On 19.01.2017 20:48, Greg KH wrote:
On Thu, Jan 19, 2017 at 02:21:26PM +0200, Mathias Nyman wrote:
Hi Greg
This series by Arnd Bergmann was originally six patches, but last two of
them were already taken to 4.10. Without the rest of them there will
be a regression in 4.10.
Is it really a regression? I thought this had never worked before in
older kernels, right?
Regression when xhci hosts in dwc3 controllers are used.
For example patch 5/6 removed setting dma mask for xhci in dwc3 host init:
+++ b/drivers/usb/dwc3/host.c
@@ -84,11 +84,7 @@ int dwc3_host_init(struct dwc3 *dwc)
return -ENOMEM;
}
- dma_set_coherent_mask(&xhci->dev, dwc->dev->coherent_dma_mask);
-
xhci->dev.parent = dwc->dev;
- xhci->dev.dma_mask = dwc->dev->dma_mask;
- xhci->dev.dma_parms = dwc->dev->dma_parms;
So now xhci platform driver prints a scary warning because of the missing dma mask:
static int xhci_plat_probe(struct platform_device *pdev)
/* Try to set 64-bit DMA first */
if (WARN_ON(!pdev->dev.dma_mask))
/* Platform did not initialize dma_mask */
ret = dma_coerce_mask_and_coherent(&pdev->dev,
DMA_BIT_MASK(64));
else
...
This is fixed in the first 4 patches.
There might be other other issues as well caused by having only the dwc3
changed applied of this series, but not the core and xhci parts
-Mathias
--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
