Unable to boot MIC after mpss_gold_update_3 upgrade

[Eric_B_]

After upgrading to update_3, I am no longer able to boot the MIC:

[plain]Shutting down MPSS Stack: [ OK ]
Starting MPSS Stack: Timeout booting MIC, check your installation
[FAILED][/plain]

From the terminal:

[plain]

[ 0.000000] SFI: Entering sfi_map_memory, phys = e0000, size = 131071
[ 0.000000] SFI: Entering sfi_map_memory, phys = ef120, size = 32
[ 0.000000] SFI: Entering sfi_map_table, pa = 92000
[ 0.000000] SFI: Entering sfi_map_memory, phys = 92000, size = 24
[ 0.000000] SFI: sfi_map_table, th = ffffffffff4ba000
[ 0.000000] SFI: Entering sfi_map_memory, phys = 92000, size = 984
[ 0.000000] SFI: Entering sfi_map_table, pa = ef168
[ 0.000000] SFI: sfi_map_table, th = ffffffffff47a168
[ 0.000000] SFI: Entering sfi_map_table, pa = eefa0
[ 0.000000] SFI: Entering sfi_map_memory, phys = eefa0, size = 24
[ 0.000000] SFI: sfi_map_table, th = ffffffffff4bafa0
[ 0.000000] SFI: Entering sfi_map_memory, phys = eefa0, size = 312
[ 0.000000] SFI: Entering sfi_map_table, pa = 92000
[ 0.000000] SFI: Entering sfi_map_memory, phys = 92000, size = 24
[ 0.000000] SFI: sfi_map_table, th = ffffffffff4ba000
[ 0.000000] SFI: Entering sfi_map_memory, phys = 92000, size = 984
[ 0.000000] SFI: Entering sfi_map_table, pa = 92000
[ 0.000000] SFI: Entering sfi_map_memory, phys = 92000, size = 24
[ 0.000000] SFI: sfi_map_table, th = ffffffffff4ba000
[ 0.000000] SFI: Entering sfi_map_memory, phys = 92000, size = 984
[ 0.000000] SFI: Entering sfi_map_table, pa = ef168
[ 0.000000] SFI: sfi_map_table, th = ffffffffff47a168
[ 0.000000] PCI: Warning: Cannot find a gap in the 32bit address range
[ 0.000000] PCI: Unassigned devices with 32bit resource registers may break!
[ 0.010000] SFI: Entering sfi_map_memory, phys = ef120, size = 48
[ 0.010000] SFI: Entering sfi_map_table, pa = 92000
[ 0.010000] SFI: Entering sfi_map_memory, phys = 92000, size = 24
[ 0.010000] SFI: sfi_map_table, th = ffffc90000000000
[ 0.010000] SFI: Entering sfi_map_memory, phys = 92000, size = 984
[ 0.010000] SFI: Entering sfi_map_table, pa = ef168
[ 0.010000] SFI: sfi_map_table, th = ffff8800000ef168
[ 0.010000] SFI: Entering sfi_map_table, pa = eefa0
[ 0.010000] SFI: Entering sfi_map_memory, phys = eefa0, size = 24
[ 0.010000] SFI: sfi_map_table, th = ffff8800000eefa0
[ 0.010000] SFI: Entering sfi_map_memory, phys = eefa0, size = 312
[ 0.010000] SFI: Entering sfi_map_table, pa = 92000
[ 0.010000] SFI: Entering sfi_map_memory, phys = 92000, size = 24
[ 0.010000] SFI: sfi_map_table, th = ffffc90000000000
[ 0.010000] SFI: Entering sfi_map_memory, phys = 92000, size = 984
[ 0.010000] SFI: Entering sfi_map_table, pa = ef168
[ 0.010000] SFI: sfi_map_table, th = ffff8800000ef168
[ 0.010000] SFI: Entering sfi_map_table, pa = eefa0
[ 0.010000] SFI: Entering sfi_map_memory, phys = eefa0, size = 24
[ 0.010000] SFI: sfi_map_table, th = ffff8800000eefa0
[ 0.010000] SFI: Entering sfi_map_memory, phys = eefa0, size = 312
[ 14.938148] i8042: Can't read CTR while initializing i8042
[ 16.427859] BUG: unable to handle kernel NULL pointer dereference at (null)
[ 16.427897] IP: [<ffffffff811c6ddb>] memcpy+0xb/0x120
[ 16.427938] PGD 1ea6ee067 PUD 1eaaad067 PMD 0
[ 16.427966] Oops: 0002 [#1] SMP
[ 16.427985] last sysfs file: /sys/devices/virtual/tty/hvc7/dev
[ 16.428009] CPU 50
[ 16.4280[ 95.131180] mpssboot: gave up waiting for init of module micscif.
[ 95.131208] mpssboot: Unknown symbol scif_close (err -16)
[ 95.132289] (INIT) Modprobe micras
[ 95.145875] Module micras loaded at 0xffffffffa007e000
[ 125.150188] micras: gave up waiting for init of module micscif.
[ 125.150217] micras: Unknown symbol scif_bind (err -16)
[ 155.150774] micras: gave up waiting for init of module micscif.
[ 155.150815] micras: Unknown symbol scif_accept (err -16)
[ 185.151170] micras: gave up waiting for init of module micscif.
[ 185.151213] micras: Unknown symbol scif_recv (err -16)
[ 215.151167] micras: gave up waiting for init of module micscif.
[ 215.151207] micras: Unknown symbol scif_open (err -16)
[ 245.151177] micras: gave up waiting for init of module micscif.
[ 245.151221] micras: Unknown symbol scif_send (err -16)
[ 275.150772] micras: gave up waiting for init of module micscif.
[ 275.150815] micras: Unknown symbol scif_listen (err -16)
[ 305.151167] micras: gave up waiting for init of module micscif.
[ 305.151193] micras: Unknown symbol scif_close (err -16)
[ 305.154446] (INIT) Set up crash kernel
[ 305.216599] (INIT) Switch to tmpfs root
[ 305.225751] (INIT) Download file system
[ 305.229346] (INIT) Rootfs download failed (exit code 1)

[/plain]

Suggestions? I've completely removed the modifications I had done to configurations (removed old configurations; started fresh) with the exception of an IP change to the 192.168 range.

[Frances_R_Intel]

When you did the install, did you first stop the previous mpss, unload the previous mpss and uninstall the previous mpss?

[Eric_B_]

I uninstalled (but failed to manually stop the service first) the previous mpss and then installed the new. I have since gone through stop/unload/uninstall/reinstall/load without improving the outcome.

I notice that the /opt/intel/mic/filesystem/mic0.image file is not being created...?

[Eric_B_]

Linux <hostname> 2.6.32-358.6.1.el6.x86_64 #1 SMP Tue Apr 23 19:29:00 UTC 2013 x86_64 x86_64 x86_64 GNU/Linux; FWIW

[Eric_B_]

Gaaa! A re-compile of the kmod did it. I'm on RHEL(CentOS) 6.4, so I was pleased to see a 6.4 package, and didn't recompile initially.

[Frances_R_Intel]

Glad that worked.

[haandradeb]

I am having similar issues.

Is the only workaround to recompiile the kmod, or is a recompiled version available somewhere for download?

Thanks.

[Frances_R_Intel]

Rebuilding the kernel modules is required only when using a version of Linux other than one of the officially supported versions. In this case, they were using CentOS rather than a straight RHEL 6.4 release.