Source-Changes archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: CVS commit: src/dist/nvi/common



christos%zoulas.com@localhost (Christos Zoulas) writes:

> The patch is correct and better since:
>
>       1. It does not produce warnings

I applied your patch and tried to cross compile a macppc nvi, and the
warning is still there: gcc seems to use the interval of the type of
(ch) and not of the (UCHAR_T) type to decide if the comparison is
always true, which from an optimization point of view is OK.

Your patch doesn't touch the litigious comparison in the USE_WIDECHAR
case, which is NetBSD's case.

>       2. It fixes a bug

True. And nice, thanks!

>       3. It generates better code in the non-wide scenario.

Uh, didn't try that (we define USE_WIDECHAR unconditionnally) but
probably gcc will optimize the test out, if it even takes the pain to
print a warning message.


How about setting MAX_FAST_KEY to 127 until we think of a better fix?
;-)

 Aymeric


Home | Main Index | Thread Index | Old Index