Topic: [ANN] cpuminer-multi v1.3.5 (Linux + Windows VStudio/MinGW64) GPL Open Source - page 14. (Read 378844 times)

Would someone be kind enough to post a compiled version for Windows of the newest release?

Thanks!

And dangerous.

i never mining its very dificult

thankyou.............

I have a couple of questions.

While mining some algos I get very few accepts and in some case none for over 30 minutes. Also the hash
rate reported at the pool is very low or zero, much lower than the miner's display.

I presume that no blocks are being found because there are no validation errors and no rejects. Is this a
difficulty issue and is there a way around it?

I also have a question about lyra2v2. I was looking forward to it due to the exceptional performance of cpuminer-multi
on lyra2 (v1). IIRC you were stuck because one of the sub-algos was non standard. Is there any progress or have you
dropped it for other projects?

I'm also working on something you might be interested in, more on that at a later time.

Edit: I sorted out the low hash rate issue, the problem was at my end.

in recent kernels and/or cpu its :

note the extra physical one.. if required, edit sysinfos.c

  CONFIG_SENSORS_CORETEMP is built-in to the kernel

    I figured the sensor reporting through lm-sensors is using the coretemp driver since it shows "coretemp-isa-0000" as I pasted in the earlier post.
    The kernel documentation says CONFIG_SENSORS_CORETEMP is good for family 6 Core2 processors.
    The /proc/cpuinfo output from my CPU would appear to meet that condition.

    Any way to debug it?
    I'll try on another system to see if temperature is reported via the API there.
    Also, any thoughts on whether the checkout from HEAD could be problematic?
    I'm using the commit from 12/27/2015, fb137363702141114c45c3e20a5aecb4c69cd919

Kind Regards,
-Chicago

    processor       : 0
    vendor_id       : GenuineIntel
    cpu family      : 6
    model           : 23
    model name      : Intel(R) Core(TM)2 Quad CPU    Q9550  @ 2.83GHz
    stepping        : 10
    microcode       : 0xa0b
    cpu MHz         : 2826.259
    cache size      : 6144 KB
    physical id     : 0
    siblings        : 4
    core id         : 0
    cpu cores       : 4
    apicid          : 0
    initial apicid  : 0
    fpu             : yes
    fpu_exception   : yes
    cpuid level     : 13
    wp              : yes
    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 lm constant_tsc arch_perfmon pebs bts rep_good nopl aperfmperf pni dtes64 monitor ds_cpl vmx smx est tm2 ssse3 cx16 xtpr pdcm sse4_1 xsave lahf_lm dtherm tpr_shadow vnmi flexpriority
    bugs            :
    bogomips        : 5652.51
    clflush size    : 64
    cache_alignment : 64
    address sizes   : 36 bits physical, 48 bits virtual
    power management:

you need to "modprobe coretemp" kernel module which is common but not loaded automatically

else the api hashrate stats are only updated on shares (you need at least one accept or reject) its there cpuminer do the sum of all threads

Hi Epsylon3,

    Okay, the cpu-affinity makes sense to me now.

    The confusing piece was having { "threads": 0 } along with any defined cpu-affinity, because in that case all of the available processors would be put to use.
    But, once I adjusted my json to have { "threads": 2 } along with { "cpu-affinity": 12 }, cpuminer ran on just the last two cores as expected.

    Thanks again for describing how the binary mask works.

    Next question is about the max-temp option.
    At idle, lm-sensors reports my Q9550 Intel Core2 Quad CPU with the following temperature.

    coretemp-isa-0000
    Adapter: ISA adapter
    Core 0:       +44.0°C  (high = +78.0°C, crit = +100.0°C)
    Core 1:       +42.0°C  (high = +78.0°C, crit = +100.0°C)
    Core 2:       +44.0°C  (high = +78.0°C, crit = +100.0°C)
    Core 3:       +44.0°C  (high = +78.0°C, crit = +100.0°C)

----------

    To test the capability of max-temp, I set { "threads": 0 } and { "cpu-affinity": 15 } in order to run on all cores and let cpuminer-multi heat up the CPU for a little bit.

