On Tuesday, December 02, 2014 07:07:47 AM Joe Perches wrote:
> net_<foo>_ratelimited uses the same form.
>
> include/linux/net.h:#define net_ratelimited_function(function, ...)
> include/linux/net.h-do {
> include/linux/net.h- if (net_ratelimit())
> include/linux/net.h- function(__VA_ARGS__);
> include/linux/net.h-} while (0)
>
Yes, I see. There are representatives of both "schools". As counter example,
netdev_<foo> uses "long" way:
#define define_netdev_printk_level(func, level) \
void func(const struct net_device *dev, const char *fmt, ...) \
{ \
struct va_format vaf; \
va_list args; \
\
va_start(args, fmt); \
\
vaf.fmt = fmt; \
vaf.va = &args; \
\
__netdev_printk(level, dev, &vaf); \
\
va_end(args); \
} \
EXPORT_SYMBOL(func);
define_netdev_printk_level(netdev_emerg, KERN_EMERG);
Problem with "short" way, it is hard to detect usage with side effect.
Example from netfilter:
if (dst_mtu(skb_dst(skb)) <= minlen) {
net_err_ratelimited("unknown or invalid path-MTU (%u)\n",
dst_mtu(skb_dst(skb)));
Now, one need to remember that dst_mtu() can't have side effects.
I agree it is bad practice to have non-trivial arguments in printk-like functions,
but it is hard to detect, specifically if one just automatically replace function with its
"ratelimited" version.
--
To unsubscribe from this list: send the line "unsubscribe linux-wireless" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
