Intel Communities : All Content - SCC Support Forum

SCC die thickness

webadmin@intel.com — Mon, 04 Mar 2013 05:40:36 GMT

Hello everybody,

I couldn't find any documentation about how much is the thickness of SCC die .. does anybody know ?

Thanks in advance

Frequency Scaling

webadmin@intel.com — Thu, 07 Jun 2012 06:30:13 GMT

Dear All,

I'm facing a problem with frequency scaling on the scc cores.

After I set the frequency to the required value for example 800 Mhz, it succeeded for just some second, and then the core return to its original frequency automatically. I knew that when I "cat" the file cpu_cur_freq. It contains the new value of the frequency just for second, and then it changes automatically. This happens with the new sccLinux 3.1.4 . Knowing that it succeeded with the older linux kernel.

So, could anyone help in this problem ?

Thanks in advance,

Heba

sccBmc errors

webadmin@intel.com — Tue, 14 Dec 2010 23:54:22 GMT

Hi,

I'm getting the following error at the end of the training run (sccBmc -i) both from the command line and sccGui:

ERROR: Training unsuccessful: Too small optimal data eye: optimal delay setting is 001 in a 006 wide range, select slower/different SIF phy operation speed ***
ERROR: Failed to trim SIF! Please try to reload the driver or powercycle SCC board in case of repeated failure!

I tried powercycling the SCC according to the instructions provided and reloading the driver after the failure, but no difference. Both the fast and extended training runs fail before the full training and the final failure. I'm running version 1.3.0 of the firmware.

The driver loads fine according to dmesg:

[   12.697324] mcedev Id: $Id: mcedev_main.c 16545 2010-06-08 14:25:34Z jbrummer $
[   12.697347] mcedev_probe found device: VID 0x8086, DID 0xc148, SVID 0x8086, SDID 0xc048, BUS 0x04, DEVFN 0x00.0
[   12.697364] mcedev 0000:04:00.0: PCI INT A -> GSI 16 (level, low) -> IRQ 16
[   12.697366] crbif_init: $Id: crbif_main.c 16585 2010-08-24 14:35:55Z jbrummer $ (debuglevel 0000)
[   12.697368] rckcrb_init: Checking and disabling ASPM on root port 8086:340e (00:07.00) for rckcrb device
[   12.697373]     Found PCIe Capability structure @ 0x90
[   12.697375]     Link Control Reg @ 0xa0: 0x30410001, setting to 0x30410000
[   12.697378]     Link Control Reg @ 0xa0 changed to: 0x30410000
[   12.697387] mcedev 0000:04:00.0: setting latency timer to 64
[   12.697468] rckcrb_init: BITSID: 0x20100608
[   12.697480] MIP/MOP size: 64/64KB (13)
[   12.697555]   alloc irq_desc for 34 on node -1
[   12.697556]   alloc kstat_irqs on node -1
[   12.697565] mcedev 0000:04:00.0: irq 34 for MSI/MSI-X
[   12.697571] IRQ 34/crbif: IRQF_DISABLED is not guaranteed on shared IRQs

The full training log is available at: http://dl.dropbox.com/u/8103741/training_error.log

Dump of lspci -vvv: http://dl.dropbox.com/u/8103741/lspci_dump.txt

Any ideas what the problem could be?

Thanks,

Hrishi

Could the 1.3.0 cable connection still work for 1.4.0

webadmin@intel.com — Tue, 21 Jun 2011 02:48:37 GMT

Hi,

I have a question about upgrading our sccKit from 1.3.0 to 1.4.0.

I am wondering could the cable connection for 1.3.0 (i.e. one PCIe cable from MCPC and one ethernet cable from MCPC to BMC)

still works for 1.4.0 setup instead of using a gigabit switch to connect MCPC, SCC and BMC all together?

We follow all the steps in the HowToInstall1.4.0.pdf, except the rewire the connections.

The Linux image could be loaded onto cores through sccGui but we can not ping or ssh to the cores.

Any advices will be apprecited. Thank you very much!

Best,

Yang

Connection to SCC breaks if I/O load is high

webadmin@intel.com — Mon, 18 Oct 2010 11:24:58 GMT

Hi,

we are again having some trouble with our SCC (RockyLake) setup. I can successfully start SCC Linux on all 48 cores, but the PCIe link becomes unresponsive if too many cores are performing too much I/O to the /shared file system at the same time.

Once in this state, sccGui reports timeouts during the DMA transfers. Sometimes, it logs a single line that an unexpected packet had been received (with varying command codes), but this message does not appear all the time. The crbif driver logs error messages to the text-mode console: MIP/MOP status 0x20000000, unable to start DMA transfer.

