Librem 5 delivery

Hardware Librem 5
21

@amosbatto

And with --cpu 4

 stress-ng --cpu 4 --tz -t 60 --metrics-brief --times --perf
stress-ng: info:  [2265] dispatching hogs: 4 cpu
stress-ng: info:  [2265] cache allocate: using defaults, can't determine cache details from sysfs
stress-ng: info:  [2265] successful run completed in 63.39s (1 min, 3.39 secs)
stress-ng: info:  [2265] stressor       bogo ops real time  usr time  sys time   bogo ops/s   bogo ops/s
stress-ng: info:  [2265]                           (secs)    (secs)    (secs)   (real time) (usr+sys time)
stress-ng: info:  [2265] cpu                1036     62.39    210.28      6.16        16.61         4.79
stress-ng: info:  [2265] cpu:
stress-ng: info:  [2265]            314,216,960,196 CPU Cycles                      4.96 B/sec
stress-ng: info:  [2265]             49,585,734,144 Instructions                    0.78 B/sec (0.158 instr. per cycle)
stress-ng: info:  [2265]              6,025,220,516 Branch Instructions            95.06 M/sec
stress-ng: info:  [2265]                139,224,400 Branch Misses                   2.20 M/sec ( 2.31%)
stress-ng: info:  [2265]            157,319,220,096 Bus Cycles                      2.48 B/sec
stress-ng: info:  [2265]              6,292,101,064 Cache References               99.27 M/sec
stress-ng: info:  [2265]                111,911,932 Cache Misses                    1.77 M/sec ( 1.78%)
stress-ng: info:  [2265]              6,073,478,780 Cache L1D Read                 95.82 M/sec
stress-ng: info:  [2265]                108,826,492 Cache L1D Read Miss             1.72 M/sec
stress-ng: info:  [2265]                333,586,320 Cache L1D Prefetch Miss         5.26 M/sec
stress-ng: info:  [2265]             32,789,797,608 Cache L1I Read                  0.52 B/sec
stress-ng: info:  [2265]                 71,246,040 Cache L1I Read Miss             1.12 M/sec
stress-ng: info:  [2265]                    206,168 Cache DTLB Read Miss            3.25 K/sec
stress-ng: info:  [2265]                     60,432 Cache ITLB Read Miss          953.41 /sec
stress-ng: info:  [2265]              9,699,408,624 Cache BPU Read                  0.15 B/sec
stress-ng: info:  [2265]                155,313,304 Cache BPU Read Miss             2.45 M/sec
stress-ng: info:  [2265]                278,173,244 Cache NODE Read                 4.39 M/sec
stress-ng: info:  [2265]                109,040,092 Cache NODE Write                1.72 M/sec
stress-ng: info:  [2265]            216,619,006,116 CPU Clock                       3.42 B/sec
stress-ng: info:  [2265]            216,730,367,076 Task Clock                      3.42 B/sec
stress-ng: info:  [2265]                      1,460 Page Faults Total              23.03 /sec
stress-ng: info:  [2265]                      1,460 Page Faults Minor              23.03 /sec
stress-ng: info:  [2265]                          0 Page Faults Major               0.00 /sec
stress-ng: info:  [2265]                      8,068 Context Switches              127.29 /sec
stress-ng: info:  [2265]                        272 CPU Migrations                  4.29 /sec
stress-ng: info:  [2265]                          0 Alignment Faults                0.00 /sec
stress-ng: info:  [2265]                          0 Emulation Faults                0.00 /sec
stress-ng: info:  [2265]                        396 System Call Enter               6.25 /sec
stress-ng: info:  [2265]                        392 System Call Exit                6.18 /sec
stress-ng: info:  [2265]                          0 Kmalloc                         0.00 /sec
stress-ng: info:  [2265]                          0 Kmalloc Node                    0.00 /sec
stress-ng: info:  [2265]                         40 Kfree                           0.63 /sec
stress-ng: info:  [2265]                         24 Kmem Cache Alloc                0.38 /sec
stress-ng: info:  [2265]                          0 Kmem Cache Alloc Node           0.00 /sec
stress-ng: info:  [2265]                        868 Kmem Cache Free                13.69 /sec
stress-ng: info:  [2265]                      1,404 MM Page Alloc                  22.15 /sec
stress-ng: info:  [2265]                         20 MM Page Free                    0.32 /sec
stress-ng: info:  [2265]                    142,320 RCU Utilization                 2.25 K/sec
stress-ng: info:  [2265]                        364 Sched Migrate Task              5.74 /sec
stress-ng: info:  [2265]                          0 Sched Move NUMA                 0.00 /sec
stress-ng: info:  [2265]                      6,508 Sched Wakeup                  102.67 /sec
stress-ng: info:  [2265]                          0 Sched Proc Exec                 0.00 /sec
stress-ng: info:  [2265]                          0 Sched Proc Exit                 0.00 /sec
stress-ng: info:  [2265]                          0 Sched Proc Fork                 0.00 /sec
stress-ng: info:  [2265]                          0 Sched Proc Free                 0.00 /sec
stress-ng: info:  [2265]                          0 Sched Proc Hang                 0.00 /sec
stress-ng: info:  [2265]                          0 Sched Proc Wait                 0.00 /sec
stress-ng: info:  [2265]                      8,064 Sched Switch                  127.22 /sec
stress-ng: info:  [2265]                         24 Signal Generate                 0.38 /sec
stress-ng: info:  [2265]                          4 Signal Deliver                  0.06 /sec
stress-ng: info:  [2265]                     57,284 IRQ Entry                     903.74 /sec
stress-ng: info:  [2265]                     57,284 IRQ Exit                      903.74 /sec
stress-ng: info:  [2265]                     55,576 Soft IRQ Entry                876.80 /sec
stress-ng: info:  [2265]                     55,576 Soft IRQ Exit                 876.80 /sec
stress-ng: info:  [2265]                          0 Writeback Dirty Inode           0.00 /sec
stress-ng: info:  [2265]                          0 Writeback Dirty Page            0.00 /sec
stress-ng: info:  [2265]                          0 Migrate MM Pages                0.00 /sec
stress-ng: info:  [2265]                          0 SKB Consume                     0.00 /sec
stress-ng: info:  [2265]                          0 SKB Kfree                       0.00 /sec
stress-ng: info:  [2265]                          0 IOMMU IO Page Fault             0.00 /sec
stress-ng: info:  [2265]                          0 IOMMU Map                       0.00 /sec
stress-ng: info:  [2265]                          0 IOMMU Unmap                     0.00 /sec
stress-ng: info:  [2265] cpu:
stress-ng: info:  [2265]          gpu-thermal   46.75 Ā°C
stress-ng: info:  [2265]     max170xx_battery   41.85 Ā°C
stress-ng: info:  [2265]          cpu-thermal   43.73 Ā°C
stress-ng: info:  [2265]          vpu-thermal   44.49 Ā°C
stress-ng: info:  [2265] for a 63.39s run time:
stress-ng: info:  [2265]     253.54s available CPU time
stress-ng: info:  [2265]     210.29s user time   ( 82.94%)
stress-ng: info:  [2265]       6.21s system time (  2.45%)
stress-ng: info:  [2265]     216.50s total time  ( 85.39%)
stress-ng: info:  [2265] load average: 3.04 1.22 0.72
3 Likes
22

