Bitcoin Forum
July 16, 2018, 09:39:29 AM *
News: Latest stable version of Bitcoin Core: 0.16.1  [Torrent]. (New!)
 
   Home   Help Search Donate Login Register  
Pages: « 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 [51] 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 ... 191 »
  Print  
Author Topic: [ANN]: cpuminer-opt v3.8.8.1, open source optimized multi-algo CPU miner  (Read 411302 times)
joblo
Legendary
*
Offline Offline

Activity: 1148
Merit: 1016


View Profile
August 06, 2016, 01:19:53 AM
 #1001


Theoretically yes if there exists any earlier executed code that contained compiler produced AVX2 instructions from regular source.
That isn't likely since the capabilities check is done ealy in main.

Let's not forget that we ask gcc to compile and optimize (all those -O2 -O3 -Ofast) for the cpu it's being run on. So regardless whether you actually include any explicit AVX/AVX2 assembler in the code, even a simple printf("hi"); may produce AVX2 instruction(s) if the compiler feels like it. That's the whole point of the compiler compiling for the given cpu (-march=native) - it's allowed to use all the capabilities (and thus instruction sets) of the cpu.


That's what I was referring to when I wrote "compiler produced AVX2". AVX(2) provides SIMD instructions an it's unlikely something
like a printf would use it. A memcpy wouldn't use it because of the overhead of loading/storing the data to/from the ymm regs.
It's only useful for vector arith, and apparently the compiler isn't smart enough to convert conventionally coded array processing
loops to AVX2. I'm not even sure *if* the compiler can optimize in this fashion, the existance of so much hand coded AVX2 suggests
otherwise.

Since we're playing semantic games would you care to explain your concerns with my use of the term cross compiling?
IMO cross compiling can mean any compilation not done on the target machine and not executable on the build machine.

And my comment about you maybe using a core2 was based on the symptoms you decribed and that some server CPUs can
be optimized for efficiency by removing/disabling unneeded features like floating point, AES or AVX.

Not true. memcpy() and friends CAN and do use SSE/AVX - if the source/dests are aligned properly.

It doesn't really matter in this context whether it crashes before or after the warning message.

I'll take your word for it, but it doesn't seem to make much sense. It is essentially load, move, store, 256 bits wide. Where
are the savings? I presume it takes longer to load data into the ymm regs than general purpose ones. The same amount of data
has to be moved around in memory. Using AVX seems to make sense if you're going to do a lot of processing of the data while
in vector format.

There's my strawman, rip it apart.

Speaking of alignment I need to fix that up in my avx code. I used all loadu/storeu for convenience.

If it's aligned, then the load/stores don't take nearly as long. Also keep in mind that there's no such thing as a mov memaddr, memaddr opcode in x86 that I know of. Therefore, it's gotta go in a register (this is simplified, I know about things like DMA, but they don't come into play for the purposes of this discussion) and if it's aligned, it makes one hell of a lot more sense to stuff it in an AVX register, because it's a lot wider than a GPR. Even better if you're doing some kind of gather-scatter shit, possibly.

OK I'm with you now, There is a significant savings in instruction count. As long as AVX instructions execute as fast as regular ones
and the code is not data bound it will be faster.

cpuminer-opt developer, https://bitcointalk.org/index.php?topic=1326803.0
BTC: 12tdvfF7KmAsihBXQXynT6E6th2c2pByTT,
ETH: 0x72122edabcae9d3f57eab0729305a425f6fef6d0
1531733969
Hero Member
*
Offline Offline

Posts: 1531733969

View Profile Personal Message (Offline)

Ignore
1531733969
Reply with quote  #2

1531733969
Report to moderator
1531733969
Hero Member
*
Offline Offline

Posts: 1531733969

View Profile Personal Message (Offline)

Ignore
1531733969
Reply with quote  #2

1531733969
Report to moderator
1531733969
Hero Member
*
Offline Offline

Posts: 1531733969

View Profile Personal Message (Offline)

