[PATCH] ath5k: don't enable MSI, we cannot handle it yet

[PATCH] ath5k: don't enable MSI, we cannot handle it yet

[Pavel Roskin]

This fixes support for AR5006 chipset, which supports MSI
---

 drivers/net/wireless/ath5k/base.c |    5 -----
 1 files changed, 0 insertions(+), 5 deletions(-)

diff --git a/drivers/net/wireless/ath5k/base.c b/drivers/net/wireless/ath5k/base.c
index 635b9ac..e57905c 100644
--- a/drivers/net/wireless/ath5k/base.c
+++ b/drivers/net/wireless/ath5k/base.c
@@ -487,9 +487,6 @@ ath5k_pci_probe(struct pci_dev *pdev,
 	/* Set private data */
 	pci_set_drvdata(pdev, hw);
 
-	/* Enable msi for devices that support it */
-	pci_enable_msi(pdev);
-
 	/* Setup interrupt handler */
 	ret = request_irq(pdev->irq, ath5k_intr, IRQF_SHARED, "ath", sc);
 	if (ret) {
@@ -567,7 +564,6 @@ err_ah:
 err_irq:
 	free_irq(pdev->irq, sc);
 err_free:
-	pci_disable_msi(pdev);
 	ieee80211_free_hw(hw);
 err_map:
 	pci_iounmap(pdev, mem);
@@ -589,7 +585,6 @@ ath5k_pci_remove(struct pci_dev *pdev)
 	ath5k_detach(pdev, hw);
 	ath5k_hw_detach(sc->ah);
 	free_irq(pdev->irq, sc);
-	pci_disable_msi(pdev);
 	pci_iounmap(pdev, sc->iobase);
 	pci_release_region(pdev, 0);
 	pci_disable_device(pdev);

Re: [PATCH] ath5k: don't enable MSI, we cannot handle it yet

[Pavel Roskin]

On Fri, 2008-07-18 at 12:28 -0400, Pavel Roskin wrote:
> This fixes support for AR5006 chipset, which supports MSI

Sorry, I found Documentation/stable_kernel_rules.txt minutes after
sending the patch.  We need to get it to the Linus' tree first, then to
the stable tree.  I hope it won't take long.

To preempt possible questions: MSI is a nice thing, but we cannot enable
it without changing the interrupt handler.  If we do it, we break MSI
capable hardware.

-- 
Regards,
Pavel Roskin

Re: [PATCH] ath5k: don't enable MSI, we cannot handle it yet

[Michael Buesch]

On Friday 18 July 2008 18:44:47 Pavel Roskin wrote:
> On Fri, 2008-07-18 at 12:28 -0400, Pavel Roskin wrote:
> > This fixes support for AR5006 chipset, which supports MSI
> 
> Sorry, I found Documentation/stable_kernel_rules.txt minutes after
> sending the patch.  We need to get it to the Linus' tree first, then to
> the stable tree.  I hope it won't take long.
> 
> To preempt possible questions: MSI is a nice thing, but we cannot enable
> it without changing the interrupt handler.  If we do it, we break MSI
> capable hardware.

Please include this into the patch changelog. It's highly confusing otherwise ;)

-- 
Greetings Michael.

[PATCH v2] ath5k: don't enable MSI, we cannot handle it yet

[Pavel Roskin]

MSI is a nice thing, but we cannot enable it without changing the
interrupt handler.  If we do it, we break MSI capable hardware,
specifically AR5006 chipset.

Signed-off-by: Pavel Roskin <proski@gnu.org>
---

 drivers/net/wireless/ath5k/base.c |    5 -----
 1 files changed, 0 insertions(+), 5 deletions(-)

diff --git a/drivers/net/wireless/ath5k/base.c b/drivers/net/wireless/ath5k/base.c
index 635b9ac..e57905c 100644
--- a/drivers/net/wireless/ath5k/base.c
+++ b/drivers/net/wireless/ath5k/base.c
@@ -487,9 +487,6 @@ ath5k_pci_probe(struct pci_dev *pdev,
 	/* Set private data */
 	pci_set_drvdata(pdev, hw);
 
-	/* Enable msi for devices that support it */
-	pci_enable_msi(pdev);
-
 	/* Setup interrupt handler */
 	ret = request_irq(pdev->irq, ath5k_intr, IRQF_SHARED, "ath", sc);
 	if (ret) {
@@ -567,7 +564,6 @@ err_ah:
 err_irq:
 	free_irq(pdev->irq, sc);
 err_free:
-	pci_disable_msi(pdev);
 	ieee80211_free_hw(hw);
 err_map:
 	pci_iounmap(pdev, mem);
@@ -589,7 +585,6 @@ ath5k_pci_remove(struct pci_dev *pdev)
 	ath5k_detach(pdev, hw);
 	ath5k_hw_detach(sc->ah);
 	free_irq(pdev->irq, sc);
-	pci_disable_msi(pdev);
 	pci_iounmap(pdev, sc->iobase);
 	pci_release_region(pdev, 0);
 	pci_disable_device(pdev);


-- 
Regards,
Pavel Roskin

Re: [PATCH v2] ath5k: don't enable MSI, we cannot handle it yet

[Nick Kossifidis]

2008/7/18 Pavel Roskin <proski@gnu.org>:
> MSI is a nice thing, but we cannot enable it without changing the
> interrupt handler.  If we do it, we break MSI capable hardware,
> specifically AR5006 chipset.
>
> Signed-off-by: Pavel Roskin <proski@gnu.org>
> ---
>
>  drivers/net/wireless/ath5k/base.c |    5 -----
>  1 files changed, 0 insertions(+), 5 deletions(-)
>
> diff --git a/drivers/net/wireless/ath5k/base.c b/drivers/net/wireless/ath5k/base.c
> index 635b9ac..e57905c 100644
> --- a/drivers/net/wireless/ath5k/base.c
> +++ b/drivers/net/wireless/ath5k/base.c
> @@ -487,9 +487,6 @@ ath5k_pci_probe(struct pci_dev *pdev,
>        /* Set private data */
>        pci_set_drvdata(pdev, hw);
>
> -       /* Enable msi for devices that support it */
> -       pci_enable_msi(pdev);
> -
>        /* Setup interrupt handler */
>        ret = request_irq(pdev->irq, ath5k_intr, IRQF_SHARED, "ath", sc);
>        if (ret) {
> @@ -567,7 +564,6 @@ err_ah:
>  err_irq:
>        free_irq(pdev->irq, sc);
>  err_free:
> -       pci_disable_msi(pdev);
>        ieee80211_free_hw(hw);
>  err_map:
>        pci_iounmap(pdev, mem);
> @@ -589,7 +585,6 @@ ath5k_pci_remove(struct pci_dev *pdev)
>        ath5k_detach(pdev, hw);
>        ath5k_hw_detach(sc->ah);
>        free_irq(pdev->irq, sc);
> -       pci_disable_msi(pdev);
>        pci_iounmap(pdev, sc->iobase);
>        pci_release_region(pdev, 0);
>        pci_disable_device(pdev);
>
>
> --
> Regards,
> Pavel Roskin
>

This is needed for all pci-e devices since enabling msi results no interrupts...

Acked-by: Nick Kossifidis <mickflemm@gmail.com>

-- 
GPG ID: 0xD21DB2DB
As you read this post global entropy rises. Have Fun ;-)
Nick

Re: [PATCH v2] ath5k: don't enable MSI, we cannot handle it yet

[Henrique de Moraes Holschuh]

On Fri, 18 Jul 2008, Nick Kossifidis wrote:
> 2008/7/18 Pavel Roskin <proski@gnu.org>:
> > MSI is a nice thing, but we cannot enable it without changing the
> > interrupt handler.  If we do it, we break MSI capable hardware,
> > specifically AR5006 chipset.
> >
> > Signed-off-by: Pavel Roskin <proski@gnu.org>
[...]
> Acked-by: Nick Kossifidis <mickflemm@gmail.com>
Cc: stable@kernel.org

So that they will pick it up automatically when it hits mainline, and it
also serves as an in-band documentation that it is something that needs to
go/went to -stable.

-- 
  "One disk to rule them all, One disk to find them. One disk to bring
  them all and in the darkness grind them. In the Land of Redmond
  where the shadows lie." -- The Silicon Valley Tarot
  Henrique Holschuh

Re: [stable] [PATCH] ath5k: don't enable MSI, we cannot handle it yet

[Greg KH]

On Fri, Jul 18, 2008 at 06:49:24PM +0200, Michael Buesch wrote:
> On Friday 18 July 2008 18:44:47 Pavel Roskin wrote:
> > On Fri, 2008-07-18 at 12:28 -0400, Pavel Roskin wrote:
> > > This fixes support for AR5006 chipset, which supports MSI
> > 
> > Sorry, I found Documentation/stable_kernel_rules.txt minutes after
> > sending the patch.  We need to get it to the Linus' tree first, then to
> > the stable tree.  I hope it won't take long.
> > 
> > To preempt possible questions: MSI is a nice thing, but we cannot enable
> > it without changing the interrupt handler.  If we do it, we break MSI
> > capable hardware.
> 
> Please include this into the patch changelog. It's highly confusing otherwise ;)

I agree.

Also, to get stuff into -stable easily, just add:
	Cc: stable <stable@kernel.org>
to the signed-off-by area in the patch and then when it goes into
Linus's tree, it will automatically be sent to us for inclusion.

Much easier on you and everyone else that way.

thanks,

greg k-h

Re: [PATCH v2] ath5k: don't enable MSI, we cannot handle it yet

[Kyle McMartin]

On Fri, Jul 18, 2008 at 08:57:39PM +0300, Nick Kossifidis wrote:
> > MSI is a nice thing, but we cannot enable it without changing the
> > interrupt handler.  If we do it, we break MSI capable hardware,
> > specifically AR5006 chipset.
> >
> > Signed-off-by: Pavel Roskin <proski@gnu.org>
> 
> This is needed for all pci-e devices since enabling msi results no interrupts...
> 
> Acked-by: Nick Kossifidis <mickflemm@gmail.com>
> 

This is likely caused by a bug being discussed on linux-pci right now.
Basically the way we're masking MSI can result in INTn interrupts
getting turned back on, with no handler attached.

http://marc.info/?t=121430463700001&r=1&w=2

regards, Kyle