Messages in this thread | | | Date | Sat, 11 Dec 2010 00:53:09 -0500 (EST) | From | "Robert P. J. Day" <> | Subject | Re: CodingStyle Problem |
| |
On Sat, 11 Dec 2010, HarryWei wrote:
> Hi all, > I just read the CodingSyle at Documentation directory in linux-2.6.23 kernel source code. > But a problem happened to me like following. > Functions in linux-2.6.23/fs/fs-writeback.c are "static int" and function name are in different lines.(function name in following line) > I can't find the rule in CodingStyle. That We often program functions are "static int" and function name are in the same line. > When do we do the first or second way? What is different between them? > > Any answer is okay. > Best regards.
the advantage to the form
static int function-name
is that, if you're looking for the actual function *definition* in the file, having the function name at the start of a new line means you can search for it with the pattern "^function-name" so that you don't have to wade through all of the invocations of that function.
i like that style; others claim that with proper cross-referencers like cscope, doing that is redundant. as far as i know, the kernel coding style doesn't take a position on that, but i'm willing to be corrected.
rday
--
======================================================================== Robert P. J. Day Waterloo, Ontario, CANADA http://crashcourse.ca
Twitter: http://twitter.com/rpjday LinkedIn: http://ca.linkedin.com/in/rpjday ========================================================================
| |