Tar is crashing on FAT32 writing

Bug #72756 reported by Michael Wohlmuther
6
Affects Status Importance Assigned to Milestone
tar (Ubuntu)
Confirmed
Undecided
Unassigned

Bug Description

I tried to create an archive which is bigger than 4GB on an FAT32 Volume, it's clear for me that it isn't possible, but tar crashed and don't give an output why it is crashing or why it's not possible to create this archive.

I think this is a important usability criteria.

Revision history for this message
Michael Wohlmuther (mwohlmuther) wrote :
Changed in tar:
status: Unconfirmed → Confirmed
Revision history for this message
Cylou68 (cylou68) wrote :

I have the same problem. I join the crash file.

Important note : I have the same problem when I want to copy with cp this file to the usb drive in fat32.

Revision history for this message
Tommy Trussell (tommy-trussell) wrote :

This happened to me on a system running Ubuntu 6.10 (edgy) -- haven't tried it yet under 7.04. I am creating the tar file on an external ieee1394 ("firewire") drive with the manufacturer's FAT32 formatting.

Under Edgy, Archive Manager put up a dialog saying an error occurred while writing the archive. Upon clicking the "Command Line Output" button, the message said "File size limit exceeded (core dumped)."

I won't bother attaching the crash report, but I will confirm that I just looked at the archive file I was trying to create, and Nautilus reports that the file is "4.0 GB (4294967295 bytes)"

I see nothing else unusual about the file.

Because Cylou68 said he also had the crash when attempting to cp the file, it seems possible this is a bug in the filesystem handling for FAT32. I'll try to confirm it and report back here.

Revision history for this message
Tommy Trussell (tommy-trussell) wrote :

Sadly, I'm afraid this is a duplicate of Bug 70535 -- you cannot put a file bigger than 4GB on a FAT32 volume.

Revision history for this message
Tommy Trussell (tommy-trussell) wrote :

It would be nice if tar didn't crash, but told you the file reached the filesize limit of FAT32.

Revision history for this message
Micah Cowan (micahcowan) wrote :

Marking this as a dupe of bug 75574, which is a bug against the kernel to get it to return an error through write() in such situations, rather than send SIGXFSZ.

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.