AbcLinuxu:/ Poradna / Hardwarová poradna / Ryzen HW bug

Štítky: AMD, Anonymous, ASUS, BIOS, bug, cat, data, dependency, download, filesystem, FTP, GCC, GNU, GNU/Linux, grep, hardware, Internet, kompilace, Linux, RAM, SMP, sudo, Tar, test, x86_64

Dotaz: Ryzen HW bug

1.1.2018 11:13 majales | skóre: 30 | blog: Majales
Ryzen HW bug

Přečteno: 1519×

Odpovědět | Admin

Zdravím, Jen by mě zajímal jak vlastně AMD vyřešilo ten Ryzen HW Bug, protože se zdá že se znovu objevuje..

Řešil někdo z vás výměnu procesoru přímo u AMD?

Můj Ryzen 5 1600, ze 41 týdne (UA1741SUS), test kill-ryzen (https://github.com/suaefar/ryzen-test) - Deska ASUS PRIME B350-PLUS, BIOS 1002 09/29/2017, RAM 16GB (2x8GB) 2400MHZ

root@RYZEN-01:/home/moje/pokus/ryzen-test# ./kill-ryzen.sh 4 4
Install required packages
Reading package lists... Done
Building dependency tree
Reading state information... Done
build-essential is already the newest version (12.4ubuntu1).
0 upgraded, 0 newly installed, 0 to remove and 137 not upgraded.
Create compressed ramdisk
64G
[KERN] pro 31 21:17:47 RYZEN-01 kernel: zram: Added device: zram0
[KERN] pro 31 21:17:47 RYZEN-01 kernel: zram0: detected capacity change from 0 to 68719476736
Download GCC sources
--2017-12-31 21:17:48--  ftp://ftp.fu-berlin.de/unix/languages/gcc/releases/gcc-7.1.0/gcc-7.1.0.tar.bz2
           => 'gcc-7.1.0.tar.bz2'
Resolving ftp.fu-berlin.de (ftp.fu-berlin.de)... 130.133.3.130
Connecting to ftp.fu-berlin.de (ftp.fu-berlin.de)|130.133.3.130|:21... connected.
Logging in as anonymous ... Logged in!
==> SYST ... done.    ==> PWD ... done.
==> TYPE I ... done.  ==> CWD (1) /unix/languages/gcc/releases/gcc-7.1.0 ... done.
==> SIZE gcc-7.1.0.tar.bz2 ... 84303533
==> PASV ... done.    ==> RETR gcc-7.1.0.tar.bz2 ... done.
Length: 84303533 (80M) (unauthoritative)

gcc-7.1.0.tar.bz2                  0%[                                                         ]       0  --.-KB/s               [KERN] pro 31 21:17:48 RYZEN-01 kernel: EXT4-fs (zram0): mounted filesystem with ordered data mode. Opts: discard
gcc-7.1.0.tar.bz2                100%[========================================================>]  80,40M  1,42MB/s    in 50s

2017-12-31 21:18:38 (1,61 MB/s) - 'gcc-7.1.0.tar.bz2' saved [84303533]