@amosbatto

The phone say quite cool. I am on kernel 5.6 on amber-staging on a Chestnut.

Here is --cpu4 over 5 minutes.

 stress-ng --cpu 4 --tz -t 300 --metrics-brief --times --perf
stress-ng: info:  [2274] dispatching hogs: 4 cpu
stress-ng: info:  [2274] cache allocate: using defaults, can't determine cache details from sysfs
stress-ng: info:  [2274] successful run completed in 304.90s (5 mins, 4.90 secs)
stress-ng: info:  [2274] stressor       bogo ops real time  usr time  sys time   bogo ops/s   bogo ops/s
stress-ng: info:  [2274]                           (secs)    (secs)    (secs)   (real time) (usr+sys time)
stress-ng: info:  [2274] cpu                5364    303.24    819.82     28.95        17.69         6.32
stress-ng: info:  [2274] cpu:
stress-ng: info:  [2274]          1,207,391,126,780 CPU Cycles                      3.96 B/sec
stress-ng: info:  [2274]            259,011,245,400 Instructions                    0.85 B/sec (0.215 instr. per cycle)
stress-ng: info:  [2274]             31,492,183,508 Branch Instructions             0.10 B/sec
stress-ng: info:  [2274]                764,978,696 Branch Misses                   2.51 M/sec ( 2.43%)
stress-ng: info:  [2274]            604,864,392,848 Bus Cycles                      1.98 B/sec
stress-ng: info:  [2274]             33,563,896,468 Cache References                0.11 B/sec
stress-ng: info:  [2274]                555,537,244 Cache Misses                    1.82 M/sec ( 1.66%)
stress-ng: info:  [2274]             33,337,685,840 Cache L1D Read                  0.11 B/sec
stress-ng: info:  [2274]                560,831,760 Cache L1D Read Miss             1.84 M/sec
stress-ng: info:  [2274]              1,809,342,480 Cache L1D Prefetch Miss         5.93 M/sec
stress-ng: info:  [2274]            169,790,602,544 Cache L1I Read                  0.56 B/sec
stress-ng: info:  [2274]                306,096,348 Cache L1I Read Miss             1.00 M/sec
stress-ng: info:  [2274]                  4,458,644 Cache DTLB Read Miss           14.62 K/sec
stress-ng: info:  [2274]                    547,516 Cache ITLB Read Miss            1.80 K/sec
stress-ng: info:  [2274]             50,276,064,004 Cache BPU Read                  0.16 B/sec
stress-ng: info:  [2274]                789,914,384 Cache BPU Read Miss             2.59 M/sec
stress-ng: info:  [2274]              1,467,937,752 Cache NODE Read                 4.81 M/sec
stress-ng: info:  [2274]                477,670,408 Cache NODE Write                1.57 M/sec
stress-ng: info:  [2274]            828,976,173,252 CPU Clock                       2.72 B/sec
stress-ng: info:  [2274]            830,383,497,784 Task Clock                      2.72 B/sec
stress-ng: info:  [2274]                      2,844 Page Faults Total               9.33 /sec
stress-ng: info:  [2274]                      2,844 Page Faults Minor               9.33 /sec
stress-ng: info:  [2274]                          0 Page Faults Major               0.00 /sec
stress-ng: info:  [2274]                     70,460 Context Switches              231.10 /sec
stress-ng: info:  [2274]                      2,616 CPU Migrations                  8.58 /sec
stress-ng: info:  [2274]                          0 Alignment Faults                0.00 /sec
stress-ng: info:  [2274]                          0 Emulation Faults                0.00 /sec
stress-ng: info:  [2274]                        396 System Call Enter               1.30 /sec
stress-ng: info:  [2274]                        392 System Call Exit                1.29 /sec
stress-ng: info:  [2274]                         12 Kmalloc                         0.04 /sec