Ignore
1531733969
Reply with quote  #2

1531733969
Report to moderator
Automated Bitcoin Fork Extraction Tool WE DO TOUGH WALLETS: BCH | BTG | BCD | SBTC | UBTC | B2X | BCX | BTF Electrum 2FA, Trezor, Ledger, SegWit, Bech32
Advertised sites are not endorsed by the Bitcoin Forum. They may be unsafe, untrustworthy, or illegal in your jurisdiction. Advertise here.
1531733969
Hero Member
*
Offline Offline

Posts: 1531733969

View Profile Personal Message (Offline)

Ignore
1531733969
Reply with quote  #2

1531733969
Report to moderator
1531733969
Hero Member
*
Offline Offline

Posts: 1531733969

View Profile Personal Message (Offline)

Ignore
1531733969
Reply with quote  #2

1531733969
Report to moderator
1531733969
Hero Member
*
Offline Offline

Posts: 1531733969

View Profile Personal Message (Offline)

Ignore
1531733969
Reply with quote  #2

1531733969
Report to moderator
johnsmithx
Hero Member
*****
Offline Offline

Activity: 588
Merit: 507

I don't buy nor sell anything here and never will.


View Profile
August 06, 2016, 02:16:41 AM
 #1002

That's what I was referring to when I wrote "compiler produced AVX2". AVX(2) provides SIMD instructions an it's unlikely something
like a printf would use it. A memcpy wouldn't use it because of the overhead of loading/storing the data to/from the ymm regs.
It's only useful for vector arith, and apparently the compiler isn't smart enough to convert conventionally coded array processing
loops to AVX2. I'm not even sure *if* the compiler can optimize in this fashion, the existance of so much hand coded AVX2 suggests
otherwise.

I wasn't talking about what is or isn't likely but what is theoretically possible. With -O flags and -march=native the compiler can use all instruction sets it has to its disposal and if and how it decides to translate a higher language into a machine code is only up to it, so if it feels like it it will use AVX2 for printf and that's that.

Since we're playing semantic games would you care to explain your concerns with my use of the term cross compiling?
IMO cross compiling can mean any compilation not done on the target machine and not executable on the build machine.

Of course you can have any opinion you want, that's your freedom, but that doesn't change the fact that cross compiling is when a different host/target is desired. Furthermore, your own personal definition doesn't fit this narrative either because we were talking about compiling on avx2 cpu and producing sse code. In such a case the code is executable on the build machine so not even by your very own definition would it be cross compiling.

And my comment about you maybe using a core2 was based on the symptoms you decribed and that some server CPUs can
be optimized for efficiency by removing/disabling unneeded features like floating point, AES or AVX.

No, core2 is not and never was a server cpu and I was very clearly talking about real servers in the data centers, not some home desktop computers that you call "servers". I started the whole discussion with complaining about Nicehash not working on my servers anymore and later I confirmed someone else's suggestion that it may be because I am using the same BTC address on all my machines, there are tens of them and each has a different IP. That in itself basically rules out any possible doubt because I could hardly have tens of computers in my living room and even then no ordinary household ISP would give me tens of public IPs. They don't even give a single public IP per household around here anymore by default, you have to ask for it and pay for it. And even furthermore, if I was to have tens of mining "servers" (i.e. desktops) in my home, full of power-hungry GPUs, I wouldn't be able to power them all because an ordinary household has a 20A fuse, that would translate into some 3500 Watts if the power supplies had 80% efficiency. You can maybe power 5 relatively decent mining machines with that but definitely not tens.


Now I don't want to sound harsh but this is enough. I very much support your efforts and try to help you and am eager to discuss any technical topic but I will not play this game, especially not at this level. I don't actually mind word games but here you have no real argument in any of what you just said and yet it seems as if you were trying to kind of get back at me regardless and that is just pointless and for you humiliating and for both of us a waste of time. I wouldn't dare to say a word about anything related to your AVX/AVX2 optimizing but when I do say something I try to pay attention on my wording so if you want to look for my errors you would have to try harder but if you really did that you would only confirm a very immature personality. I merely, totally innocently, rightfully and truthfully pointed out, in the brackets even, that cross compiling is something else than what you said and that triggered you and turned you into a rogue word warrior trying to reclaim his supremacy? Come on, kids do that, not adults.

