GOTO, What would you think of reassigning this bug to the kernel, asking for mroe speciific error numbers to be returned by its file system?
Thanks, Kingsley On Wed:16:47, GOTO Masanori wrote: > Kingsley, > > Could I close this bug? > > Regards, > -- gotom > > At Wed, 16 Apr 2003 00:17:02 +0900, > GOTO Masanori wrote: > > > > At Mon, 14 Apr 2003 21:33:05 -0700, > > Kingsley G. Morse Jr. <[EMAIL PROTECTED]> wrote: > > > > > > On Mon:16:09, Ben Collins wrote: > > > > Kingsley wrote: > > > > > glibc currently returns errno 28, which is > > > > > defined as > > > > > > > > > > "No space left on device" > > > > > > > > > > and is misleading. It should mention inodes. > > > > > > > > Inodes are a specific term that does not apply > > > > to all cases where ENOSPC is used. What happens > > > > if ENOSPC starts to be returned when an A/V > > > > filesystem runs out of "media blocks", should we > > > > then change it to "No space, inodes or media > > > > blocks left on device"? Not likely. ENOSPC is > > > > not specific to filesystems either. > > > > > > Good point. > > > > Yes, actually this problem is sometimes occured to develop the > > enhanced file system, and it's difficult to modify. From my > > experience, I can say that there is no generic way to resolve this > > issue on the novel unix. Only we can do is printing out actual kernel > > error message to syslog. > > > > > > > > Thanks. > > > > > > In light of your considerable experience in these > > > matters, do you happen to have any thoughts on how > > > more helpful error messages could be provided to > > > newbie users? > > > > > > For example, is there a way that glibc could > > > report specifically whether it's disk space, > > > inodes or media blocks that have been exhausted? > > > > > > Unique errnos??? > > > > It might be filesystem issue which is implemented into kernel, not > > glibc. > > > > Glibc follows the error number which is returned from filesystem. If > > you define unique errnos, then your application should be aware such > > error, and such application becomes non-portable for various > > filesystems. > > > > Regards, > > -- gotom > > > > > > > > Thanks, > > > Kingsley > > > > > > > > > > > At most, this is a FAQ. > > > > > > > > -- Debian - http://www.debian.org/ Linux > > > > 1394 - http://www.linux1394.org/ Subversion - > > > > http://subversion.tigris.org/ Deqo - > > > > http://www.deqo.com/ > > > > > > > > Content preview: > glibc currently returns > > > > errno 28, which is defined as > > > > > > "No space left on device" > > and is > > > > > > misleading. It should mention > > > > inodes. Inodes are a specific term that does > > > > not apply to all cases where ENOSPC is used. > > > > What happens if ENOSPC starts to be returned > > > > when an A/V filesystem runs out of "media > > > > blocks", should we then change it to "No > > > > space, inodes or media blocks left on device"? > > > > Not likely. ENOSPC is not specific to > > > > filesystems either. [...] > > > > > > > -- > > > To UNSUBSCRIBE, email to [EMAIL PROTECTED] > > > with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED] > > -- > > To UNSUBSCRIBE, email to [EMAIL PROTECTED] > > with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
