When variable needcolon is assigned a value of false, it must be
assigned a value of true later on, which is redundant code.
lib/vsprintf.c:1411:4: warning: Value stored to 'needcolon' is never read.
Reported-by: Abaci Robot <[email protected]>
Closes: https://bugzilla.openanolis.cn/show_bug.cgi?id=6909
Signed-off-by: Jiapeng Chong <[email protected]>
---
lib/vsprintf.c | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)
diff --git a/lib/vsprintf.c b/lib/vsprintf.c
index afb88b24fa74..9a9086885da8 100644
--- a/lib/vsprintf.c
+++ b/lib/vsprintf.c
@@ -1406,10 +1406,9 @@ char *ip6_compressed_string(char *p, const char *addr)
i += longest - 1;
continue;
}
- if (needcolon) {
+ if (needcolon)
*p++ = ':';
- needcolon = false;
- }
+
/* hex u16 without leading 0s */
word = ntohs(in6.s6_addr16[i]);
hi = word >> 8;
--
2.20.1.7.g153144c
On Wed, Oct 18, 2023 at 02:48:17PM +0800, Jiapeng Chong wrote:
> When variable needcolon is assigned a value of false, it must be
> assigned a value of true later on, which is redundant code.
>
> lib/vsprintf.c:1411:4: warning: Value stored to 'needcolon' is never read.
...
> @@ -1406,10 +1406,9 @@ char *ip6_compressed_string(char *p, const char *addr)
> i += longest - 1;
> continue;
> }
> - if (needcolon) {
> + if (needcolon)
> *p++ = ':';
> - needcolon = false;
> - }
> +
> /* hex u16 without leading 0s */
> word = ntohs(in6.s6_addr16[i]);
> hi = word >> 8;
Logically you may remove then the assignment to true as well.
But I would double check if it's not a continue missing somewhere or so.
--
With Best Regards,
Andy Shevchenko
On 18/10/2023 08.48, Jiapeng Chong wrote:
> When variable needcolon is assigned a value of false, it must be
> assigned a value of true later on, which is redundant code.
NAK on any patch modifying these things without first filling the blank
space in ip6() in lib/test_printf.c with tests for all the various ip6
printing modes.
Adding test cases there is much more useful than code-golfing, and would
help verify that these changes are actually ok. Or may reveal that the
current code is buggy.
Rasmus
On Thu, Oct 19, 2023 at 01:51:45PM +0200, Rasmus Villemoes wrote:
> On 18/10/2023 08.48, Jiapeng Chong wrote:
...
> Or may reveal that the current code is buggy.
...and hopefully didn't become a part of any ABIs...
--
With Best Regards,
Andy Shevchenko
On Wed 2023-10-18 14:48:17, Jiapeng Chong wrote:
> When variable needcolon is assigned a value of false, it must be
> assigned a value of true later on, which is redundant code.
>
> lib/vsprintf.c:1411:4: warning: Value stored to 'needcolon' is never read.
>
> Reported-by: Abaci Robot <[email protected]>
> Closes: https://bugzilla.openanolis.cn/show_bug.cgi?id=6909
> Signed-off-by: Jiapeng Chong <[email protected]>
> ---
> lib/vsprintf.c | 5 ++---
> 1 file changed, 2 insertions(+), 3 deletions(-)
>
> diff --git a/lib/vsprintf.c b/lib/vsprintf.c
> index afb88b24fa74..9a9086885da8 100644
> --- a/lib/vsprintf.c
> +++ b/lib/vsprintf.c
> @@ -1406,10 +1406,9 @@ char *ip6_compressed_string(char *p, const char *addr)
> i += longest - 1;
> continue;
> }
> - if (needcolon) {
> + if (needcolon)
> *p++ = ':';
> - needcolon = false;
> - }
> +
> /* hex u16 without leading 0s */
> word = ntohs(in6.s6_addr16[i]);
> hi = word >> 8;
I am against this change. The assignment should be there from the
logical POV. IMHO, it helps people to verify that the code works
as expected. The removal might even cause regression when the logic
gets modified in the future.
Anyway, this is a slow path. The code readability is more important
then the speed. Let compiler do these micro optimizations.
Best Regards,
Petr