On Mon, Jun 30, 2008 at 04:18:25PM -0700, Andrew Morton wrote:
> On Fri, 13 Jun 2008 14:09:58 +0100
> David Howells <[email protected]> wrote:
>
> > Provide __ucmpdi2() for MN10300 so that allmodconfig can be built.
>
> Well... x86 has this problem as well, and when it pops up we address
> it by going in and looking at the codesite which is generating the call
> and fixing it. And I do mean "fixing", because it's often the case
> that the caller was doing something silly and inefficient.
In this case I'd blame the gcc 3.4 fork used by MN10300.
> So are you sure that we should do this, rather than massaging the
> callers?
Patch to fix the callers (if wanted) is below.
cu
Adrian
<-- snip -->
The gcc 3.4 fork used to compile the MN10300 port emits unwanted
__ucmpdi2() calls for this switch on a 64bit value.
Fix it by transforming the switch to equivalent "if ... else if ..."
statements.
Signed-off-by: Adrian Bunk <[email protected]>
---
a89c72d736986777b2cc4e07589aa46e957c15a5
diff --git a/drivers/net/wireless/b43legacy/dma.c b/drivers/net/wireless/b43legacy/dma.c
index c990f87..794dc83 100644
--- a/drivers/net/wireless/b43legacy/dma.c
+++ b/drivers/net/wireless/b43legacy/dma.c
@@ -1028,19 +1028,15 @@ int b43legacy_dma_init(struct b43legacy_wldev *dev)
enum b43legacy_dmatype type;
dmamask = supported_dma_mask(dev);
- switch (dmamask) {
- default:
- B43legacy_WARN_ON(1);
- case DMA_30BIT_MASK:
+
+ if (dmamask == DMA_30BIT_MASK)
type = B43legacy_DMA_30BIT;
- break;
- case DMA_32BIT_MASK:
+ else if (dmamask == DMA_32BIT_MASK)
type = B43legacy_DMA_32BIT;
- break;
- case DMA_64BIT_MASK:
+ else if (dmamask == DMA_64BIT_MASK)
type = B43legacy_DMA_64BIT;
- break;
- }
+ else
+ B43legacy_WARN_ON(1);
err = ssb_dma_set_mask(dev->dev, dmamask);
if (err) {