Newsgroups: php.internals
Path: news.php.net
Xref: news.php.net php.internals:51417
Mailing-List: contact internals-help@lists.php.net; run by ezmlm
Received-SPF: pass (pb1.pair.com: domain gmail.com designates 209.85.216.177 as permitted sender)
DomainKey-Status: bad
DomainKey-Signature: a=rsa-sha1; c=nofws;
        d=gmail.com; s=gamma;
        h=mime-version:reply-to:from:date:message-id:subject:to:content-type;
        b=o+NXXqGeJOwTRKKY9sVh96q43dRnAlQHKR0DmU8F4Gw2pv9SltDNrYamm16DtYFidU
         Ce+R+NtkN61dn8ilbUDOaZsCHC6ABthq0jbv2kq+nP47hup1SrxL1ib7/nEqoseECEDy
         j1knENUhe9Hm2aB80pVaQM7JRVkA2IQgEqotc=
MIME-Version: 1.0
Reply-To: RQuadling@googlemail.com
Date: Tue, 1 Feb 2011 13:35:26 +0000
Message-ID: <AANLkTinghCx4acSckWxiz_a8oXJnRy1QmAOgr4dZzb6a@mail.gmail.com>
To: PHP internals <internals@lists.php.net>
Content-Type: text/plain; charset=UTF-8
Subject: size_t, casting and expression is always true (or false)
From: rquadling@gmail.com (Richard Quadling)

Hi.

I was looking at the cause for a test failing.

It came down to ...

size_t n;

if (n < 0) {
 // Code never reached as n is unsigned and cannot be negative.
}


I'm not a strong c developer, but from reading it seems that size_t is
always unsigned.

Using the windows compiler (and turning on all the warnings via /Wall
though /W4 should be enough), this warning reports as ...

xxx\yyy.c(nnn) : warning C4296: '>=' : expression is always false

I then ran a full build and extracted all the C4926 warnings.

main\network.c(1175) : warning C4296: '>=' : expression is always true
main\streams\filter.c(363) : warning C4296: '<' : expression is always false
main\streams\streams.c(921) : warning C4296: '>=' : expression is always true
ext\mysqlnd\mysqlnd.c(1006) : warning C4296: '>=' : expression is always true
ext\mysqlnd\mysqlnd.c(1031) : warning C4296: '>=' : expression is always true
ext\mysqlnd\mysqlnd_loaddata.c(215) : warning C4296: '<' : expression
is always false
ext\mysqlnd\mysqlnd_loaddata.c(223) : warning C4296: '<' : expression
is always false
ext\mysqlnd\mysqlnd_ps.c(1480) : warning C4296: '<' : expression is always false
ext\mysqlnd\mysqlnd_ps.c(1620) : warning C4296: '<' : expression is always false
ext\standard\ftp_fopen_wrapper.c(645) : warning C4296: '>=' :
expression is always true
ext\standard\image.c(645) : warning C4296: '<' : expression is always false
ext\standard\streamsfuncs.c(637) : warning C4296: '>=' : expression is
always true
ext\standard\streamsfuncs.c(677) : warning C4296: '>=' : expression is
always true
ext\standard\var.c(403) : warning C4296: '<' : expression is always false
ext\pdo\pdo_stmt.c(463) : warning C4296: '<' : expression is always false
ext\soap\php_encoding.c(2794) : warning C4296: '<' : expression is always false

In the case of the bug [1], the incorrect type comparison resulted in
the suppression of an error which cased the test to fail, leading me
here.

I think other comparisons in the above list are also suppressing
errors (I think mysqlnd will fail to report a loss of connection
during mysqlnd_handle_local_infile() for example).

Regards,

Richard.

[1] http://bugs.php.net/bug.php?id=49072

-- 
Richard Quadling
Twitter : EE : Zend
@RQuadling : e-e.com/M_248814.html : bit.ly/9O8vFY