Quote from: decreped on April 11, 2018, 11:12:30 PM

Quote from: dodgertc on April 10, 2018, 11:56:49 AM

Quote from: dodgertc on April 06, 2018, 08:51:02 PM

Quote from: Sledge0001 on April 06, 2018, 08:18:02 PM

Quote from: magicsys.nw on April 06, 2018, 05:47:12 PM

I got this on 4 rigas with nevermore-brian112358-v0.2.2 with XR16. What this can be ?

============ Starting Miner ===============
*** nevermore 0.2.2 for nVidia GPUs by brian112358@github bin run sbin Built with the nVidia CUDA Toolkit 8.0 64-bits
Originally based on Christian Buchner and Christian H. project
Include some kernels from alexis78, djm34, djEzo, tsiv and krnlx.
Dev donation set to 1.0%. Thanks for supporting this project!
[2018-04-06 19:45:47] Starting on stratum tcp://stratum.threeeyed.info:3331
[2018-04-06 19:45:47] NVML GPU monitoring enabled.
[2018-04-06 19:45:47] 5 miner threads started, using 'x16r' algorithm.
/root/xminer.sh: line 69: 16587 Illegal instruction $MINER_PATH $MINER_OPTIONS
Miner ended/crashed. Restarting miner in 10 seconds --------------

Running 980Ti's by chance? I get the same issue on a 980Ti rig... Enemy however does work...

got the same issue (i reported some pages back already) with all rigs. 1050ti´s, 1060´s, 1070(ti)´s, 1080(ti)´s. all no difference.
enemy miner works.
nevermore, suprminer or tpruvot-v2.2.5 dont.

literally same command line (all these miners are ccminer derivates) just copied over. same error as you posted above. for RVN only enemy works. everything else (RVN x16r Nvidia) crashes with /root/xminer.sh: line 69: 16587 Illegal instruction $MINER_PATH $MINER_OPTIONS

What did i tried:

I entered the complete command line manually.
I analyzed in HEX the ascii codes, iE that the spaces are spaces (ascii 32)
i copied the exact enemy line to nevermore (its identical as both are ccminer derivates)
i copied the exacty suprminer 1.6 line which works under windows without issue to SmOS : error as above.
I created a NEW usb stick with SmOS
i checked if there is enough room on the USB stick (it is)
i turned off any overclocking
restarted the rigs after switching from enemy instead just reloading the miner.
i tried ravencoin mining using ccminer-tpruvot-v2.2.5-x16s-cuda8.0 and suprminer after they got added. It leads to the same error as with nevermore. But the same config works for enemy.

anyway i sent the report in this morning (CET) to admin@simplemining.net including acc+rigID´s.

hope they can find out why it affects some pps while others dont have any issue.

thanks to Tytanick and Mirelle we could figure it out today.
The fix was to add the -t behind. iE in my case using suprnova:

-a x16r -o stratum+tcp://rvn.suprnova.cc:6667 -u name.$rigName -p x -i 19 -t 5

[please note in case you missed this: these issues were in case enemy was working but nevermore, ccminer and suprminer were crashing. these are nvidia miners]

Funny enough under windows i didnt had to use -t

For some reason this fix didn't work for me. I still get the same error on both nevermore and suprminer. Like others I was only able to get enemy to work for RVN, but no update to enemy yet to allow x16s. Cry

Here is the code that I am using.
-a x16s -o stratum+tcp://pign.suprnova.cc:4096 -u decreped -p x -t 6
Any ideas?

unfortunately i can confirm that not all systems get "fixed" by this. (suprminer determines on most systems the threads required automatically (iE amount of GPU´s installed. in case it has issues doing so the -t is "fixing" it. maybe its related to weather theres an onboard graphics installed and the way its turned off/inactive) the -t parameter worked great for two rigs and i could switch them from enemy to suprminer after initially having thosde crashes. however one rig which is incomplete and thus work in progress (only 2 cards installed yet) is still crashing. even with -t. (tried also specifying devices etc)

this is how its invoked for the rig that is still crashing (the 2 card one that is incomplete).

to exclude any smos ui specific issues i went directly to the smos command line and invoked from there. the general command line is as follows for the 2 card rig

Code:

miner@simpleminer:/root/miner_org/suprminer-ocminer-v1.6$ ./ccminer -t 2 -a x16r -o stratum+tcp://rvn.suprnova.cc:6667 -u dodgertc.$rigName -p x

in addition i started it with -D debug parameter to see if i get a bit more information about whats going on. this is what it gives:

Quote

[2018-04-12 10:44:05] POOL 0: rvn.suprnova.cc:6667 USER dodgertc. -s 10
[2018-04-12 10:44:05] Starting on stratum+tcp://rvn.suprnova.cc:6667
[2018-04-12 10:44:05] restart_threads
[2018-04-12 10:44:05] CUDA GPU 0 matches NVML GPU 0 by busId 2
[2018-04-12 10:44:05] CUDA GPU 1 matches NVML GPU 1 by busId 3
[2018-04-12 10:44:05] NVML GPU monitoring enabled.
[2018-04-12 10:44:05] 2 miner threads started, using 'x16r' algorithm.
[2018-04-12 10:44:05] Binding thread 0 to cpu 0 (mask 1)
[2018-04-12 10:44:05] Binding thread 1 to cpu 1 (mask 2)
[2018-04-12 10:44:06] Stratum session id: deadbeefcafebabee524010000000000
Illegal instruction
miner@simpleminer:/root/miner_org/suprminer-ocminer-v1.6$

i may look to see how suprminer was compiled (what flags). as the system which is work in progress features an old pcie system built around some amd low power cpu (i use 1to4 adapters to attach up to 5 miners to it. I have multiple of these low-roi systems which dont give any issues otherwise). maybe it was compiled with some unsupported flags.

in any case it doesnt looks like it would be an smOS related issue.

dodgertc

Posts

Merits