stress-ng: info:  [2274]                          0 Kmalloc Node                    0.00 /sec
stress-ng: info:  [2274]                      1,956 Kfree                           6.42 /sec
stress-ng: info:  [2274]                         60 Kmem Cache Alloc                0.20 /sec
stress-ng: info:  [2274]                          0 Kmem Cache Alloc Node           0.00 /sec
stress-ng: info:  [2274]                     26,628 Kmem Cache Free                87.33 /sec
stress-ng: info:  [2274]                      2,784 MM Page Alloc                   9.13 /sec
stress-ng: info:  [2274]                      1,952 MM Page Free                    6.40 /sec
stress-ng: info:  [2274]                    767,568 RCU Utilization                 2.52 K/sec
stress-ng: info:  [2274]                      3,844 Sched Migrate Task             12.61 /sec
stress-ng: info:  [2274]                          0 Sched Move NUMA                 0.00 /sec
stress-ng: info:  [2274]                     47,012 Sched Wakeup                  154.19 /sec
stress-ng: info:  [2274]                          0 Sched Proc Exec                 0.00 /sec
stress-ng: info:  [2274]                          0 Sched Proc Exit                 0.00 /sec
stress-ng: info:  [2274]                          0 Sched Proc Fork                 0.00 /sec
stress-ng: info:  [2274]                        120 Sched Proc Free                 0.39 /sec
stress-ng: info:  [2274]                          0 Sched Proc Hang                 0.00 /sec
stress-ng: info:  [2274]                          0 Sched Proc Wait                 0.00 /sec
stress-ng: info:  [2274]                     70,460 Sched Switch                  231.10 /sec
stress-ng: info:  [2274]                         44 Signal Generate                 0.14 /sec
stress-ng: info:  [2274]                          4 Signal Deliver                  0.01 /sec
stress-ng: info:  [2274]                    232,408 IRQ Entry                     762.26 /sec
stress-ng: info:  [2274]                    232,408 IRQ Exit                      762.26 /sec
stress-ng: info:  [2274]                    283,132 Soft IRQ Entry                928.62 /sec
stress-ng: info:  [2274]                    283,132 Soft IRQ Exit                 928.62 /sec
stress-ng: info:  [2274]                          0 Writeback Dirty Inode           0.00 /sec
stress-ng: info:  [2274]                          0 Writeback Dirty Page            0.00 /sec
stress-ng: info:  [2274]                          0 Migrate MM Pages                0.00 /sec
stress-ng: info:  [2274]                         12 SKB Consume                     0.04 /sec
stress-ng: info:  [2274]                          0 SKB Kfree                       0.00 /sec
stress-ng: info:  [2274]                          0 IOMMU IO Page Fault             0.00 /sec
stress-ng: info:  [2274]                          0 IOMMU Map                       0.00 /sec
stress-ng: info:  [2274]                          0 IOMMU Unmap                     0.00 /sec
stress-ng: info:  [2274] cpu:
stress-ng: info:  [2274]          gpu-thermal   49.75 Ā°C
stress-ng: info:  [2274]     max170xx_battery   44.41 Ā°C
stress-ng: info:  [2274]          cpu-thermal   46.52 Ā°C
stress-ng: info:  [2274]          vpu-thermal   47.33 Ā°C
stress-ng: info:  [2274] for a 304.90s run time:
stress-ng: info:  [2274]    1219.58s available CPU time
stress-ng: info:  [2274]     819.84s user time   ( 67.22%)
stress-ng: info:  [2274]      29.04s system time (  2.38%)
stress-ng: info:  [2274]     848.88s total time  ( 69.60%)
stress-ng: info:  [2274] load average: 5.04 3.86 2.04
4 Likes
23

