Re: 5.10 regression caused by: "uas: fix sdev->host->dma_dev": many XHCI swiotlb buffer is full / DMAR: Device bounce map failed errors on thunderbolt connected XHCI controller

2020-11-30 Thread Hans de Goede
Hi, On 11/28/20 6:15 PM, Christoph Hellwig wrote: > Can you give this one-liner a spin? > > diff --git a/drivers/iommu/intel/iommu.c b/drivers/iommu/intel/iommu.c > index c6622011d4938c..e889111b55c71d 100644 > --- a/drivers/iommu/intel/iommu.c > +++ b/drivers/iommu/intel/iommu.c > @@ -4007,6 +40

Re: 5.10 regression caused by: "uas: fix sdev->host->dma_dev": many XHCI swiotlb buffer is full / DMAR: Device bounce map failed errors on thunderbolt connected XHCI controller

2020-11-28 Thread Christoph Hellwig
Can you give this one-liner a spin? diff --git a/drivers/iommu/intel/iommu.c b/drivers/iommu/intel/iommu.c index c6622011d4938c..e889111b55c71d 100644 --- a/drivers/iommu/intel/iommu.c +++ b/drivers/iommu/intel/iommu.c @@ -4007,6 +4007,7 @@ static const struct dma_map_ops bounce_dma_ops = {

Re: 5.10 regression caused by: "uas: fix sdev->host->dma_dev": many XHCI swiotlb buffer is full / DMAR: Device bounce map failed errors on thunderbolt connected XHCI controller

2020-11-28 Thread Hans de Goede
Hi Tom, On 11/28/20 2:25 AM, Tom Yan wrote: > Should we still be clamping max_sectors to dma_max_mapping_size(dev) > (for now)? with dev being us->pusb_dev->bus->sysdev and > devinfo->udev->bus->sysdev respectively (i.e. revert only > scsi_add_host_with_dma() to scsi_add_host())? I would expect t

Re: 5.10 regression caused by: "uas: fix sdev->host->dma_dev": many XHCI swiotlb buffer is full / DMAR: Device bounce map failed errors on thunderbolt connected XHCI controller

2020-11-27 Thread Tom Yan
Should we still be clamping max_sectors to dma_max_mapping_size(dev) (for now)? with dev being us->pusb_dev->bus->sysdev and devinfo->udev->bus->sysdev respectively (i.e. revert only scsi_add_host_with_dma() to scsi_add_host())? On Sat, 28 Nov 2020 at 02:12, Hans de Goede wrote: > > Hi, > > On 11

Re: 5.10 regression caused by: "uas: fix sdev->host->dma_dev": many XHCI swiotlb buffer is full / DMAR: Device bounce map failed errors on thunderbolt connected XHCI controller

2020-11-27 Thread Hans de Goede
Hi, On 11/27/20 5:19 PM, Christoph Hellwig wrote: > On Fri, Nov 27, 2020 at 01:32:16PM +0100, Hans de Goede wrote: >> I ran some more tests, I can confirm that reverting: >> >> 5df7ef7d32fe "uas: bump hw_max_sectors to 2048 blocks for SS or faster >> drives" >> 558033c2828f "uas: fix sdev->host->

Re: 5.10 regression caused by: "uas: fix sdev->host->dma_dev": many XHCI swiotlb buffer is full / DMAR: Device bounce map failed errors on thunderbolt connected XHCI controller

2020-11-27 Thread Christoph Hellwig
On Fri, Nov 27, 2020 at 01:32:16PM +0100, Hans de Goede wrote: > I ran some more tests, I can confirm that reverting: > > 5df7ef7d32fe "uas: bump hw_max_sectors to 2048 blocks for SS or faster drives" > 558033c2828f "uas: fix sdev->host->dma_dev" > > Makes the problem go away while running a 5.10

Re: 5.10 regression caused by: "uas: fix sdev->host->dma_dev": many XHCI swiotlb buffer is full / DMAR: Device bounce map failed errors on thunderbolt connected XHCI controller

2020-11-27 Thread Hans de Goede
Hi, On 11/27/20 12:41 PM, Hans de Goede wrote: > Hi, > > On 11/24/20 11:27 AM, Christoph Hellwig wrote: >> On Mon, Nov 23, 2020 at 03:49:09PM +0100, Hans de Goede wrote: >>> Hi, >>> >>> +Cc Christoph Hellwig >>> >>> Christoph, this is still an issue, so I've been looking around a bit and >>> th