I have run x86-64-level on all three of my machines and it seems to work fine.
Iām not a pro, my 12 y.o. desktop does what I need it to do. I realize itās old but it still runs vBox. A new Apple Silicon will not run vBox.
vBox is the biggest resource user I run. Itās slower now but does what I need. So reading this thread, which is a bit over my head, this is what I have:
CPU:
Info: model: Intel Core i7-3770 bits: 64 type: MT MCP arch: Ivy Bridge
gen: core 3 level: v2 built: 2012-15 process: Intel 22nm family: 6
model-id: 0x3A (58) stepping: 9 microcode: 0x21
Topology: cpus: 1x cores: 4 tpc: 2 threads: 8 smt: enabled cache:
L1: 256 KiB desc: d-4x32 KiB; i-4x32 KiB L2: 1024 KiB desc: 4x256 KiB
L3: 8 MiB desc: 1x8 MiB
Speed (MHz): avg: 2220 high: 3812 min/max: 1600/3900 scaling:
driver: intel_pstate governor: powersave cores: 1: 1656 2: 1652 3: 3155
4: 1777 5: 3812 6: 1656 7: 1666 8: 2390 bogomips: 54276
Flags: avx ht lm nx pae sse sse2 sse3 sse4_1 sse4_2 ssse3 vmx
It does say core 3 level: v2
Will the above still meet the requirements of Leap 16 (Iām at 15.6)?
V2 + additions.
.
If CPU supports SSE4.2 (flag āsse4_2ā), then it supports x86-64-v2.
.
If CPU supports AVX2 (flag āavx2ā), then it supports x86-64-v3.
Tumbleweed supports x86-64-v1 = AMD64.
What else people want?
It is possible to patch code instead of recompiling.
Some x86-64-v2 commands are already supported by users CPUs.
As I wrote in
https://en.opensuse.org/openSUSE:X86-64-Architecture-Levels#Possible_option:_Making_small_steps
Phenom (AMD K10) supports x86-64-v1d = x86-64-v1 + SSE3 + CMPXCHG16B + LAHF/SAHF + POPCNT.
So it is needed to replace SSSE3, SSE4.1 and SSE4.2 with machine (assembly) code.
For broader CPU support use x86-64-v1c = x86-64-v1 + SSE3 + CMPXCHG16B + LAHF/SAHF and patch other CPU instructions.
@Svyatko That is all well and good for some niche hardware a user may want to keep running and maintain the security implications as well as support.
For the average openSUSE end user, not a good idea, I suspect there are more systems which wil be v2+
When it is about hardware that includes me.
Is it correct that those users can test doing
boven:~ # inxi -C -a | grep level
arch: Skylake-S gen: core 6 level: v3 note: check built: 2015
boven:~ #
and then look at what is following level: to check if Leap 15.6 will run on their system?
In the past inxi had errors with determining x86-64 levels.
It is better to check sse4_2 flag:
lscpu | grep sse4_2
boven:~ # lscpu | grep sse4_2
Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc art arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc cpuid aperfmperf pni pclmulqdq dtes64 monitor ds_cpl est tm2 ssse3 sdbg fma cx16 xtpr pdcm pcid sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand lahf_lm abm 3dnowprefetch cpuid_fault epb pti ssbd ibrs ibpb stibp fsgsbase tsc_adjust bmi1 avx2 smep bmi2 erms invpcid mpx rdseed adx smap clflushopt intel_pt xsaveopt xsavec xgetbv1 xsaves dtherm ida arat pln pts hwp hwp_notify hwp_act_window hwp_epp md_clear flush_l1d arch_capabilities
boven:~ #
Nice, but what has the technical noob now to understand from this?
In the other one there was v1, v2 or v3 if I am correct, where v1 is the āno Leap 16.0ā. But here is a mass of strange words without any significance for me.
Please understand that I try to get a recipe here for a certain type of openSUSE users so they can find out if it is OK for Leap 16.0 or not. Nothing more, nothing less.
If flag is available, you get string "Flags: ā¦ " with searched item coloured in red.
If flag is unavailable, you get blank output. You can test it with
lscpu | grep sse4_3
So
lscpu | grep sse4_2 || echo "not Leap 16.0 ready"
will do.
Thanks, I hope this will make the thread useful for those who wonder about their systems.
This topic already listed two ways to correctly detect x86_64 microarchitecture level. Please explain why you dismissed them in favour of the third, icomplete and potentially wrong, one?
I did a quick search but apparently missed them. Maybe you can provide the numbers of the posts.
In any case when there were already two, I still would be interested which one is the best and most telling. You say that now there is a third, which the more let me ask: can you experts please come to a definitive conclusion and provide a, at best 100%, certain way that an openSUSE end-user can see at a glance if the system can be used by Leap 15.6. This to make it possible for everybody, pro or not, to start searching for alternatives in time.
Sorry, but I fail to see there a straight, to be copied/pasted, command to be executed nor how to interpret the output that it delivers.
I assume that that is what people like to be offered. And I tried to get such a thing posted.
1 Like
Script tests flags. And it doesnāt test them for sse.
