lkml.org 
[lkml]   [2010]   [Apr]   [30]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [PATCH 13/30] viafb: Separate global and fb-specific data
On Wed, 28 April 2010 Jonathan Corbet <corbet@lwn.net> wrote:
> diff --git a/drivers/video/via/via-core.c b/drivers/video/via/via-core.c
> index cda4de4..6f8f8e2 100644
> --- a/drivers/video/via/via-core.c
> +++ b/drivers/video/via/via-core.c

<snip>

> +/*
> + * Figure out and map our MMIO regions.
> + */
> +static int __devinit via_pci_setup_mmio(struct viafb_dev *vdev)
> +{
> + /*
> + * Hook up to the device registers.
> + */
> + vdev->engine_start = pci_resource_start(vdev->pdev, 1);
> + vdev->engine_len = pci_resource_len(vdev->pdev, 1);
> + /* If this fails, others will notice later */
> + vdev->engine_mmio = ioremap_nocache(vdev->engine_start,
> + vdev->engine_len);

Shouldn't this ioremap_nocache() have error-checking
as the one below (instead of relying on later code to bail out)?

> +
> + /*
> + * Likewise with I/O memory.
> + */
> + vdev->fbmem_start = pci_resource_start(vdev->pdev, 0);
> + vdev->fbmem_len = viafb_get_fb_size_from_pci(vdev->chip_type);
> + if (vdev->fbmem_len < 0)
> + return vdev->fbmem_len;
> + vdev->fbmem = ioremap_nocache(vdev->fbmem_start, vdev->fbmem_len);
> + if (vdev->fbmem == NULL)
> + return -ENOMEM;

IMHO it would be better to iounmap(vdev->engine_mmio) here
instead of letting our caller run via_pci_teardown_mmio()

> + return 0;
> +}
> +
> +static void __devexit via_pci_teardown_mmio(struct viafb_dev *vdev)
> +{
> + iounmap(vdev->fbmem);
> + iounmap(vdev->engine_mmio);
> +}
> +
>
> static int __devinit via_pci_probe(struct pci_dev *pdev,
> const struct pci_device_id *ent)
> @@ -33,9 +199,19 @@ static int __devinit via_pci_probe(struct pci_dev *pdev,
> if (ret)
> return ret;
> /*
> + * Global device initialization.
> + */
> + memset(&global_dev, 0, sizeof(global_dev));
> + global_dev.pdev = pdev;
> + global_dev.chip_type = ent->driver_data;
> + spin_lock_init(&global_dev.reg_lock);
> + ret = via_pci_setup_mmio(&global_dev);
> + if (ret)
> + goto out_teardown;

Why goto out_teardown here?
If via_pci_setup_mmio() failed it should undo its successful
actions itself... (see also above)

> + /*
> * Set up subsidiary devices
> */
> - ret = via_fb_pci_probe(pdev, ent);
> + ret = via_fb_pci_probe(&global_dev);
> if (ret)
> return ret;

In this case we should goto out_teardown so the mmio setup earlier gets
undone.

> /*
> @@ -48,12 +224,17 @@ static int __devinit via_pci_probe(struct pci_dev *pdev,
> return ret;
> }
> return 0;
> +
> +out_teardown:
> + via_pci_teardown_mmio(&global_dev);
> + return ret;
> }

Thanks,
Bruno


\
 
 \ /
  Last update: 2010-04-30 19:31    [from the cache]
©2003-2011 Jasper Spaans