    A short while later, lm-sensors reported the following.

    coretemp-isa-0000
    Adapter: ISA adapter
    Core 0:       +60.0°C  (high = +78.0°C, crit = +100.0°C)
    Core 1:       +53.0°C  (high = +78.0°C, crit = +100.0°C)
    Core 2:       +55.0°C  (high = +78.0°C, crit = +100.0°C)
    Core 3:       +58.0°C  (high = +78.0°C, crit = +100.0°C)

    Within my json configuration I have set { "max-temp": 45 }, expecting cpuminer-multi to shutoff nearly as quickly as it started.
    Unfortunately, it appears cpuminer-multi does not register my temperature.

    The summary report from the miner's API outputs a TEMP=0.0.

    telnet 127.0.0.1 4048
    Trying 127.0.0.1...
    Connected to 127.0.0.1.
    Escape character is '^]'.
    summary
    NAME=cpuminer-multi;VER=1.2-dev;API=1.1;ALGO=x15;CPUS=4;KHS=0.00;SOLV=0;ACC=0;REJ=0;ACCMN=0.000;DIFF=2.450681;TEMP=0.0;FAN=0;FREQ=0;UPTIME=354;TS=1452097962|Connection closed by foreign host.

    I don't know if this is related, but note above KHS=0.00 as well.

    However, the threads report from the miner's API outputs actual decimal values.

    telnet 127.0.0.1 4048
    Trying 127.0.0.1...
    Connected to 127.0.0.1.
    Escape character is '^]'.
    threads
    CPU=0;KHS=17.45|CPU=1;KHS=17.78|CPU=2;KHS=16.07|CPU=3;KHS=16.27|Connection closed by foreign host.

    Any suggestions would be greatly appreciated.

    Thanks so much for your kind support.

Kind Regards,
-Chicago

Thanks Epsylon3,

    I will test the cpu-affinity with the binary mask as you have described.

    Just 4 cores accessible on this box but I can get on a few larger systems where it would be more complex and I can put together a table.

Kind Regards,
-Chicago

for the cpu-affinity, its a binary mask

so for cores 0 and 2, it's 2^0 + 2^2 = 1 + 4 = 5
and for 1 and 7 its 2^1 + 2^7 = 2 + 128 = 130

ive seen some people doing a commit about that, maybe there is something to fix but seems ok here.


For neoscrypt, i never used the N factor, its for possible variants of the algo (never seen)

the "=" symbol is optional on the command line

Hi Epsylon3,

    Thanks for your work on this cpuminer.

    Would you kindly share a couple of more examples for --cpu-affinity usage.
    Also, sorry for being ignorant here, but the --help shows the --nfactor as "neoscrypt N-Factor" and I was figuring if it were to be followed by a value it my look better as --nfactor=N.

    Is that right?

Kind Regards,
-Chicago

tx for the report, i fixed the chmod and the asm/ path in the script

not sure that fix the neoscrypt problem but its a start :p

I could not get it to build on OS X. I've been trying to fix this for over 6 hours, so maybe this could help someone.

I'm on Mac OS X 10.11.1. Here is what I'm doing, exactly a per README instructions:


I am presented with an error about macros. You moved the asm sources into an "asm" folder, but you didn't change the nomacro.pl to parse files in that folder. So I try again:


Then I get this following error (full compile log is here: http://pastebin.com/nj0hQnTa)


It seems like this ASM code isn't working for neoscrypt, for some reason... So I try again once more now disabling ASM manually


Now it worked

Good Job! Dude thank you for the information.

Hi Epsylon,

greetings for great job around cpu-miner - this is very good result for me..

I also have little troubles with compiling x86 version, who can work with automatic start from .json config file..

Can you compile one copy for me please?

regards


Edit: My x64 version also no work with .json file.. maybe I miss something...

Anyone can help me to compile two wersions with work .json option file please?

Thanks

nope, do it if you find that useful :p

Hi Epsylon3
Is it possible to support encrypted cpuminer-conf.json, such as base64?

Try and learn... A forum is not a chat.. Check the vcxproj content

Hmm, I will be able to compile x86 miner without adding to project x86 libs? Just to place them in folder, yes?