Does testing for flags rather than for sse affect the validity of the result? If so, what are the conditions when a test for flags will give a false positive? I ask because the script came up with the expected level for my three machines, v1 for the one from 2007, v2 for the one with the Celeron J3060 chip from 2018 and v3 for the one with the Core i7 chip from the same period.
Ah, donāt bother yourself. It is just objections for āold pepperboxā (āDer Alte peppermannā) Arvidjaar.
Also if ālmā flag (AMD64) is available, then SSE and SSE2 are also available, so there is no need to check them.
Iām afraid, this seems to be correct:
pluto:~ # inxi -Ca
CPU:
Info: model: Intel Core i7-8550U socket: BGA1356 (U3E1) note: check bits: 64
type: MT MCP arch: Coffee Lake gen: core 8 level: v3 note: check built: 2017
process: Intel 14nm family: 6 model-id: 0x8E (142) stepping: 0xA (10)
microcode: 0xF6
Topology: cpus: 1x cores: 4 tpc: 2 threads: 8 smt: enabled cache:
L1: 256 KiB desc: d-4x32 KiB; i-4x32 KiB L2: 1024 KiB desc: 4x256 KiB
L3: 8 MiB desc: 1x8 MiB
Speed (MHz): avg: 1075 high: 1473 min/max: 400/1800 base/boost: 2700/8300
scaling: driver: intel_pstate governor: powersave volts: 0.9 V
ext-clock: 100 MHz cores: 1: 1445 2: 1473 3: 832 4: 814 5: 954 6: 875
7: 1101 8: 1109 bogomips: 31999
Flags: avx avx2 ht lm nx pae sse sse2 sse3 sse4_1 sse4_2 ssse3 vmx
Inxi confirms level 3 on my laptop. Itās not too old, Iām ready to believe it.
But thenā¦
pluto:~ # lscpu | grep sse4_3
pluto:~ # lscpu | grep sse4_2
Markierungen: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc art arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc cpuid aperfmperf pni pclmulqdq dtes64 monitor ds_cpl vmx est tm2 ssse3 sdbg fma cx16 xtpr pdcm pcid sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave avx f16c rdrand lahf_lm abm 3dnowprefetch cpuid_fault epb invpcid_single pti ssbd ibrs ibpb stibp tpr_shadow vnmi flexpriority ept vpid ept_ad fsgsbase tsc_adjust bmi1 avx2 smep bmi2 erms invpcid mpx rdseed adx smap clflushopt intel_pt xsaveopt xsavec xgetbv1 xsaves dtherm arat pln pts hwp hwp_notify hwp_act_window hwp_epp md_clear flush_l1d arch_capabilities
pluto:~ #
Well, itās not yet a problem, level 2 is still fine, but still a bit unsettling.
A sidenote: My homeserver I thought to be dying is still running, just due to a lack of a) time and b) disaster. Now, I assume itās realling doomed:
kasi@lanserv:~> inxi -Ca
CPU:
Info: model: AMD Athlon II X2 250 bits: 64 type: MCP arch: K10 level: v1
built: 2009-13 process: AMD 45nm family: 0x10 (16) model-id: 6 stepping: 2
microcode: 0x1000098
Topology: cpus: 1x cores: 2 smt: <unsupported> cache: L1: 256 KiB
desc: d-2x64 KiB; i-2x64 KiB L2: 2 MiB desc: 2x1024 KiB
Speed (MHz): avg: 800 min/max: 800/3000 boost: disabled scaling:
driver: acpi-cpufreq governor: ondemand cores: 1: 800 2: 800 bogomips: 12054
Flags: ht lm nx pae sse sse2 sse3 sse4a svm
Itās doing its job and Iād prefer to keep it running for sake of sustainability and - even more - bother. However, I do understand the points mentionend against it.
Just out of interest, for my understanding:
If it can be done for Tumbleweed using whcaps, I assume it could be done for Leap. Itās not being done
Right?
This is really meant without any judgement. Iām not in the position to step up and give a lot of support for testing, I just canāt. In a few hours, Iām going to leave for my next business trip to India, Thailand and Taiwan. The issue here prevented me from upgrading my laptop. When I was ready to start, a productive (Win 11) machine of my wifeās office took priority seriously, followed by my laptop choosing this time suitable for a fatal hardware failure. It freezes even in the EFI menu.
Well, when I get back home, spring time is going to loom. Hm, Winter is my computer fixing time. Iāll be glad to get everything back to status quo and forget about anything I was planning to accomplish before gardening season keeps me outside for about 8 months.
Sorry, for the prose. I just thought of adding that being able to maintain stuff as long as possible has itās perk. 
(The main point was about inxi vs. lscpu.)
kasi
My idea was that there are a lot of people that know what they are talking about with respect to the subject and that thus would be able to offer a simple way for noobs to determine what their situation is.
IMHO the subject is scary for many and the discussion is not very enlightening to point to the one and most important factor for most: is my system for for Leap 15.6 or not, a simple āyesā or ānoā.
I am aware that this is Open Chat, thus maybe I should not have expected a useful helping technical answer for those, but thanks for the suggestion to everybody to find out on their own.
1 Like