On Fri, 02 May 2008 17:15:55 -0700
Harvey Harrison <harvey.harrison@gmail.com> wrote:

quoted text
> int hex_char_to_int(char ch)
> {
> 	return -1 if not [a-fA-F0-9]
> }
> 
> int hex_to_u8(const char *buf, u8 *val)
> int hex_to_u16(const char *buf, u16 *val)
> int hex_to_u32(const char *buf, u32 *val)
> int hex_to_u64(const char *buf, u64 *val)
> 
> Which return the number of chars read and reads up to a maximum of
> 2,4,8,16 chars respectively and returns early if a non-hex char
> is encountered.
> 
> Thoughts?

I dunno.  Return-by-reference is a bit nasty and forces the compiler to
generate extra code to put local storage into a stack slot rather than
keeping it in a register.


Does anyone actually test whether hex_to_foo returns -1?  Probably not
many.

In which case it might be better to do

u64 hex_to_u64(const char *buf);

and, if we encounter a non-hex char, do a WARN_ON(1).


One could legitimately do

	int c = hex_to_int(some_user_input);
	if (c < 0) {
		printk("user: you're an idiot\n");
	}

to perform validation and conversion in a single operation but I expect
that such code is in the minority, and could be covered by running a separate

bool is_hex_string(const char *bug, size_t nchars);

beforehand?
--
unsubscribe notice
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/