On Tue, Apr 21, 2015 at 01:52:09PM +0100, Ian Abbott wrote: > >Is that really an improvement? The original code was actually defined. > > > >1U << 32 is actually defined. It is zero. Which works for us. > > According to the C standards it is undefined (for 32-bit unsigned > int). See the late draft of ISO/IEC 9899:2011 (dubbed C11) at > <http://www.open-std.org/jtc1/sc22/wg14/www/docs/n1570.pdf>, §6.5.7, > Semantics: > > 3. The integer promotions are performed on each of the operands. The > type of the result is that of the promoted left operand.If the value > of the right operand is negative or is greater than or equal to the > width of the promoted left operand, the behavior is undefined. > Yeah. You're right. > > It would be better to use (1U << b_chans) instead of (1 << b_chans) > in the code above, or possibly the slightly more obtuse: > > b_mask = ((b_chans < 32) ? 1 << b_chans : 0) - 1; I like just switching Hartley's 1 to 1U. regards, dan carpenter _______________________________________________ devel mailing list devel@xxxxxxxxxxxxxxxxxxxxxx http://driverdev.linuxdriverproject.org/mailman/listinfo/driverdev-devel
