https://jira.whamcloud.com This file is an XML representation of an issue en-us 9.4.14 940014 05-12-2023 https://jira.whamcloud.com/browse/LU-15944 Lustre <p>This has been found and reported by <a href="https://jira.whamcloud.com/secure/ViewProfile.jspa?name=hornc" class="user-hover" rel="hornc">hornc</a>:</p> <p>The following sequence of commands causes a crash:</p> <div class="code panel" style="border-width: 1px;"><div class="codeContent panelContent"> <pre class="code-java"> # lnetctl peer del --prim_nid=10.1.0.60@o2ib1 # &lt;-- make sure there no record of <span class="code-keyword">this</span> peer # lnetctl udsp add --dst tcp --prio 1 # lnetctl discover 192.168.122.60@tcp</pre> </div></div> <p>The trace is as follows:</p> <div class="code panel" style="border-width: 1px;"><div class="codeContent panelContent"> <pre class="code-java"> [5449781.397300] BUG: unable to handle kernel NULL pointer dereference at 0000000000000028 [5449781.399193] IP: [&lt;ffffffffc0c36ddb&gt;] lnet_udsp_apply_rule_on_lpni+0xbb/0x7b0 [lnet] [5449781.400130] PGD 8000000055a7f067 PUD 4964e067 PMD 0  [5449781.400717] Oops: 0000 [#1] SMP  [5449781.418329] Call Trace: [5449781.419109]  [&lt;ffffffffc0c35844&gt;] lnet_udsp_apply_single_policy+0xf4/0x540 [lnet] [5449781.419881]  [&lt;ffffffffc0c35cce&gt;] lnet_udsp_apply_policies_helper.part.8+0x3e/0x70 [lnet] [5449781.420644]  [&lt;ffffffffc0c37db6&gt;] lnet_udsp_apply_policies_on_lpni+0x56/0x80 [lnet] [5449781.421386]  [&lt;ffffffffc0c36d20&gt;] ? lnet_udsp_apply_rte_rule_on_nets+0x130/0x130 [lnet] [5449781.422228]  [&lt;ffffffffc0c28231&gt;] lnet_peer_attach_peer_ni+0x161/0x600 [lnet] [5449781.422987]  [&lt;ffffffffc0c2883e&gt;] lnet_peer_ni_traffic_add+0x16e/0x2b0 [lnet] [5449781.423761]  [&lt;ffffffffc0c2de25&gt;] lnet_peerni_by_nid_locked+0xe5/0x140 [lnet] [5449781.424521]  [&lt;ffffffffc0c2df5e&gt;] lnet_nid2peerni_locked+0xde/0xf0 [lnet] [5449781.425281]  [&lt;ffffffffc0bf8713&gt;] LNetCtl+0x14d3/0x1c80 [lnet] [5449781.426061]  [&lt;ffffffffc0bf59fb&gt;] ? LNetNIInit+0x8b/0xd50 [lnet] [5449781.426818]  [&lt;ffffffffc0c18a33&gt;] lnet_ioctl+0x63/0x270 [lnet] [5449781.427581]  [&lt;ffffffff8ad90b6f&gt;] notifier_call_chain+0x4f/0x70 [5449781.428345]  [&lt;ffffffff8a6cc15d&gt;] __blocking_notifier_call_chain+0x4d/0x70 [5449781.429083]  [&lt;ffffffff8a6cc196&gt;] blocking_notifier_call_chain+0x16/0x20 [5449781.429837]  [&lt;ffffffffc0bbc3ad&gt;] libcfs_psdev_ioctl+0x43d/0x5c0 [libcfs] [5449781.430580]  [&lt;ffffffff8a863590&gt;] do_vfs_ioctl+0x3a0/0x5b0 [5449781.431319]  [&lt;ffffffff8a863841&gt;] SyS_ioctl+0xa1/0xc0 [5449781.432065]  [&lt;ffffffff8ad95f92&gt;] system_call_fastpath+0x25/0x2a </pre> </div></div> LU-15944

LNet: adding dst udsp rule before peer is discovered causes oops on peer discovery

Bug Minor Resolved Fixed Cyril Bordage Serguei Smirnov lnet udsp Tue, 14 Jun 2022 19:37:44 +0000 Wed, 8 Mar 2023 04:08:40 +0000 Wed, 8 Mar 2023 04:08:40 +0000 Lustre 2.16.0 0 5 <p>Temporary fix applied by Chris locally:</p> <div class="code panel" style="border-width: 1px;"><div class="codeContent panelContent"> <pre class="code-java"> diff --git a/lnet/lnet/udsp.c b/lnet/lnet/udsp.c index 08c1a7fccc..1f55b9289f 100644 --- a/lnet/lnet/udsp.c +++ b/lnet/lnet/udsp.c @@ -536,6 +536,9 @@ lnet_udsp_apply_rule_on_lpni(struct udsp_info *udi)          &amp;lp_match-&gt;ud_net_id.udn_net_num_range,          &amp;lp_match-&gt;ud_addr_range); +    <span class="code-keyword">if</span> (!udi-&gt;udi_lpn) +        udi-&gt;udi_lpn = lpni-&gt;lpni_peer_net; +      <span class="code-comment">/* check <span class="code-keyword">if</span> looking <span class="code-keyword">for</span> a net match */</span>      <span class="code-keyword">if</span> (!rc &amp;&amp;          (lnet_get_list_len(&amp;lp_match-&gt;ud_addr_range) ||</pre> </div></div> <p>This prevents the crash, but causes nid priority to be inherited from the previously set net priority for the peer.</p> <p>"Cyril Bordage &lt;cbordage@whamcloud.com&gt;" uploaded a new patch: <a href="https://review.whamcloud.com/c/fs/lustre-release/+/48801" class="external-link" target="_blank" rel="nofollow noopener">https://review.whamcloud.com/c/fs/lustre-release/+/48801</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-15944" title="LNet: adding dst udsp rule before peer is discovered causes oops on peer discovery" class="issue-link" data-issue-key="LU-15944"><del>LU-15944</del></a> lnet: remove crash with UDSP<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: 1<br/> Commit: cea355d4837c945a3f0193fb92331abb65c13d5c</p> <p>"Oleg Drokin &lt;green@whamcloud.com&gt;" merged in patch <a href="https://review.whamcloud.com/c/fs/lustre-release/+/48801/" class="external-link" target="_blank" rel="nofollow noopener">https://review.whamcloud.com/c/fs/lustre-release/+/48801/</a><br/> Subject: <a href="https://jira.whamcloud.com/browse/LU-15944" title="LNet: adding dst udsp rule before peer is discovered causes oops on peer discovery" class="issue-link" data-issue-key="LU-15944"><del>LU-15944</del></a> lnet: remove crash with UDSP<br/> Project: fs/lustre-release<br/> Branch: master<br/> Current Patch Set: <br/> Commit: c56b9455f05f760aea6785c47061761bbc76f3b6</p> <p>Landed for 2.16</p> Development Rank 1|i02s47: Rank (Obsolete) 9223372036854775807 Severity