My list of 44(+1) reviewed Bitcoin forks | You don't have to download the pre-fork blockchain again for each fork! | Beware of fraudulent AWS accounts sellers and dangerous edu AWS codes! + My personal list of legit sellers and scammers | Never publicly reveal your btc addresses, ownership or any other details and stay very far away from anybody who asks you to! | The general rule of safe buying is: if the seller is a newbie, with no reputation, with no topic nor trust feedback, offering no vouches and/or selling from a locked or self-moderated topic and unwilling to go first or use escrow => AVOID. Always check the trust feedback first and make sure that you have enabled "Show untrusted feedback by default" in "Profile / Forum Profile Information".
joblo
Legendary
*
Offline Offline

Activity: 1148
Merit: 1016


View Profile
August 06, 2016, 03:51:42 AM
 #1003

[rambling rant removed]

Holy crap calm down. What does Nicehash and IP limiting have to do with cpuminer-opt? On second thought don't
answer that.

I've spent considerable time trying to help you with your compile problem and you've been needling me
from day one about conflating old CPUs with server class CPUs. I know damn well what a server class CPU is.
Some server class CPUs, not necessarily Intel or x86, have the ability to disable functional units if they aren't needed
in a specific environment. It was all irrelevant because all I wanted was to rule out a CPU issue and did so with your first
post with the miner output identifying your CPU's capabilities. I didn't know or care whether your CPU was bigger than
my CPU I just wanted to confirm it had AVX2. And I thought we had moved on.

Then you needle me again about cross compiling. I also let that go at the time but decided to bring up these issues
in a lighthearted way once the main issue was mostly understood. But you respond with an incoherent rant.

There was no point in bringing up semantics, and the strict definition of cross compiling or that I didn't use it correctly
or in context. There was no ambiguity in my language, a rarity it seems even with native english speakers. Iam not
referring to you, you have been very clear.

What was your point in mentioning these trivial issues if you didn't expect me to respond to your jabs?

cpuminer-opt developer, https://bitcointalk.org/index.php?topic=1326803.0
BTC: 12tdvfF7KmAsihBXQXynT6E6th2c2pByTT,
ETH: 0x72122edabcae9d3f57eab0729305a425f6fef6d0
joblo
Legendary
*
Offline Offline

Activity: 1148
Merit: 1016


View Profile
August 06, 2016, 03:42:52 PM
 #1004

Back to our regularly scheduled program. In the last episode it was discovered there are two copies of the
scryptjane code, one for the scryptjane algo itself and one for argon2 algo. The former compiles, the latter doesn't using
gcc 5.4.0 using the -flto option.

The main difference betwen these two file sets is in the presence of AVX2 code in the argon2 copy. The technique
for selecting a target at compile time is the same in both copies. There are some small code differences between the duplicated files
but it is unclear if they are functional or if they would affect compilation, specifically LTO optimization.

The plan forward is to create a build environment using gcc 5.4.0 to do more testing and comparing of the two scryptjane
versions with the goals of finding the cause of the compile error, merging the two copies of scryptjane and using the AVX2
code to optimize the scryptjane algo.

In the meantime users of gcc 5.4.0 can either compile without -flto or apply the workaround to disable argon2 and the GRS
SSE2 code as previusly described.

cpuminer-opt developer, https://bitcointalk.org/index.php?topic=1326803.0
BTC: 12tdvfF7KmAsihBXQXynT6E6th2c2pByTT,
ETH: 0x72122edabcae9d3f57eab0729305a425f6fef6d0
MrPlazu
Jr. Member
*
Offline Offline

Activity: 54
Merit: 0


View Profile
August 07, 2016, 05:36:39 PM
 #1005

I'm tring to get this version to Work on Ubuntu 16 with algo HMQ1725.