Extract GCC sources
Download prerequisites
2017-12-31 21:19:09 URL: ftp://gcc.gnu.org/pub/gcc/infrastructure/gmp-6.1.0.tar.bz2 [2383840] -> "./gmp-6.1.0.tar.bz2" [1]
2017-12-31 21:19:18 URL: ftp://gcc.gnu.org/pub/gcc/infrastructure/mpfr-3.1.4.tar.bz2 [1279284] -> "./mpfr-3.1.4.tar.bz2" [1]
2017-12-31 21:19:24 URL: ftp://gcc.gnu.org/pub/gcc/infrastructure/mpc-1.0.3.tar.gz [669925] -> "./mpc-1.0.3.tar.gz" [1]
2017-12-31 21:19:32 URL: ftp://gcc.gnu.org/pub/gcc/infrastructure/isl-0.16.1.tar.bz2 [1626446] -> "./isl-0.16.1.tar.bz2" [1]
gmp-6.1.0.tar.bz2: OK
mpfr-3.1.4.tar.bz2: OK
mpc-1.0.3.tar.gz: OK
isl-0.16.1.tar.bz2: OK
All prerequisites downloaded successfully.
cat /proc/cpuinfo | grep -i -E "(model name|microcode)"
model name      : AMD Ryzen 5 1600 Six-Core Processor
microcode       : 0x8001129
model name      : AMD Ryzen 5 1600 Six-Core Processor
microcode       : 0x8001129
model name      : AMD Ryzen 5 1600 Six-Core Processor
microcode       : 0x8001129
model name      : AMD Ryzen 5 1600 Six-Core Processor
microcode       : 0x8001129
model name      : AMD Ryzen 5 1600 Six-Core Processor
microcode       : 0x8001129
model name      : AMD Ryzen 5 1600 Six-Core Processor
microcode       : 0x8001129
model name      : AMD Ryzen 5 1600 Six-Core Processor
microcode       : 0x8001129
model name      : AMD Ryzen 5 1600 Six-Core Processor
microcode       : 0x8001129
model name      : AMD Ryzen 5 1600 Six-Core Processor
microcode       : 0x8001129
model name      : AMD Ryzen 5 1600 Six-Core Processor
microcode       : 0x8001129
model name      : AMD Ryzen 5 1600 Six-Core Processor
microcode       : 0x8001129
model name      : AMD Ryzen 5 1600 Six-Core Processor
microcode       : 0x8001129
sudo dmidecode -t memory | grep -i -E "(rank|speed|part)" | grep -v -i unknown
        Speed: 2400 MT/s
        Part Number: BLS8G4D240FSE.16FBD
        Rank: 2
        Configured Clock Speed: 1200 MT/s
        Speed: 2400 MT/s
        Part Number: BLS8G4D240FSE.16FBD
        Rank: 2
        Configured Clock Speed: 1200 MT/s
uname -a
Linux RYZEN-01 4.13.0-16-generic #19-Ubuntu SMP Wed Oct 11 18:35:14 UTC 2017 x86_64 x86_64 x86_64 GNU/Linux
...¨
[loop-1] Ne pro 31 21:26:30 CET 2017 build failed
[loop-1] TIME TO FAIL: 417 s
[loop-0] Ne pro 31 21:26:30 CET 2017 build failed
[loop-0] TIME TO FAIL: 417 s
[loop-2] Ne pro 31 21:26:33 CET 2017 build failed
[loop-2] TIME TO FAIL: 420 s
[loop-3] Ne pro 31 21:26:49 CET 2017 build failed
[loop-3] TIME TO FAIL: 436 s

Řešení dotazu:

Komentář #1 (majales, 1 hlasů)

Nástroje: Začni sledovat (2) ?

Odpovědi

Řešení 1× (majales (tazatel))

2.1.2018 14:31 majales | skóre: 30 | blog: Majales
Rozbalit Rozbalit vše Re: Ryzen HW bug

Tak zpět.. Byla to moje chyba. Originální kill-ryzen.sh je pouze pro ubuntu 17.04, pod 17.10 se z kvůli nějakému konfliktu gcc-7.1.0 nezkompiluje a vyhodí chybu. Upravil jsem kill-ryzen.sh skripty a změnil to na gcc-7.2.0. Kompilace běžela více jak 1h 30m bez problému.

./kill-ryzen.sh 4 4
Install required packages
Reading package lists... Done
Building dependency tree
Reading state information... Done
build-essential is already the newest version (12.4ubuntu1).
0 upgraded, 0 newly installed, 0 to remove and 137 not upgraded.
Create compressed ramdisk
64G
[KERN] led 01 21:29:18 RYZEN-01 kernel: zram: Added device: zram0
[KERN] led 01 21:29:18 RYZEN-01 kernel: zram0: detected capacity change from 0 to 68719476736
[KERN] led 01 21:29:18 RYZEN-01 kernel: zram: Added device: zram0
[KERN] led 01 21:29:18 RYZEN-01 kernel: zram0: detected capacity change from 0 to 68719476736
Download GCC sources
--2018-01-01 21:29:19--  ftp://ftp.fu-berlin.de/unix/languages/gcc/releases/gcc-7.2.0/gcc-7.2.0.tar.gz
           => 'gcc-7.2.0.tar.gz'
Resolving ftp.fu-berlin.de (ftp.fu-berlin.de)... 130.133.3.130
Connecting to ftp.fu-berlin.de (ftp.fu-berlin.de)|130.133.3.130|:21... connected.
Logging in as anonymous ... Logged in!
==> SYST ... done.    ==> PWD ... done.
==> TYPE I ... done.  ==> CWD (1) /unix/languages/gcc/releases/gcc-7.2.0 ... done.
==> SIZE gcc-7.2.0.tar.gz ... 111997123
==> PASV ... done.    ==> RETR gcc-7.2.0.tar.gz ... done.
Length: 111997123 (107M) (unauthoritative)

gcc-7.2.0.tar.gz                   0%[                                                         ]       0  --.-KB/s               [KERN] led 01 21:29:19 RYZEN-01 kernel: EXT4-fs (zram0): mounted filesystem with ordered data mode. Opts: discard
[KERN] led 01 21:29:19 RYZEN-01 kernel: EXT4-fs (zram0): mounted filesystem with ordered data mode. Opts: discard
gcc-7.2.0.tar.gz                 100%[========================================================>] 106,81M  1,58MB/s    in 1m 46s

