On Mon, Jan 12, 2009 at 05:35:28PM +0200, Tony Lindgren wrote: > Well at least you could remove some parens. > How about (dma_trigger / 32) << 19 instead? Be very careful. gcc 4.3.2 ignores the cost of calling its support functions, so it could well decide to _actually_ do that as a divide rather than the obvious bit clear. Not only that, but the guys at Simtec found that gcc 4.3.2 sometimes generates the optimal code followed by the library call, followed by code overwrite the result of the library call with the version from the optimal code! This issue is the subject of an open gcc bug; and I believe is the reason that kautobuild's compiler doesn't have ARMv7 support at the moment. -- To unsubscribe from this list: send the line "unsubscribe linux-omap" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
