<div dir="ltr"><div><div><div><div>In the end I used kernel 4.2.x since we don't have time or expertise to permit us to take the risk of editing the lustre source.<br></div>I believe HEAD is actually compatible with newer kernels since someone else on the list recently reported installing it on Ubuntu 16.04 which is kernel 4.4.x<br><br></div>Only one small thing needed to be changed for kernel 4.2.x though I expect it was already done in HEAD I am still including it below.<br><br></div>Regards,<br></div>Eli<br><div><div><br>commit ea0e7ce4f5031de4dfa8598a3185dca40ca175c5<br>Author: Keeper-of-the-Keys <<a href="mailto:esr%2Blustre@mail.hebrew.edu">esr+lustre@mail.hebrew.edu</a>><br>Date: Tue May 31 20:43:22 2016 +0300<br><br> asm/i387.h was renamed fpu/api.h as described here:<br> <a href="https://github.com/torvalds/linux/commit/df6b35f409af0a8ff1ef62f552b8402f3fef8665">https://github.com/torvalds/linux/commit/df6b35f409af0a8ff1ef62f552b8402f3fef8665</a><br><br>diff --git a/libcfs/libcfs/linux/linux-crypto-crc32c-pclmul.c b/libcfs/libcfs/linux/linux-crypto-crc32c-pclmul.c<br>index 9858db4..3c707bf 100644<br>--- a/libcfs/libcfs/linux/linux-crypto-crc32c-pclmul.c<br>+++ b/libcfs/libcfs/linux/linux-crypto-crc32c-pclmul.c<br>@@ -30,7 +30,7 @@<br> #include <crypto/internal/hash.h><br> #include <linux/crc32.h><br> #include <asm/cpufeature.h><br>-#include <asm/i387.h><br>+#include <asm/fpu/api.h><br> #include <libcfs/libcfs.h><br> <br> #define CHKSUM_BLOCK_SIZE 1<br><br></div></div></div><div class="gmail_extra"><br><div class="gmail_quote">On Mon, May 23, 2016 at 5:41 AM, Drokin, Oleg <span dir="ltr"><<a href="mailto:oleg.drokin@intel.com" target="_blank">oleg.drokin@intel.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Hello!<br>
<span class=""><br>
On May 22, 2016, at 2:38 PM, E.S. Rosenberg wrote:<br>
<br>
> Internal functions have changed and as a result it is currently not compiling, what is the accepted style for fixing these things?<br>
><br>
> Right now I am looking at this error:<br>
> lustre-release/lustre/fid/lproc_fid.c: In function \u2018lprocfs_client_fid_space_seq_show\u2019:<br>
> lustre-release/lustre/fid/lproc_fid.c:542:5: error: void value not ignored as it ought to be<br>
> rc = seq_printf(m, "["LPX64" - "LPX64"]:%x:%s\n",<br>
><br>
> which may be straight forward to fix or may not be depending on how lprocfs_client_fid_space_seq_show is used.<br>
><br>
> seq_printf used to return 0 or -1 which would in turn be returned by this function, now seq_printf does not return anything anymore.<br>
<br>
</span>Just open a ticket in <a href="http://jira.hpdd.intel.com" rel="noreferrer" target="_blank">jira.hpdd.intel.com</a> describing the problem either especifically for every problem, or in general like "support for Linux kernel 4.5"<br>
Then submit patch to <a href="http://review.hpdd.intel.com" rel="noreferrer" target="_blank">review.hpdd.intel.com</a> with the fix, reference the ticket.<br>
Don't forget to add necessary compat logic to configure and elsewhere when needed (though I guess we can just irnore return value from<br>
seq_printf everywhere with no ill effects?<br>
<br>
See <a href="https://wiki.hpdd.intel.com/display/PUB/Submitting+Changes" rel="noreferrer" target="_blank">https://wiki.hpdd.intel.com/display/PUB/Submitting+Changes</a> and <a href="https://wiki.hpdd.intel.com/display/PUB/Patch+Landing+Process+Summary" rel="noreferrer" target="_blank">https://wiki.hpdd.intel.com/display/PUB/Patch+Landing+Process+Summary</a><br>
<br>
Bye,<br>
Oleg</blockquote></div><br></div>