2018-01-01 21:31:05 (1,01 MB/s) - 'gcc-7.2.0.tar.gz' saved [111997123]

Extract GCC sources
Download prerequisites
2018-01-01 21:31:24 URL: ftp://gcc.gnu.org/pub/gcc/infrastructure/gmp-6.1.0.tar.bz2 [2383840] -> "./gmp-6.1.0.tar.bz2" [1]
2018-01-01 21:31:31 URL: ftp://gcc.gnu.org/pub/gcc/infrastructure/mpfr-3.1.4.tar.bz2 [1279284] -> "./mpfr-3.1.4.tar.bz2" [1]
2018-01-01 21:31:36 URL: ftp://gcc.gnu.org/pub/gcc/infrastructure/mpc-1.0.3.tar.gz [669925] -> "./mpc-1.0.3.tar.gz" [1]
2018-01-01 21:31:47 URL: ftp://gcc.gnu.org/pub/gcc/infrastructure/isl-0.16.1.tar.bz2 [1626446] -> "./isl-0.16.1.tar.bz2" [1]
gmp-6.1.0.tar.bz2: OK
mpfr-3.1.4.tar.bz2: OK
mpc-1.0.3.tar.gz: OK
isl-0.16.1.tar.bz2: OK
All prerequisites downloaded successfully.
cat /proc/cpuinfo | grep -i -E "(model name|microcode)"
model name      : AMD Ryzen 5 1600 Six-Core Processor
microcode       : 0x8001129
model name      : AMD Ryzen 5 1600 Six-Core Processor
microcode       : 0x8001129
model name      : AMD Ryzen 5 1600 Six-Core Processor
microcode       : 0x8001129
model name      : AMD Ryzen 5 1600 Six-Core Processor
microcode       : 0x8001129
model name      : AMD Ryzen 5 1600 Six-Core Processor
microcode       : 0x8001129
model name      : AMD Ryzen 5 1600 Six-Core Processor
microcode       : 0x8001129
model name      : AMD Ryzen 5 1600 Six-Core Processor
microcode       : 0x8001129
model name      : AMD Ryzen 5 1600 Six-Core Processor
microcode       : 0x8001129
model name      : AMD Ryzen 5 1600 Six-Core Processor
microcode       : 0x8001129
model name      : AMD Ryzen 5 1600 Six-Core Processor
microcode       : 0x8001129
model name      : AMD Ryzen 5 1600 Six-Core Processor
microcode       : 0x8001129
model name      : AMD Ryzen 5 1600 Six-Core Processor
microcode       : 0x8001129
sudo dmidecode -t memory | grep -i -E "(rank|speed|part)" | grep -v -i unknown
        Speed: 2400 MT/s
        Part Number: BLS8G4D240FSE.16FBD
        Rank: 2
        Configured Clock Speed: 1200 MT/s
        Speed: 2400 MT/s
        Part Number: BLS8G4D240FSE.16FBD
        Rank: 2
        Configured Clock Speed: 1200 MT/s
uname -a
Linux RYZEN-01 4.13.0-16-generic #19-Ubuntu SMP Wed Oct 11 18:35:14 UTC 2017 x86_64 x86_64 x86_64 GNU/Linux
cat /proc/sys/kernel/randomize_va_space
2
/ /mnt/ramdisk/workdir
/mnt/ramdisk/workdir
Using 4 parallel processes
[KERN] -- Logs begin at Mon 2018-01-01 19:35:25 CET. --
[KERN] led 01 21:16:28 RYZEN-01 kernel: zram0: detected capacity change from 0 to 68719476736
[KERN] led 01 21:16:29 RYZEN-01 kernel: EXT4-fs (zram0): mounted filesystem with ordered data mode. Opts: discard
[KERN] led 01 21:24:18 RYZEN-01 kernel: zram: Removed device: zram0
[KERN] led 01 21:24:37 RYZEN-01 kernel: zram: Added device: zram0
[KERN] led 01 21:24:37 RYZEN-01 kernel: zram0: detected capacity change from 0 to 68719476736
[KERN] led 01 21:24:38 RYZEN-01 kernel: EXT4-fs (zram0): mounted filesystem with ordered data mode. Opts: discard
[KERN] led 01 21:29:10 RYZEN-01 kernel: zram: Removed device: zram0
[KERN] led 01 21:29:18 RYZEN-01 kernel: zram: Added device: zram0
[KERN] led 01 21:29:18 RYZEN-01 kernel: zram0: detected capacity change from 0 to 68719476736
[KERN] led 01 21:29:19 RYZEN-01 kernel: EXT4-fs (zram0): mounted filesystem with ordered data mode. Opts: discard
[loop-0] Po led 1 21:31:48 CET 2018 start 0
[loop-1] Po led 1 21:31:49 CET 2018 start 0
[loop-2] Po led 1 21:31:50 CET 2018 start 0
[loop-3] Po led 1 21:31:51 CET 2018 start 0
[KERN] led 01 22:31:41 RYZEN-01 kernel: perf: interrupt took too long (2503 > 2500), lowering kernel.perf_event_max_sample_rate to 79750
[KERN] led 01 22:31:41 RYZEN-01 kernel: perf: interrupt took too long (2503 > 2500), lowering kernel.perf_event_max_sample_rate to 79750
[KERN] led 01 22:31:41 RYZEN-01 kernel: perf: interrupt took too long (2503 > 2500), lowering kernel.perf_event_max_sample_rate to 79750
[loop-2] Po led 1 22:44:19 CET 2018 start 1
[loop-1] Po led 1 22:44:40 CET 2018 start 1
[loop-3] Po led 1 22:47:11 CET 2018 start 1
[loop-0] Po led 1 22:49:06 CET 2018 start 1

