AMD finally solves Ryzen linux bug

[NumLock21]

Quote

AMD’s Ryzen 7 has been generally well-received by the enthusiast community, but there’s been one low-level problem that we’ve been watching but haven’t previously reported on. In early June, Ryzen users running Linux began reporting segmentation faults when running multiple concurrent compilation workloads using multiple different versions of GCC. LVVM/Clang was not affected, and the issue appears confined to Linux.

 

Not to long ago there has been some linux bug that affect AMD Ryzen processors. Now AMD has finally solve this problem by, making a new AMD Ryzen cpu. AMD sent Michael Larabel from Phoronix, the new AMD Ryzen cpu. He took it for a test, ran the same script that originally discovered the bug, and this new Ryzen cpu passed with flying colors. To confirm it's working, he installed the old cpu and the problem came back, and when he swapped it out again with the new one, the problem disappeared. The old Ryzen is manufacture before Week 25 and the new one is manufactured on Week 30. AMD is doing replacement for those that are affected by it and all users, will need to do is request a RMA. Buying it new, then users will have to find out if their cpu is the old version or new one, by looking at the label on their cpu box.

 

 

https://www.extremetech.com/computing/254750-amd-replaces-ryzen-cpus-users-affected-rare-linux-bug

[Shiv78]

lolwut

cool of them to do it but loooooool

[ARikozuM]

Took long enough and it's nice that AMD is allowing a replacement for those affected. I wonder how many are affected though...

[JoeCoke]

Glad I didn't buy mine yet, even though this is 99.999% likely to not affect me in any way.

[NumLock21]

2 minutes ago, Shiv78 said:

lolwut

cool of them to do it but loooooool

Not sure if the linux users who have encounter this bug are loooooooling.

 

 

[Guest]

I was immediately terrified by the "Threadripper 2" in your sig lol. Glad to see it's fake.  Good one.

I know it's coming eventually, but much like winter I don't want to think about it yet.

[Windows7ge]

I've been having thoughts of moving to Threadripper. Since it's essentially two 1800X's I'd expect them to be affected as well. I may want to play with Linux so I'm glad I haven't bought anything yet.

[Shiv78]

10 minutes ago, NumLock21 said:

Not sure if the linux users who have encounter this bug are loooooooling.

 

 

true. 

 

just thought it was an atypical solution, but there must've been a reason it had to be a hardware and not a software fix.

[2FA]

Is it a new stepping or just an erratum fix?

[ARikozuM]

24 minutes ago, DeadEyePsycho said:

Is it a new stepping or just an erratum fix?

Wouldn't a new stepping be needed for an erratum fix? Some can be done through the BIOS or microcode, but I thought majority was stepping? 

[2FA]

2 minutes ago, ARikozuM said:

Wouldn't a new stepping be needed for an erratum fix? Some can be done through the BIOS or microcode, but I thought majority was stepping? 

Minor architecture fixes don't always get a new stepping.

[AnonymousGuy]

5 minutes ago, DeadEyePsycho said:

Minor architecture fixes don't always get a new stepping.

Having to send a new chip means it's a new stepping.  And all silicon fixes will get a new stepping.

[2FA]

1 minute ago, AnonymousGuy said:

Having to send a new chip means it's a new stepping.

No, it doesn't. It's only a new stepping if they call it a new one. It's literally akin to version numbers.

[Vode]

1 hour ago, ARikozuM said:

Took long enough and it's nice that AMD is allowing a replacement for those affected. I wonder how many are affected though...

3,5 nerds.

[W-L]

@NumLock21 please don't forget to include a quote related to the topic, as of currently it has been moved to the CPUs, Motherboards, and Memory section until updated.

 

 

[NumLock21]

9 hours ago, W-L said:

@NumLock21 please don't forget to include a quote related to the topic, as of currently it has been moved to the CPUs, Motherboards, and Memory section until updated.

 

 

 

Can I update it with this?

Quote

Even though this quote, has nothing to do with the article, at least it's still a quote. Therefore it meets LTT's topic posting guide lines. This topic has now been approved and certified to be in, Tech News and Reviews! Yay!

 

[Humbug]

On 9/2/2017 at 9:27 AM, ARikozuM said:

I wonder how many are affected though...

Most users (windows or Linux) don't run these kind of workloads so...

[Tenelia]

Let us take a moment to recall that Intel has a track record of refusing such things...

[Dabombinable]

24 minutes ago, Tenelia said:

Let us take a moment to recall that Intel has a track record of refusing such things...

Well they didn't have competition that frequently, and they do make shitloads more money than AMD ever has so they really don't care if people get pissed off.

[FratStar]

On 9/2/2017 at 0:08 AM, Windows7ge said:

I've been having thoughts of moving to Threadripper. Since it's essentially two 1800X's I'd expect them to be affected as well. I may want to play with Linux so I'm glad I haven't bought anything yet.

Theadripper and Epyc have not been effected by the issue. Only normal Ryzen silicon have experienced the issues.

[LehransMedallion]

I haven't run into this bug yet and I hope I don't during the course of uni. I don't have any of my other builds with me in my dorm though so I would be stuck with my smartphone while I waited for a replacement (assuming of course I wouldn't need explicit proof of a workload that was affected by the bug to begin with).

[raphidy]

Welp, time to RMA my 1700.

[HisEvilness]

I have to ask how can i check if my CPU is effected?

[Shiv78]

On 9/1/2017 at 11:54 PM, NumLock21 said:

The old Ryzen is manufacture before Week 25 and the new one is manufactured on Week 30. AMD is doing replacement for those that are affected by it and all users, will need to do is request a RMA. Buying it new, then users will have to find out if their cpu is the old version or new one, by looking at the label on their cpu box.

@HisEvilness

[HisEvilness]

1 minute ago, Shiv78 said:

@HisEvilness

Yeh I did read that part, where can I see what week my chip was produced?