Please do. If weā€™re going to have Lollypop letā€™s have the whole thing. Iā€™m ok with sending just sound to a bt speaker but that may no be good enough later! Thanks for this update!

2 Likes
24

@joadri , @FamousJameous - It would be even better if you posted a request upstream at

I very much doubt that Purism would make changes that are Purism-specific - that kind of changes are better suited upstream.

10 Likes
25

Hey FamousJameous thanks for all the info,

So just to be totally clear, phone calls work without making use of any internet based apps?

For example you make calls off the phone SIM which uses regular masts to connect the call, you do not have to be connected to the internet to make calls is that correct?

26

So what do we all reckon for the release of Evergreen & Fir?

Will Evergreen be out for August 2020? And will it be a much more ā€œuse out of the boxā€ experience?

PinePhoone are now / soon releasing the UBports version of the pinephone, has any looked at that?

For me the Librem 5 seems the best product on the market, I am hoping lots more issues have been resolved for Evergreen, but should I purchase an Evergreen or a Fir?

27

Hey guys something I forgot to add to my last post,

All the apps on the Librem phone, such as calculator, calander, and so on, can you freely use them ok?

What I mean is, with IOS & Android you have to be singed into GMAIL or another email account in order to have access to these very basic apps, are they totally ā€œfreeā€ on the Librem 5, or do you have to sign in using an Purism email account?

Thanks in advance

LinuxNew

28

Sorry, I probably confused things by talking about that FreedomPop stuff. To answer your question directly: Yes, phone calls work using the included Calls app when using a standard SIM card, no extra internet based app is required.

And for your follow up question, I did not have to sign into any accounts on the phone to be able to use everything. I could still use all of the included apps as well as using the Software Store to install additional apps. No account is required.

2 Likes
29

FamousJameous,

You are a gentleman, thank you for all the info thatā€™s great. I am conflicted between getting a Evergreen or Fir, as the Fir will have a better CPU, what are your / everyones thoughts?

Hope you guys are all staying safe.

Thanks in advance

LinuxNew

30

I donā€™t know yet about the Evergreen vs Fir. I think it will depend on how far Purism can optimize power use/battery life before Evergreen. If the battery life is improved significantly, it may be worth going with Evergreen just to get it sooner. I am sure Fir will provide significant improvements when Purism implements everything they have learned in a new version where they have more time than they are getting between batches now. But that is more time to wait for a new phone.

