Is there any chance to see Hawaii(gfx7) kernels for amdgpu-pro linux driver?
Thanks.
We already support Hawaii with optimized driver in both Windows and Linux (and we are always using amdgpu-pro drivers). Most probably you are using unsupported driver. Please check the first few pages of your log file for something like "Unknown OpenCL driver version! Hashrate and stale shares may suffer". In any case please let us know what Linux distro/version and which AMD driver version you are using.
Im running MSOS (ubuntu 16.04, 4.15.0 kernel, 17.5 amd driver) and ever since the ETH upgrade I have had 10 times the amount of stale and invalid shares at the pool. I tried everything including dialing back the OC/UV settings and nothing could get my previous stability back. No mater what I tried it was always 5% stales and 1% invalid shares (normal was 0.5% or less stales and a random invalid every so often). I switched to claymore and while its running at a slower hash Im getting a way higher overall usable hashrate and little to no stales/ invalid. this is only at the pool side the miner side shows way less stales and invalids than what the pool shows. Im thinking this is a linux only issue as my windows rig running 4.1c is fine no unusual stales or invalids, and the linux/MSOS rig running claymore is fine but phoenix on linux/MSOS is just trash on stability now.... If you got any tricks to fix it let me know if not maybe its something you guys can look into on the next release?
thanks for any help
Please check in your log file if the driver is properly recognized (it should be recognized as 17.50-511655). If you can find a line in the log file that looks like this:
Unknown OpenCL driver version! Hashrate and stale shares may suffer then the driver is not recognized. In such case please check the exact version of the AMD driver. We support 16.40, 16.60, 17.10, 17.20, 17.30, 17.40, 18.10, 18.20, 18.30, and 18.40 Linux drivers.
Understood, that makes sense, no worries.
Great to see the new beta, and trying it out right now. It still has the problem that I and a couple of others have reported about solo mode with GetWork - it still dumps out saying that a Stratum thread is not responding. This version is still not usable for solo mining using geth client. Version 4.0b works fine.
~# ~/phoenixminer/PhoenixMiner -gwtime 50 -rate 0 -pool http://10.69.1.10:8547 -coin clo -mi 14 -cdm 0 -amd -gpus 1 -acm -clKernel 0 -gpureset 1 -rmode 0 -log 0
Phoenix Miner 4.2a Linux/gcc - Release build
--------------------------------------------
Unknown OpenCL driver version! Hashrate and stale shares may suffer
OpenCL platform: OpenCL 2.1 AMD-APP (2766.4)
Available GPUs for mining:
GPU1: Vega 10 XTX [Radeon Vega Frontier Edition] (pcie 3), OpenCL 1.2, 16 GB VRAM, 64 CUs
Eth: the pool list contains 1 pool (1 from command-line)
Eth: primary pool: http://10.69.1.10:8547
Starting GPU mining
Eth: Connecting to http://10.69.1.10:8547 (GetWork/Solo)
Eth: New job #db436429 from http://10.69.1.10:8547; diff: 14057125MH
GPU1: Starting up... (0)
GPU1: Generating ethash light cache for epoch #73
Eth: Connected to ethash pool http://10.69.1.10:8547
GPU1: 44C 46%
Eth: New job #3f23092f from http://10.69.1.10:8547; diff: 14057125MH
Light cache generated in 1.0 s (24.5 MB/s)
GPU1: Using generic OpenCL kernels
GPU1: Allocating DAG for epoch #73 (1.57) GB
GPU1: Generating DAG for epoch #73
GPU1: DAG 39%
GPU1: DAG 78%
Eth speed: 0.000 MH/s, shares: 0/0/0, time: 0:00
GPU1: DAG generated in 3.8 s (419.4 MB/s)
Eth: New job #1e921adf from http://10.69.1.10:8547; diff: 14063984MH
Eth speed: 35.978 MH/s, shares: 0/0/0, time: 0:00
Eth: New job #ca64d685 from http://10.69.1.10:8547; diff: 14063984MH
Eth: New job #30a7b50e from http://10.69.1.10:8547; diff: 14070850MH
Eth speed: 36.000 MH/s, shares: 0/0/0, time: 0:00
Eth: New job #c1c2944c from http://10.69.1.10:8547; diff: 14070850MH
Eth: New job #68e9b9a2 from http://10.69.1.10:8547; diff: 14070850MH
Eth speed: 36.006 MH/s, shares: 0/0/0, time: 0:00
Eth: New job #905adf93 from http://10.69.1.10:8547; diff: 14077722MH
Eth: New job #da4b152d from http://10.69.1.10:8547; diff: 14077722MH
Eth speed: 42.021 MH/s, shares: 0/0/0, time: 0:00
Eth: New job #593c043d from http://10.69.1.10:8547; diff: 14077722MH
GPU1: 56C 75%
Eth: New job #c9a1abfe from http://10.69.1.10:8547; diff: 14077722MH
Eth: New job #a3a841f9 from http://10.69.1.10:8547; diff: 14077722MH
Eth speed: 42.021 MH/s, shares: 0/0/0, time: 0:00
Eth: New job #d541fa9f from http://10.69.1.10:8547; diff: 14084601MH
Eth: New job #2326c892 from http://10.69.1.10:8547; diff: 14084601MH
Eth speed: 42.021 MH/s, shares: 0/0/0, time: 0:00
Eth: New job #b95e9046 from http://10.69.1.10:8547; diff: 14084601MH
Eth: New job #e5503f0e from http://10.69.1.10:8547; diff: 14084601MH
Eth speed: 42.023 MH/s, shares: 0/0/0, time: 0:00
Eth: New job #3616fc00 from http://10.69.1.10:8547; diff: 14084601MH
*** 0:00 *** 3/12 11:42 **************************************
Eth: Mining CLO on http://10.69.1.10:8547 for 0:00
Eth speed: 42.021 MH/s, shares: 0/0/0, time: 0:00
Eth: Accepted shares 0 (0 stales), rejected shares 0 (0 stales)
Eth: Incorrect shares 0 (0.00%), est. stales percentage 0.00%
Eth: Average speed (5 min): 36.015 MH/s
Eth: New job #7c5650d1 from http://10.69.1.10:8547; diff: 14084601MH
Eth: New job #06b785bd from http://10.69.1.10:8547; diff: 14084601MH
Eth speed: 42.018 MH/s, shares: 0/0/0, time: 0:00
Stratum thread not responsive for 51 s
Thread(s) not responding. Restarting.
Thank you for the log! We have found the problem and it will be fixed in PhoenixMiner 4.2b, which will be out in a few days.
Can report my evga nvidia 2070s are not working with 4.2a keep getting error device kernel image is invalid (300), same error with -nvkernel set to 1 or 2 .. using latest nvidia drivers (417.xx)..
Thank you for reporting this, it will be fixed in PhoenixMiner 4.2b, which will be out in a few days.
Fans are controlled with overdriventool. Should I do anything in phoenix?
If you are happy with the way they work now, there is no need to change anything.