3.1.2018 01:06 pc2005 | skóre: 38 | blog: GardenOfEdenConfiguration | liberec
Rozbalit Rozbalit vše Re: Ryzen HW bug

Není 1:30 málo? Jsem měl za to, že to muselo běžet tak týden.

BTW chybnej ryzen by ti měli vyměnit, pokud máš ten s chybou a při výměně poukážeš na to že si shazuje kompilaci.

Jestli tu chybu opravili vydáním novýho firmware to si nejsem jistý.

Intel meltdown a = arr[x[0]&1]; karma | 帮帮我，我被锁在中国房

3.1.2018 01:33 majales | skóre: 30 | blog: Majales
Rozbalit Rozbalit vše Re: Ryzen HW bug

Podle reportu to shazovalo spolehlivě ty Ryzeny s HW bugem do 10 minut.. pravda mohl jsem to nechat běžet déle.. ještě to vyzkouším.

3.1.2018 02:13 pc2005 | skóre: 38 | blog: GardenOfEdenConfiguration | liberec
Rozbalit Rozbalit vše Re: Ryzen HW bug

S vypnutým SMT byl ten čas delší. Ale myslím, že mám tu časovou informaci z nějakého původního testu nebo přímo z reálné kompilace v gentoo.

Jinak použitý ryzen je podle phoronixu asi OK.

it appears to affect Ryzen CPUs manufactured prior to week 25

Intel meltdown a = arr[x[0]&1]; karma | 帮帮我，我被锁在中国房

9.1.2018 22:46 Lukáš Zapletal | skóre: 42 | blog: lzapův svět | Olomouc
Rozbalit Rozbalit vše Re: Ryzen HW bug

Opatchoval jsem mu (autorovi kill-ryzen.sh skriptu) Fedoru, na jakým distru to nejde?

Každopádně já jsem měl kus co padal po 5 minutách, po 10 dnech se supportem a 25 dnech bez CPU mi poslali novej, kterej jede jak má. Ale SMT mám zatím vypnutý, nechci laborovat je to můj pracovní stroj a 8 vláken mi stačí.

Later --- Lukáš Zapletal

3.2.2018 10:19 Vladimír Čunát | skóre: 19
Rozbalit Rozbalit vše Re: Ryzen HW bug

To jste tedy šel přes klasický reklamační proces u prodejce? Já jsem dlouho doufal ve vyřešení microcodem, ale zřejmě marně.

Založit nové vlákno • Nahoru

Tiskni Sdílej: