| Message ID | 1548831714-3706-1-git-send-email-guohanjun@huawei.com |
|---|---|
| State | New |
| Headers | show |
| Series | [RFC] USB: PCI: set 32bit DMA mask for PCI based USB controllers | expand |
On Wed, Jan 30, 2019 at 03:01:54PM +0800, Hanjun Guo wrote: > This is the RFC version, I'm not sure this is the best solution, > comments are warmly welcomed. > > Thanks > Hanjun > > drivers/usb/core/hcd-pci.c | 4 ++++ > 1 file changed, 4 insertions(+) > > diff --git a/drivers/usb/core/hcd-pci.c b/drivers/usb/core/hcd-pci.c > index 0343246..a9c33e6 100644 > --- a/drivers/usb/core/hcd-pci.c > +++ b/drivers/usb/core/hcd-pci.c > @@ -188,6 +188,10 @@ int usb_hcd_pci_probe(struct pci_dev *dev, const struct pci_device_id *id) > if (pci_enable_device(dev) < 0) > return -ENODEV; > > + retval = dma_coerce_mask_and_coherent(&dev->dev, DMA_BIT_MASK(32)); > + if (retval) > + return retval; dma_coerce_mask_and_coherent is only for platform devices (and I'm not sure it is a good idea to start with, but that is a different story). PCI device should have the dma_mask pointer set already, so you should use dma_set_mask_and_coherent here.
Hi John, On 2019/1/31 17:54, John Garry wrote: > On 30/01/2019 07:01, Hanjun Guo wrote: >> From: Hanjun Guo <hanjun.guo@linaro.org> [...] >> >> drivers/usb/core/hcd-pci.c | 4 ++++ >> 1 file changed, 4 insertions(+) >> >> diff --git a/drivers/usb/core/hcd-pci.c b/drivers/usb/core/hcd-pci.c >> index 0343246..a9c33e6 100644 >> --- a/drivers/usb/core/hcd-pci.c >> +++ b/drivers/usb/core/hcd-pci.c >> @@ -188,6 +188,10 @@ int usb_hcd_pci_probe(struct pci_dev *dev, const struct pci_device_id *id) >> if (pci_enable_device(dev) < 0) >> return -ENODEV; >> >> + retval = dma_coerce_mask_and_coherent(&dev->dev, DMA_BIT_MASK(32)); >> + if (retval) >> + return retval; > > Hi Hanjun, > > You're missing tidy-up upon failure. Good catch, needs to disable pci for failure, I will send a updated version to address the comments from you and Christoph. Thanks Hanjun
On 2019/2/1 17:13, Hanjun Guo wrote: > On 2019/2/1 13:55, Hanjun Guo wrote: >> Hi John, >> >> On 2019/1/31 17:54, John Garry wrote: >>> On 30/01/2019 07:01, Hanjun Guo wrote: >>>> From: Hanjun Guo <hanjun.guo@linaro.org> >> [...] >>>> >>>> drivers/usb/core/hcd-pci.c | 4 ++++ >>>> 1 file changed, 4 insertions(+) >>>> >>>> diff --git a/drivers/usb/core/hcd-pci.c b/drivers/usb/core/hcd-pci.c >>>> index 0343246..a9c33e6 100644 >>>> --- a/drivers/usb/core/hcd-pci.c >>>> +++ b/drivers/usb/core/hcd-pci.c >>>> @@ -188,6 +188,10 @@ int usb_hcd_pci_probe(struct pci_dev *dev, const struct pci_device_id *id) >>>> if (pci_enable_device(dev) < 0) >>>> return -ENODEV; >>>> >>>> + retval = dma_coerce_mask_and_coherent(&dev->dev, DMA_BIT_MASK(32)); >>>> + if (retval) >>>> + return retval; >>> >>> Hi Hanjun, >>> >>> You're missing tidy-up upon failure. >> >> Good catch, needs to disable pci for failure, I will send >> a updated version to address the comments from you and Christoph. > > There is a _DMA method which was introduced in ACPI 6.2 to cover > this, I will try that solution then report back. We add a _DMA method under the hostbridge which the EHCI/OHCI being connected to, the calltrace is gone and EHCI works as expected. No need for kernel change, so this patch is dropped. Thanks Hanjun
diff --git a/drivers/usb/core/hcd-pci.c b/drivers/usb/core/hcd-pci.c index 0343246..a9c33e6 100644 --- a/drivers/usb/core/hcd-pci.c +++ b/drivers/usb/core/hcd-pci.c @@ -188,6 +188,10 @@ int usb_hcd_pci_probe(struct pci_dev *dev, const struct pci_device_id *id) if (pci_enable_device(dev) < 0) return -ENODEV; + retval = dma_coerce_mask_and_coherent(&dev->dev, DMA_BIT_MASK(32)); + if (retval) + return retval; + /* * The xHCI driver has its own irq management * make sure irq setup is not touched for xhci in generic hcd code