<!-- 
RSS generated by JIRA (9.4.14#940014-sha1:734e6822bbf0d45eff9af51f82432957f73aa32c) at Sat Feb 10 02:45:31 UTC 2024

It is possible to restrict the fields that are returned in this document by specifying the 'field' parameter in your request.
For example, to request only the issue key and summary append 'field=key&field=summary' to the URL of your request.
-->
<rss version="0.92" >
<channel>
    <title>Whamcloud Community JIRA</title>
    <link>https://jira.whamcloud.com</link>
    <description>This file is an XML representation of an issue</description>
    <language>en-us</language>    <build-info>
        <version>9.4.14</version>
        <build-number>940014</build-number>
        <build-date>05-12-2023</build-date>
    </build-info>


<item>
            <title>[LU-11626] mdc: obd might go away while referenced by code in mdc_changelog</title>
                <link>https://jira.whamcloud.com/browse/LU-11626</link>
                <project id="10000" key="LU">Lustre</project>
                    <description>&lt;p&gt;In &lt;tt&gt;lustre/mdc/mdc_changelog.c&lt;/tt&gt;:&lt;/p&gt;

&lt;div class=&quot;preformatted panel&quot; style=&quot;border-width: 1px;&quot;&gt;&lt;div class=&quot;preformattedContent panelContent&quot;&gt;
&lt;pre&gt;/**
 * Find the OBD device associated to a changelog character device.
 * @param[in]  cdev  character device instance descriptor
 * @return corresponding OBD device or NULL if none was found.
 */
static struct obd_device *chlg_obd_get(dev_t cdev)
{
	int minor = MINOR(cdev);
	struct obd_device *obd = NULL;
	struct chlg_registered_dev *curr;

	mutex_lock(&amp;amp;chlg_registered_dev_lock);
	list_for_each_entry(curr, &amp;amp;chlg_registered_devices, ced_link) {
		if (curr-&amp;gt;ced_misc.minor == minor) {
			/* take the first available OBD device attached */
			obd = list_first_entry(&amp;amp;curr-&amp;gt;ced_obds,
					       struct obd_device,
					       u.cli.cl_chg_dev_linkage);
			break;
		}
	}
	mutex_unlock(&amp;amp;chlg_registered_dev_lock);
	return obd;
}
&lt;/pre&gt;
&lt;/div&gt;&lt;/div&gt;

&lt;p&gt;The &quot;take the first available OBD device attached&quot; approach is broken as that OBD might go away while the changelog is being read.&lt;/p&gt;

&lt;p&gt;Here is an example of how things can go wrong:&lt;/p&gt;
&lt;ul class=&quot;alternate&quot; type=&quot;square&quot;&gt;
	&lt;li&gt;mount lustre at /mnt/lustre;&lt;/li&gt;
	&lt;li&gt;mount the same lustre filesystem at /mnt/lustre2;&lt;/li&gt;
	&lt;li&gt;open /dev/changelog-lustre-MDT0000 in writing mode;&lt;/li&gt;
	&lt;li&gt;unmount /mnt/lustre;&lt;/li&gt;
	&lt;li&gt;write &apos;clear:cl0:0&apos; to the opened chardevice.&lt;/li&gt;
&lt;/ul&gt;


&lt;p&gt;That triggers an LBUG.&lt;/p&gt;</description>
                <environment></environment>
        <key id="53930">LU-11626</key>
            <summary>mdc: obd might go away while referenced by code in mdc_changelog</summary>
                <type id="1" iconUrl="https://jira.whamcloud.com/secure/viewavatar?size=xsmall&amp;avatarId=11303&amp;avatarType=issuetype">Bug</type>
                                            <priority id="4" iconUrl="https://jira.whamcloud.com/images/icons/priorities/minor.svg">Minor</priority>
                        <status id="5" iconUrl="https://jira.whamcloud.com/images/icons/statuses/resolved.png" description="A resolution has been taken, and it is awaiting verification by reporter. From here issues are either reopened, or are closed.">Resolved</status>
                    <statusCategory id="3" key="done" colorName="success"/>
                                    <resolution id="1">Fixed</resolution>
                                        <assignee username="hongchao.zhang">Hongchao Zhang</assignee>
                                    <reporter username="cealustre">CEA</reporter>
                        <labels>
                    </labels>
                <created>Tue, 6 Nov 2018 10:40:33 +0000</created>
                <updated>Wed, 5 Aug 2020 13:50:22 +0000</updated>
                            <resolved>Mon, 16 Sep 2019 23:21:52 +0000</resolved>
                                                    <fixVersion>Lustre 2.13.0</fixVersion>
                    <fixVersion>Lustre 2.12.4</fixVersion>
                                        <due></due>
                            <votes>0</votes>
                                    <watches>9</watches>
                                                                            <comments>
                            <comment id="236478" author="pjones" created="Tue, 6 Nov 2018 18:36:55 +0000"  >&lt;p&gt;Quentin&lt;/p&gt;