./cpuminer -a hmq1725 -o stratum+tcp://esp.maxminers.net:5872 -u user.worker -p 123

All I get is this

CPU:            Intel(R) Xeon(R) CPU @ 2.30GHz
CPU features: SSE2 AES AVX AVX2
SW built on Aug  7 2016 with GCC 5.4.0
SW features: SSE2 AES AVX AVX2
Algo features: SSE2 AES AVX AVX2
Start mining with SSE2 AES AVX AVX2

[2016-08-07 17:33:33] Starting Stratum on stratum+tcp://esp.maxminers.net:5872
[2016-08-07 17:33:33] 8 miner threads started, using 'hmq1725' algorithm.
[2016-08-07 17:33:37] Stratum difficulty set to 1
[2016-08-07 17:33:37] esp.maxminers.net:5872 hmq1725 block 521805
[2016-08-07 17:33:38] CPU #1: 6630 H, 25.98 kH/s
[2016-08-07 17:33:38] accepted: 0/1 (0.0%), 6630 H, 25.98 kH/s nooooo
[2016-08-07 17:33:38] reject reason: low difficulty share of 0.000022975398271283902
[2016-08-07 17:33:38] factor reduced to : 0.67
[2016-08-07 17:33:38] CPU #2: 17.57 kH, 25.69 kH/s
[2016-08-07 17:33:38] accepted: 0/2 (0.0%), 24.20 kH, 51.67 kH/s nooooo
[2016-08-07 17:33:38] reject reason: low difficulty share of 0.00007275697698863166
[2016-08-07 17:33:38] factor reduced to : 0.44
[2016-08-07 17:33:39] CPU #4: 32.69 kH, 26.84 kH/s
[2016-08-07 17:33:39] accepted: 0/3 (0.0%), 56.89 kH, 78.51 kH/s nooooo
[2016-08-07 17:33:39] reject reason: low difficulty share of 0.00001634549809734534
[2016-08-07 17:33:39] factor reduced to : 0.30
[2016-08-07 17:33:39] CPU #3: 35.29 kH, 23.53 kH/s
[2016-08-07 17:33:39] CPU #6: 39.41 kH, 26.12 kH/s
[2016-08-07 17:33:39] accepted: 0/4 (0.0%), 131.60 kH, 128.16 kH/s nooooo
[2016-08-07 17:33:39] reject reason: low difficulty share of 0.000017716213142858426
[2016-08-07 17:33:39] factor reduced to : 0.20
[2016-08-07 17:33:39] accepted: 0/5 (0.0%), 131.60 kH, 128.16 kH/s nooooo
[2016-08-07 17:33:39] reject reason: low difficulty share of 0.00004261769515069645
[2016-08-07 17:33:39] factor reduced to : 0.13
[2016-08-07 17:33:39] CPU #1: 34.05 kH, 24.98 kH/s
[2016-08-07 17:33:39] accepted: 0/6 (0.0%), 159.02 kH, 127.16 kH/s nooooo
[2016-08-07 17:33:39] reject reason: low difficulty share of 0.00005388752541995255
[2016-08-07 17:33:39] factor reduced to : 0.09
[2016-08-07 17:33:39] esp.maxminers.net:5872 hmq1725 block 521806
[2016-08-07 17:33:39] CPU #2: 24.05 kH, 23.01 kH/s
[2016-08-07 17:33:39] CPU #6: 3348 H, 15.04 kH/s
[2016-08-07 17:33:39] CPU #1: 1804 H, 15.04 kH/s
[2016-08-07 17:33:39] CPU #4: 10.57 kH, 20.45 kH/s
[2016-08-07 17:33:39] CPU #7: 41.26 kH, 23.86 kH/s
[2016-08-07 17:33:39] CPU #0: 40.93 kH, 23.52 kH/s
[2016-08-07 17:33:39] CPU #3: 3664 H, 15.51 kH/s
[2016-08-07 17:33:39] CPU #5: 43.05 kH, 24.83 kH/s

