Summary: | x11-base/xorg-server-1.11.4-r1 fails compilation xf86EdidModes.c error: array subscript is above array bounds | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | Corvus Corax <corvuscorax> |
Component: | [OLD] Unspecified | Assignee: | Gentoo Linux bug wranglers <bug-wranglers> |
Status: | RESOLVED DUPLICATE | ||
Severity: | major | CC: | kjackie |
Priority: | Normal | ||
Version: | unspecified | ||
Hardware: | AMD64 | ||
OS: | Linux | ||
Whiteboard: | |||
Package list: | Runtime testing required: | --- | |
Attachments: |
complete build log
environment file from build directory |
Description
Corvus Corax
2012-06-07 15:44:50 UTC
Created attachment 314599 [details]
complete build log
Created attachment 314601 [details]
environment file from build directory
Change -O3 to -O2 can get rid of these errors. Thanks for the hint, yeah knowing the way GCC does checks, I think with -O2 or less that error would indeed disappear. Although I'd rephrase that as "changing to -O2 will make the compiler too dump to notice the flaw". They'd still be there, being an unchecked array out of bounds error waiting to happen at runtime (aka buffer overflow, segfault, ...) won't they? I built my entire system with -O3 for ages. Previous versions of xorg worked fine with -O3. While there are some ebuilds that I know enforce lower optimization settings for particular packages, I'm not sure if that would be the apropriate action here, or if that would just mask apossible upstream bug that should rather justify masking the entire xorg version as unsafe? I mean just look at all these thousands of warnings about incompatible function prototypes, I'd say something is seriously wrong upstream or on behalf of some patches here - as far as coding standards are concerned, isn't it? And Xorg is a pretty critical core package at least on desktop systems. I just reproduced the bug on a second system, same package, same error during emerge -u world *** This bug has been marked as a duplicate of bug 402285 *** |