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-1599 Lustre <p>After building the master branch (Jul. 4/12), I wanted to run an LNet Selftest. When I enter "modprobe lnet-selftest", there is about a 10 second pause before the load completes. When I look at the logs to see why, I see this:</p> <blockquote> <p>Jul 4 13:31:50 localhost kernel: LNet: HW CPU cores: 2, npartitions: 1<br/> Jul 4 13:31:50 localhost kernel: alg: No test for crc32 (crc32-table)<br/> Jul 4 13:31:50 localhost kernel: alg: No test for adler32 (adler32-zlib)<br/> Jul 4 13:31:50 localhost kernel: alg: No test for crc32 (crc32-pclmul)<br/> Jul 4 13:31:54 localhost kernel: padlock: VIA PadLock Hash Engine not detected.<br/> Jul 4 13:31:54 localhost modprobe: FATAL: Error inserting padlock_sha (/lib/modules/2.6.32.lustre23/kernel/drivers/crypto/padlock-sha.ko): No such device<br/> Jul 4 13:31:59 localhost kernel: LNet: Added LNI 10.211.55.16@tcp <span class="error">&#91;8/256/0/180&#93;</span><br/> Jul 4 13:31:59 localhost kernel: LNet: Accept secure, port 988</p></blockquote> <p>It seems to be tripping up on loading padlock_sha.ko.</p> <p>This bug appears to have been introduced by <a href="https://jira.whamcloud.com/browse/LU-1201" title="Lustre crypto hash cleanup" class="issue-link" data-issue-key="LU-1201"><del>LU-1201</del></a>.</p> CentOS 6.2 LU-1599

Error inserting padlock_sha.ko when load lnet-selftest

Bug Major Resolved Not a Bug WC Triage Doug Oucharek Wed, 4 Jul 2012 16:39:40 +0000 Mon, 29 May 2017 04:39:41 +0000 Mon, 29 May 2017 04:39:41 +0000 Lustre 2.3.0 1 5 <p>&gt;&gt;there is about a 10 second pause before the load completes<br/> because libcfs run performance test for supported crypto hash algo. This data should be used for best default algorithm selection for osc-ost interaction.<br/> &gt;&gt;padlock: VIA PadLock Hash Engine not detected.<br/> The messages come up because you don't have hardware support for encryption.<br/> &gt;&gt;Jul 4 13:31:54 localhost modprobe: FATAL: Error inserting padlock_sha (/lib/modules/2.6.32.lustre23/kernel/drivers/crypto/padlock-sha.ko): No such device<br/> I don`t know who use sha algo for cksum. If you insmod only libcfs.ko no padlock-sha.ko error at dmesg log for my SL6.1.</p> <p>I have one assumption:<br/> when performance test starts for sha algo, kernel crypto hash api try to load all *sha.ko, including padlock-sha.ko but this is wrong if you get VIA PadLock Hash Engine not detected. May be this relate to distro problem.</p> <p>We are seeing a similar issue.</p> <p>Lustre-2.4.1<br/> RHEL-6.4</p> <p>Nov 18 07:03:31 st61 modprobe: FATAL: Error inserting crc32c_intel<br/> (/lib/modules/2.6.32-358.18.1.el6_lustre.x86_64/kernel/arch/x86/crypto/crc32c-intel.ko):<br/> No such device<br/> Nov 18 07:03:31 st61 kernel: alg: No test for crc32 (crc32-table)<br/> Nov 18 07:03:31 st61 kernel: alg: No test for adler32 (adler32-zlib)<br/> Nov 18 07:03:35 st61 kernel: padlock: VIA PadLock Hash Engine not detected.<br/> Nov 18 07:03:35 st61 modprobe: FATAL: Error inserting padlock_sha<br/> (/lib/modules/2.6.32-358.18.1.el6_lustre.x86_64/kernel/drivers/crypto/padlock-sha.ko):<br/> No such device<br/> Nov 18 07:03:39 st61 kernel: Lustre: Lustre: Build Version:<br/> 2.4.1-RC2--PRISTINE-2.6.32-358.18.1.el6_lustre.x86_64</p> <p>This happens when we are loading Lustre on MDS / OST servers.<br/> Also we are using IB for Lnet. </p> <p>Nothing bad really happens. padlock-sha.ko was not loaded cause you have no VIA PadLock Hash Engine hardware. If this fatal message confuse, you can blacklist padlock-sha module.</p> Development Rank 1|hzv3en: Rank (Obsolete) 4051 Severity