And after a min or to

[2016-08-07 17:34:39] esp.maxminers.net:5872 hmq1725 block 521810
[2016-08-07 17:34:39] CPU #2: 153.68 kH, 15.52 kH/s
[2016-08-07 17:34:39] CPU #6: 123.07 kH, 15.50 kH/s
[2016-08-07 17:34:39] CPU #4: 151.63 kH, 15.31 kH/s
[2016-08-07 17:34:39] CPU #5: 153.09 kH, 15.46 kH/s
[2016-08-07 17:34:39] CPU #7: 151.47 kH, 15.30 kH/s
[2016-08-07 17:34:39] CPU #0: 152.34 kH, 15.39 kH/s
[2016-08-07 17:34:39] CPU #1: 149.37 kH, 15.09 kH/s
[2016-08-07 17:34:39] CPU #3: 149.26 kH, 15.07 kH/s
[2016-08-07 17:34:48] esp.maxminers.net:5872 hmq1725 block 521811
[2016-08-07 17:34:48] CPU #5: 139.25 kH, 15.46 kH/s
[2016-08-07 17:34:48] CPU #7: 138.15 kH, 15.34 kH/s
[2016-08-07 17:34:48] CPU #0: 138.86 kH, 15.42 kH/s
[2016-08-07 17:34:48] CPU #2: 140.02 kH, 15.54 kH/s
[2016-08-07 17:34:48] CPU #1: 136.06 kH, 15.10 kH/s
[2016-08-07 17:34:48] CPU #6: 139.75 kH, 15.51 kH/s
[2016-08-07 17:34:48] CPU #4: 138.34 kH, 15.36 kH/s
[2016-08-07 17:34:48] CPU #3: 136.03 kH, 15.10 kH/s
[2016-08-07 17:34:59] esp.maxminers.net:5872 hmq1725 block 521812
[2016-08-07 17:34:59] CPU #7: 168.57 kH, 15.31 kH/s
[2016-08-07 17:34:59] CPU #4: 168.87 kH, 15.34 kH/s
[2016-08-07 17:34:59] CPU #6: 170.61 kH, 15.50 kH/s
[2016-08-07 17:34:59] CPU #5: 170.04 kH, 15.44 kH/s
[2016-08-07 17:34:59] CPU #2: 170.90 kH, 15.52 kH/s
[2016-08-07 17:34:59] CPU #0: 169.72 kH, 15.42 kH/s
[2016-08-07 17:34:59] CPU #1: 166.12 kH, 15.09 kH/s
[2016-08-07 17:34:59] CPU #3: 166.18 kH, 15.09 kH/s

And that just go on and on and on

I have another machine with a older version where it's working. What I'm doing wrong here??
joblo
Legendary
*
Offline Offline

Activity: 1148
Merit: 1016


View Profile
August 07, 2016, 07:44:14 PM
 #1006

I'm tring to get this version to Work on Ubuntu 16 with algo HMQ1725.


I can't test this anymore at Suprnova. I have an idea what's wrong, it would help if you can tell me the last version
where it worked. I should easilly be able to restore it.

cpuminer-opt developer, https://bitcointalk.org/index.php?topic=1326803.0
BTC: 12tdvfF7KmAsihBXQXynT6E6th2c2pByTT,
ETH: 0x72122edabcae9d3f57eab0729305a425f6fef6d0
MrPlazu
Jr. Member
*
Offline Offline

Activity: 54
Merit: 0


View Profile
August 07, 2016, 07:53:58 PM
 #1007

I'm tring to get this version to Work on Ubuntu 16 with algo HMQ1725.


I can't test this anymore at Suprnova. I have an idea what's wrong, it would help if you can tell me the last version
where it worked. I should easilly be able to restore it.

You can use https://esp.maxminers.net.

I'm not sure of the last working version
felixbrucker
Hero Member
*****
Offline Offline

Activity: 630
Merit: 500


View Profile WWW
August 08, 2016, 07:04:45 AM
 #1008

hi joblo, im having a rather unusual problem

when running the miner (latest version in linux + 3.3.8 on windows tested) and using a socks proxy, after some time the miner loses the connection with

Quote
Stratum connection failed: Could not resolve host: lyra2re.eu.nicehash.com
...retry after 10 seconds

my proxy reports this:

