Bug 116471
Summary: | Core in ext4 filesystem on a Dell Server | ||
---|---|---|---|
Product: | File System | Reporter: | kparamasivam |
Component: | ext4 | Assignee: | fs_ext4 (fs_ext4) |
Status: | NEW --- | ||
Severity: | high | CC: | kernel, sandeen |
Priority: | P1 | ||
Hardware: | x86-64 | ||
OS: | Linux | ||
Kernel Version: | 3.10.0-123.20.1.el7.x86_64 | Subsystem: | |
Regression: | No | Bisected commit-id: |
Description
kparamasivam
2016-04-15 21:00:58 UTC
First of all, this looks like a rebuilt centos kernel based on RHEL7.0 - i.e a modified, old, distro kernel. So upstream bugzilla may not be the best place for it. That said, this is down a do_coredump() path, hitting BUG_ON(irqs_disabled()) We've seen similar things on xfs, also down the do_coredump() path, in an upstream 3.14 kernel (http://oss.sgi.com/archives/xfs/2014-02/msg00325.html). That launched a long thread, and nobody seemed to come to an answer of why interrupts were disabled at that point. A similar message has been reported in the CentOS bugtracker last year: https://bugs.centos.org/view.php?id=9080, which also shows ext4 involved. Wow, I even found a posting of my own: https://lkml.org/lkml/2004/5/21/242 - but back then fs/buffer.c:1270 probably meant something completely different :-) Thanks Christian and Eric. This place https://bugs.centos.org/view.php?id=9080 has no resolution for the issue. Consistently we have hit this same issue on 2 of our box. |