On Fri, Feb 09, 2007 at 10:55:39AM -0500, James Bottomley wrote:
> On Thu, 2007-02-08 at 19:40 +0000, Christoph Hellwig wrote:
> > The logic here seems odd to me.  Shouldn't we first check
> > dma_get_required_mask and the do dma_set_mask?
> 
> Yes ... I picked up the mask problem, but not that.  And actually
> there's a missing zero check on one of the dma_set_mask()s.
> 
> Does this look like the right patch then?

This looks functionally correct to me, but there's some cosmetic issues
left:

> Index: scsi-misc-2.6/drivers/scsi/aic7xxx/aic79xx_osm_pci.c
> ===================================================================
> --- scsi-misc-2.6.orig/drivers/scsi/aic7xxx/aic79xx_osm_pci.c 2007-02-09 
> 09:12:55.000000000 -0500
> +++ scsi-misc-2.6/drivers/scsi/aic7xxx/aic79xx_osm_pci.c      2007-02-09 
> 09:21:32.000000000 -0500
> @@ -132,6 +132,7 @@ ahd_linux_pci_dev_probe(struct pci_dev *
>       struct           ahd_pci_identity *entry;
>       char            *name;
>       int              error;
> +     struct device   *dev = &pdev->dev;
>  
>       pci = pdev;
>       entry = ahd_find_pci_device(pci);
> @@ -161,20 +162,16 @@ ahd_linux_pci_dev_probe(struct pci_dev *
>       pci_set_master(pdev);
>  
>       if (sizeof(dma_addr_t) > 4) {
> -             uint64_t   memsize;
> -             const uint64_t mask_39bit = 0x7FFFFFFFFFULL;
> -
> -             memsize = ahd_linux_get_memsize();
> -
> -             if (memsize >= 0x8000000000ULL
> -              && pci_set_dma_mask(pdev, DMA_64BIT_MASK) == 0) {
> +             if (dma_get_required_mask(dev) > DMA_39BIT_MASK
> +                 && dma_set_mask(dev, DMA_64BIT_MASK) == 0)
>                       ahd->flags |= AHD_64BIT_ADDRESSING;
> -             } else if (memsize > 0x80000000
> -                     && pci_set_dma_mask(pdev, mask_39bit) == 0) {
> +             else if (dma_get_required_mask(dev) > DMA_32BIT_MASK
> +                      && dma_set_mask(dev, DMA_39BIT_MASK) == 0)
>                       ahd->flags |= AHD_39BIT_ADDRESSING;
> -             }
> +             else
> +                     dma_set_mask(dev, DMA_32BIT_MASK);
>       } else {

I'd rather do the dma_get_required_mask only once, and we we want
&& at the end of the first instead of at the beginning of the second
line.  So this block should look like:

                u64 required_mask = dma_get_required_mask(dev);

                if (required_mask > DMA_39BIT_MASK &&
                    dma_set_mask(dev, DMA_64BIT_MASK) == 0)
                        ahd->flags |= AHD_64BIT_ADDRESSING;
                else if (required_mask > DMA_32BIT_MASK &&
                         dma_set_mask(dev, DMA_39BIT_MASK) == 0)
                        ahd->flags |= AHD_39BIT_ADDRESSING;
                else
                        dma_set_mask(dev, DMA_32BIT_MASK);
-
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Reply via email to