATCH v4 1/2] libahci: Add support to handle HOST_IRQ_STAT as edge trigger latch. Date: Tue, 5 May 2015 13:04:40 -0400 Message-ID: <20150505170440.GW1971@htj.duckdns.org> References: <1430838822-17507-1-git-send-email-stripathi@apm.com> <1430838822-17507-2-git-send-email-stripathi@apm.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Content-Disposition: inline In-Reply-To: <1430838822-17507-2-git-send-email-stripathi@apm.com> Sender: linux-scsi-owner@vger.kernel.org To: Suman Tripathi Cc: olof@lixom.net, arnd@arndb.de, linux-scsi@vger.kernel.org, linux-ide@vger.kernel.org, devicetree@vger.kernel.org, linux-arm-kernel@lists.infradead.org, mlangsdo@redhat.com, jcm@redhat.com, patches@apm.com List-Id: devicetree@vger.kernel.org Hello, Suman. Looks pretty. Some nitpicks. On Tue, May 05, 2015 at 08:43:41PM +0530, Suman Tripathi wrote: > diff --git a/drivers/ata/ahci.h b/drivers/ata/ahci.h > index 71262e0..2df2237 100644 > --- a/drivers/ata/ahci.h > +++ b/drivers/ata/ahci.h > @@ -238,6 +238,8 @@ enum { > AHCI_HFLAG_MULTI_MSI = (1 << 16), /* multiple PCI MSIs */ > AHCI_HFLAG_NO_DEVSLP = (1 << 17), /* no device sleep */ > AHCI_HFLAG_NO_FBS = (1 << 18), /* no FBS */ > + AHCI_HFLAG_EDGE_TRIG_IRQ = (1 << 19), /* HOST_IRQ_STAT behaves as > + Edge Triggered */ Let's just name it ahci_HFLAG_EDGE_IRQ. > +static irqreturn_t ahci_level_trig_intr(int irq, void *dev_instance) ahci_single_level_irq_intr() > +{ ... > @@ -1877,13 +1885,51 @@ static irqreturn_t ahci_single_irq_intr(int irq, void *dev_instance) > * Also, use the unmasked value to clear interrupt as spurious > * pending event on a dummy port might cause screaming IRQ. > */ > + Extra newline which wasn't there before. > + writel(irq_stat, mmio + HOST_IRQ_STAT); > + > + spin_unlock(&host->lock); > + > + VPRINTK("EXIT\n"); > + > + return IRQ_RETVAL(rc); > +} > + > +static irqreturn_t ahci_edge_trig_intr(int irq, void *dev_instance) ahci_single_edge_irq_intr() And let's please separate refactoring of intr routine and addition of edge irq handling into two patches. Thanks. -- tejun