Quote
sockd[11167]: info: block(3): tcp/accept ]: 10.1.0.47.36480 10.1.0.40.1080: error after reading 4 bytes in 0 seconds: eof from local client
sockd[11168]: info: block(3): tcp/accept ]: 10.1.0.47.36482 10.1.0.40.1080: error after reading 4 bytes in 0 seconds: eof from local client

such an error only happend before and directly when starting the miner if i did not specify my proxy was socks by "socks5://"

maybe you can look into it Smiley

br

joblo
Legendary
*
Offline Offline

Activity: 1148
Merit: 1016


View Profile
August 08, 2016, 01:56:42 PM
 #1009

hi joblo, im having a rather unusual problem

when running the miner (latest version in linux + 3.3.8 on windows tested) and using a socks proxy, after some time the miner loses the connection with

Quote
Stratum connection failed: Could not resolve host: lyra2re.eu.nicehash.com
...retry after 10 seconds

my proxy reports this:

Quote
sockd[11167]: info: block(3): tcp/accept ]: 10.1.0.47.36480 10.1.0.40.1080: error after reading 4 bytes in 0 seconds: eof from local client
sockd[11168]: info: block(3): tcp/accept ]: 10.1.0.47.36482 10.1.0.40.1080: error after reading 4 bytes in 0 seconds: eof from local client

such an error only happend before and directly when starting the miner if i did not specify my proxy was socks by "socks5://"

maybe you can look into it Smiley

br

I'm not sure how much I can help, I know nothing about socks. I also don't think it's a problem with the miner because that
has never been reported by anyone else.

I have seen a single failure on startup at times with other miners so I assume this is an issue with nicehash, or a bug common
to ccminer and cpuminer, either way protocol incompatibility of some kind that only occurs with nicehash..

If using socks causes new symptoms I would suspect a problem with socks and/or nicehash. I you don't see the disconnects at
other pools it may be the same nicehash incompatibility causing both errors.

cpuminer-opt developer, https://bitcointalk.org/index.php?topic=1326803.0
BTC: 12tdvfF7KmAsihBXQXynT6E6th2c2pByTT,
ETH: 0x72122edabcae9d3f57eab0729305a425f6fef6d0
felixbrucker
Hero Member
*****
Offline Offline

Activity: 630
Merit: 500


View Profile WWW
August 08, 2016, 02:44:59 PM
 #1010

hmmm i only had this issue if using socks, when connected directly the miner was running fine, socks is setup correctly and working fine with other services as well, will test if this also happens with other miners with nicehash to determine if its a miner or nh problem

will take some time as im leaving for vacations today though

br

joblo
Legendary
*
Offline Offline

Activity: 1148
Merit: 1016


View Profile
August 08, 2016, 02:57:28 PM
 #1011

hmmm i only had this issue if using socks, when connected directly the miner was running fine, socks is setup correctly and working fine with other services as well, will test if this also happens with other miners with nicehash to determine if its a miner or nh problem

will take some time as im leaving for vacations today though

br

Test with socks at other pools as well. If socks is protocol aware (not sure if it is) it may be specific to the stratum+tcp
protocol. But since I have seen stratum errors at nicehash with different miners and without socks I'm leaning more
toward nicehash. This is assuming the problem I'm seeing and yours have the same root cause.

In my case it reconnects promptly so it's not too disruptive.

cpuminer-opt developer, https://bitcointalk.org/index.php?topic=1326803.0
BTC: 12tdvfF7KmAsihBXQXynT6E6th2c2pByTT,
ETH: 0x72122edabcae9d3f57eab0729305a425f6fef6d0
felixbrucker
Hero Member
*****
Offline Offline

Activity: 630
Merit: 500


View Profile WWW
August 08, 2016, 03:12:39 PM
 #1012

for me it stays disconnected until restarted (that is killed and started again)

socks is just proxying tcp and udp connections, so its not protocol aware, will test when im back


br

MrPlazu
Jr. Member
*
Offline Offline

Activity: 54
Merit: 0


View Profile
August 08, 2016, 05:15:31 PM
 #1013

I'm tring to get this version to Work on Ubuntu 16 with algo HMQ1725.


I can't test this anymore at Suprnova. I have an idea what's wrong, it would help if you can tell me the last version
where it worked. I should easilly be able to restore it.

You can use https://esp.maxminers.net.

I'm not sure of the last working version

Any update on this?
joblo
Legendary
*
Offline Offline

Activity: 1148
Merit: 1016


View Profile
August 08, 2016, 07:06:50 PM
 #1014

I'm tring to get this version to Work on Ubuntu 16 with algo HMQ1725.


