Project

General

Profile

Bug #10004

Segmentation fault in /usr/sbin/format when performing verify on disk larger than 2Tb

Added by Granville Moore about 1 year ago.

Status:
New
Priority:
Normal
Assignee:
-
Category:
-
Start date:
2018-11-22
Due date:
% Done:

0%

Estimated time:
Difficulty:
Medium
Tags:
needs-triage

Description

On OpenIndiana Hipster 201810, /usr/sbin/format segfaults when performing an analyze>verify operation on a disk larger than 2Tb.

The operation proceeds as expected, writing a test pattern, until the 2Tb point is reached, at which point a segmentation error occurs (after about 10 hrs runtime).

System is an HP Microserver G8, disk is an external USB3 8Tb drive: <Seagate-Backup+ Hub BK-D781-7.28TB>.

Before the test, the disk was completely zeroed (i.e. no data, no disk labels).
After the segfault, the disk was examined and the first 2Tb only were found to have been written with test patterns. Above 2Tb, the disk was still zeroed.

Output from "pstack core" and output from "truss format" are attached. Truss output shows start of verify operation only, not the segrault itself.


Files

pstack core.txt (1.56 KB) pstack core.txt Output from "pstack core" Granville Moore, 2018-11-22 02:42 PM
truss output during verification startup.txt (4.36 KB) truss output during verification startup.txt Output from "truss format" (start of verify operation only) Granville Moore, 2018-11-22 02:46 PM

Also available in: Atom PDF