458728 - gcc warning: src/nsAttrAndChildArray.cpp:169: warning: comparison is always true due to limited range of data type

Status: RESOLVED FIXED

(Developer Infrastructure :: Source Code Analysis, defect)

Description

[georgi - hopefully not receiving bugspam]

+++ This bug was initially created as a clone of Bug #458491 +++

Comment 1

[Ed Morley [:emorley]]

http://mxr.mozilla.org/mozilla-central/source/content/base/src/nsAttrAndChildArray.cpp#169

167   PRUint32 childCount = ChildCount();
168 
169   NS_ENSURE_TRUE(childCount < ATTRCHILD_ARRAY_MAX_CHILD_COUNT,
170                  NS_ERROR_FAILURE);

Comment 2

[Ed Morley [:emorley]]

For future reference...

ChildCount() is defined here:
http://mxr.mozilla.org/mozilla-central/source/content/base/src/nsAttrAndChildArray.h#83

ATTRCHILD_ARRAY_MAX_CHILD_COUNT here:
http://mxr.mozilla.org/mozilla-central/source/content/base/src/nsAttrAndChildArray.h#66

Comment 3

[:aceman]

Is this a valid problem?
Maybe today the comparison is unneeded with the current value of ATTRCHILD_ARRAY_MAX_CHILD_COUNT. But if that constant changes, it may be necessary to catch it here.

Comment 4

[Marco Castelluccio [:marco]]

Attachment: Patch

Comment 5

[Ed Morley [:emorley]]

Comment on attachment 620444 [details] [diff] [review]
Patch

Sorry, I'm not a peer; redirecting to someone who is :-)

Comment 6

[Boris Zbarsky [:bzbarsky]]

Why is this warning at all?  childCount is a PRUint32.  ATTRCHILD_ARRAY_MAX_CHILD_COUNT should be 0x3fffff.  There are certainly values of PRUint32 that are bigger than that!

Marco, I assume you reproduced the warning?

Comment 7

[Boris Zbarsky [:bzbarsky]]

Note that back when the bug was filed, a 64-bit build would have warned here.  But that should have been fixed in bug 480655.

Comment 8

[Marco Castelluccio [:marco]]

(In reply to Boris Zbarsky (:bz) from comment #6)
> Marco, I assume you reproduced the warning?

No, I just assumed it was reproducible. I didn't see the bug was so old.