I can't test this anymore at Suprnova. I have an idea what's wrong, it would help if you can tell me the last version
where it worked. I should easilly be able to restore it.

You can use https://esp.maxminers.net.

I'm not sure of the last working version

Any update on this?

Maybe you could be a little more patient considering you declined my request for more info.
It will increase the chances I actually fix it next release. I'n not creating an account at maxminers
just for this. If I know what release broke it I will know exactly what broke it and how to fix it.

I'm working on some other optimizations, It'll be a couple of days before the next release.

cpuminer-opt developer, https://bitcointalk.org/index.php?topic=1326803.0
BTC: 12tdvfF7KmAsihBXQXynT6E6th2c2pByTT,
ETH: 0x72122edabcae9d3f57eab0729305a425f6fef6d0
MrPlazu
Jr. Member
*
Offline Offline

Activity: 54
Merit: 0


View Profile
August 08, 2016, 08:58:05 PM
 #1015

I'm tring to get this version to Work on Ubuntu 16 with algo HMQ1725.


I can't test this anymore at Suprnova. I have an idea what's wrong, it would help if you can tell me the last version
where it worked. I should easilly be able to restore it.

You can use https://esp.maxminers.net.

I'm not sure of the last working version

Any update on this?

Maybe you could be a little more patient considering you declined my request for more info.
It will increase the chances I actually fix it next release. I'n not creating an account at maxminers
just for this. If I know what release broke it I will know exactly what broke it and how to fix it.

I'm working on some other optimizations, It'll be a couple of days before the next release.

I'm happy to help fix this. I know that Windows version 3-3-8 and Linux version 3-3-5 is working perfect. I haven't tested all Windows and Linux version.

Let me know if you want me to test something specific.
joblo
Legendary
*
Offline Offline

Activity: 1148
Merit: 1016


View Profile
August 08, 2016, 09:56:01 PM
 #1016

I'm tring to get this version to Work on Ubuntu 16 with algo HMQ1725.


I can't test this anymore at Suprnova. I have an idea what's wrong, it would help if you can tell me the last version
where it worked. I should easilly be able to restore it.

You can use https://esp.maxminers.net.

I'm not sure of the last working version

Any update on this?

Maybe you could be a little more patient considering you declined my request for more info.
It will increase the chances I actually fix it next release. I'n not creating an account at maxminers
just for this. If I know what release broke it I will know exactly what broke it and how to fix it.

I'm working on some other optimizations, It'll be a couple of days before the next release.

I'm happy to help fix this. I know that Windows version 3-3-8 and Linux version 3-3-5 is working perfect. I haven't tested all Windows and Linux version.

Let me know if you want me to test something specific.

Thanks for the info. That's not what I expected. Are you sure 3.3.8 works? I made a change to hmq1275 in v3.3.8 and that is what
I suspected broke it. I have made no changes to this algo since. Which version is broken for you?

Edit: I made a change that affected many algos, including hmq1275, in 3.4.0. I tested several of the algos and none of them were
broken by it. I don't think this is what broke hm1275, I still suspect the 3.3.8 change.

cpuminer-opt developer, https://bitcointalk.org/index.php?topic=1326803.0
BTC: 12tdvfF7KmAsihBXQXynT6E6th2c2pByTT,
ETH: 0x72122edabcae9d3f57eab0729305a425f6fef6d0
felixbrucker
Hero Member
*****
Offline Offline

Activity: 630
Merit: 500


View Profile WWW
August 08, 2016, 09:56:56 PM
 #1017

@joblo

just remembered i was using your miner with optiminers hodl pool with the proxy just fine for several days straight, so it seems indeed to be nh related, i will try to get in touch with them, as a temporary resolution i will most likely just restart the miner if my app detects it is "hanging"

thanks for your help though

MrPlazu
Jr. Member
*
Offline Offline

Activity: 54
Merit: 0


View Profile
August 09, 2016, 04:22:00 AM
 #1018

I'm tring to get this version to Work on Ubuntu 16 with algo HMQ1725.


