Home
| Calendar
| Mail Lists
| List Archives
| Desktop SIG
| Hardware Hacking SIG
Wiki | Flickr | PicasaWeb | Video | Maps & Directions | Installfests | Keysignings Linux Cafe | Meeting Notes | Linux Links | Bling | About BLU |
On Tue, May 14, 2013 at 02:01:27PM -0400, Bill Bogstad wrote: > Actually, I was expecting that GCC would treat EACCES and ENONET > identically. i.e. Silently keep looking for the requisite files and > only failing > if it couldn't find them in its list of directories. I admit that I > wasn't actually > thinking at the time of the exact value of errno. I was thinking more > conceptually, but I still feel that this is bad behavior on the part of GCC. gcc should be thought of as right in this case. ENOENT is the operating system's way of saying "there really isn't anything here," whereas EACCES variously means "this path may or may not be there, but we won't tell you." gcc rightfully considers ENOENT as definitively meaning the include files aren't present and it should carry on, but it can't reach that conclusion with EACCES. An alternative suggestion to the problem of hiding /usr/local temporarily would be mounting a tmpfs on top of /usr/local. -- Alex Pennace, alex at pennace.org, http://osiris.978.org/~alex/
BLU is a member of BostonUserGroups | |
We also thank MIT for the use of their facilities. |