The thermal improvements from the Dogwood CPU flip are probably also worth watching. I havenā€™t had a problem with the thermals since several months ago, so that isnā€™t a huge deal to me. But better thermals likely mean longer life from the processor and other components, so that is a welcome change.

So, I donā€™t really have a concrete answer for you, but I would also be interested in hearing what other people are thinking about Evergreen vs Fir.

2 Likes
31

but a less beefier GPU ā€¦ also will there even be a 2021 ?

1 Like
32

The Librem 5 has a history of delays, but Purism might start shipping by August. Even if it ships, there will almost certainly still be major software issues to resolve. The camera app is still in the early design stages and Purism just got code in the kernel so it can detect one the cameras. I doubt that all the power management issues will be resolved, since there are still a lot of components to work on to conserve energy. Being able to put components to sleep, but then wake them up when a notification or call is received will take quite a bit of work.

The major issue with power management in the PinePhone has been resolved and it can now get up to 15 hours in sleep mode while waking up to receive phone calls. Canonical spent years working on Ubuntu Touch, so UBports is more polished and complete than GTK+Phosh as a phone interface. I personally prefer the design of Phosh over UBports. If you want to try out Phosh, then download the Librem 5 image and run it in Qemu on your PC to try out the interface.

If you want a working Linux phone as soon as possible or are on a tight budget, then you probably want the PinePhone. Also, keep in mind that the Librem 5 is 6.3 mm thicker than the PinePhone and 45 grams heavier. My article might help you decide between the two phones.

As for Evergreen vs Fir, see this post about the i.MX 8M Quad vs Plus. If you donā€™t mind being an early adopter and dealing with bugs, then you probably want Evergreen. If you just want stuff to work, than you will probably be happier waiting for Fir, since it will have a more energy-efficient processor and most of the major software issues will probably be resolved by the time it is released, but youā€™ll have to wait until next year. You can be first in the queue for Fir by ordering now and then sending an email to Purism saying that you want Fir.

1 Like
33

I ordered my L5 on Jan. 11 , 2018. Iā€™m waiting as long as I have to till my Evergreen comes. In the meantime I ordered a Pinephone thatā€™s supposed to ship later this month. That will be interesting too. And if I got the cash when Fir comes out Iā€™m getting that too. A buddy of mine is a Linux master who guides me as i donā€™t know RAM from ROM. I run Linux Mint 19.3 on my NUC desktop. Linux is so much cheaper , more reliable and forgiving than any of the Windows/IOS junk and iPhones were. An Apple update fried my Mac Pro and that was the straw that broke my camels back.
You guys know what youā€™re doing and will be my tech support. So I got nothing to lose. I hate Google , Apple and data harvestors in general. So Iā€™m all in for the L5! That Purism Mini looks good too but Iā€™m not sure about PureOs yet.

4 Likes
34

@purismforum, Thanks for posting the stress-ng numbers. I ran the same test on a PC with an Intel Core i5-5200U (2.20GHz) and it is roughly 12.5 times more powerful. However, the documentation for stress-ng says that it shouldnā€™t be used for benchmarking, because it isnā€™t very reliable.

Would you mind running these tests?
sudo apt install sysbench
sysbench --test=cpu run
sysbench --test=memory run
sysbench --test=fileio --file-test-mode=seqwr run

Iā€™m going to also run them on my Raspberry Pi 4, so we can do a comparison. If anyone has a PinePhone, it would be cool to see the results.

3 Likes
35

@amosbatto

I got sysbench from https://github.com/akopytov/sysbench since it is not available in PureOS repo.

sysbench results

2 Likes
36

Thanks @purismforum. Would you mind running it for 4 threads?
sysbench --threads=4 cpu run

1 Like
37

Yea I had that already, sorry I pasted the wrong link for the ghist.Here is the right link to all the tests

5 Likes
38

Thats awesome thanks amosbatto

39

Is this result as expected? This means by executing xrandr command output will be of approximately 30 Hz refresh rate (720p @ 30 Hz). @purismforum, please :blush: run for us vblank_mode=0 glxgears as well.

Is it GPU locked into a lower frame rate to be gentle toward battery, could someone give any kind of explanation if this was usual and necessary for Chestnut, user adjustable and how, related to DP alt mode over USB-C (upscaling ready)?

40

Wow this is quite a good performance, isnā€™t it?
I can only compare it to a Raspberry Pi 3B+ which is significantly slowerā€¦

1 Like