cancel
Showing results for 
Search instead for 
Did you mean: 

Drivers & Software

drescherjm
Adept II
Adept II

Re: gcc segmentation faults on Ryzen / Linux

After fixing the script to work again on ubuntu-18.04 (uncomment the section that installs the build tools in ubuntu). I was able to run it on my Ryzen 2700 / ASUS Prime PRO + 16GB of Crucial DDR4 Unbuffered ECC. After ~36 hours of testing (24 contiguous), all is well (at least when I test from installed ubuntu). I had two lockups when testing from a USB stick. One in kill-ryzen.sh one in memtest86. Not sure why for either. Overall, I am very impressed at this CPU so far. That is as long as the lockups don't happen (my purpose for this system is for a 24/7 fileserver / pvr device that needs to be 100% stable). My thoughts are that it could be that a microcode update is delivered in the installed ubuntu but not delivered in the live usb stick for either. I have not updated the UEFI/BIOS yet and I do see that there have been several updates on that even though my board is very new.

drseehas
Adept I
Adept I

Re: gcc segmentation faults on Ryzen / Linux

drescherjm wrote:

... ASUS Prime PRO + 16GB of Crucial DDR4 Unbuffered ECC. ...

Does ECC work?

0 Kudos
drescherjm
Adept II
Adept II

Re: gcc segmentation faults on Ryzen / Linux

I have not verified that yet. The specification page for the ASUS Prime PRO X470 board says it supports ECC depending on the CPU. I assumed that meant APUs don't have ECC (since other boards mention that). However I did not see any mention of ECC in the manual or the BIOS. memtest86 7.5 did mention that I had ECC but not sure if they were just reading that from the RAM or whether or not it was actually active. I plan to check on that after further testing. I am still running the kill-ryzen script.

yojimbo87
Journeyman III
Journeyman III

Re: gcc segmentation faults on Ryzen / Linux

Has anyone RMA'd a 1000 series and gotten a 2000 series for replacement?

drescherjm
Adept II
Adept II

Re: gcc segmentation faults on Ryzen / Linux

I believe the following means ECC is enabled and in Single bit correction / double bit detection mode.

Here is the linux kernel version

jmd1 ~/shell-scripts # uname -a

Linux jmd1.comcast.net 4.16.13-gentoo-20180603-1145-jmd1.comcast.net #3 SMP Sun Jun 3 11:52:55 EDT 2018 x86_64 AMD Ryzen 7 2700 Eight-Core Processor AuthenticAMD GNU/Linux

This tells me ECC is enabled.

jmd1 ~/shell-scripts # dmesg | grep ECC

[    8.557846] systemd[1]: systemd 238 running in system mode. (+PAM -AUDIT -SELINUX +IMA -APPARMOR +SMACK -SYSVINIT +UTMP -LIBCRYPTSETUP +GCRYPT -GNUTLS +ACL -XZ +LZ4 +SECCOMP +BLKID -ELFUTILS +KMOD -IDN2 -IDN +PCRE2 default-hierarchy=hybrid)

[    9.132922] EDAC amd64: Node 0: DRAM ECC enabled.

This tells me there have been 0 errors (I expect that from server experience ECC errors should be rare)

        jmd1 ~/shell-scripts # edac-util -v

mc0: 0 Uncorrected Errors with no DIMM info

mc0: 0 Corrected Errors with no DIMM info

mc0: csrow0: 0 Uncorrected Errors

mc0: csrow0: mc#0csrow#0channel#0: 0 Corrected Errors

mc0: csrow0: mc#0csrow#0channel#1: 0 Corrected Errors

edac-util: No errors to report.

This tells me the mode of error correction for the first rank

jmd1 ~/shell-scripts # cat /sys/devices/system/edac/mc/mc0/rank0/dimm_edac_mode

SECDED

Same goes for the 2nd rank

jmd1 ~/shell-scripts # cat /sys/devices/system/edac/mc/mc0/rank1/dimm_edac_mode

SECDED

Here is what SECDED means

EDAC_SECDED

Single bit error correction, Double detection

constantinx
Adept II
Adept II

Re: gcc segmentation faults on Ryzen / Linux

I have just replaced my Ryzen 1700, bought in August 2017, from batch 1714 PGT .

The new processor is from batch 1738 SUS.

The segfault error seems to be gone, after mild testing. I'll test more, but that batch is likely error free.

I delayed so much until I started the RMA because I didn't have a spare processor, and I really hoped for a microcode fix.

The RMA was very fast, and took about 1 week. I live in Europe, and their local warehouse is in Netherlands.

----------

I made an online service request, warranty category, on Friday, 01 June, in the afternoon.

They replied on Monday, 04 June, in the morning.

A quick exchange of emails followed, and next day, 05 June, they approved the RMA and sent me a DHL account number with free shipping.

I spent the next few days getting a temporary processor.

I sent the processor by DHL on 11 June. They received it on 12th and "test/inspection passed" the same day. On 13th they sent the new processor, and today, 14th of June, I received it.

They were supposed to send a tracking number, but I was simply phone called by the courier 10 minutes before the delivery. Good thing I was at home.

----------

Overall a great experience. Big thanks to AMD!

0 Kudos
patrikx3
Journeyman III
Journeyman III

Re: gcc segmentation faults on Ryzen / Linux

For me, what was the solution, was to keep the temperature under 70 Celsius, then I got no segfault, when it goes over 70 Celsius I always had segfaults.

It is an Asus laptop ASUS-ROG-Strix-GL702ZC, I could only set the fan in Windows to 80% and then the problem got lost. On auto fan, always breaking the compiling. 

0 Kudos