The SCC performance meter no longer displays the current CPU load; the graphs are advancing, but only plot the same (last-known) value. Also, network traffic to the SCC is no longer possible, and "ping rck**" times out.

Occasionally, after some time, I have seen the screen of the MCPC go black (as if the X server was shut down), and the system then locking up completely. If this is the case, I cannot even reboot the MCPC without a hard power cycle on the SCC first; it displays the BIOS screen, but Linux never comes up. It just sits at a black screen. If the MCPC is still responsive, I can call sccPowercycle, but it usually (roughly 9 out of 10 tries) does not fix the problem; even re-training the SIF fails, until the SCC has been hard-powercycled.

To generate I/O load, we use a recompiled SCC Linux and some scripts. The Linux code is identical to the one in the public repository, with one change: in rckos/fs/etc/init.d/tc-config, I appended the following:

<code>

# HPI: Additional init script
if [ -f /shared/coreInit/init.sh ]
then
echo "%%%% Invoking /shared/coreInit/init.sh for PID$core" > /shared/coreInit/results/PID$core.log
chmod 777 /shared/coreInit/results/PID$core.log
/shared/coreInit/init.sh $core < /dev/null >> /shared/coreInit/results/PID$core.log 2>&1
echo "%%%% Done for PID$core" >> /shared/coreInit/results/PID$core.log
fi

</code>

This executes a script from the shared file system after each core has been booted. To reproduce my problem, it is now sufficient to insert some copy commands (for roughly 20MB of data) from /shared to the SCC Linux ramdisk, then start all 48 cores in parallel. For up to ~30 cores, this works fine and the files are copied correctly; once too many cores are started simultaneously (and thus producing I/O), the PCIe connection breaks.

The MCPC uses the recommended Intel DX58SO mainboard.

Has anybody experienced similar issues? Any help is appreciated.

Kind regards,

Jan-Arne Sobania

sccBmc MEMRD errors, sccBoot/sccGui freezes MCPC

webadmin@intel.com — Thu, 16 Sep 2010 14:50:07 GMT

I just got our RockyLake system running and recognized by the MCPC, but I'm having some other issues now. First of all, "sccBmc -i" sometimes displays errors during the MEMRD test after configuring the memory controllers the first time (please see attached file sccBmc.i.err.jpg). After configuring them again, no further errors are displayed, and the program exists normally; however, I wonder whether this behaviour is expected and other users are getting the same messages? Or does this indicate a more serious problem with the MCPC or SCC board?

After sccBmc seems to have run successfully, there is another more severe problem: after invoking "sccBoot -l 0" to start Linux (default image from sccKit 1.3.0), the MCPC freezes immediately after "INFO: Found object for MC x=0, y=0" is displayed. From this point on, the message "BUG: soft lockup - XPU#0 stuck for 61s! [crbifd:735]" is displayed once every minute (please see attached screenshot). The same lockup happens regardless of the SCC core id (and even when trying to start Linux on all cores at once), and regardless of whether I try sccBoot or sccGui. I also tried the memory and controller test from within sccGui, which freezes the MCPC as well.

Does anybody have an idea on how this can be resolved?

For your information: the MCPC is a Dell Optiplex 755 with BIOS A17 (April 2010), running Ubuntu 10.04.1 64bit. The PCIe card is installed in the PEG slot, as this is the only PCIe slot available on the mainboard.

Interestingly, this system does not boot if the SCC is powered on (just stays at a blank screen without any video output; it doesn't even show the Dell logo!). To get the system into booting Linux, I need to start with the SCC powered down, then halt at the GRUB2 menu. I then start the SCC via another computer, wait until the BMC sends the "Done" message, then continue booting Linux on the MCPC.

PCIe driver not found

webadmin@intel.com — Wed, 15 Sep 2010 00:35:40 GMT

I'm trying to follow the How to Install Ubuntu and sccKit directions, but when I try to run the SIF training using "sccBmc -i" it appears to fail to find (load?) the driver for the OSS PCIe card:

> # sccBmc -i

> INFO: Packet tracing is disabled...

> INFO: Opened "/tmp/sccKit_root/trace.log" for writing...

> INFO: openBMCConnection(192.168.2.127:5010): You are participant #1

> INFO: Initializing System Interface (SCEMI setup)....

> ERROR: initializeSysIF(): Failed to open PCIe driver device. Driver

> devices not installed/available...

> ERROR: Unable to connect to PCIe driver...

> Make sure that PCIe driver is installed and

> not in use by someone else... Aborting!

Suggestions?