On Tue, 2021-08-17 at 20:49 +0200, Greg KH wrote:

On Tue, Aug 17, 2021 at 11:36:09AM -0700, Joe Perches wrote:

quoted

On Tue, 2021-08-17 at 19:57 +0200, Greg KH wrote:

quoted

On Mon, Aug 16, 2021 at 09:31:25PM +0200, Michael Straube wrote:

quoted

Refactor functions rtw_is_cckrates_included() and
rtw_is_cckratesonly_included(). Add new helper function rtw_is_cckrate()
that allows to make the code more compact. Improves readability and
slightly reduces object file size. Change the return type to bool to
reflect that the functions return boolean values.

[]

quoted

quoted

diff --git a/drivers/staging/r8188eu/core/rtw_ieee80211.c b/drivers/staging/r8188eu/core/rtw_ieee80211.c

[]

quoted

quoted

+bool rtw_is_cckratesonly_included(u8 *rate)
 {
-	u32 i = 0;
+	u8 r;
 

-	while (rate[i] != 0) {
-		if  ((((rate[i]) & 0x7f) != 2) && (((rate[i]) & 0x7f) != 4) &&
-		     (((rate[i]) & 0x7f) != 11)  && (((rate[i]) & 0x7f) != 22))
+	while ((r = *rate++)) {

Ick, no.

While it might be fun to play with pointers like this, trying to
determine the precidence issues involved with reading from, and then
incrementing the pointer like this is crazy.

The original was obvious as to how it was walking through the array.

It's sad to believe *ptr++ is not obvious to you as it's very commonly
used in the kernel sources (over 10,000 instances).

There's lots of sad things in life :(

Your difficulty reading very standard c is relatively low on my list.

Still, it's you not this particular patch to the realtek staging code.

What's really poor about this code is using a 0 terminated list rather
than passing the array size.

But then again, almost all of the realtek code in staging is really poor.

cheers, Joe