Bug 169287
| Summary: | program reads uninited local variable | ||
|---|---|---|---|
| Product: | [Fedora] Fedora | Reporter: | David Binderman <dcb314> |
| Component: | qt | Assignee: | Than Ngo <than> |
| Status: | CLOSED UPSTREAM | QA Contact: | Ben Levenson <benl> |
| Severity: | medium | Docs Contact: | |
| Priority: | medium | ||
| Version: | rawhide | ||
| Target Milestone: | --- | ||
| Target Release: | --- | ||
| Hardware: | All | ||
| OS: | Linux | ||
| Whiteboard: | |||
| Fixed In Version: | Doc Type: | Bug Fix | |
| Doc Text: | Story Points: | --- | |
| Clone Of: | Environment: | ||
| Last Closed: | 2005-09-27 11:13:47 UTC | Type: | --- |
| Regression: | --- | Mount Type: | --- |
| Documentation: | --- | CRM: | |
| Verified Versions: | Category: | --- | |
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |
| Cloudforms Team: | --- | Target Upstream Version: | |
| Embargoed: | |||
which gcc version have you used here? I cannot reproduce it here with
gcc-4.0.1-14. Taken a look at the source, the error is initialized before using
if ( gdef->GlyphClassDef.loaded )
{
error = Get_Class( &gdef->GlyphClassDef, glyphID, &klass, &index );
if ( error && error != TTO_Err_Not_Covered )
return error;
}
else
{
klass = 0;
index = 0;
}
/* we don't accept glyphs covered in `GlyphClassDef' */
if ( !error )
return TTO_Err_Not_Covered;
>which gcc version have you used here? The one that is standard in FC4. I think it is gcc 4.0.0 >I cannot reproduce it here The error is obvious by visual inspection. Please try again. >the error is initialized before using Local variable "error" is only initialise under some, not all, execution paths. Suggest initialise "error" on _all_ execution paths. Hm, by visual inspection the local variable "error" is always initialised here if
condition gdef->GlyphClassDef.loaded is true!
if ( gdef->GlyphClassDef.loaded )
{
error = Get_Class( &gdef->GlyphClassDef, glyphID, &klass, &index );
>local variable "error" is always initialised
What happens if gdef->GlyphClassDef.loaded is false ?
I suggest again to have another look.
Both I and the compiler think there are execution paths where "error"
is not initialised.
yes, it's obviously that "error" is not initialised before using. Sorry i have been blind! ok, it seems that this 3dparty code is not used by default. I have already sent the fix to trolltech for including in next upstream release. Thanks for your report. |
Description of problem: I just tried to compile Fedora development package qt-3.3.5-2 with the extra compiler flag -Wall. The compiler said 3rdparty/opentype/ftxgdef.c:1084: warning: 'error' is used uninitialized in this function The source code is /* we don't accept glyphs covered in `GlyphClassDef' */ if ( !error ) return TTO_Err_Not_Covered; Suggest initialise local variable "error" before first use. Version-Release number of selected component (if applicable): How reproducible: Steps to Reproduce: 1. 2. 3. Actual results: Expected results: Additional info: