linux-kernel - Re: [PATCH] ALSA: hda-intel - Disable MSI support by default

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Wed, 15 Nov 2006 19:58:13 +0100
From:	Takashi Iwai <tiwai@...e.de>
To:	Jeff Garzik <jeff@...zik.org>
Cc:	David Miller <davem@...emloft.net>, torvalds@...l.org,
	linux-kernel@...r.kernel.org
Subject: Re: [PATCH] ALSA: hda-intel - Disable MSI support by default

At Wed, 15 Nov 2006 13:32:02 -0500,
Jeff Garzik wrote:
> 
> Takashi Iwai wrote:
> > The snd-hda-intel driver has a test of MSI, but it seems not working
> > on every machine.  It caused non-cared interrupts and the kernel
> > disabled that irq.
> 
> Possibly the test was broken.  Did you have IRQF_DISABLED and 
> IRQF_SHARED flags set?

I think IRQF_* is irrelevant there.
It looks like that the hardware issues INTX regardless whether MSI is
enabled or not.  Thus it ends up with unexpected irqs which are never
caught by the driver (the driver expects a different irq from MSI),
and eventually the kernel kills this irq.

Possibly calling pci_intx() as you suggested might help to avoid this
situation.  Can anyone test the patch below?


Takashi

---
diff --git a/sound/pci/hda/hda_intel.c b/sound/pci/hda/hda_intel.c
index e35cfd3..bdb92b3 100644
--- a/sound/pci/hda/hda_intel.c
+++ b/sound/pci/hda/hda_intel.c
@@ -544,6 +544,7 @@ static unsigned int azx_rirb_get_respons
 		free_irq(chip->irq, chip);
 		chip->irq = -1;
 		pci_disable_msi(chip->pci);
+		pci_intx(chip->pci, 1);
 		chip->msi = 0;
 		if (azx_acquire_irq(chip, 1) < 0)
 			return -1;
@@ -830,14 +831,15 @@ static irqreturn_t azx_interrupt(int irq
 {
 	struct azx *chip = dev_id;
 	struct azx_dev *azx_dev;
+	unsigned long flags;
 	u32 status;
 	int i;
 
-	spin_lock(&chip->reg_lock);
+	spin_lock_irqsave(&chip->reg_lock, flags);
 
 	status = azx_readl(chip, INTSTS);
 	if (status == 0) {
-		spin_unlock(&chip->reg_lock);
+		spin_unlock_irqrestore(&chip->reg_lock, flags);
 		return IRQ_NONE;
 	}
 	
@@ -867,7 +869,7 @@ static irqreturn_t azx_interrupt(int irq
 	if (azx_readb(chip, STATESTS) & 0x04)
 		azx_writeb(chip, STATESTS, 0x04);
 #endif
-	spin_unlock(&chip->reg_lock);
+	spin_unlock_irqrestore(&chip->reg_lock, flags);
 	
 	return IRQ_HANDLED;
 }
@@ -1380,7 +1382,8 @@ static int __devinit azx_init_stream(str
 
 static int azx_acquire_irq(struct azx *chip, int do_disconnect)
 {
-	if (request_irq(chip->pci->irq, azx_interrupt, IRQF_DISABLED|IRQF_SHARED,
+	if (request_irq(chip->pci->irq, azx_interrupt,
+			chip->msi ? 0 : IRQF_SHARED,
 			"HDA Intel", chip)) {
 		printk(KERN_ERR "hda-intel: unable to grab IRQ %d, "
 		       "disabling device\n", chip->pci->irq);
@@ -1435,9 +1438,12 @@ static int azx_resume(struct pci_dev *pc
 		return -EIO;
 	}
 	pci_set_master(pci);
-	if (chip->msi)
+	if (chip->msi) {
 		if (pci_enable_msi(pci) < 0)
 			chip->msi = 0;
+		else
+			pci_intx(pci, 0);
+	}
 	if (azx_acquire_irq(chip, 1) < 0)
 		return -EIO;
 	azx_init_chip(chip);
@@ -1561,9 +1567,12 @@ static int __devinit azx_create(struct s
 		goto errout;
 	}
 
-	if (chip->msi)
+	if (chip->msi) {
 		if (pci_enable_msi(pci) < 0)
 			chip->msi = 0;
+		else
+			pci_intx(pci, 0);
+	}
 
 	if (azx_acquire_irq(chip, 0) < 0) {
 		err = -EBUSY;
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/