&lt;p&gt;Is this something that you plan to work on a fix for?&lt;/p&gt;

&lt;p&gt;Peter&lt;/p&gt;</comment>
                            <comment id="236535" author="bougetq" created="Wed, 7 Nov 2018 10:21:11 +0000"  >&lt;p&gt;I would rather not. Neil Brown is working on it, I am trying to help, but I am not available enough to try and write a whole fix for it.&lt;/p&gt;

&lt;p&gt;The discussion is also happening on lustre-devel, anyone interested is welcome to chime in. I&apos;ll report here if we reach some kind of consensus on how to fix this.&lt;/p&gt;</comment>
                            <comment id="236536" author="bougetq" created="Wed, 7 Nov 2018 10:21:32 +0000"  >&lt;p&gt;I think someone familiar with the internals of llogs could help quite a lot here.&lt;/p&gt;</comment>
                            <comment id="236549" author="pjones" created="Wed, 7 Nov 2018 13:41:08 +0000"  >&lt;p&gt;&lt;a href=&quot;https://jira.whamcloud.com/secure/ViewProfile.jspa?name=simmonsja&quot; class=&quot;user-hover&quot; rel=&quot;simmonsja&quot;&gt;simmonsja&lt;/a&gt; is this something that you are monitoring along with other upstream tasks?&lt;/p&gt;</comment>
                            <comment id="236552" author="simmonsja" created="Wed, 7 Nov 2018 14:01:50 +0000"  >&lt;p&gt;Yes I&apos;m involved.&lt;/p&gt;</comment>
                            <comment id="251729" author="simmonsja" created="Fri, 19 Jul 2019 19:43:25 +0000"  >&lt;p&gt;The one change log device to N mdc obd devices seems strange to me. If you have multiple mounts on a client of the same file system do each obd device point to the same obd_import?&lt;/p&gt;</comment>
                            <comment id="252410" author="adilger" created="Thu, 1 Aug 2019 19:36:41 +0000"  >&lt;blockquote&gt;
&lt;p&gt;The one change log device to N mdc obd devices seems strange to me. If you have multiple mounts on a client of the same file system do each obd device point to the same obd_import?&lt;/p&gt;&lt;/blockquote&gt;

&lt;p&gt;There is already a problem in &lt;a href=&quot;https://jira.whamcloud.com/browse/LU-12506&quot; title=&quot;Client unable to mount filesystem with very large number of MDTs&quot; class=&quot;issue-link&quot; data-issue-key=&quot;LU-12506&quot;&gt;&lt;del&gt;LU-12506&lt;/del&gt;&lt;/a&gt; with the client trying to mount a filesystem with a lot of MDTs, because it is trying to create a misc character device for each MDT.  That is limited to 64 misc devices per client, so if we create a separate character device for each mount it would cause mount failures with half the number of MDTs in half or fewer.  &lt;/p&gt;

&lt;p&gt;It makes more sense to multiplex a single character device across multiple MDTs, named &lt;tt&gt;changelog-$fsname&lt;/tt&gt;.  To track the MDT index on the open file handle (default = 0, which will work for most filesystems without any change) add an &lt;tt&gt;ioctl()&lt;/tt&gt; to change the MDT index for that file handle if needed.&lt;/p&gt;

&lt;p&gt;That avoids the need to create so many character devices, avoids the need to share a single &lt;tt&gt;chlg_registered_dev&lt;/tt&gt; between multiple OBDs (one for each opener), and this interface change can be encapsulated inside the llapi code.  This will also avoid the complexity in &lt;tt&gt;chlg_registered_dev_find_by_obd()&lt;/tt&gt; if we have only a single &lt;tt&gt;chlg_registered_dev&lt;/tt&gt; per OBD.&lt;/p&gt;

&lt;p&gt;There would need to be some small changes to &lt;tt&gt;liblustreapi_chlg.c&lt;/tt&gt; to open the &lt;tt&gt;changelog-$fsname&lt;/tt&gt; device and call the &lt;tt&gt;ioctl()&lt;/tt&gt; to change the MDT index instead of opening a different device for each MDT, with a fallback to the old behavior if the new device name doesn&apos;t exist.  Probably the best is to change &lt;tt&gt;chlg_dev_path()&lt;/tt&gt; to &lt;tt&gt;chlg_dev_open()&lt;/tt&gt; and return the open file handle or an error instead of the pathname.&lt;/p&gt;

&lt;p&gt;On the kernel side in &lt;tt&gt;mdc_changelog_cdev_init()&lt;/tt&gt;, we should consider still creating some limited number of &lt;tt&gt;changelog-$fsname-MDTnnnn&lt;/tt&gt; devices (maybe max 16?) for compatibility with userspace applications/libraries that are opening the old devices and are statically linked to &lt;tt&gt;liblustreapi.a&lt;/tt&gt; (under &lt;tt&gt;LUSTRE_VERSION_CODE&lt;/tt&gt; checks so they go away eventually).  However, it shouldn&apos;t be an error if the &lt;b&gt;compat&lt;/b&gt; devices cannot be created if there are many MDTs, since most clients will not be changelog consumers.&lt;/p&gt;</comment>
                            <comment id="252411" author="adilger" created="Thu, 1 Aug 2019 20:01:13 +0000"  >&lt;p&gt;&lt;a href=&quot;https://jira.whamcloud.com/secure/ViewProfile.jspa?name=bougetq&quot; class=&quot;user-hover&quot; rel=&quot;bougetq&quot;&gt;bougetq&lt;/a&gt;,&lt;br/&gt;
I see that in patch &lt;a href=&quot;https://review.whamcloud.com/18900&quot; class=&quot;external-link&quot; target=&quot;_blank&quot; rel=&quot;nofollow noopener&quot;&gt;https://review.whamcloud.com/18900&lt;/a&gt; &quot;&lt;a href=&quot;https://jira.whamcloud.com/browse/LU-7659&quot; title=&quot;Replace KUC by more standard mechanisms&quot; class=&quot;issue-link&quot; data-issue-key=&quot;LU-7659&quot;&gt;LU-7659&lt;/a&gt; mdc: expose changelog through char devices&quot; one of the motivations is &quot;... to allow non-llapi to read them and to make delivery more efficient.&quot;  Does RobinHood open and access the &lt;tt&gt;changelog-$fsname-MDTxxxx&lt;/tt&gt; devices directly, or does it still use &lt;tt&gt;llapi_changelog_start()&lt;/tt&gt; to open the char device?&lt;/p&gt;</comment>
                            <comment id="252413" author="adilger" created="Thu, 1 Aug 2019 20:21:23 +0000"  >&lt;p&gt;Hongchao, assigning this to you since I think the fix will also resolve &lt;a href=&quot;https://jira.whamcloud.com/browse/LU-12506&quot; title=&quot;Client unable to mount filesystem with very large number of MDTs&quot; class=&quot;issue-link&quot; data-issue-key=&quot;LU-12506&quot;&gt;&lt;del&gt;LU-12506&lt;/del&gt;&lt;/a&gt;.&lt;/p&gt;</comment>
                            <comment id="252422" author="adilger" created="Thu, 1 Aug 2019 22:53:02 +0000"  >&lt;p&gt;Actually, having looked at this code in more detail, it seems that the obvious first step would be to either get a device reference between &lt;tt&gt;chlg_open()&lt;/tt&gt; and &lt;tt&gt;chlg_close()&lt;/tt&gt;, or a device reference each &lt;tt&gt;chlg_load()&lt;/tt&gt; and &lt;tt&gt;chlg_write()&lt;/tt&gt;.&lt;/p&gt;

&lt;p&gt;For both cases:&lt;/p&gt;
&lt;ul class=&quot;alternate&quot; type=&quot;square&quot;&gt;
	&lt;li&gt;have &lt;tt&gt;chlg_obd_get()&lt;/tt&gt; actually return a reference &lt;tt&gt;class_incref(obd)&lt;/tt&gt; so that it is not a dangling pointer to a structure that might be freed.  There should be a &lt;tt&gt;chlg_obd_put()&lt;/tt&gt; that drops the reference using &lt;tt&gt;class_decref(crs_obd)&lt;/tt&gt;.&lt;/li&gt;
&lt;/ul&gt;


&lt;p&gt;For open-to-close reference:&lt;/p&gt;
&lt;ul class=&quot;alternate&quot; type=&quot;square&quot;&gt;
	&lt;li&gt;the reference on &lt;tt&gt;crs_obd&lt;/tt&gt; should be gotten in &lt;tt&gt;chlg_open()&lt;/tt&gt; and dropped in &lt;tt&gt;chlg_release()&lt;/tt&gt;.  One drawback is that an open Changelog reader would leave a danging OBD device configured when the client is unmounted, though &quot;&lt;tt&gt;lsof&lt;/tt&gt;&quot; would at least show the &quot;&lt;tt&gt;/dev/changelog-$fsname-MDTxxxx&lt;/tt&gt;&quot; device as open, and hopefully the cleanup would still work properly when the device is closed after the client is unmounted.&lt;/li&gt;
&lt;/ul&gt;


&lt;p&gt;For read/write references:&lt;/p&gt;
&lt;ul class=&quot;alternate&quot; type=&quot;square&quot;&gt;
	&lt;li&gt;&lt;tt&gt;obd_device&lt;/tt&gt; would be looked up each time via &lt;tt&gt;chlg_obd_get()&lt;/tt&gt;, by saving &lt;tt&gt;crs_rdev = i_rdev&lt;/tt&gt; in &lt;tt&gt;chlg_open()&lt;/tt&gt; instead of &lt;tt&gt;crs_obd&lt;/tt&gt;, maybe after validating that the device actually exists.  This adds some overhead, but may allow the Changelog code to be more robust, as the client could be unmounted/mounted cleanly if the changelog consumer is not actively using the device.&lt;/li&gt;
	&lt;li&gt;&lt;tt&gt;chlg_write-&amp;gt;chlg_clear()&lt;/tt&gt; is not as performance critical as the &lt;tt&gt;chlg_read()&lt;/tt&gt; path, so doing a lookup each time is probably fine.&lt;/li&gt;
	&lt;li&gt;&lt;tt&gt;chlg_read()&lt;/tt&gt; depends on &lt;tt&gt;chlg_load()&lt;/tt&gt; in the background, which uses &lt;tt&gt;crs_obd&lt;/tt&gt; to lookup &lt;tt&gt;llog_ctxt&lt;/tt&gt;, which is tied to a specific OBD device.  Storing a direct reference &lt;tt&gt;crs_llog_ctxt&lt;/tt&gt; would avoid the &lt;tt&gt;crs_obd&lt;/tt&gt; dereference, but this would still prevent the OBD device to be cleaned up, so is not better than getting a reference on &lt;tt&gt;obd_device&lt;/tt&gt; directly in &lt;tt&gt;chlg_open()&lt;/tt&gt;.  Since &lt;tt&gt;chlg_load()&lt;/tt&gt; is already getting a loop around the llog operations in patch &lt;a href=&quot;https://review.whamcloud.com/35262&quot; class=&quot;external-link&quot; target=&quot;_blank&quot; rel=&quot;nofollow noopener&quot;&gt;https://review.whamcloud.com/35262&lt;/a&gt; &quot;&lt;tt&gt;&lt;a href=&quot;https://jira.whamcloud.com/browse/LU-12553&quot; title=&quot;polling mode for changelog driver&quot; class=&quot;issue-link&quot; data-issue-key=&quot;LU-12553&quot;&gt;&lt;del&gt;LU-12553&lt;/del&gt;&lt;/a&gt;  mdc: polling mode for changelog reader&lt;/tt&gt;&quot; it isn&apos;t much harder to expand that loop a few more lines and include &lt;tt&gt;chlg_obd_get()&lt;/tt&gt; and &lt;tt&gt;chlg_obd_put()&lt;/tt&gt; on each loop.&lt;/li&gt;
&lt;/ul&gt;
</comment>
                            <comment id="252539" author="bougetq" created="Mon, 5 Aug 2019 16:04:03 +0000"  >&lt;blockquote&gt;&lt;p&gt;Does RobinHood open and access the changelog-$fsname-MDTxxxx devices directly, or does it still use llapi_changelog_start() to open the char device?&lt;/p&gt;&lt;/blockquote&gt;

&lt;p&gt;It uses &lt;tt&gt;llapi_changelog_start()&lt;/tt&gt;.&lt;/p&gt;</comment>
                            <comment id="252728" author="adilger" created="Thu, 8 Aug 2019 08:49:22 +0000"  >&lt;p&gt;It might make sense to also allow writing to the open file descriptor like &quot;&lt;tt&gt;mdt:mdt_idx&lt;/tt&gt;&quot; to set the specific MDT to be used.  That allows userspace scripts to potentially read and manage the changelog file, instead of having to find some way to call &lt;tt&gt;ioctl()&lt;/tt&gt;.&lt;/p&gt;</comment>
                            <comment id="252988" author="gerrit" created="Tue, 13 Aug 2019 10:58:14 +0000"  >&lt;p&gt;Hongchao Zhang (hongchao@whamcloud.com) uploaded a new patch: &lt;a href=&quot;https://review.whamcloud.com/35784&quot; class=&quot;external-link&quot; target=&quot;_blank&quot; rel=&quot;nofollow noopener&quot;&gt;https://review.whamcloud.com/35784&lt;/a&gt;&lt;br/&gt;
Subject: &lt;a href=&quot;https://jira.whamcloud.com/browse/LU-11626&quot; title=&quot;mdc: obd might go away while referenced by code in mdc_changelog&quot; class=&quot;issue-link&quot; data-issue-key=&quot;LU-11626&quot;&gt;&lt;del&gt;LU-11626&lt;/del&gt;&lt;/a&gt; mdc: hold obd while processing changelog&lt;br/&gt;
Project: fs/lustre-release&lt;br/&gt;
Branch: master&lt;br/&gt;
Current Patch Set: 1&lt;br/&gt;
Commit: f866eac40cc761b960cc92907d1d5546b6e24e38&lt;/p&gt;</comment>
                            <comment id="253114" author="simmonsja" created="Thu, 15 Aug 2019 13:58:33 +0000"  >&lt;p&gt;Patch&#160;&lt;/p&gt;

&lt;p&gt;&lt;a href=&quot;https://review.whamcloud.com/35668&quot; class=&quot;external-link&quot; target=&quot;_blank&quot; rel=&quot;nofollow noopener&quot;&gt;https://review.whamcloud.com/35668&lt;/a&gt; which landed resolved this. It would be nice to push a reproducer to the test suite from the information in this ticket. That was missing in the patch.&lt;/p&gt;</comment>
                            <comment id="253303" author="hongchao.zhang" created="Tue, 20 Aug 2019 02:18:45 +0000"  >&lt;p&gt;Hi James,&lt;br/&gt;
The issue still exists on master branch with the patch &lt;a href=&quot;https://review.whamcloud.com/35668&quot; class=&quot;external-link&quot; target=&quot;_blank&quot; rel=&quot;nofollow noopener&quot;&gt;https://review.whamcloud.com/35668&lt;/a&gt;.&lt;/p&gt;
&lt;div class=&quot;preformatted panel&quot; style=&quot;border-width: 1px;&quot;&gt;&lt;div class=&quot;preformattedContent panelContent&quot;&gt;
&lt;pre&gt;[  106.920737] general protection fault: 0000 [#1] SMP 
[  106.927417] Modules linked in: lustre(OE) ofd(OE) osp(OE) lod(OE) ost(OE) mdt(OE) mdd(OE) mgs(OE) osd_zfs(OE) zfs(PO) zunicode(PO) zavl(PO) icp(PO) zcommon(PO) znvpair(PO) spl(O) lquota(OE) lfsck(OE) obdecho(OE) mgc(OE) lov(OE) mdc(OE) osc(OE) lmv(OE) fid(OE) fld(OE) ptlrpc(OE) obdclass(OE) ksocklnd(OE) lnet(OE) libcfs(OE) intel_powerclamp iosf_mbi crc32_pclmul ghash_clmulni_intel ppdev aesni_intel lrw gf128mul glue_helper ablk_helper cryptd sg parport_pc pcspkr parport i2c_piix4 video i2c_core ip_tables xfs libcrc32c sr_mod cdrom ata_generic sd_mod crc_t10dif crct10dif_generic pata_acpi ahci libahci ata_piix libata e1000 crct10dif_pclmul crct10dif_common crc32c_intel serio_raw dm_mirror dm_region_hash dm_log dm_mod
[  106.941934] CPU: 0 PID: 1287 Comm: bash Kdump: loaded Tainted: P           OE  ------------   3.10.0-862.9.1.el7.x86_64 #1
[  106.944006] Hardware name: innotek GmbH VirtualBox/VirtualBox, BIOS VirtualBox 12/01/2006
[  106.945492] task: ffff8f61898b5ee0 ti: ffff8f61803e4000 task.ti: ffff8f61803e4000
[  106.946817] RIP: 0010:[&amp;lt;ffffffffc09ed510&amp;gt;]  [&amp;lt;ffffffffc09ed510&amp;gt;] chlg_write+0x450/0x6c0 [mdc]
[  106.948338] RSP: 0018:ffff8f61803e7e70  EFLAGS: 00010206
[  106.949300] RAX: ffff8f6047e741e0 RBX: ffff8f6171eeac00 RCX: 0000000000000000
[  106.950582] RDX: 0000000000000000 RSI: ffffffffc09f8d74 RDI: 0000000000000000
[  106.951766] RBP: ffff8f61803e7ec0 R08: ffff8f6171eeac0a R09: 0000000000000000
[  106.953063] R10: 000000000000000a R11: f000000000000000 R12: 000000000000000b
[  106.954279] R13: 00007f9075b49000 R14: 5a5a5a5a5a5a5a5a R15: 0000000000000000
[  106.955473] FS:  00007f9075b43740(0000) GS:ffff8f619fc00000(0000) knlGS:0000000000000000
[  106.956802] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[  106.957723] CR2: 00005619854aa528 CR3: 0000000216e52000 CR4: 00000000000606f0
[  106.958892] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[  106.960067] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[  106.961216] Call Trace:
[  106.961613]  [&amp;lt;ffffffffac41b490&amp;gt;] vfs_write+0xc0/0x1f0
[  106.962440]  [&amp;lt;ffffffffac9206e1&amp;gt;] ? system_call_after_swapgs+0xae/0x146
[  106.963499]  [&amp;lt;ffffffffac41c2bf&amp;gt;] SyS_write+0x7f/0xf0
[  106.964300]  [&amp;lt;ffffffffac9206e1&amp;gt;] ? system_call_after_swapgs+0xae/0x146
[  106.965349]  [&amp;lt;ffffffffac920795&amp;gt;] system_call_fastpath+0x1c/0x21
[  106.966311]  [&amp;lt;ffffffffac9206e1&amp;gt;] ? system_call_after_swapgs+0xae/0x146
[  106.967362] Code: 27 7e 01 00 ef 01 00 00 c7 05 21 7e 01 00 01 00 00 00 48 c7 05 1e 7e 01 00 00 00 00 00 e8 89 d2 90 ff 4d 85 f6 0f 84 33 01 00 00 &amp;lt;49&amp;gt; 8b 86 f8 00 00 00 48 85 c0 0f 84 a3 00 00 00 48 8b 10 48 85 
[  106.970783] RIP  [&amp;lt;ffffffffc09ed510&amp;gt;] chlg_write+0x450/0x6c0 [mdc]
[  106.971791]  RSP &amp;lt;ffff8f61803e7e70&amp;gt;
[    0.794662] mce: Unable to init device /dev/mcelog (rc: -5)
[    0.800919] Failed to execute /init
[    0.801546] Kernel panic - not syncing: No init found.  Try passing init= option to kernel. See Linux Documentation/init.txt for guidance.
[    0.803450] CPU: 0 PID: 1 Comm: swapper/0 Not tainted 3.10.0-862.9.1.el7.x86_64 #1
[    0.804651] Hardware name: innotek GmbH VirtualBox/VirtualBox, BIOS VirtualBox 12/01/2006
[    0.805965] Call Trace:
[    0.806399]  [&amp;lt;ffffffff8170e84e&amp;gt;] dump_stack+0x19/0x1b
[    0.807277]  [&amp;lt;ffffffff81708b50&amp;gt;] panic+0xe8/0x21f
[    0.808041]  [&amp;lt;ffffffff8122ccdd&amp;gt;] ? putname+0x3d/0x60
[    0.808799]  [&amp;lt;ffffffff816fd780&amp;gt;] ? rest_init+0x80/0x80
[    0.809627]  [&amp;lt;ffffffff816fd862&amp;gt;] kernel_init+0xe2/0xf0
[    0.810741]  [&amp;lt;ffffffff817205f7&amp;gt;] ret_from_fork_nospec_begin+0x21/0x21
[    0.811744]  [&amp;lt;ffffffff816fd780&amp;gt;] ? rest_init+0x80/0x80
&lt;/pre&gt;
&lt;/div&gt;&lt;/div&gt;</comment>
                            <comment id="254811" author="gerrit" created="Mon, 16 Sep 2019 23:04:47 +0000"  >&lt;p&gt;Oleg Drokin (green@whamcloud.com) merged in patch &lt;a href=&quot;https://review.whamcloud.com/35784/&quot; class=&quot;external-link&quot; target=&quot;_blank&quot; rel=&quot;nofollow noopener&quot;&gt;https://review.whamcloud.com/35784/&lt;/a&gt;&lt;br/&gt;
Subject: &lt;a href=&quot;https://jira.whamcloud.com/browse/LU-11626&quot; title=&quot;mdc: obd might go away while referenced by code in mdc_changelog&quot; class=&quot;issue-link&quot; data-issue-key=&quot;LU-11626&quot;&gt;&lt;del&gt;LU-11626&lt;/del&gt;&lt;/a&gt; mdc: hold obd while processing changelog&lt;br/&gt;
Project: fs/lustre-release&lt;br/&gt;
Branch: master&lt;br/&gt;
Current Patch Set: &lt;br/&gt;
Commit: d7bb6647cd4dd26949bceb6a099cd606623aff2b&lt;/p&gt;</comment>
                            <comment id="254822" author="pjones" created="Mon, 16 Sep 2019 23:21:52 +0000"  >&lt;p&gt;Landed for 2.13&lt;/p&gt;</comment>
                            <comment id="255711" author="gerrit" created="Tue, 1 Oct 2019 14:09:32 +0000"  >&lt;p&gt;Minh Diep (mdiep@whamcloud.com) uploaded a new patch: &lt;a href=&quot;https://review.whamcloud.com/36338&quot; class=&quot;external-link&quot; target=&quot;_blank&quot; rel=&quot;nofollow noopener&quot;&gt;https://review.whamcloud.com/36338&lt;/a&gt;&lt;br/&gt;
Subject: &lt;a href=&quot;https://jira.whamcloud.com/browse/LU-11626&quot; title=&quot;mdc: obd might go away while referenced by code in mdc_changelog&quot; class=&quot;issue-link&quot; data-issue-key=&quot;LU-11626&quot;&gt;&lt;del&gt;LU-11626&lt;/del&gt;&lt;/a&gt; mdc: hold obd while processing changelog&lt;br/&gt;
Project: fs/lustre-release&lt;br/&gt;
Branch: b2_12&lt;br/&gt;
Current Patch Set: 1&lt;br/&gt;
Commit: 24935d83103d44dee84a16a7e304bb13180ec22d&lt;/p&gt;</comment>
                            <comment id="256074" author="simmonsja" created="Tue, 8 Oct 2019 17:06:53 +0000"  >&lt;p&gt;Peter, any chance this can land for the 2.12.3 release &lt;img class=&quot;emoticon&quot; src=&quot;https://jira.whamcloud.com/images/icons/emoticons/smile.png&quot; height=&quot;16&quot; width=&quot;16&quot; align=&quot;absmiddle&quot; alt=&quot;&quot; border=&quot;0&quot;/&gt;&lt;/p&gt;</comment>
                            <comment id="256087" author="pjones" created="Tue, 8 Oct 2019 17:51:11 +0000"  >&lt;p&gt;James&lt;/p&gt;

&lt;p&gt;We tagged RC1 this morning so, while there&apos;s still a chance, it is looking unlikely as things stand&lt;/p&gt;

&lt;p&gt;Peter&lt;/p&gt;</comment>
                            <comment id="258607" author="gerrit" created="Thu, 21 Nov 2019 07:33:41 +0000"  >&lt;p&gt;Oleg Drokin (green@whamcloud.com) merged in patch &lt;a href=&quot;https://review.whamcloud.com/36338/&quot; class=&quot;external-link&quot; target=&quot;_blank&quot; rel=&quot;nofollow noopener&quot;&gt;https://review.whamcloud.com/36338/&lt;/a&gt;&lt;br/&gt;
Subject: &lt;a href=&quot;https://jira.whamcloud.com/browse/LU-11626&quot; title=&quot;mdc: obd might go away while referenced by code in mdc_changelog&quot; class=&quot;issue-link&quot; data-issue-key=&quot;LU-11626&quot;&gt;&lt;del&gt;LU-11626&lt;/del&gt;&lt;/a&gt; mdc: hold obd while processing changelog&lt;br/&gt;
Project: fs/lustre-release&lt;br/&gt;
Branch: b2_12&lt;br/&gt;
Current Patch Set: &lt;br/&gt;
Commit: 9d88030b8a532de3bec3f75d01ec108bfba97082&lt;/p&gt;</comment>
                    </comments>
                <issuelinks>
                            <issuelinktype id="10010">
                    <name>Duplicate</name>
                                                                <inwardlinks description="is duplicated by">
                                        <issuelink>
            <issuekey id="56446">LU-12566</issuekey>
        </issuelink>
                            </inwardlinks>
                                    </issuelinktype>
                            <issuelinktype id="10011">
                    <name>Related</name>
                                            <outwardlinks description="is related to ">
                                        <issuelink>
            <issuekey id="56268">LU-12506</issuekey>
        </issuelink>
            <issuelink>
            <issuekey id="34089">LU-7659</issuekey>
        </issuelink>
                            </outwardlinks>
                                                        </issuelinktype>
                    </issuelinks>
                <attachments>
                            <attachment id="31419" name="reproducer.bash" size="857" author="bougetq" created="Tue, 6 Nov 2018 10:37:39 +0000"/>
                    </attachments>
                <subtasks>
                    </subtasks>
                <customfields>
                                                                                                                                                                                            <customfield id="customfield_10890" key="com.atlassian.jira.plugins.jira-development-integration-plugin:devsummary">
                        <customfieldname>Development</customfieldname>
                        <customfieldvalues>
                            
                        </customfieldvalues>
                    </customfield>
                                                                                                                                                                                                                                                                                                                                                        <customfield id="customfield_10390" key="com.pyxis.greenhopper.jira:gh-lexo-rank">
                        <customfieldname>Rank</customfieldname>
                        <customfieldvalues>
                            <customfieldvalue>1|i005sn:</customfieldvalue>

                        </customfieldvalues>
                    </customfield>
                                                                <customfield id="customfield_10090" key="com.pyxis.greenhopper.jira:gh-global-rank">
                        <customfieldname>Rank (Obsolete)</customfieldname>
                        <customfieldvalues>
                            <customfieldvalue>9223372036854775807</customfieldvalue>
                        </customfieldvalues>
                    </customfield>
                                                                                            <customfield id="customfield_10060" key="com.atlassian.jira.plugin.system.customfieldtypes:select">
                        <customfieldname>Severity</customfieldname>
                        <customfieldvalues>
                                <customfieldvalue key="10022"><![CDATA[3]]></customfieldvalue>

                        </customfieldvalues>
                    </customfield>
                                                                                                                                                                                                                                                                                                                                                        </customfields>
    </item>
</channel>
</rss>