I can't test this anymore at Suprnova. I have an idea what's wrong, it would help if you can tell me the last version
where it worked. I should easilly be able to restore it.

You can use https://esp.maxminers.net.

I'm not sure of the last working version

Any update on this?

Maybe you could be a little more patient considering you declined my request for more info.
It will increase the chances I actually fix it next release. I'n not creating an account at maxminers
just for this. If I know what release broke it I will know exactly what broke it and how to fix it.

I'm working on some other optimizations, It'll be a couple of days before the next release.

I'm happy to help fix this. I know that Windows version 3-3-8 and Linux version 3-3-5 is working perfect. I haven't tested all Windows and Linux version.

Let me know if you want me to test something specific.

Thanks for the info. That's not what I expected. Are you sure 3.3.8 works? I made a change to hmq1275 in v3.3.8 and that is what
I suspected broke it. I have made no changes to this algo since. Which version is broken for you?

Edit: I made a change that affected many algos, including hmq1275, in 3.4.0. I tested several of the algos and none of them were
broken by it. I don't think this is what broke hm1275, I still suspect the 3.3.8 change.

Let me test 3.3.8 on Linux and see if it Works. Can you link to Google docs where I can download 3.3.8. Can only find 3.4.0 now
ctgiant
Sr. Member
****
Offline Offline

Activity: 298
Merit: 250


View Profile WWW
August 09, 2016, 06:35:17 AM
 #1019

I'm tring to get this version to Work on Ubuntu 16 with algo HMQ1725.


I can't test this anymore at Suprnova. I have an idea what's wrong, it would help if you can tell me the last version
where it worked. I should easilly be able to restore it.

You can use https://esp.maxminers.net.

I'm not sure of the last working version

Any update on this?

Maybe you could be a little more patient considering you declined my request for more info.
It will increase the chances I actually fix it next release. I'n not creating an account at maxminers
just for this. If I know what release broke it I will know exactly what broke it and how to fix it.

I'm working on some other optimizations, It'll be a couple of days before the next release.

I'm happy to help fix this. I know that Windows version 3-3-8 and Linux version 3-3-5 is working perfect. I haven't tested all Windows and Linux version.

Let me know if you want me to test something specific.

Thanks for the info. That's not what I expected. Are you sure 3.3.8 works? I made a change to hmq1275 in v3.3.8 and that is what
I suspected broke it. I have made no changes to this algo since. Which version is broken for you?

Edit: I made a change that affected many algos, including hmq1275, in 3.4.0. I tested several of the algos and none of them were
broken by it. I don't think this is what broke hm1275, I still suspect the 3.3.8 change.

The endian change broke hmq1725.  I just tested it on linux and replaced with the old code and it worked again.  I don't know what version you made that change on but I hope this helps isolate the issue for you.

the code I commented out:
Code:
       uint32_t endiandata[20] __attribute__((aligned(64)));
        uint32_t hash64[8] __attribute__((aligned(32)));

        flipend32_array( endiandata, pdata, 19 );
And the code I reverted:
Code:
uint32_t _ALIGN(32) hash64[8];
uint32_t endiandata[32];

int kk=0;
for (; kk < 32; kk++)
{
be32enc(&endiandata[kk], ((uint32_t*)pdata)[kk]);
};

MiningAlts Block Explorers and Windows/Mac Wallet Building Services - http://miningalts.com
Epsylon3
Legendary
*
Offline Offline

Activity: 1386
Merit: 1075


ccminer/cpuminer developer


View Profile WWW
August 09, 2016, 08:40:55 AM
 #1020

dont waste your time with this bad algo, their wallet is half broken at all level, and the price is reaching 1 LTC satoshi Wink not tradable so... soon only DOGE markets will be possible

BTC: 1FhDPLPpw18X4srecguG3MxJYe4a1JsZnd - My Projects: ccminer - cpuminer-multi - yiimp - Forum threads : ccminer - cpuminer-multi - yiimp
Pages: « 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 [51] 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 ... 191 »
  Print  
 
Jump to:  

Sponsored by , a Bitcoin-accepting VPN.
Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2006-2009, Simple Machines Valid XHTML 1.0! Valid CSS!